Dvoretsky I Yusupov •

Secrets of Positional Play

PmgressinCriess

Volume 25 of the ongoing series

Editorial board
GM Victor Korchnoi
GM Helmut Pfleger
GM Nigel Short
GM Rudolf Teschner

2008
EDITION OLMS

m
Mark Dvoretsky and Artur Yusupov

Secrets of Positional
Play

School of Future Champions 4

Edited and translated

by Ken Neat

2008
EDITION OLMS

m
4

Books by the same authors:

Mark Dvoretsky. Artur Yusupov. School of Future Champions

Vol. 1 : Secrets of Chess Training ISBN 978-3-283-00515-3 Available

Vol. 2: Secrets of Opening Preparation ISBN 978-3-283-00516-0 Available

Vol. 3: Secrets of Endgame Technique ISBN 978-3-283-00517-7 Available

Vol. 4: Secrets of Positional Play ISBN 978-3-283-00518-4 Available

Vol. 5: Secrets of Creative Thinking ISBN 978-3-283-00519-1 In Preparation

Mark Dvoretsky. School of Chess Excellence

Vol. 1: Endgame Analysis ISBN 978-3-283-00416-3 Available

Vol. 2: Tactical Play ISBN 978-3-283-00417-0 Available

Vol. 3: Strategic Play ISBN 978-3-283-00418-7 Available

Vol. 4: Opening Developments ISBN 978-3-283-00419-4 Available

Bibliographic information published by Die Deutsche Bibliothek

Die Deutsche Bibliothek lists this publication in the Deutsche

Nationalbibliografie; detailed bibliographic data is available in
the internet at http://dnb.ddb.de.

© 2008 Edition Olms AG

Willikonerstr. 1 0 · CH 861 8 Oe1wil a. S./Zurich

-

E-mail: info@edition-olms.com
Internet: www.edition-olms.com

All rights reserved. This book is sold subject to the condition that it shall not. by way of trade
or otherwise. be lent. re-sold, hired out or otherwise circulated in any form of binding or cover
other than that in which it is published and without a similar condition including this condition
being imposed on the subsequent purchaser.

Printed in Germany

Editor and translator: Ken Neat

Typeset: Arno Nickel · Edition Marco. D- 1 0551 Berlin

Printed by: Druckerei Friedr. Schmucker GmbH. D-49624 Loningen

Cover: Eva Konig. D-22769 Hamburg

ISBN 978-3-283-00518-4
 5

Contents

P reface (Mark Dvoretsky) . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . 6

PART I METHO DS OF IM PROVING IN POS ITIONAL P LAY

The Improvement of Positional Mastery ( Mark Dvoretsky) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

Prophylactic Thinking (Mark Dvoretsky) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
A Novelty is born (Max Dlugy) . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Positional Exercises (Mark Dvoretsky) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

PART II WAYS OF LOOKING FOR POSITIONAL SO LUTIONS

Manoeuvri ng (Artur Yusupov) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71

How to draw up a P l a n (Aiexey Kosikov) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
Sensing the Tempo (Aiexey Kosikov) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
Transformation of a Position (Mark Dvoretsky) . . . . . . . . . . . .... . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
. . . . . . . . . . . . . . . . . . . . 113

PART Il l T Y PICA L POSITIONS

Opposite-Colour Bishops in the Middlegame (Mark Dvoretsky) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 30

. . .

You can't get by without a Combination ! (Mark Dvoretsky) . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 57

Modern I nterpretation of the Dutch Defence (Igor Khenkin, Vladimir Kramnik) . . . . . . . . . . . . . . . . 1 62

PART IV COM P LICATED STRATEGY IN PRACTICE

Crux of the Position (Artur Yusupov) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 83
Strategy in Grandmaster Games (Evgeny Bareev) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 94
Whose Strategy will triumph? (Mark Dvoretsky) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216

PARTV

From Games by Pupils of the School (Artur Yusupov) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230

I ndex of Players and Analysts . . . . . . . . . . . . . . . . . . . . . . . . . . .. . ....

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . .. . . . . . . . . . . . . . . . . . . . 238
Index of Openings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
6
Mark Dvoretsky
P reface
I the series based on
am pleased to present the fou rth book in
material from the
Dvoretsky-Yusupov school for talented young
players. For those who are not familiar with
the previous volumes ( Secrets of Chess
Training, Secrets of Opening Preparation
and Secrets of Endgame Technique), I
should explain that we held several thematic
sessions of the schoo l , devoted to the most
important directions of chess improvement.
We did not have sufficient time to pass on all
the necessary specific knowledge, and in­
deed , th is could not have happened - the
process of chess development is practically
without limit. We set ourselves the aim of
disclosing the deficiencies in the pupils' play,
helping them to eliminate them , demonstrat­
ing the most effective ways of studying
chess, and acquainting them with the most
general mechanisms, ideas and methods of
playing. All the books in this series are based
on th is approach , the one before you being
no exceptio n . It is devoted to the improve­
ment of positional mastery.
Even adults sometimes naively bel ieve in the
existence of some mysterious key to rapid
success. The authors of many books happily
exploit this delusio n , asserti ng that they
know of such a single correct way - new,
original , and also h itherto secret. I n fact there
are many ways to the goal , but not one of
them is easy. You need to master various
methods of working on chess , and skilfully
combine them depending on your tastes and
individual traits, strength and style of play. I
hope that the present book, like the previous
ones, will help you to do this.
I n the first and second parts of the book the
authors acquaint the readers with various
aspects of positional play, approaches to the
development of positional mastery, and
methods of looking for positional solutions.
You will see that sometimes it even makes
sense to consider one and the same problem
in different ways - like , for example, the
conceptions of play on opposite wings i n the
lectures of Artur Yusu pov and Alexey Kosik­
ov.
Among the ideas developed in my own
lectu res , I advise you to pay particu lar
attention to the topic 'Prophylactic thinking ' .
Why this topic i s exceptionally important for
the over-the-board player is someth ing you
will understand after read ing the correspond­
ing lectu re.
Chess is a practical ski l l . Here theory alone
is i nsufficient - pu rposefu l tra i n i ng work is
also necessary (a very important principle of
effective work on chess ! ) . The prog ram of
each session of the school invariably includ­
ed not only lectu res, but also training
exercises. You will find a description of these
exercises in the first and th ird parts of the
book.
The session of the schoo l , on which the
material i n the g iven book is based , was held
i n early 1 992. Among those who took part in
it were the talented young masters (soon to
become grandmasters) Vladimir Kramnik
and Igor Khenkin . They not only attended
many classes, but also themselves read a
lecture. At first sight their lecture seems to be
purely about the opening. However, while
explaining the theory of certai n variations of
the Dutch Defence, at the same time
Kramnik and Khenkin expressed their under­
standing of the situations a rising here , and
the in herent strateg ic ideas. This is how
 Preface
modern players master typical positions,
characteristic of the openings they employ.
Another approach to the study of typical
middlegame positions is employed in my
lecture, included in the same th ird part of the
book.
The fou rth part is devoted to the purely
practical implementation of various princi­
ples of positional play. In it an analysis is
given of some strategically complicated
games, played in top-level events. Here too it
is interesting to compare the ways of th inking
and the approaches to the taking of deci­
sions of such outstanding g rand masters as
Artur Yusupov and Evgeny Bareev.
Incidentally, I should mention that Yusupov,
who in 1 99 1 took up residence in Germany,
was unfortunately unable to take part in the
final sessions of the school . His lectu res
were written later - when the book was being
prepared for publ ication . This factor allowed
Artu r to make use of games played two or
three years later, and in particular, two
brilliant wi ns by Viswanathan Anand over
Gata Kamsky in the Final Candidates Match ,
and some impressive games played by
Yusu pov hi mself at a tou rnament in Switzer­
land in 1 994 .
As for Bareev's material - this is indeed a
lecture given at the schoo l . It made a strong
impression on the pupils, not only through its
purely chess virtues, but also its u nusual
presentation - ironic, at times even caustic.
In my view, such a manner of del ivery, fully
reflecting the character of the g randmaster,
was an embellishment to the lectu re. It could
be perceived as being offensive only by a
person totally lacking a sense of humour.
7
After a l l , Bareev's irony is not at all malicious,
and in add ition it is di rected not only at his
opponents or l isteners, but also at hi mself.
Many years ago I saw a cartoon in which a
g rand mother was saying to a l ittle boy: 'And
now, g randso n , let's repeat some words
which you should never say. ' I remembered
it, when I looked through the traditional
concluding material by Yusu pov, analysing
fragments from games played by pupils of
the schoo l . On this occasion the g rand mas­
ter focused his attention on instructive
positional mistakes made by the young
players. In chess teach ing such a n approach
is q u ite appropriate. Not without reason do
they say: ' Learn from your mistakes. '
I n conclusion I have given several opin ions
expressed by legendary chess players,
emphasising the exceptional importance, for
any player, of the problems considered in the
book:
A sensible plan makes heroes of us all; the
absence of a plan makes us faint-hearted
fools. ( E m . Lasker)
Contrary to general opinion, generated by
ignorance, Morphy's main strength was not
his combinative gift, but his positional play
and general style. After all, a combination
can be carried out only when the position
permits it. (J . R . Capablanca)
The ability to evaluate a position is just as
necessary as the ability to calculate varia­
tions. ( M . Botvinnik)
Endeavour to memorise as few variations as
possible! Positional feeling should become
your release from the slavery of 'variations '.
And therefore: try to develop your positional
feeling. (A. N imzowitsch )
8
PART I
Methods of Improving
in Positional Play
Mark Dvoretsky
The Improveme nt of Pos ition a l Maste ry
Whereas by combination values are transformed, they are
proved and confirmed by 'position play'.
'
H game' - this is how we usually cha­
e has a subtle understanding of the
racterise a strong positional player. It is very
flattering to hear such a comment about our
own play, but, alas, not all of us can boast of
th is. And yet positional understanding is a
very important condition for our com petitive
success. What should be studied by a player
wishing to ach ieve serious prog ress in this
field , and what are the main directions and
forms of such work? These are the questions
that will he answered in this lecture.
You , of course, know that grandmaster
Yusu pov is renowned as a skilful strateg ist. I
will illustrate these thoug hts with some
examples from his games.
I . Some genera l recommendations
A number of books have been written about
positional play. By no means all of them
deserve attention, but some of them should
Emanuel Lasker
defin itely be studied. And above all Aaron
N imzowitsch 's remarkable book My System.
Recently I looked at some of the games
played in my youth , when I had fi rst category
rati ng, and I was staggered by the great
number of crude, obvious positional mis­
takes. But at the time they weren't obvious to
me. On the contrary, I was sure that I was
playing q u ite strongly - the only th ings that
h i ndered me were 'accidental' oversig hts,
and a lack of opening knowledge. For a time
I was qu ite unable to make any prog ress .
And then , on the advice of my trainer
(Alexander Roshal) I stud ied My System. I
didn't read it, but studied it, and I even copied
out the most importa nt ideas and exam ples.
This work qu ickly told on both the qual ity of
my play, and my resu lts - I won several fi rst
category tournaments , became a candidate
master, and then ach ieved the master norm .
I also recommend the study of collections of
games played by outstanding positional
 The Improvement of Positional Mastery
players, preferably with their own com ments.
You will realise, of course, that d ifferent
players have their own way of playing - a
universal positional style does not exist.
Some grandmasters ('strateg ists') are char­
acterised by their logical manner of th inking
(for example: Rubinstei n , Botvinnik, Par­
tisch ), while for others it is intu itive (Ca­
pablanca, Smyslov, Karpov). Also present in
their play are other disti nctions: they may
adhere to an attacki ng or defensive style,
they may aim for classical (with seizu re of
space) or less orthodox set-ups, and so o n .
Study the games o f a player w h o is closest to
you in style, or, on the contrary, one who is
especially skilled in what you are not good at.
I should like to remind you of the method with
which you should record the episodes that
are the most interesting and usefu l for you -
'positional sketches'. We recommended it in
the first session of our school (cf. the book
Secrets of Chess Training). Draw a diagra m ,
and write a commentary expressing the
essence of the position , the minimum neces­
sary for the understanding of its moves and
variations. As a result, the ideas and evalua­
tions related to this instance will be fi rmly
engraved in your memory. In exactly the
same way you can record instructive exam­
ples from your own games.
II . Positional operations
When assessing a position , an experienced
player never tries straight away to take
account of all its featu res, and does not
compile a complete list of all its pl uses and
minuses (although this is what is called for in
some books). If such work is done, then it is
done subconsciously. The Russian master
Ben iamin Blu menfeld, who made a deep
investigation of the problems of chess
th inking, wrote: 'Evaluation relates to per­
ception of the position and is largely a
subconscious action, in the sense that, to a
significant degree, if not fully, the intermedi-
l[j g
ate links do not pass through the centre of
consciousness. '
The art of evaluation is the abil ity to pick out
the essence of a position - those and only
those featu res of it, which should be taken
i nto consideration when search ing for the
strongest move. In training sessions it is
useful to express in words you r perception of
the essence of a position , i n order to note
and subsequently correct mistakes in posi­
tional understanding,
After picking out the most important assess­
ment considerations (most often - sub­
consciously) and checking the variations, we
find the move (or a small number of moves)
which corresponds to our perception of the
situation . Usually this move pursues some
defi nite aim, i . e . it is part of a specific
operation.
Of cou rse, often our decisions prove to be
difficult and complex, based on the calcula­
tion of lengthy variations and on numerous
different evaluative considerations. But nev­
ertheless, at the basis of any strateg ic
decision are simple positional operations,
and it is very important to be able to fi nd them
q u ickly and confidently. You can assi milate
them , by studying commentaries by grand­
masters, and it is very useful also to try
solving special exercises (for example, the
form of training games for the development
of intuition , which was discussed in a
previous session of the school - it will be
described in the next book).
We will pick out the basic types of positional
operations.
Improving the placing of the pieces,
manoeuvres, regroupings
Yusu pov - Ke ngis
Moscow 1 983
(see diagram)
10 w The I mprovement of Positional Mastery

White has excellent prospects on the king­ 36 . . . l:te8!

side. The attack will become irresistible if the 37 .l:lxd5
queen's rook, cu rrently out of play, can be
If 37 .l:tb7+ there follows not 37 .. .<.ti>f6? 38
included in it.
.l:tb6 , but 37 .. .<it> g8! 38 �h6 ltJf5 39 �g5
2 1 .l:.a3 ! 'it>h8 .l:.a8.
22 .l:.g3 37 . . . ltJe4
Th reatening 23 ltJxh6 gxh6 24 'iWh5 . If 38 �e1 l:i.h8
22 . . . .ltf8 , then 23 ltJxh6 gxh6 24 'i!i'g4 , while
The operation carried out by Black has
22 . . . .lta6 is hopeless i n view of 23 'ii'e4 .
secured h i m a n obvious advantage. His
22 . . . .l:!.g8 minor pieces have occupied active and at the
23 ltJc4 .ltd5 same time invul nerable posts, and his rook is
23 . . . .lta6 24 'iie4 . intending to i nvade the opponent's position .
24 .Uh3 .ltf8 39 'it>f1
25 b3 i.xc4? Better practical chances of saving the game
26 bxc4 dxe5 were offered by the piece sacrifice suggest­
27 'ii'x e5 ed by Robert H ubner: 39 ltJf3!? gxf3+ 40
'it>xf3 .l:.e8 41 b5. bpk
Black resigned .
39 . . . .l:!.h2
40 b5 .l:lb2
Djuri c - Yusupov
A standard elementary operation: the rook is
Sarajevo 1 984
placed behind the passed pawn .
(see diagram) 41 ttJe2 .l:tb1 !
Zugzwang! If 42 .l:lf5 , then 42 . . . i.g3 .
At first sight the situation is not easy to
42 b6 .Uxb6
assess. Black is more active, but his d5-
pawn is under attack. Both sides' pieces are 43 .lth4 .l:. b 1 +
somehow chaotically placed . 44 'it>g2 .l:.b2
Yusupov finds a regrouping which coord i­ The rook has again retu rned to its position of
nates his small force and enables him to maximum activity. It only remains to i nclude
create real threats. the king i n the play.
 The Improvement of Positional Mastery 11

45 \t>g1 \t>g6 move made by Black, but the assessment

46 l:ta5 g3! behind it. If 25 'ii'e 3 A rtu r was planning
47 l:ta8 �h5 25 . . . 4Jxe6 (not changing the pawn struc­
ture), yet after 26 'ii'x g5 hxg5 27 4Je2 the
48 .Jtxg3 4Jxg3+
position is roughly equal. Meanwhile Black
49 4Jxg3 �xg3
has the right to fig ht for an advantage, by
50 .l:!.c8 �g4 playing 24 . . . fxe6! (and possibly . . . e6-e5) or
51 .l:txc4 �f3 24 . . . 4Jf4 25 'i!ke3 fxe6 ! with the threat of
52 .l:tb4 llh2 26 . . . h5.
53 'it>g1 l:th8 However, the other, tactical idea, concealed
54l:ib2 l:.e8 behind 24 . . . 4Jf4 , was com pletely correct.
White resigned in view of 55 .Ub 1 �f2+ 56 This was the 'bait' that Rafik Gabdrakhanov
'i!tf1 l:!. h8. swallowed . He was tempted by the possibili­
ty of winning a pawn .

Pawn play; the formation of 25 .Ue8+? l:!.xe8

a favourable pawn structure 26 'i!Vxa6 h5!
Yusupov correctly judged that Black's attack
Many years ago, when Yusu pov was stil l just
would become extremely dangerous.
a candidate master, I commented in my
27 'ifxc6 l:te6
notebook, that 'Artu r finds it hard to decide
on sharp changes in the character of the 27 . . JJ.e7 was perhaps stronger, since now
play, and in particu lar on sharp pawn moves Wh ite could have played 28 'ifxc7 ! h4 29
and changes in the pawn structu re. ' To try �h2. After 29 . . . hxg3 30 fxg3 4Jxh3 31 gxh3
and correct this deficiency of his, we careful­ lie2+ (or 31 . . . 'il¥d2+) Black is guaranteed a
ly analysed every such episode from his d raw, but it is not clear whether he can hope
games. for more.
28 h4 'i!Vxh4
Gabdrakhmanov - Yusu pov
29 'ii'a 8+ �h7
Podolsk 1 976
30 b5?! 'ii'g 5
Wh ite resigned .

Of course, Artu r g radually rid h imself of the

weakness mentioned , and his play became
more dynamic.

(see diagram)

24 .l:txe6 4Jf4?!
In the given instance the 'dubious' sign
reflects not the objective strength of the
12 � The Improvement of Positional Mastery

Timman - Yusu pov

Candidates Match , 4th Game,
Tilburg 1 986

25 . . . l:tc5?
A vexing mistake. By playing 25 . . . l:te3! Black
would have retained his extra pawn. For
example: 26l"Llf1 l"Llf3+ ! 27 'itf2 l:.e5 28 l:tad 1
(28 l"Lle3 l:tgg5 ! ) 28 . . . l:txd5 29 l:txd5 l:lg5 ! , or
How should the position be assessed? For
26 .U.c1 'ite7 (26 .. .f6 ! ? ) 27 l"Llf1 .l:te2 28 l"Llg3
the knight Black has a sufficient equivalent ­
'iti>e6! 29 .U.cd 1 .l:.e 1 + ! .
three pawns. But if White should succeed i n
consolidating (l"Llf3 , 0-0 etc . ) , he w i l l have an 26 l:.xc5 bxc5
obvious positional advantage. Someth ing 27 l"Lle4 'ite7
must be urgently done. 28 l"Llxc5
19 . . . g5! And the game soon ended in a draw.
An excellent counter! If 20 g3 there follows
20 . . . gxf4 21 gxf4 l1g8. Both black rooks Spraggett - Yusu pov
become extremely active, and in addition the
Candidates Match, 9th Game,
opponent constantly has to reckon with . . . f7-
Quebec 1 989
f6.
20 0-0 gxf4
21 l:txf4 l"Llx e5
The wh ite centre has been completely
destroyed .
22 l:txd4 l:tg8
23 l"Llb4 l:!.c3
In the event of 24 .U.e 1 f6 Wh ite's position is
worse. In the endgame a knight which has
no strong points often proves to be
weaker than pawns. Therefore Jan Tim man
hurries to return the piece.
24 l"Llxd5!? exd5
25 i.ixd5 25 . . . b5!
 The Improvement of Positional Mastery ctJ 13

A typical blow at a centre with hanging

pawns. After the removal of the c4-pawn
Yusupov's pieces will dominate on the light
squares.
26 axb5 axb5
27 c5
In the event of 27 cxb5 Black has a pleasant
choice between 27 . . . �d5 and 27 .. J�xc2 28
'i'xc2 lZ:lxb5 .
27 . . . ltJc4
28 .l:ta2
28 i.f2 �xf3 .
28 . . . "iix a2 ! !
17 . . . �b41
29 'iVxa2 ltJxe3
1 8 �g1 ?1
The positional q ueen sacrifice has enabled
If 1 8 ltJe2 , then 18 ... 'ii h 4 1 9 �g 1 'il'e4 is
Black to destroy the wh ite centre. If now 30
u npleasant, but 18 ltJa2 or 18 'ii'd 2 should
'i'a5, then 30 . . . �xd4! .
have been preferred .
30 l:r.c1 �xd4
18 . . . �xc3 !
31 �h 1 ltJxc5
1 9 bxc3 a5
32 ltJxc5 J:txc5
What has Black achieved , by exchanging his
33 J:txc5 �xc5 bishop for the opponent's knight?
On the board it looks al most completely Firstly, his backward a6-pawn has been
black, and soon Kevi n Spraggett curtailed transformed i nto a dangerous passed pawn .
his resistance. The win i n this game brought At the same time the problem of his other
Yusupov overall victory i n their Cand idates backward pawn on c7 has been solved -
match . since now the c-file is closed .
Secondly, the remai n i ng black bishop is
Exchanges obviously stronger than its white opponent. It
In our analysis of the preced ing examples we will soon be switched to e4 and will help to
have already encountered elementary ex­ create th reats to the enemy king (with
chang ing operations. Even the positional opposite-colour bishops, playing for an at­
queen sacrifice made by Yusupov against tack is the correct strategy). By contrast, the
Spraggett was also essentially an exchange, wh ite bishop is ru nning up agai nst its own
although an unusual one. pawns and is not capable of attacking
anyth ing.
20 'iid 3?!
G h i nda - Yusu pov
A loss of time - all the same it is not possible
Olympiad , Dubai 1 986
to prevent the activation of the black bishop.
20 . . . 'il'd7
(see diagram) 21 �c1 �f5
22 'ii'd 2 h6
23 �a3 .l:.a6!
14 � The Improvement of Positional Mastery
Black has an obvious advantage.
Unfortunately, after 24 l:l.f1 Yusupov commit­
ted a serious inaccuracy: 24 . . . .l:.g6? (24 . . .
�e4! was correct, and only then . . . l:!.g6). His
opponent promptly exchanged the rooks:
25 l:.f3! �e4 26 l:tg3 llxg3 27 hxg3, and
thereby sign ificantly improved his chances of
savi ng the game. However, su bseq uently in
the endgame he nevertheless went wrong
and lost.
In the next example the problem that had to
be solved combined the problems of ex­
changing and the choice of pawn structure.
Yusu pov - Spasov
European Team Championship,
Skara 1 980
Yusu pov plan ned to exchange the bishop on
d5 - the opponent's only well-placed piece.
But the way he chose to carry out this
exchange was not the best.
He should have opened l ines in the centre, in
order to widen the front and press on the
vulnerable e6-pawn . The strongest conti nua­
tion was 1 3 'ii'e 2 ! t'Lld7 (after 1 3 . . . �xf3 1 4
'ii'xf3 d 5 1 5 l:tfc1 the white pieces control the
entire board) 14 e4 ! fxe4 1 5 �xe4 �xe4
( 1 5 . . . t'Llf6 1 6 �xd5 exd5 1 7 'ii'e 6+ 'it>h8 1 8
t'Llg5 'it'd7 1 9 .l:.fe 1 also offers Black l ittle joy)
1 6 'ii'xe4 Ilea 1 7 Ilac1 with an obvious
positional advantage.
U nfortunately, i n the game Yusu pov did not
play so dynamically, and did not change the
pawn structu re (perhaps he suffered a
recu rrence of the afore-mentioned 'child­
hood ill ness').
13 Ilc1 ? ! t'Lld7
1 4 �c4 t'Ll f6
1 5 'i!i'a4 a6
1 6 �xd5 t'Llxd5
1 7 .l:r.c2 h6?
Now Wh ite's idea is justified - he gains
control of the c-file. Without wasting precious
time, Black should have prepared to ex­
change the heavy pieces: 1 7 . . . b5! 1 8 'ii' b 3
'iVd7 1 9 Ilfc1 .Ufc8 . In this case Black could
have cou nted on eq ualising .
1 8 .l:.fc1 b5
1 9 'ii' b 3
1 9 'ii'a 3!?.
19 . . . l:tf7
Now if 1 9 . . .'it'd7 Wh ite has the un pleasant 20
t'Lld2! l:.fc8 2 1 e4 .
20 t'Lle1 !
This manoeuvre emphasises Wh ite's advan­
tage. The knight goes via d3 to b4 or f4 , in
order to exchange the knight on d5 - the
mainstay of the opponent's defences. Wh ite
went on to win .
I should mention that the inaccu racy on the
1 3th move was d iscovered in home analysis
by Yusu pov himself. It is very important to
instil in you rself a critical approach even to
your won games, to analyse them carefu lly
and, without being flattered by the decisive
result, look for errors in your own actions.
This will help you to avoid similar mistakes in
the futu re.
Prophylaxis
In previous sessions of the school I have
already emphasised several times the very
 The I mprovement of Positional Mastery lZ:J 15

important role of prophylaxis, 'prophylactic 1 8 e5!

thinking' in positional play. In this discussion A positional pawn sacrifice typical of such
about elementary positional operations it positions. Otherwise the opponent would
also cannot be avoided . After a l l , moves may have gained a comfortable game by . . . 'file?
be aimed not only at carrying out you r own and . . . tt:Jc5.
ideas ( manoeuvres, exchanges, and so on), 18 . . . dxe5
but also at preventing the opponent's ideas.
1 9 d6 .l:.c8
20 f5
20 'ika4? would have been a mistake in view
Yusu pov - Ti mman of 20 . . . exf4 ! 2 1 gxf4 (2 1 i.. xf4 tt:Jc5)
Candidates Match , 5th Game, 21 . . . .l:.xe3 ! . It was also bad to play 20 tt:Jd5?!
Tilburg 1 986 tt:Jc5 21 tt:Je7+ l:Ixe7 , but 20 l::t a d 1 ! ? came
i nto consideration.
20 . . . tt:Jc5
21 ..txc5!
Of course, the knight can not be allowed to go
to d 3 .
21 . . . .l:txc5
22 tt:Je4
22 l:tad 1 was also possible. By skilfully
combining offensive actions with prophylax­
is, Wh ite has obtained a promising position .

N u n n - Yus u pov
Linares 1 988
Yusupov is plan ning a pawn offensive in the
centre. However, he sees the manoeuvre
planned by the opponent in reply to e2--e4 :
. . . l2Jf6-g4--e5 , and the knight gains control
of the d3- and c4-squares. This manoeuvre
must be prevented .
1 3 h3! tt:Je5
An important improvement for Black was
demonstrated in the game Gelfand-Kasparov
(Novgorod 1 997): 1 3 . . . b5! 1 4 e4 'ikc8 ! ! , and
the threat of 1 5 . . . b4 is highly un pleasant.
1 4 tt:Jxe5 .l:lxe5
1 5 e4 .l:te8
1 6 .i.e3 lt:Jd7 Black is a pawn u p , but the enemy pieces are
The knight again dreams of going to e5. threateningly trained on the kingside. Both
1 7 f4! c4 wh ite rooks ca n q u ickly end up there. Now
1 9 i.. h 7+ is threatened .
16 � The I mprovement of Positional Mastery

The first desire is to remove the q ueen from Engl ish g ra n d master: after the n atu ral
the danger zone by 1 8 . . . 'it'c7 . But then there 25 ... l:tfb8?! there follows 26 l:tb3! l:t2xb3
follows 1 9 'ifh5, and the white q ueen joins (forced ) 27 cxb3, and the wh ite q ueen
the attack on the king. 20 .l:tg3 or 20 .l:.g4 is obtains use of the c-file.
threatened , and 1 9 .. .f5?! is dangerous: 20 26 h3 l:tfb8
.l:tg3 iixe5 2 1 ltJf3 'i!i'c3 22 'ili'xh6. Black is Now 27 l:.b3 is pointless - Black replies
forced to retu rn the pawn: 1 9 . . . ltJf5 20 .ll xf5 27 . . . a5! . Less convincing is 27 . . . ltJgxe5 28
exf5 2 1 'ifxf5 ltJc4 , but in this case too Wh ite ltJxe5 ltJxe5 29 'it'e3.
retains the better chances.
27 'ith2! a5
18 . . . 'ith8!
28 'itg3
Of course, on c3 the queen feels uncomfort­
But what does White want now? Obviously,
able, but from there, by attacking the knight
h3-h4-h5.
on d2, it restricts the mobil ity of the wh ite
queen and in general it rivets the opponent's 28 . . . .l:.c4!
attention, diverting him from the kingside. 29 h4? is no longer possible in view of
Remember: this defensive idea - restraining 29 . . . l:tbb4.
the opponent's activity with a far-advanced 29 c3
queen - was recommended in a lectu re by Another trap: the tempting 29 . . . d4 is refuted
Mikhail Shereshevsky, which he read at the by 30 ltJxd4 ltJxd4 3 1 'ii'f 1 ! . Even so, 29 ltJd2
first session of the school ( Secrets of Chess was more tenacious.
Training). 29 . . . a4!
19 g4? 1 30 ltJd2 ltJcxe5! !
1 9 ltJf3 was better. 3 1 1lla 1
19 . . . ltJac6! 3 1 ltJxc4 ltJxc4 3 2 l:te2 a 3 i s hopeless for
By attacki ng the e5-pawn , Black prevents Wh ite.
the opening of lines on the kingside by g4- 31 . . . l:!.c7
g5.
32 1t'xa4 .l:tbc8
20 ltJf3 !:tabS !
33 'ifa5 ltJc4
Again prophylaxis - this time against the
34 ltJxc4 .U.xc4
threat of 21 .ll c4 'ii'b 2 22 l:tb3. And if 21 g5
there follows 2 1 . . . l:tb4 ! . Black clearly has a g reat advantage and
subseq uently be successfully converted it.
21 ..tc4! 'ifb2
22 .ll b3 ltJg6!
Ill. Typical Positions
23 l:ta2
The black queen is nevertheless trapped , but In our games we constantly reach positions
the compensation for it will be more than with the same pawn structure , the same
sufficient. material balance and roughly the same
23 . . . l:txb3 arrangement of the pieces , as i n n umerous
games played earlier. It is useful to know
24 l:txb2 .:!.xb2
how strong players handled such positions,
25 'ii'c 1 ?! what plans they carried out and what ideas
25 .l:.b3 was better. they implemented .
25 . . . l:tb4! Many rules, relating to particular types of
Yusupov sees through the trap set by the positions, are well-known and are clearly
 The Improvement of Positional Mastery lZJ 17

formulated . For example: 'in open positions, 1 5 �h1 'ii' b 8

bishops are stronger than knig hts'; 'the The opening stage has developed log ically.
presence of opposite-colour bishops in the Wh ite has constructed a powerfu l pawn­
middlegame strengthens an attack' ; 'the side piece centre, and in reply Yusupov has
fighting against an isolated pawn i n the chosen a flexible set-up known as the
centre should aim for simplificatio n , to ' Hedgehog ' . In his position there are no
transpose into an endgame . ' weaknesses, and all the time the opponent
However, i n the arsenal o f grandmasters a n d has to reckon with the undermining pawn
masters there are also nu merous more moves . . . d6-d5 and . . . b6-b5. The chances
subtle, non-formal assessments. We under­ are roughly eq ual, and the outcome of the
stand that 'in this type of position you should su bsequent struggle depends on the strate­
act in this way' , but sometimes we fi nd it hard gic skill of the two players.
to formulate what exactly is 'th is type' of A year earlier Yusupov obtained the same
position. position with Wh ite in a game against Vitaly
The question of working on typical middle­ Tsesh kovsky (49th U S S R Championship,
game positions has already been touched on Fru nze 1 98 1 ). The events in it took a rather
in the second session of the school (cf. instructive course: 1 6 i.g 1 i.f8 1 7 b4! ? tt:Je5!
Secrets of Opening Preparation). There an ( 1 7 . . . d5? 1 8 cxd5 i.. x b4 1 9 dxe6 fxe6 20
important principle was emphasised : nowa­ 'ii' b 2 is u nfavourable for Black) 1 8 tt:Ja4 tt:Jfd7
days such work is unthinkable without linking 1 9 a3 Ji.a8 20 tt:Jb3 (20 'ike3!?). Now
it with the theory of the open ing from which Tsesh kovsky should have chosen between
the typical position arises. The reverse is 20 . . . 1i.c6 and 20 . . . tt:Jxc4!? 21 Ji.xc4 b5. But
also true: open ing studies will be effective he was tempted by the thematic advance
only if a deep acquaintance is made with the 20 . . . b5?, which in the given situation proved
ideas of the forthcoming middlegame. bad and led to a clear advantage for Wh ite
after 2 1 cxb5 axb5 22 tt:Jc3 tt:Jc4 23 'ii'a 2!
Taimanov - Yusu pov (th reatening 24 tt:Jxb5 ! ) 23 . . . tt:Jde5 24 tt:Jd4.
Len ing rad-Moscow Match 1 982 Yusupov's success in this game was due in
no small measure to the fact that he carefully
English Opening
watched for . . . d6-d5 and . . . b6-b5, and
1 d4 tt:Jf6 always had a worthy reply to the m .
2 c4 c5 1 6 'iif2 i.. d 8!?
3lt:\ f3 cxd4 An u nusual bishop manoeuvre. More often in
4lt:\ xd4 b6 such positions Black retreats it to f8 and then
5lt:\ c3 i.. b 7 after . . . g7-g6 develops it at g 7 .
6 f3 e6 1 7 tt:Jb3?
7 e4 d6 A passive move , which makes it hard for
8 i.e2 a6 Wh ite to carry out the usual plan of play on
9 i.e3 tt:Jbd7 the q ueenside: b2-b4, tt:Jb3 and at some
10 0-0 Ji.e7 point c4-c5. True, the i m mediate 1 7 b4?!
was prematu re because of 1 7 ... tt:Je5 18 tt:Ja4
1 1 11i'd2 0-0
d 5 ! . But it was possible, for example, to play
1 2 l1fd 1 l:r.c8
1 7 'iig 1 tt:Je5 ( 1 7 . . . 1i.c7 1 8 b4) 1 8 b3 h6 1 9
1 3ltac1 'iic 7 tt:Ja4, nevertheless i ntending to play b3-b4!
1 41i.f1 l:tfe8 at a convenient moment. I n the game
18 � The Improvement of Positional Mastery

An ikaev-Merkulov (USSR 1 982), where 24 tt:Jf2 was better.

Wh ite made these moves, Black carried out 24 . . . g4!
the central break 1 9 . . . d5? (Yu ri Anikaev 25 fxg4 e5
recommends 1 9 . . . tt:Jfd7 20 b4 g5!? 21 a3
26 ..te3 tt:Jxg4
'i!tg7). Alas, like the advance on the q ueen­
side made by Tsesh kovsky, it proved prema­ 27 tt:Jd5?
ture. After 20 exd5 exd5 21 c5 b5 22 tt:Jb6 This natural move is i n fact simply a loss of
..txb6 23 cxb6 'i!Vd6 24 'ii'f2 'iVxb6 25 tt:Jxb5 time, helping the opponent to include his
.l:.xc1 26 .l::tx c1 'ii'd 8 27 .i.b6! Wh ite gained dark-square bishop in the assault. 27 tt:Jf2
the advantage. was correct, aiming for simpl ificatio n .
17 . . . ..tc7 27 . . . .i.d8
18 'ili'g1 28 tt:Jf2 .i.h4!
29 :ee2
29 tt:Jxb6 tt:Jxb6 30 .i.xb6 tt:Jxh2 ! .
29 . . . tt:Jxe3
30 tt:Jxe3
30 !txe3 ..tg5.
30 . . . .i.xf2
3 1 'ii'xf2 ..txe4
The game is decided - Black has both an
extra pawn , and an attack.
32 tt:Jts tt:Jcs
33 tt:Jg3 .i.a8
33 .. .f5 ! , prepari ng 34 . . . tt:Jd 3 , was more ener­
getic. The inaccuracies by both players in the
18 . . . 'it>h8 ! ! concluding stage of the game a re explained
The start of a flexible plan of attack on the by the time scramble.
king side. 34 l:r.d1 tt:Je6!
19 .U.c2 .l:.g8! 35 'ili'xb6 tt:Jf4
20 .l:tcd2 g5! 36 .l:tf2 'il'h6?!
21 ..td4 :g6 Sergey Shipov recommends 36 . . . h5! 37
21 . . . h5 was probably more accu rate . .l:txf4 (forced ) 37 . . . exf4 38 'ii'd 4+ 'it'g7 39
'ii'xf4 'ii' e 5! 40 'ii' h 4 .l:!.g4 41 'i!i'h3 h4 and
22 tt:Jc1 ?!
wins.
Wh ite underestimates the danger facing h i m .
37 'i!t g1 'ii' h 4
He should have halted t h e black pawns by
playing 22 g4! . 38 'ii' b 3?
22 . . . .l:tcg8 If 38 l:.xd6 Yusupov had prepared 38 . . . tt:Jh3+!
39 gxh3 �g3+ 40 hxg3 �xg3+ 4 1 .i.g2
23 tt:Jd3
..txg2. However, as S h i pov pointed out,
23 g4 h5! 24 gxh5 :h6.
instead of 4 1 .i.g2? Wh ite can play 41 .ll g 2!
23 . . . 'ili'f8 ! ..txg2 42 'iif2 ! .i.xf1 + 43 'ili'xg3 .l:lxg3+ 44
24 .l:te1 'i!t xf1 with an u nclear rook endgame.
 The Improvement of Positional Mastery
38 . . . llh6
39 I:i.xf4 exf4
40 'ir'c3+ f6
41 tiJf5 .l:.xg2+!
42 i.xg2 'ifxh2+
White resigned .
Would it have been easy at the board to
devise such a plan: . . .<it> h8, . . . .l:tg8 and g7-
g5 ? Of course it wou ldn't, but in the g iven
instance there was also no need . Yusupov
was well familiar with a game by Bobby
Fischer, in which such a plan was apparently
carried out for the first time.
Fischer - Andersson
Siegen 1 970
Simagin-Larsen Opening
1 b3 e5 2 i.b2 tLic6 3 c4 tiJf6 4 e3 i.e7 5 a3
0-0 6 'i'c2 .l:te8 7 d3 i.f8 8 lLif3 a5 9 i.e2 d5
10 cxd5 tLixd 5 11 tLibd2 f6 12 0-0 i.e6
After playing the open ing of the game
experimentally, Fischer has set up (with
reversed colours) the same 'Hedgehog'
construction . Incidentally, this is a good
illustration of the u niversal natu re of many
strategic ideas - after studying the m , you will
be able to use them in the most varied
open ing lines.
lZJ 19
1 3 'it>h1 !! 'ii'd 7 1 4 .l:lg1 ! .l:tad8 1 5 ttJe4 'iff7?!
16 g4! g6?! ( 1 6 . . . t'Db6) 17 .l:tg3 i.g7 1 8
llag 1 t'Db6 1 9 t'Dc5 i.c8 2 0 t'Dh4 t'Dd7 21
t'De4 t'Df8 22 t'Df5 ! i.e6 23 t'Dc5 t'De7?
(23 . . . i.c8) 24 t'Dxg7 'it>xg7 25 g5! t'Df5 26
llf3 b6 27 gxf6+ 'it>h8 28 t'Dxe6 llxe6 29 d4!
exd4 30 i.c4 d3 31 i.xd3 llxd3 32 'ikxd3
l:.d6 33 'ii'c4, and soon Black resig ned .
IV. Typical Situations
Typical positions are cha racterised by a
purely chess pattern - by a particular
balance of force and its arra ngement on the
board , whereas typical situations are charac­
terised by the content of the forthcoming
struggle, derived from a general evaluation
of the position. For example, you can study
the laws of attack and defence, the convert­
ing of an advantage or manoeuvring in a
roughly equal position , playing for a block­
ade, the struggle for the in itiative . . . Also of
i nterest is the analysis of purely competitive
situations, such as playi ng for a win or for a
d raw, time-trouble, the encounteri ng of a
su rprise in the openi n g , the avoidance of
oversig hts, and so o n . I n previous sessions
we have already discussed many such
situations, and some of them have been very
thoroughly stud ied - for example, the prob­
lem of converting an advantage.
The rules and regularities that we set for
ou rselves in such an analysis are no longer
purely chess , but are rather concerned with
chess psychology and behaviour. Remem­
ber, for instance, the need to fi nd a concrete
solution at a key moment in the conversion of
an advantage and the essentially contradic­
tory recommendation 'do not h u rry' . Such
principles do not tell us directly what move
should be made, but they suggest the correct
direction of the search and help to create the
frame of mind which is most appropriate to
the situation .
20 � The Improvement of Positional Mastery

Yusupov - Timoshchenko 'i!i'xb5 g6. A new situation now arose - the

Kislovodsk 1 982 search for the correct response to an
Caro-Kann Defence opening novelty by the opponent.

1 c4 c6 1 3 dxe6 l2Jxe6
2 e4
This game was played in the USSR Team
Championsh ip, and the first challenge, in­
volvi ng the choice of open ing weapon , arose
before it began. When in his preparations
Yusupov looked to see how the opponent
repl ied to the English Opening, he noticed
that there was a possibil ity of transposing
into the Caro-Kan n Defence. Sergey Dolma­
tov, who was playing for the same team as
Artur, had played successfu lly aga inst this
open ing, he had h is own prepared l i nes in
the Panov Attack, and he was prepared to
share them with his friend . The problem was
that Yusupov hardly ever played 1 e4 and he 1 4 .ie3?
had no experience in the resulting positions.
A normal developing move - this would
Yet he had made a deep study of chess as a
probably have been played without th inking
whole and had analysed games played with
by almost everyone. But now Black gains
all the openings. If you possess a broad
time, by attacking the bishop on b5, to force
chess erudition , you can permit you rself (and
favourable exchanges.
sometimes it is even usefu l ) to take a risk in
the open ing. Cast an unprej ud iced glance at the position ­
2 . . . d5 doesn't it remind you of anyth ing? Wouldn't
you agree that we have seen someth ing
3 exd5 cxd5
similar in the old games of Paul Morphy and
4 d4 l2Jf6 Adolf Anderssen? The centre is ope n , the
5 l2Jc3 l2Jc6 black king is stuck in the centre (true, at the
6 l2Jf3 .ig4 moment Wh ite's is also there), and there is a
7 cxd5 l2Jxd5 pin on the a4-e8 diagonal as in the famous
8 'ikb3 .ixf3 Morphy-AII ies game. But do you remember
what you should do i n such situations?
9 gxf3 l2Jb6
Sacrifice if necessary, i n h ibit the opponent's
Another possibility for Black is the endgame development and at the same time complete
arising after 9 . . . e6 1 0 'i!i'xb7 l2Jxd4 1 1 .ib5+ your own development as quickly as possi­
l2Jxb5 1 2 'ikc6+! rl;;e 7 1 3 'ii'x b5 'i!i'd7 1 4 ble, bring you r rooks to the open files in the
l2Jxd5+ �xd5. centre and create a mating attack on the
1 0 d5 l2Jd4 enemy king .
11 .ib5+ l2Jd7 Alas, such opening strategy is someth ing
1 2 'iia 4 e5? that we have now half-forgotten , since in
In his preparations Yusupov had only reck­ modern set-ups the play is usually in a q u ite
oned with 1 2 . . . l2Jxf3+ and 1 2 . . . l2Jxb5 1 3 different key. I am in no doubt that without
 The Im provement of Positional Mastery
thinking Morphy would have played 1 4
� g5! ! , i n order t o place his rook on d 1 a s
quickly as possible. Black's position would
immed iately have become hopeless:
14 .. .<�J xg5 1 5 0-0-0;
14 . . . �e7 1 5 .i.xe7 �xe7 1 6 0-0-0 �d8 1 7
tt'ld5;
14 . .lL:lec5 1 5 �xd8 lL:lxa4 1 6 .i.xa4 .l::!.x d8 1 7
.
0-0-0 �e 7 1 8 :I he 1 (or 1 8 llxd7);
1 4 . . . '�xg5 1 5 .i.xd7+ rJ;;e 7. N ow 1 6 lL:le4 is
not bad , but the striving for rapid castl ing
may suggest an even more effective way: 1 6
f4! lLlxf4 1 7 0-0-0! .
14 . . . a6
1 5 .i.xd7+
If 1 5 .i.e2 there fol lows 1 5 . . . .i.c5 or 1 5 . . . b5
16 "it'e4 I1 c8 and 1 7 . . . �c5.
15 . . . 'ifxd7
1 6 "i*'xd7+ rJ;;x d7
A new situation has arise n , one that is rather
difficult to play. Yusupov worked it out
splendid ly.
1 7 0-0 !
I t i s wel l known that t h e favourable factors
operati ng in a position can be sub-divided
into constant (end uring) and temporary. It is
obvious that Wh ite has no constantly operat­
ing advantages - on the contrary, his pawn
structure is worse than the opponent's. His
ctJ 21
only chance consists i n the vulnerability of
the enemy king , which comes under attack
by the wh ite rooks. But this is a temporary
advantage, and if the opponent should
manage to consolidate, it will evaporate. The
flame of the i n itiative, which is about to go
out at any moment, must be skilfully fanned ,
a n d this demands extremely accurate and
dynamic play.
I n the endgame one is supposed to keep
the king closer to the centre , and therefore
1 7 0-0-0+ or 1 7 rJ;;e 2 suggests itself. But, as
you will already have real ised , here the
position should not be approached as if it is
a n endgame. The wh ite king reti res to the
kingside, in order to leave free for the rooks
all th ree c-, d- and e-files, on which the
opponent's king might be able to h ide.
17 . . . .i.d6
The next move seems obvious: 1 8 llfd 1 . But
what will the opponent do i n reply? -
Yusupov asked h imself. After some thought
he found a strong defensive manoeuvre:
. . . .l:ta8-c8-c6 followed by . . . 'it>c8 . For exam­
ple, 1 8 .l::tfd 1 ? ! I1ac8 ! 1 9 lLle4 I1c6 20 I1 ac1
.l:1d8! , and Black successfully com pletes his
development.
1 8 lLld5!
The g randmaster frustrates the suggested
arrangement of the forces. As you see, even
when fig hting for the i nitiative , prophylactic
operations are sometimes g iven preference
over attacking ones.
18 . . . .l::!. a d8
What to do now? In the event of 1 9 .i.b6 after
1 9 . . . .l:!.c8 Black will nevertheless play his
rook to c6 . And if 1 9 .l:i.fd 1 , then 1 9 . . . rJ;;c8 20
.i.b6 (20 �a7 b5!) 20 . . . �de8! (20 . . . l:!.d7? 2 1
l:!.ac1 + 'it>b8 22 lLle3 ! , a n d lLlf5 o r l:Ic2-d2 is
threatened ) 21 .i.a7 b5! 22 lL:l b6+ 'it>b7 23
l:txd6 ctixa7, and Black, at the least, stands
no worse. I n stead of 21 �a7 it is preferable
to play 2 1 lLlf6 ! ? gxf6 22 l:Ixd6 or 21 . . . .i.xh2+
22 'it'xh2 gxf6 23 l:Iac 1 + 'it>b8 24 nd7
'
22 � The I mprovement of Positional Mastery

retaining some pressure, but to Yusupov this
evidently seemed insufficient.
19 l:tac1 1
Again prophylaxis: Wh ite prevents the re­
treat of the ki ng. Now it does not have the c8-
square, while if 1 9 .. .<iot e8 the simple 20 .l::tfd 1
is strong (20 �b6 is less convincing:
20 . . . �xh2+ 2 1 'lt>xh2 .l::tx d5 22 l:tc8+ l2Jd8).
19 . . . l2Jc7
20 l2Jb6+ 'lt>e6
Perhaps at last it is time to place a rook on
d 1 ? No, it is still too early - Black will then
consolidate his position with 21 . . . g5! fol­
lowed by . . . f7-f6 and . . . �e5.
21 llfe1 !
I n the event of 2 1 f4 !? Wh ite would have had
to reckon with 21 . . . 'lt>f5 22 .l::tfd 1 l2Je6 23
lld5+ 'lt>g4, and the f4-pawn is attacked .
21 . . . 'iii> f5
22 .l::te d1
23 .U.xd6 l:txd6 24 .l:txc7 is threatened .
22 . . . 'lt>e6? is bad in view of 23 I:.xd6+! �xd6
24 �f4+, while after the knight moves the
wh ite rook invades with gain of tempo on d 5 .
22 . . . l2Je6
23 lld5+ 'it>f6
23 . . . 'lt>g6!? or 23 . . . ii.e5 !? 24 l2Jd7 f6 came
into consideration .
24 l:tcd 1 �c7
25 l2Jd7+ 'lt>e7
The battle to retai n the i n itiative has conclud­
ed successfully for Wh ite. With a series of
precise prophylactic moves Yusupov has
cramped the opponent's pieces and gained a
defin ite spatial advantage.
Now the character of the play changes -
there comes a phase of manoeuvring with
the aim of accu mulating add itional positional
pluses and g radually breaking up the enemy
defences. I should mention that for the
moment Wh ite's advantage is not yet decis­
ive and for success he requ i res a certai n
'cooperation' on t h e p a rt o f t h e opponent.
But, of course, he won't go wrong of his own
free will - i n this he must be hel ped .
26 b4!
The plan is clear: a2-a4 and then at an
appropriate moment b4-b5-b6 . Of course,
this is not fata l , but it is rather u n pleasant.
I n cidentally, the attempt to win a pawn did
not work: 26 ii.c5+? t2J xc5 27 t2Jxc5 l:.xd5 28
.l::tx d5 .l::td 8 29 .Uxd8 ii.xd8 30 t2J xb7 ii.b6.
26 . .. f6
27 a4 g5?!
By strengthening his control of the f4-point,
Timoshchenko fights against the advance of
the b-pawn . If 28 b5 he had prepared
28 . . . axb5 29 axb5 l:lhe8 30 b6 �f4 . Howev­
er, the cure proves worse than the ail ment ­
in Black's position there is now a real
weakness: the f6-pawn . The restrai ned
27 . . . l:.he8 followed by . . . 'lt>f7 was better.
28 ii.c1 ! ?
The th reat o f 2 9 �b2 i s highly u npleasant,
and in some cases .ta3 is also possible.
However, 28 ii.d4!? also deserved serious
consideration , forcing an advantageous ex­
change of minor pieces (28 . . . .:txd7? is not
possible on account of 29 llxd7+ 'lt> xd7 30
.txf6+ and 31 �xh8).
28 . . . l2Jf4?
Another mistake, provoked by Yusupov's last
move. 28 . . . .l::t h g8! was essential, preparing to
defend the f6-pawn with the rook from g6.
 The Improvement of Positional Mastery CtJ 23

29 i.xf4 .ixf4?! good knowledge of what to do in it. Also in

N ow Black loses a pawn. But also after the endgame a complicated plan may be put
29 . . . gxf4 White has a clear positional su peri­ into effect - if we know how the theoretical
ority. reference books recommend playing it. But a
30 l2Jxf6 ! multi-stage program of action found directly
at the board is a g reat rar;•y.
The concluding phase of the game is
How then in fact do players plan their play?
the conversion of the adva ntage already
achieved . Usually they decide only on a direction, a
very general cou rse of action. For example,
30 . . . .l:tc8?!
we real ise that the position req uires an
30 . . . h6 was more tenacious. offensive on the q ueenside, and we draw up
3 1lLlh5 an approximate outline of such an offensive.
3 1 ltJxh 7 ! ? was also possible. I n d ividual (as a rule - small) strategic opera­
31 . . . .ic7 tions are plan ned in more deta i l . If the
32l:Ie 1 + �f7 operations carried out are timely and good ,
33l:Id7+ 'itt g 6 and they contribute to eventual success,
when we look through the game in our eyes
34l2Jg7 ! ?
they unite into a consistent plan. Here is what
34 ltJg3 . grand master Alexander Kotov writes about
34 . . . .idS this i n his book Think like a Grandmaster.
34 . . . l:!.hf8 would have lost q u ickly: 35 .l:te6+ 'A unified plan in a game of chess is a series
J:l.f6 36 .l:!.xf6+ �xf6 37 .l:txc7 l:txc7 38 tLle8+ . of strategic operations, following one after
3 5 .l::Lx b7 .if6 another and each time carrying out a
36.l:l.e6 .l:.hf8 separate idea which arises from the de­
36 . . . .!:!.b8 37 .l:!.xb8 (37 .l:.a7! ? ) 37 . . . l:!.xb8 38 mands of the position on the board. '
ltJe8 'iitf7 39 .l:txf6+ 'it>xe8 40 l:txa6 was also I n the game which we will now examine, at
hopeless. first sight the outline of a un ified plan is
37l:Ixa6 visible: Wh ite closed the centre and the
q ueenside, after which he successfully con­
And Wh ite won easily.
ducted an attack on the king . How such a
plan is in fact constructed is someth ing that I
V. Planning will endeavour to demonstrate in the notes.
There is a widely-held opinion that the
highest strategic skill is the abil ity to encom­ Yusu pov - Rubi netti
pass virtually the enti re game with in a deep I nterzonal Tournament, Toluca 1 982
and integ ral plan, and that is how the lead ing Old Indian Defence
grandmasters th ink. Of course, this is a
1 d4 tLlf6
delusio n . It is pointless to decide on an ultra­
long program of actio n , when the very next 2 c4 d6
move may completely change the situation 3 tLlc3 tLlbd7
on the board and g ive the play a q u ite 4 tLlf3 c6
different directio n . Black is intending the development scheme
I t can happen that a position reached from . . . e7-e5, . . . .ie7 and . . . 0-0 . In carrying out
the open ing has been studied in detail by his ideas, a player usually endeavours to
chess theory and that we do indeed have a take i nto account, and if possible prevent the
24 � The Improvement of Positional Mastery
opponent's plans that are u npleasant for
him. Jorge Rubinetti did not play 4 . . . e5
immediately, in order not to al low a well­
known (and rather dangerous) wh ite devel­
opment scheme: 5 �g5 �e7 6 'ifc2 followed
by e2-e3 and �d3. Now if 5 ..ig5 he has the
good reply 5 . . . h6 6 ..ih4 g5 7 �g3 l2Jh5.
5 e4 e5
6 ..ie2 ..ie7
7 0-0 0-0
8 h3
The first small strategic idea (not cou nting
the choice of arranging the pieces in the
opening). Wh ite prepares �e3, in order then
to play d4-d5 and after the reply . . . l2Jc5 to
defend the e4-pawn by l2Jd2 , without sh ut­
ting in the bishop. It is with such operations
that any experienced competitor plans his
play. But, of course , this is not the only
possible approach to the position - accord­
ing to theory, the immed iate 8 d5 is also
strong.
8 . . . a6
Black wants to play 9 . . . b5, creati ng the th reat
of 1 O . b4 . Should White forestall the oppo­
. .
nent's activity by 9 a4 , in retu rn conced ing to
him the c5-point, or is the simple 9 ..ie3 b5
1 0 a3 better? I don't know the correct answer
to this question. Chess is a complicated and
imprecise game, and the choice is often a
matter of a player's style and tastes.
9 a4 a5
Otherwise Wh ite would have seized space
on the queenside with 1 0 a5.
1 0 �e3
In the event of 1 0 d5 l2Jc5 1 1 'i¥c2 Black
replies 1 1 . . . 'ikc7 , creating the th reat of
12 . . . cxd5 1 3 cxd5? l2Jcxe4 .
10 . . . lieS
A questionable move. It is useful in the event
of the pawn exchange on d4, but after
White's intended d4-d5 the rook would do
better to remain on f8 . 1 0 . . . exd4 1 1 l2Jxd4
l2Jc5, attacking the e4-pawn , looks more
logical (to defend it by f2-f3 with the pawn
already on h3 is very u ndesirable). Yusu pov
would most probably have replied 1 2 'ii'c 2 ,
1 3 l:tad 1 (with the threat of 1 4 l2Jxc6 bxc6 1 5
..ixc5), and then either f2-f4 followed by
�f3 , or l2Jf5 . Again I would not ventu re to
judge which is more important: Wh ite's
spatial advantage or the opponent's control
of the dark squares .
1 1 d5
Fully consistent (White has prepared this
move) and also in the style of Yusupov, who
l i kes positions with a spatial advantage.
11 . . . "ikc7
1 2 l2Jd2 l2Jc5
Wh ite has largely completed the develop­
ment of his pieces; now the time has come to
decide what to do next. With the pawn on a2
the usual plan is an offensive on the
q ueenside: l:. b 1 , b2-b3, a2-a3 and b3-b4.
But the pawn is on a4 and there is no play on
the q ueenside. The other standard advance
f2-f4 will further weaken the dark squares,
g ranting the opponent the excellent central
square e5.
The grand master found what was probably
the only promising plan (or more precisely ­
next strategic operation ). The preceding
operations of both sides were standard ones,
and the only d ifficulty was choosing between
 The Improvement of Positional Mastery
various possibil ities. But the idea fou nd now
by Yusupov was very u n u su a l , and wou l d
appear n o t t o have occu rred before in similar
positions - it is this that makes the game
special.
13 'it>h2! ..tdB
1 4 l:tg1 ! !
Wh ite prepares g2-g3 and then f2-f4 . (After
the immediate 1 4 g3 there was the reply
14 . . .'ii' d 7, whereas now it will be possible to
defend the h3-pawn with the bishop from f1 . )
I n the event o f the exchange on f4 , witho u t
conced ing any central sq uares Wh ite i n ­
creases his spatial advantage, and his r o o k
is excellently placed o n t h e newly-opened g ­
file. And i f Black avoids t h e exchange o f
pawns, there follows f4-f5 and g3-g4-g5 ,
and again t h e roo k stands where i t i s
needed .
This is the main plan of action , but Yusupov
also envisaged another deve l opment of
events. Wh ite can also play g2-g4 , and then
either o rgan ise a pawn storm on the king side
with g4-g5 and h3-h4-h5 (the rook will
support it from g1 ), or manoeuvre his knight
via the vacated f1 -square to f5 . In the event
of the exchange on f5 the rook will end up on
an open file.
We see that again o n e is not talking about a
clear-cut plan, since depending o n his
opponent's reactio n Wh ite can choose this or
that line of play. But at any event we are right
to ad mire the versatil ity of Yusupov's idea -
the moves he has made will prove u sefu l
whatever the development of events.
For h is part, Black did not manage to cou nter
his opponent's excellent play. However, the
plan , beg u n with his last move , of playing his
bisho p to b6 is q u ite logica l . Also after
1 3 . . ..tf8 1 4 l:tg 1 !! or 1 3 . . . h6 1 4 l:.g 1 !! lt:Jh7 1 5
.
g3 Wh ite would have stood better.
14 . . . 'it'b6?
It turns out that, i n o rder to prepare . . . ..tb6,
Black has decided to play his q ueen to a7 -
lLJ 25
an obviously u nsou n d idea . The suggestion
of Vlad i m i r Kra m n i k was fa r stronger:
14 . . . 'ii'd 7 ! , not only cleari ng the way for the
bishop to g o to b6, but also preventi ng g2-g 3
for t h e m oment and intending to meet 1 5 g4
with 1 5 . . . lt:Jh5! .
1 5 llb1
As the general plan is put into effect, the
placing of the individual pieces is made more
precise. 1 5 'il'c2 is weaker - from d 1 the
q ueen threatens the d6-pawn , and in addi­
tion it may later be needed on the d 1 -h5
diagonal .
15 . . . 'iia 7
In order to carry out his pla n , Black has had
to s h ut his q u een o ut of the game - a clear
sign that his idea is faulty.
1 6 g4!
Taking into account the poor placi ng of the
opponent's pieces, Yusupov chooses the
more agg ressive of his two plan ned offen­
sive o ptions.
16 . . . l:tf8?
Too passive (Black wants to support his d6-
pawn with his knight from e8). 1 6 . . . ..tb6 1 7
g 5 li:Jfd7 was more consistent. Apparently,
Rubinetti was afraid of losing a pawn after 1 8
li:Jf1 li:Jf8 1 9 dxc6 bxc6 20 'ii'x d6, but even by
playing simply 20 . . . ..tb7!? followed by . . .
:ad8 , . . . lt:Jce6 and . . . lt:Jg6, Black would
have gai ned fi ne cou nter-chances, and with
20 . . . Itd8 ! 21 'i¥xc6 (2 1 'ii'x e5 ..tc7) 21 . . . ..td7
he would have trapped the enemy q u een .
Yusupov was intending to prevent the ap­
pearance of the knight at g6 by 1 8 h4! li:Jf8 1 9
h 5 , and if 1 9 . . .lt:Ja6, then 20 l:.g3, preparing
an attack on the kingside and intending, of
cou rse, to captu re on e3 with the f-pawn .
1 7 li:Jf1 !
1 7 g5?! lt:Je8 was now inadvisable - Black
would have advanced his f-pawn , obtaining
cou nterplay. We once again see how flexibly
you should vary you r plans, depending on
the opponent's actions.
26 � The Improvement of Positional Mastery

17 . . . tt:Je8 fxg6 does not work because of 23 .. .f5 . The

1 8 lt:Jg3 f6 sacrifice can be prepared by 22 Wg4 l:td7 23
1 9 ttJf5 il.xf5 'i!i'h4 . Yusu pov was afraid that after 23 . . . ttJd3
24 Ji.g6 the opponent would prolong the
Positional capitu lation, but all the same
resistance by giving u p the exchange:
Black's position was already difficult.
24 . . . hxg6 25 fxg6 .Uff7 . However, as was
20 gxf5 Ji.b6 shown by g randmaster Patrick Wolff, Wh ite
wins by force with 26 'ilfh7+ 'it>f8 27 'ii' h 8+
'it>e7 28 gxf7 'it>xf7 29 l:txg7+! tt:Jxg7 30 .ll g 1 .
The winning method chosen in the game is in
no way worse.
22 il.h6! .ll d 7
23 Ji.xe8 :xeS
24 'it'h5
There was no need to calculate the i m medi­
ate sacrifice on g7, since the q ueen can be
brought u p with gain of tempo.
24 . . . 'it'b8
25 Ji.xg7!
Black resigned .
21 Ji.h5!
The entire game is an excellent, textbook
Rubinetti has only just completed his planned example on the theme of 'strategy'. The
bishop manoeuvre, but Yusupov is already events in it were explained not by variations,
embarking on the final attack. His bishop but by the plans of the two sides. Wh ite's
itself attacks something and at the same time rapid success was determ ined , on the one
clears the way for the q ueen . hand, by the deep plan found by Yusupov on
21 . . . lidS the 1 3th move, and on the other hand - by
Now a concrete path to the goal has to be the fau lty strategic idea, which his opponent
chosen. For the moment 22 Ji.g6 hxg6 23 began carrying out at the same time.

Mark Dvoretsky
P rophylactic Thi n ki n g
he idea of prophylaxis was put fo �ard
Tby Ni mzowitsch . He defines the a 1 m of
prophylaxis as being 'to blunt the edge of
certain possibilities which in a positional
sense would be undesirable. ' I n his famous
book My System Aaron N imzowitsch makes
a detailed exam ination of forms of prophylax­
is such as the over-protection of strateg ically
important poi nts and the prevention of
freeing pawn moves.
In N imzowitsch's opinion , the role of prophy­
laxis in chess is extremely important. He
writes: 'Neither attack nor defence is, in my
opinion, a matter properly pertaining to
position play, which is rather an energetic
and systematic application of prophylactic
measures. '
This opinion seems paradoxica l . I n position­
al play there are so many different aspects ­
is it admissible to reduce it to just one
element, even though a very important one?
Nimzowitsch's idea became closer and more
understandable to me after I stud ied the
following example.
l2J 27
Many years ago, when I was still a school­
boy, g randmaster Simagin set u p this posi­
tion and invited me to fi nd a winning move for
Wh ite . After some thought, I said that the
problem had no solution .
Vlad imir Pavlovich retreated the bishop to
a2.
'Can such a move really be winning?' I asked
in perplexity.
Try and fi nd a satisfactory reply. '
I tried , but I couldn't fi nd one. I remember that
this episode made a very strong impression
on me - for the first time I was able to sense
the power and beauty of qu iet positional
moves.
A few years later, tu rning over the pages of
the book by Paul Keres on the 1 948 World
Championship Match-Tournament, I came
across the familiar position . It occu rred there
in two of the games. The initial moves were 1
d4 d5 2 c4 e6 3 tt:'!f3 tt:'!f6 4 ttJc3 c6 5 e3
tt:'!bd7 6 i.d3 i.b4 7 a3 ..tas 8 'ilc2 'ile7 9
i.d2 dxc4 1 0 ..txc4 e5 1 1 0-0 0-0.
Reshevsky-Euwe : 1 2 d5 c5? 1 3 d6! with
adva ntage to White , since 1 3 . . . 'ii'x d6? 1 4
tt:'!b5 'ii' b 6 1 5 b4 i s bad for Black. However,
as Keres pointed out, Black could have
ach ieved an acceptable game in at last th ree
ways: a) 1 2 . . . ..tc7 1 3 dxc6 bxc6 ; b) 1 2 . . . ..txc3
1 3 ..txc3 cxd5 1 4 i.b4 tt:'!c5 1 5 i.xd5 tt:'!xd5
1 6 i.xc5 'ii'c 7; c) 1 2 . . . tt:'!b6 1 3 ..ta2 i.xc3 1 4
..txc3 tt:'!bxd5 1 5 i.xe5 i.g4 . I n the fi rst two
cases Wh ite's position would nevertheless
have remai ned preferable, but i n the th ird
one the chances are roughly eq ual.
Botv i n n i k-Euwe : Wh ite chose the logical
move 1 2 llae 1 , prepari ng the advance of his
e- and f-pawns, which is typical of such
28 � Prophylactic Thinking
positions (after the exchange on e5). The
opponent replied 1 2 . . . ..ic7, th reate n i n g
1 3 . . . e4 .
Now Wh ite ca n try to conti nue his plan with
1 3 ..id3 .l:te8 14 dxe5 tt:'lxe5 1 5 tt:'lxe5 'i!lxe5
1 6 f4 , but then 1 6 . . . 'ii' h 5, and Black is alright.
The prophylactic move 1 3 h3!? deserved
serious consideration, in order after 1 3 . . . e4
1 4 tt:'lg5 to parry the th reat of 1 4 . . . ..ixh2+.
Black would have repl ied 14 . . . tt:'lb6! 1 5 ..ia2
'i!Vd6 or 15 . . . �f5 with a complicated game.
Mikhail Botvi nnik chose 1 3 tt:'le4 tt:'lxe4 1 4
'ii'xe4 , and after 1 4 . . . a5? ! 1 5 �a2 tt:'lf6 1 6
'ii'h 4 e4 1 7 tt:'le5! he seized the in itiative .
Sounder was 1 4 . . . �d6 1 5 ..ic3 exd4 1 6
'ii'x d4 tt:'lf6 or 1 4 . . . tt:'lf6 1 5 'ii' h 4 e4 1 6 tt:'le5
..ie6 with roughly equal cha nces.
Thus in neither of these two games did Wh ite
demonstrate a convincing way to ach ieve an
advantage. The strongest continuation was
pointed out by Keres.
12 �a2 ! !
How can i t b e arrived at? Let us ask
ourselves what Black wants, and how he is
now intending to play. Obviously, not 1 2 . . . e4?
1 3 tt:'lxe4 . 12 . . . �xc3? is also clearly bad : 1 3
..ixc3 e4 1 4 tt:'le5 with the th reats of 1 5 ..ib4
or 1 5 f3 . The open ing of li nes in the centre by
1 2 . . . exd4?! 1 3 exd4 ! is also to Wh ite's
advantage - he qu ickly develops dangerous
pressure by .l:tfe 1 and tt:'le5 (or tt:'lg5). After
1 2 . . . .l:te8? the f7-point is weakened , while in
the event of 12 . . . h6? there is the un pleasant
reply 1 3 tt:'lh4, when the knight goes to f5 or
g6.
The only sensible move is 12 . . . �c7 ! , which
prepares 1 3 . . . e4 and thereby provokes the
opponent into releasing the tension i n the
centre. If it were possible to prevent this
move, Black would encou nter serious prob­
lems.
Now we can appreciate the true worth of the
modest bishop retreat. It is usefu l as regards
Wh ite's subsequent plans (the bishop is not
attacked after . . . tt:'lb6, and it can possibly go
to b1 ). But the main th ing is that now in reply
to 1 2 . . . �c7 he has 1 3 tt:'lb5 ! . For example,
1 3 ... �b8 14 �b4 c5 1 5 ..ixc5 ! , winning a
pawn . At the same time Wh ite creates the
positional threat of gaining the advantage of
the two bishops by 1 3 tLld5 and 1 4 ..ixa5. If
this is parried , say, by 1 2 . . . ..ib6, then he
strengthens his position with 1 3 l:.ae 1 , when
what Black should do is not clear. Perhaps it
makes sense to nevertheless try 1 2 . . . ..ic7 ! ?
('if i t is impossible, b u t you very much want
to , then you can ! ' ) 1 3 tt:'lb5 ..ib6! 1 4 ..ib4 c5.
However, after 1 5 dxc5 ..ixc5 1 6 l:tfd 1 ! and
the u navoidable tt:'ld6 his position is unenvi­
able.
The strong impression made by the g iven
example induced me to make a serious
study of the problem of prophylaxis in
genera l . Soon I switched my attention from
actual prophylactic moves to the process of
search ing for them - from the practical point
of view this seemed more important.
It became clea r that there was the need to
develop an approach to the position, which I
call 'prophylactic thinking' - the habit of
constantly asking yourself what the oppo­
nent wants to do, what he would play if it
were h i m to move, the abil ity to find an
answer to this question and to take account
of it i n the process of coming to a decisio n .
Developing t h e skill o f prophylactic th i n king
enables a player to make an enormous step
forward , and to g reatly raise his standard of
play. Why? I will single out two main reasons.
1 ) The range of positions in which prophylac­
tic th inking can be used is extremely broad.
Any sign ificant positional decision is bound
to combine the implementation of your own
plans with actions against the opponent's
(th is is how I understand the sense of
N imzowitsch 's statement g iven above about
the role of prophylaxis). The main principle of
converting an advantage is the restriction of
the opponent's possibilities; it is clear that
 Prophylactic Thinking lLJ 29

here one can n ot manage without prophylac-
.
tic th inking . By developing this trait, you also
become stronger in tactics and you make
fewer blunders. When defending a d ifficult
position, you must all the time see with what
you are th reatened ; and when attacking you
must reckon with the opponent's defensive
resou rces. Thus a possession of the skill of
prophylactic th inking exerts a favourable
influence o n practically all aspects of your
play.
2) Chess is a battle between two players with
equal rights, and your opponent's ideas may
be no worse than yours. It is logically clear
that the o ptimal strategy should harm o n i ous­
ly combine the implementation of your own
ideas and the preventio n of your o pponent's.
Of course, it is bad to remain passive and
merely destroy, but the opposite tendency is
also extremely dangero u s , and liable to lead
to constant failures.
Meanwh ile, players often forget to th i n k
about their opponent's p l a n s . T h i s is under­
standable: concentration on o ne's own feel­
ings is typical of human natu re . After a l l ,
sometimes i n life too , u nfortunately, w e d o
not take t o o much account of t h e tho u ghts
and feelings of others.
Thus it is not a matter, of cou rse, of giving
priority to destructive actio n s over creative
ones, but simply that the important skill of
prophylactic thinking is most probably insuffi­
ciently well developed i n us. By improving
this aspect i n which we are backward , and
making our th inking more harmonious, we
will certainly raise sign ificantly our overall
standard of play.
Of my own games on the theme of prophy­
laxis, the o n e that left the strongest impres­
sion was my encounter with Vitaly Tseshko­
vsky from the 1 975 USSR Spartakiad .
This was a famous match , in which the
Moscow team lost to Russia with a score of
%-8Y:z! A certai n folklore even developed
aro u n d it. 'We will g ive u p M oscow, but save
Russia', the spectators q uoted M . I . Kutuzov*
after the end of the match . And g rand master
Bukh uti Gurgen idze spread his hands in
astonishment: ' E ight and half, was this
deliberate? After all, there is a fi l m of this
name by Fel l i n i . ' With Wh ite against Lev
Polugayevsky, the Moscow team captain
Vasily S myslov made the opening moves 1
d4 lLlf6 2 c4 c5 3 d 5 . After losing the game,
Vasily Vasil ievich lamented : ' I was wrong to
play 3 d 5 ; I got carried away and over-rated
my positi o n . I should have played 3 lLlf3 ! ' .
B u t a t a team meeti ng after t h e match
Smyslov tried to reassure everyone: ' Never
m i n d , the most important thing is that
everyone is still alive . '
I a l s o made my 'contribution' to the defeat of
the M oscow tea m , but alas, it was my
opponent who demonstrated the strength of
prophylactic th inking .
Tsesh kovsky - Dvoretsky
Riga 1 975
French Defence
1 e4 e6
2 d4 d5
3 lLld2 c5
4 lbgf3 lbc6
5 exd5 exd5
6 ..tbS i.. d 6
7 dxc5 ..txc5
8 lbb3 ..td6
9 0-0 lbge7
In 1 974 I spent several days at a training
session with Victor Korch n o i , who was

*Russian a rmy com m ander M i khai l Kutuzov, po pu l arl y cred ited with saving Russia a g a i nst N a poleon's
invasion i n 1 8 1 2 . (translator's n ote)
30 � Prophylactic Thinking

preparing for his final Candidates match
against Anatoly Karpov. I remember that we
were analysing a similar position from the
same open ing variation, and I asked Victor
Lvovich why he developed his knight at f6,
and not at e7. The grandmaster looked at me
in su rprise.
'Let's stop and th ink about th is. How should
the pieces be arranged when you have an
isolated pawn? The place for the knight is at
f6 , and later - at e4 . And the bishop is best
kept on the g 1 -a7 diagonal - from there it
puts pressure on the f2-poi nt. In the main
variation Black plays 8 . . . ii.d6 and 9 . . . tt:'le7
simply on account of concrete factors (if
9 . . . tt:'lf6 there is the un pleasant reply 1 0
.l:r.e 1 + , while in the event of 8 . . . ii.b6 Wh ite will
immediately offer the advantageous ex­
change of bishops: 9 l:!.e 1 + and 1 0 ii.e3). But
if you can develop the pieces on their lawful
sq uares without being pun ished , you should
do so . '
Such evaluations, heard from the lips (or
read in the commenta ries) of top players,
sometimes help you to sense the subtleties
of opening strategy far better than lengthy
articles and specialised books .
1 0 .l:te1 0-0
queen on d8 Black would have captu red with
the bishop and the d5-pawn would have
been defended ) 1 6 tt:'lfd4 ii.xe2 1 7 Vxe2 a6
1 8 'ii'f3 llad8 1 9 .l:!.ad 1 . By carrying out
advantageous excha nges, Karpov has em­
phasised the weakness of the d5-pawn . He
now plans to i ntensify the pressu re on it, by
doubling rooks on the d-file and playing his
knight to e3.
12 h3 ii.h5
13 ii.xc6
Tseshkovsky 'takes the bull by the horns' -
he immediately tries to refute the opponent's
opening set-u p. Other, more restrai ned pos­
sibilities , are 1 2 i.. e 2 and 1 2 i.. h 4 (without
the inclusion of 1 2 h3) .
13 . . . bxc6
1 4 tt:'lbd4 l:tcB
1 5 c4 l:teB
Black prepares . . . f7-f6 . Later, also against
Tsesh kovsky (Sochi 1 975), Boris Gulko
played 1 5 . . . h6, and after 1 6 .ih4 g5
( 1 6 . . . 'i!i'c7 ! ? ) 1 7 ii.g3 .ixg3 he obtai ned a
good position . Wh ite chose a more critical
conti nuation in the game Peters-Ervin (Lone
Pine 1 978): 1 6 i.. x e7 !? i.. x e7 1 7 g4 ii.g6 1 8
tt:'le5 .

1 1 ii. g 5 ii. g 4 1 6 l:tc1 ? !

Many years later Artur Yusupov played this in It was probably better t o exchange pawns on
the first game of his Candidates match d5. S ubsequently the black bishop migh1
against Vasily lvanch uk (Brussels 1 99 1 ). have been able to go to d 5 .
lvanchuk repl ied 12 i.. h 4, preparing 13 i.. g 3. 16 . . . f6
After the game Korch noi asked Yusupov: 1 7 .ie3 'ii'd 7
'What, old chap, did you al low the bishops to 1 8 'ifa4 i.. f7
be exchanged? Back in 1 974 I realised that 1 8 . . . dxc4!? followed by . . . ii. h5-f7-d5 came
1 1 .. .'ii' c7 should be played . ' into consideration .
However, in the 1 8th game of the Karpov­ 1 9 c5 i.. b B
Korch noi match (Moscow 1 974) Wh ite was
20 b4 i.. h 5 ! ?
able to demonstrate what is in my view a
convincing plan for gaining an advantage: 1 2 I n t h e event o f 20 . . . tt:'l g 6 (intending . . . tt:'l e 5 ) I

c3 ( 1 2 i.. h 4? tt:'lf5) 1 2 . . . .ig4 1 3 h 3 ii.h5 1 4 was concerned about the reply 2 1 b5.
ii.e2 h6 1 5 .ixe7! tt:'Jxe7 (the drawback of the
(see diagram)
queen 's position at c7 is felt - the knight has
to be placed in a passive position ; with the
 Prophylactic Thinking
I took an optimistic view of this position . By
playing 21 . . . lZ:lf5 Black intends to develop
pressure on the opponent's kingside. If 22
lLlxf5 there follows either 22 . . . ..txf3 23 lZ:ld4
.te4 (intending 24 .. .'ii' c 7), or 22 . . . 'ii'xf5 23
lLld4 'ii'e 5 24 g3 (24 f4 'ii' c 7 with the th reat of
25 . . .l:be3) 24 . . . ..tg6 ! ? 25 lZ:lxc6 'i¥h5 .
I will make a slight digression . I g reatly
respected my opponent Vitaly Tsesh kovsky.
He was a deep, creative player, and the first
encounters with him in championships of the
country did not go my way. Tseshkovsky
played better, saw more and constantly
outplayed me. But our most recent meeting
in the Premier League of the 1 97 4 U S S R
Championship h a d followed a d ifferent sce­
nario. There I shared fifth place, but I could
even have fi nished th ird if in the last round I
had beaten Tseshkovsky with Black. And , it
has to be said , I was close to success. I n a
complicated strategic battle I outplayed h i m .
In a n overwhelming position and with m y
opponent in severe time-trouble, i t was only
a vexing, over-hasty move, involving an
oversight, which prevented me from achiev­
ing my goal (cf. the lectu re Transformation
of a position ' ) .
After this g a m e I decided that t h e era o f his
superiority over me had ended and that now
we were fig hting as equals. At the Spartakiad
I was no longer afraid of my opponent.
t2J 31
Having coped successfully with my opening
problems and seen how badly th ings were
going on the other boards, I decided to play
for a wi n .
I thought that Tseshkovsky's main strength
was in a fight for the in itiative, in dynamic,
attacking play, and that in positional play he
felt less confident than me. But he demon­
strated that he had a splendid mastery of all
types of chess weapons.
2 1 ..td2 ! !
A n excellent prophylactic move. Now 2 1 . . .
lLlf5?! i s pointless because of 2 2 l:txe8+,
when Black has to recaptu re with the bishop
(22 . . . l:txe8? 23 'ifxc6 , or first 23 lLlxf5).
Black's idea is frustrated , and he has to
readj ust and develop a new pla n . I n such
situations the probability of a mistake usually
increases sharply.
I probably should have played 2 1 . . . ..tg6!
followed by . . . ..te4 . But I conceived another
plan , involving the manoeuvre of my knight
to c4 . U nfortunately, I often used to carry out
my ideas too hastily, without due verification.
21 . . . ..tf3?!
22 lZ:lxf3 lLlg6
22 . . . lLlf5 ! ? .
23 .l:i.xe8+ l:.xe8
24 lLld4 lZ:le5?
A serious mistake, which effectively decides
the game. The correct move order was
24 . . . l:.c8 ! , intending . . . lZ:le5-c4 followed by
. . . ..te5, or, if the knight manoeuvre does not
prove feasible, then 25 . . . ..te5.
(see diagram)
However, even now it is not immed iately
apparent how to forestal l Black's plan
(25 . . . lZ:lc4 , then 26 . . . l:tc8 and 27 . . . ii.e5). 25
..tf4? fails to 25 . . . lZ:ld3. There is a pretty reply
to 25 f!c2 - 25 . . . lZ:lc4 ! 26 'ii'x c6? lZ:lxd2 ! . If 25
l::te 1 , then 25 . . . f!c8 . It is probable that here
32 � Prophylactic Thinking
- position after 24 lL:Je5? -
. . .
too White retains the better chances after 26
'ii'a 6! (but not 26 �f4? tLld3) 26 . . . <it>f7
(26 . . . ttJc4? 27 .l:.e 7 ! ) 27 ..tf4 tLlf3+ 28 ttJxf3
..txf4 , but the conti nuation found by Tseshko­
vsky is far more convincing.
25 .l:.d1 ! !
A second brilliant prophylactic move! The
bishop is defended , and if 25 . . .ttJc4 there
follows 26 'ii'xc6. But perhaps I should
nevertheless have gone into the rook ending
a pawn down , which arises after 26 .. .'ii' xc6
27 tt:Jxc6 ttJxd2 28 ttJxb8 l:txb8 29 11xd2
.U.xb4 30 .l:Ixd5 i1c4 .
25 o o o l:tc8
26 ..tf4!
Now 26 . . . lLlc4 is bad in view of 27 �xb8 tLlb2
(27 . . . .l:.xb8 28 'i!Vxc6) 28 'i!Vb3 ttJxd 1 29 �g3
ttJxf2 30 �xf2 . 26 . . . tLlf3+ 27 ttJxf3 �xf4 28
b5! is also unfavourable for Black. Thus my
plan has been disrupted , and White can
calmly strengthen his position. Apart from
the objective difficulties , one should not
forget about the psycholog ical effect of your
own plans being wrecked . It is not surprising
that Black loses without a fight.
26 0 0 0 'ii' b 7
Probably Black should have looked for
counter-chances in the variation 26 . . . g5!? 27
�g3 f5 .
27 'ii' b 3 a6
28 l:te1 g6
29 ..tg3 ctm
30 'iWe3
There is no defence against the i nvasion on
the e-file.
30oo .'ii'd 7 31 f4 ttJc4 32 We6+ 'ii'x e6 33 l:lxe6
lLlb2 34 1:lxc6 .Uxc6 35 ttJxc6 �c7 36 <it>f2
a5 37 bxa5 ttJd3+ 38 �e3 ttJxc5 39 <it>d4
ttJe4 40 <it>xd5 ttJxg3 41 a6 �b6 42 a4 �e8
43 a5 �f2 44 ttJd4 Black resigns.
How can you develop prophylactic th inking?
I n the first instance it is important simply to
pay attention to this problem . Remember
instructive episodes from your own play, and
from the games of other players; look for
them in books and magazines. Try to think
prophylactically in you r coming tou rnament
battles. Concentrati ng your attention on a
pa rticular problem helps you to g radually
resolve it, and ach ieve serious prog ress in
the given field . I remember how, after
attending a lectu re of mine about prophylax­
is, Boris Gelfand beca me interested in this
topic and later he proudly showed me some
su btle prophylactic moves which he had
been able to fi nd at the board . In 1 988 Smbat
Lputian won the second prize at an excep­
tionally strong international tou rnament i n
S a i n t Joh n . He admitted that d u ring t h e play
he had repeatedly used the idea of prophy­
laxis, about which he had been chatti ng with
me before the event.
Clearly, it is very usefu l to study the games
and commentaries of g reat players, who are
especially skilled in the sphere of prophyl­
axis. In particular - Tigran Petrosian and
Anatoly Karpov. In contrast to the majority of
their colleagues, it would appear that from
natu re they were endowed with highly
developed prophylactic thinking .
The followi ng game by Ka rpov (and to no
lesser extent his comments on it) vividly
 Prophylactic Thinking
illustrate the style of play and way of th i n king
of the then world champion. It should be said
that when I saw the game for the fi rst time it
did not make much of an impression ,
because Black lost without a struggle. And
only later, when I read Karpov's comments,
did I real ise how much su btle work was
concealed beh ind outward simplicity.
Karpov - Ti mman
Montreal 1 979
Pirc-Ufimtsev Defence
1 e4 d6
2 d4 tZ'lf6
3 lZ'lc3 g6
4 g3 .ig7
5 i.. g 2 0-0
6 lZ'lge2 e5
7 0-0 lZ'la6?!
Theory recommends 7 . . . lZ'lc6 or 7 . . . c6 . After
7 . . . c6 the usual reply is 8 a4 (preventing
. . . b7-b5, a useful move for Black), but then
the development of the knight at a6 becomes
more justified , since the b4-square is availa­
ble to it.
8 l:te1 c6
9 h3
'A typical prophylactic move i n such situa­
tions,' writes Karpov. 'Wh ite restricts the
opponent's bishop, and at the same time he
also creates a ' no-go' area on the kingside
for the remaining minor pieces . '
The remaining pieces' - this obviously
means the knight on f6 . The wh ite bishop
wants to got to e3, and it must be safeguard­
ed against the attack by . . . lZ'lg4 . The
advance of the pawn to h3 is typical in such
positio n s , and it is played without even
thinking about the opponent's possibilities .
9 . . . .l::!. e 8
'All Black's hopes of obtaining counterplay
are associated with pressure on the e4-
pawn . '
4:J 33
This is a l ready prophylactic th inking. Karpov
i m med iately defines the opponent's main
idea , on which he will keep a careful watch
th roughout the entire game.
10 i.. g 5
A standard idea - Wh ite provokes . . . h7-h6,
i n order to then develop his q ueen at d2 with
an attack on the h6-pawn.
10 . . . h6
1 0 . . . exd4 1 1 lZ'lxd4 h6 is tempti ng, hoping for
1 2 i.. e 3 lZ'lc5 , when it is awkward to defend
the e4-pawn ( 1 3 i..f4 tZ'lh5). But Wh ite replies
1 2 .if4 ! g5 1 3 .ic1 , and the weakness of the
f5-square will su bsequently tel l .
'After 1 0 . . . 'i!Vb6 Wh ite could have gained a n
advantage b y the simple 1 1 .l:!.b1 , since
1 1 . . . exd4 1 2 lZ'lxd4 lZ'lc5 1 3 b4 lZ'lcxe4 is
clearly unsatisfactory in view of 14 i.. xe4!
lZ'lxe4 1 5 lZ'lxe4 i.. x d4 1 6 lZ'lf6 + . '
A not altogether convincing comment! Of
cou rse, the e4-pawn should not be captu red
- the normal reply is 1 3 . . . lZ'le6 . Besides,
instead of 1 2 . . . lZ'lc5 Tim man could have tried
either 1 2 . . . d5, or 1 2 . . . lZ'lg4 !? 1 3 hxg4 i.. x d4
1 4 i.. e 3 i.. x e3 1 5 l:txe3 d5! with unclear play.
This variation has its i n ner logic: the bishop
at g5 is hardly participati ng in the battle for
the centre , and Black exploits this factor.
1 1 i.. e 3
Wh ite obviously wants t o complete his
 Prophylactic Thinking
development by 1 2 'ili'd2 (with gain of
tempo! ) 1 2 . . . 'it>h7 1 3 l:lad 1 . Is it possible to
hinder the implementation of his plan?
I n the event of 11 . . . exd4 12 i.. x d4 ! �c5?!
( 1 2 ... �c7 is better) there is the u npleasant
1 3 e5! �h7 1 4 exd6 'ii'x d6 1 5 b4 (or 1 5 i.. x g7
'ii'x d 1 16 l:.axd 1 'iti>xg7 1 7 b4 �a6 1 8 b5)
15 . . .�a6 16 b5 cxb5 1 7 �xb5 with advan­
tage to White.
The prophylactic move suggested by Karpov
- 1 1 . . . 'iti>h7 ! ? , defending the h6-pawn in
advance, deserved serious consideration .
The point is that if 1 2 'ii'd 2 there follows
1 2 . . . exd4! 1 3 i.xd4 ( 1 3 �xd4 �c5 1 4 f3 is
stronger, and if 1 4 . . . d5, then 1 5 �xc6) 1 3 . . .
�c5 , and 1 4 e5? i s not possible because of
the pin on the bishop after 1 4 . . . dxe5.
By conti nuing 1 2 g4 (instead of 1 2 \\Vd2),
Wh ite would have retained the better chanc­
es. But at the least Black would not have lost
anything: while making a useful move , he
has hindered the most natu ral arrangement
of the opponent's pieces, and set him a
problem. A good example of the solving of a
strategic problem in the opening on the basis
of prophylactic th inking .
11 . . . "ikc7?
Jan Timman does not even try to fight
against his opponent's plans. I n addition , the
black queen takes away a square from the
knight, which is now stuck on the edge of the
board .
1 2 'ii'd 2 'iti>h7
After 12 . . . exd4 1 3 �xd4 Black has to spend
a tempo on parrying the th reat of 14 i.. x h6.
1 3 llad 1 i.d7
1 3 . . . i.e6 came into consideration, and if 1 4
g4 l:tad8 1 5 f4 i.. c4 1 6 �g3 (Karpov), then
16 . . . exf4 1 7 i..xf4 c5.
(see diagram)
Wh ite has successfully consolidated . Now
14 f4 or 14 g4 suggests itself. How did
Karpov make his choice?
- position after 1 3 i.d7 -
. . .
'On the threshold of the midd legame it is
always useful to weight u p the resou rces of
the two sides, and to make adj ustments to
one's initial plans. Here I sank into thought,
and I soon came to the conclusion that
straightforward play in the centre would not
get me anywhere . Now, with the aim of
seizi ng fresh territory, the kingside pawns
must be advanced . But I did not want to play
1 4 f4 i m mediately. It would be illog ical to
increase the tension straight away - all the
same White will subsequently have to play
g3-g4 , so why not fi rst utilise a resou rce for
strengthening the position, such g3-g4 and
�e2-g 3 . At the same time Wh ite also solves
his main strategic problem - in the event of
the centre being opened , he supports his e4-
pawn . '
I n cidentally, the game i s an excellent illustra­
tion of N imzowitsch 's idea of the over­
protection of strategically important points.
Karpov all the time remembers about this,
endeavouring to g ive maximum support to
his central e4-pawn .
1 4 g4 l:tad8
1 5 �g3 i.c8
16 f4 b5
Knowing Karpov's style, it is very easy to
guess his next move .
1 7 a3! b4? !
 P rophylactic Thinking
Timman is an active player, sometimes
excessively active . It was hardly good to
weaken the q ueenside pawns.
18 axb4 lt'lxb4
1 9 lt'lce2
'Black's idea was to . . ' For heavens' sake,
.
what does it matter about Black's idea! On a
first glance at the position it is clear that
White has gained a solid positional advan­
tage, and we would probably have been
thinking about how to exploit it as soon as
possible - whether to play f4-f5, prepare g4-
g5, and so o n . But Karpov th inks completely
differently - even in such situations he fi rst of
all mon itors the opponent's possible inten­
tions and endeavours to forestal l them .
So, once more: 'Black's idea was to some­
how bri ng his pieces together, by playing
. . . a7-a5, . . . ..tc8-a6, . . . e5xd4 and . . . c6-c5.
But this is a lengthy process, and Wh ite
succeeds in h indering his opponent's pla n . '
N ow i f 1 9 . . . a5 there follows 20 c3 , d riving
away the knight and supporting the centre.
Black also has a bad position after 1 9 ... c5 20
fxe5 dxe5 2 1 d5.
19 . . . exd4
20 lt'lxd4 aS
'White's su bsequent play essentially reduc­
es to preventi ng the opponent's pieces from
coming into play. '
21 c3 lt'la6
l2J 35
Karpov's next move is probably the best in
the game. How did he find it? Obviously he
asked himself what the opponent wanted .
The answer is clear: to bring the knight into
play via c5. How can this be prevented?
22 'ii'c 2 ! 1
'A subtle move, which, firstly, prevents the
black knight from moving to c5 (in view of the
reply b2-b4 ! ) , and secondly, continues the
previous strateg ic policy - the supporti ng of
the e4-pawn . '
A s you see , here two forms o f prophylaxis
mentioned by N imzowitsch are combined -
prevention of the opponent's plans and over­
protectio n .
It is curious that Ka rpov does not even
examine 22 lt'lxc6!? 'i¥xc6 23 e5 - a
continuation which players nowadays would
certainly seriously consider. Why? - in this
case the enemy pieces would have become
active. However, after 23 . . . 'ilkc7 24 exf6 .i.xf6
25 lt'le4 followed by .i.d4, or 23 . . . d5 25 exf6 ,
Wh ite would still have retained t h e better
chances.
22 . . . .i.d7
Agai n Black wants to play . . . lt'lc5 (23 . . . lt'lc5
24 b4 axb4 25 cxb4 lt'le6, and the c6-pawn is
defended ), and again Wh ite prevents th is.
23 lt'lf3 ! l:l.e7?!
If 23 . . . lt'lc5! ? Ka rpov had prepared 24 e5.
Timman bel ieves his opponent, but he was
apparently wrong to do so. In difficult
positions you should ca refully check any
possibil ity of sharpening the play. I do not
see how 24 . . . lt'ld5 is refuted . If 25 exd6
'ilkxd6 26 'ilkf2 (26 c4? lt'lxe3 ), then 26 . . . l:!.xe3!
27 l:l.xe3 'ii'xf4 28 l:tee 1 lt'le6, and Black has
defi n ite compensation for the sacrificed
exchange. Wh ite would probably have had to
restrict himself to the qu iet 25 .i.xc5 dxc5,
but here too it is not easy for him to
demonstrate his advantage: 26 lt'lh5 (Adian­
to) 26 . . . ..th8, or 26 'i¥c 1 ( N u n n ) 26 . . . c4 27
lt'le4 f6 28 lt'ld6 l:!.f8 .
36 � Prophylactic Thinking

Karpov's recommendation 23 . . . cS was also
preferable to the move in the game.
24 �f2 !
'One of the last prophylactic moves. Prior to
his decisive offensive, Wh ite places his
pieces in the most harmonious way possible,
and . . . once again reinforces his central
outpost at e4! 24 'ii'd 3 is premature in view of
24 . . . �c8 . '
Here i t i s perhaps t h e last com ment by
Karpov that for me is the most informative.
Using it, I will try to reconstruct his approxi­
mate train of thought:
'The double attack 24 'ii'd 3 is tempting, but
after 24 . . . �c8 2S eS ttJdS 26 exd6 l:Ixd6 the
move 27 c4? does not work because of
27 . . . l1xe3! (but not 27 . . . tt:Jxe3? 28 'ifxd6
tt'lxd 1 29 'ii'x e7 ) 28 lhe3 tt:Jxe3 29 'ifxd6
tt'lxd 1 (poi nted out by N u n n ) . No, there's no
point in going in for such adventures!
'But what does Timman want to play; why
did he make his last move? It can hardly be
24 . . . .l:tde8 - then simply 2S 'iid 3 , and he
loses a pawn . Perhaps 24 . . . i.e8, to support
the d6-pawn , and vacate the d 7-square for
his knight or rook. But then 2S 't!fd3, and the
bishop is no longer defending the knight. If
2S ... 'ii'b 7 it is possible to attack the aS-pawn .
'Well, th is means that for the moment I
should simply wait. What useful move can I
make? Probably 24 ..tf2 - it would not be bad
to defend the e4-pawn once more , and also
the th reat of the e4-eS breakthrough (after
2S 'ili'd3 ) gains i n strength . '
A s you see , prophylactic th i n king is by n o
means synonymous with passivity; i t i n ­
volves simultaneously taking account o f both
the opponent's, and you r own resou rces,
with the accu rate calcu lation of short varia­
tions. Opposing such a manner of play is not
easy - you must be not inferior to the
opponent i n abil ity to g uess the other side's
plans. I n this respect Timman was not fu lly
Karpov's equal .
24 . . . �e8?!
24 . . . .tea was more tenacious.
25 'ii'd 3 1 'ii' b 7
2S . . . tt'lb8 26 eS! .
26 l:ta 1 !
The modest manoeuvres of the wh ite pieces
with i n their own territory have led to material
gains - the aS-pawn ca nnot be defended .
26 . . . tt:J c 7
27 .l:lxa5 .l:ldd7
27 . . . 'ii'x b2? 28 .l:lb1 .
28 b4 tt'le6
Now, of course, 29 fS? ltJf4 is not good for
Wh ite. He must defend his f4-pawn , but
how? Karpov is vigilant right to the end.
 P rophylactic Thinking ftJ 37

The position is completely won for White, Petrosi a n . I n their manner of play Karpov
but a certain accu racy is stil l requ i red . Thus, and Petrosian have much in common, but
after 29 'iid 2 d5!? Black could have obtained even so it seems to me that Karpov is a more
someth ing resembling cou nterplay: 30 e5 aggressive player. For h i m prophylactic
ttJe4 31 ttJxe4 dxe4 32 t2Jd4 c5, with th i n king was always a reliable weapon in
complication s . ' playing for a wi n , whereas Petrosian usually
2 9 �e3 ! c5 ai med above all to safeg uard h imself against
30 f5 tiJdB defeat and his prophylaxis sometimes looked
excessive .
31 b5
Of cou rse, the knight should not be allowed
to go to c6 . ' For "complete happiness", it only Petrosian - G ufeld
remains for Wh ite to play c3-c4 , in order to 28th USSR Championship, Moscow 1 96 1
achieve domination over the entire board . ' King's Indian Defence
31 . . . 'it>h8 1 c4 g6
32 .llf2 2 d4 .ll g 7
The bishop has done its work on e3 and it 3 ttJc3 ttJf6
again retreats, in order to support the e4- 4 e4 0-0
point.
5 �g5 d6
32 . . . 'ili'c7
6 'ili'd2
33 lla4 'ii b 8
Wh ite has chosen an u n usual move order in
34 c4 the opening.
Apart from his material advantage, Wh ite
Once I was observing a joint analysis by
also has an enormous positional advantage
Petrosian and Gufeld . The former world
- the opponent's pieces are completely
champion was constantly outplayi ng his
starved of oxygen. The decisive break­
opponent.
through is not far off.
' How come , ' asked Eduard Gufeld i n per­
34 . . . .l:.a7
plexity, 'surely I have the better position?'
35 .l:.xa7 I:!.xa7
'Yes, but I have the better bra i n , ' Petrosian
36 e5 dxe5 explai ned .
37 ttJxe5 .l:la2 lig ran Vartanovich did not attach too g reat
38 .ll x c5 importance to opening theory and he often
Black resigned . took certa in l iberties , in order to take his
I repeat once more: on a superficial exami­ opponent away from familiar paths and make
nation the game does not attract attention, use of his 'better bra i n ' . I ndeed , why allow
and the impression is created (generally the young Gufeld , who was considered an
speaking, justified ) that the play was 'all at expert on the King's I ndian Defence, to
one end ' . And only after a serious study do demonstrate his knowledge? Wouldn't it be
you beg in to sense the g reat mastery behind better to test his positional understanding, in
Wh ite's seemingly modest moves, a mastery which he was su rely lacking? Especially
largely con nected with prophylactic th inking . since Petrosian himself had an excellent
feeling for such situations: as he himself put
And now we will analyse a game by another it, he 'fed his family for many years thanks to
legend in the field of prophylaxis - ligran King's I ndian set-ups. '
38 � Prophylactic Thinking

6. . . c5 I n other words, there w i l l hardly b e a

7 d5 'ii'a 5 conven ient opportun ity to play . . . b7-b5 or
This position can also be interpreted i n other . . . f7-f5. Whereas White can easily prepare
ways: 7 ... e6 (after which Petrosian was play on the q ueenside with a2-a3 and b2-
planning 8 dxe6 followed by ..td3 and b4.
tLlge2), 7 . . . b5!? 8 cxb5 a6, or 7 . . . a6 8 a4 What then should Black have played? The
'ifa5. logical consequence of his preced ing moves
8 ..td3 a6 was the active 9 . . . b5! . After 1 0 cxb5 the
position can be handled like a Volga Gambit:
9 tLlge2
1 O . . . tLlbd 7 ! ? , but the more forcing 1 O . . . axb5!?
1 1 ..txb5 tLlxe4 also comes i nto considera­
tio n . Petrosian examines the variation 1 2
ttJxe4 'i!kxb5 1 3 ..txe7 l:te8 1 4 ttJxd6 'ii'x b2 1 5
'ilfxb2 ..txb2 1 6 lib 1 ( 1 6 ttJxe8 ! ? ..txa 1 1 7
tLlc1 tLld7 ) 1 6 . . . llxe7 1 7 ttJxc8 l:lb7 1 8 tLld6
.l:tb4 1 9 tLlc4 , and reckons that here Black
faces a struggle for a d raw. In fact after
1 9 . . . .l::tx a2 it is exactly the other way rou n d .
1 0 0-0 tLlbd7
11 a3 tLlh5

9. . . e5?
A serious positional mistake . By blocki ng his
bishop's diagonal, Black deprives h imself of
any counterplay, and now Wh ite's spatial
advantage guarantees h i m an enduring
initiative. Petrosian gave an instructive as­
sessment of the situation:
'Outwardly the position appears highly prom­
ising for Black. By blocking the pawn chain in
the centre he has transferred the weight of
the struggle to the wings, and the possibility
Black is prepari ng 1 3 .. .f5. Of cou rse , he is
of playing . . . b7-b5 and . . . f7-f5 would seem
not afraid of 1 3 ..te 7?! l:te8 14 ..txd6?? 'ii' b 6,
to give him the better chances. But if Black,
while 1 3 g4 allows a standard pawn sacrifice:
in reasoning this way, was attaching the
1 3 . . . tLlf4 14 tLlxf4 exf4 1 5 ..txf4 tLle5 1 6 ..te2
greatest importance to the mobil ity of his
..td7 followed by . . . b7-b5. And in genera l ,
pawn structu re on the wings, he should not
sharp moves such as t h i s are not in
have forgotten that it is the job of the pawns
Petrosian's style.
to clear the way for the pieces. Then he
would not have overlooked the fact that 12 f3 !
White's pieces are much better placed i n the 'A good prophylactic move , d irected in
event of the position being opened . ' particular against . . . f7-f5. It transpires that
 Prophylactic Thinking ltJ 39

12 .f5 is bad in view of 1 3 exf5 gxf5 14 'ii'c2 ! ,

.. I t is clear that with the bishop on g7 there
when there i s n o conven ient way of defend­ was no point in even thinking about f3-f4 .
ing the f5-pawn . ' But here Black i ntends to retreat his bishop
White's move not only i n h ibits the oppo­ to e7, and then he will have to reckon with
nent's cou nterplay, but is also useful in itself this advance . However, in the game things
- now the idea of g2-g4 has to be more do not come to this.
seriously reckoned with . 14 . . . .l:!.b8
12 . . . ii.f6
1 3 i.h6 tt:'l g7

It was probably better to retreat the bishop to

g7. Then Wh ite has a choice: a ) 1 4 ii.e3; b)
14 i.g5 i.f6 1 5 i.e3; c) the exchange of
bishops (in one version or a nother). Since if
he wishes Wh ite can repeat moves, he had
no need to come to a decision beforehand -
it is required only when (and if) the position
arises on the board . S uch reasoning is
typical and important for the practical player,
enabling him to save time on the clock. I am
sure that Petrosian will have played 1 3 i.h6
quite qu ickly. After a l l , perhaps (as in fact
happened in the game) this problem will not What does Black want? Most probably,
have to be solved at a l l , or it will be possible 1 5 . . . b5. But is he prepared for the opening of
to do this during the opponent's time, while the position? Let us check: 1 5 . . . b5 1 6 cxb5
he is considering which piece to place on g7. axb5 1 7 b4! cxb4 ( 1 7 . . . 'ii'a 6 1 8 tt:'lxb5! l:txb5
1 9 a4) 1 8 axb4 , and 1 8 . . . 'ii'x b4 1 9 i.e3
14 g3!?
followed by 20 l:tfb 1 is bad for h i m .
I t was perhaps o n l y Petrosian w h o used to
T h u s 1 5 . . . b5 is n o t a threat. B u t does the
play in this man ner. The point of this move is
opponent real ise this? He must be helped to
not easy to understand without his own
go wrong, by making some neutral move.
explanation.
Say, 1 5 'it>h 1 , especially since if l i nes are
'White's position is so good that he can opened it will be useful to remove the king
choose between different plans. The ad­ from the g 1 -a7 d iagonal .
vance of the g-pawn by two squares is for the
Take note: prophylactic th inking impl ies
moment replaced by the more modest g2-
constant mon itori ng of the opponent's ideas,
g3, but now Black has to reckon with a
but this by no means sign ifies that they
possible f3-f4. I n situations where one side
should defi n itely be disru pted . Sometimes
has no possibility of active cou nterplay,
(as i n the g iven instance) it makes sense, on
whereas the other, with a sign ificant spatial
the contrary, to provoke activity, if you j udge
advantage, has several ways of strengthen­
its consequences to be i n your favou r.
ing his position, such a method of play is
1 5 'it>h1 1 ? iic7
sometimes more u npleasant and dangerous
than direct action . After a l l , it is hard for the But now White seriously has to reckon with
defender to guess from which d i rection 1 6 . . . b5 1 7 cxb5 c4 .
danger will stri ke. ' 16 b3!
40 � Prophylactic Thinking

'A continuation of the same unhurried strate­ Vyacheslav Chebanenko, the wel l-known
gy. Before playing his pawn to b4 Wh ite trainer from Moldova). By moving his knight
prepares to double rooks on the b-file, which to the kingside, Black has weakened his
sooner or later will be opened . At the same defence of the important sq uares c5 and b6.
time another problem is solved : it is no It makes sense for Wh ite to exploit the factor
longer necessary to watch out for . . . b7-b5 . ' i mmed iately and to open up the game on the
When studying the present game (and also queenside.
the previous one) you should not attach too 19 b4! tt:'lg8
much sign ificance to individual moves, or try 20 ..ie3 f5
to establish whether or not they were The lesser evil was 20 . . . b6 2 1 bxc5 bxc5 22
objectively the strongest - this is not the .:txb8 'ilt'xb8, although then too after 23 .l:!.b 1
point. It is more important to follow and and 24 'iib 2 Wh ite has a clear advantage.
sense how the taking of decisions was
21 bxc5 dxc5
approached by these great players, with
whom prophylactic thinking was fully devel­ 22 l:.fb1 tt:'lf6
oped (in the case of Tigran Vartanovich , The wh ite rook should not have been
perhaps even slightly more than neces­ allowed to go to b6. Of cou rse , 22 . . . b6 would
sary!). have opened new possibil ities for Wh ite,
16 . . . ..ie7 involvi ng the advance of his a-pawn, but
nevertheless it should have been played . '
1 7 .l:!.ab1 <ith8
2 3 :b6 ..id6
1 8 l:lb2 tt:'lf6
23 . . . tt:'ld7? did not work in view of 24 d6!
'Black embarks on a new reg rouping , which
..ixd6 25 tt:'ld5. However, 23 . . . fxe4! ? was
in the end enables him to advance his pawn
more tenacious, trying to obtain the sq uares
to f5. Had he tried to do this a move or two
f5 or g4 for his pieces.
earlier, then after the exchange on f5 Wh ite
would have created unpleasant pressure on
the b 1 -h7 diagonal by 'ii'd 2--c2 . In combina­
tion with g3-g4 and the opening (after b3-
b4) of a 'second front' on the queenside, this
would have led to a difficult situation for
Black. It would have been most sensible for
him to stick to passive tactics, and to wait for
Wh ite to clarify his intentions. Moves such as
18 . . . b6 followed by 1 9 . . . .l:tb7 would to some
extent have improved his defensive resou rc­
es. '
How should Wh ite now combat a standard
plan in such positions: . . . tt:'lf6-g8 , and after
the retreat of the bishop - . . . f7-f5 ? He can ,
of course, resort to g3-g4. 24 ..ih6!
But let's remember a typical idea when ' Despite the strong pressure, Black still
seeking a solution - a question which in such retains hopes of obtaining cou nter-chances
cases it is useful to ask you rself: 'What is the by exchanging on e4 , when after the
drawback of my opponent's move?' (It would recapture with the pawn he can eliminate the
appear that this was first recommended by dark-square bishop by . . . tt:'lf6-g4 , while after
 Prophylactic Thinking l2J 41

tt:lc3xe4 or ..lii. d 3xe4 he has . . . ltJg7-f5. It
should be mentioned that, with his over­
whelming positional su periority, Wh ite would
have retained an obvious advantage even if
the 'threats' mentioned above were to be
carried out, but there is no reason for h i m to
part unnecessarily with even a portion of his
positional gains.'
Now if 24 ... fxe4 there follows 25 ltJxe4 ltJxe4
26 ..be4 l:!.f7 27 g4! , restricting both the
opponent's knight, and his bishop.
24 . . . �f7
25 lL'lg1
Another purely Petrosian-l ike move. The
grandmaster does not h urry to take specific
action , but prefers , just in case , to support
his f3-point beforehand . If 25 . . . fxe4 he was
intending 26 ..lii. x g7+!? (after 26 ltJxe4 ltJf5 2 7
tt:lxd6 ltJxd6 Black g a i n s cou nter-chances)
26 . . . <.t>xg7 27 ltJxe4 ltJxe4 28 ..lii. xe4, followed
by g3-g4 and lbh3.
Black should probably have evicted the
dangerous rook from b6 by 25 . . . ltJd7 . I n ­
stead o f this Gufeld tries t o i n itiate complica­
tions, which tu rn out to be not i n his favour.
25 . . . f4?
26 gxf4 lL'ld7
26 . . . exf4 27 J:ixd6! and wins.
'Now on the retreat of the rook there follows
28 . . . lbh5!, when the black pieces obtain
convenient posts at d6, e5 and f4. But if
Black's plan could have been realised, this
would naturally have called into question
White's play in the middlegame. After all,
from the opening Black stood badly. '
28 J:ie6 !
'All is in order! Black is prevented from
playing 28 . . . ltJh5 by the th reat of 29 Ite8+.
Simultaneously an attack is made on the
bishop at e5 - the sole barrier i n the path of
the passed pawns in the centre. By eliminat­
ing this obstacle after 28 . . . lDf8 29 �xe5! ,
White would have won without difficulty. '
( I ndeed , 29 . . . 'tli'xe5 30 f4 'ii' h 5 3 1 f5 !? looks
th reatening, but nevertheless Black should
have gone i n for this position - most
probably he had noth ing better. )
'The conclusion of the game, which my
young opponent conducted i n time-trouble,
resembles 'g ive-away' chess . '
28 . . . b5
29 cxb5
Thanks!
29 . . . c4
30 �c6 'iVd8

27 fxe5 ..lii. x e5 31 ..txc4

27. . .lt.Jxb6 28 exd6 'iVxd6 29 e5 was hope­ Thanks again!
less for Black. 31 . . . 'ii' h 4
32 J:ic1
A familiar trait! Wh ite frees his q ueen from
having to defend his c3-kn ight and hopes in
the future to worry the c-bishop along the c­
file.
32 . . . lbh5
33 ..tg5 lbg3+
34 'itg2 ltJxe4
35 ltJxe4 'iVxh2+
36 Wf1 J:ixf3+
37 lbxf3 �h 1 +
38 'it>f2
42 � Prophylactic Thinking

'Here Black fi nally remembered that, apart pawn to c6? The point is that Wh ite has not
from mate , in chess it is also possible simply yet determi ned the position of his king . After
to resign a game, which he very belatedly q ueens ide castling one of Black's best plans
did.' is the preparation of . . . c7-c5 ( . . . ..te6 , . . . a7-
a6 and perhaps . . . .l:tc8). It is clear that in this
We will now examine a few examples, case . . . c7-c6 will tu rn out to be a clear loss
showing how prophylactic th inking helps in of tempo. A typical example of prophylaxis in
taking decisions in various stages of the the opening - taking account of the op­
game. ponent's possible plans, in order to arrive at
the most accu rate move order.
Opening 1 0 0-0 c6
When analysing the following game we will 1 1 .Ua b1
at the same time recall certain strategic Wh ite is preparing the standard minority
ideas, typical of the Carlsbad structu re in the pawn attack on the q ueenside. Here 1 1 llae 1
Queen's Gambit. has also occurred , and later the rather
dangerous variation 1 1 f3! ? came into
Botvi n n i k - Keres fashion. For example, the th ird game of the
20th USSR Championsh i p , Moscow 1 952 lvanch u k-Yusupov Candidates match (Brus­
sels 1 99 1 ) went 1 1 . . . ttJh5 1 2 ..txe7 'ifxe7 1 3
Queen 's Gambit
e4 dxe4 1 4 fxe4 Sl.g4? ! ( 1 4 . . . Sl.e6 ! ) 1 5 e5!
1 d4 ttJf6
.l:tad8 1 6 ttJe4 , reach ing a difficult position for
2 c4 e6 Black, roughly similar to that which occu rred
3 ttJc3 d5 in the game we have j ust begu n examining.
4 cxd5 exd5 Mikhail Moiseevich , who was present at the
s ..tgs ..te7 match , caustically remarked : 'Aha, Yusu pov
doesn't know the Botvinnik-Keres game.
6 e3 0-0
That's bad ! ' In fact Artur knew this game, of
7 Si.d3 ttJbd7
course, but at the board he was unable to
8 'ifc2 lieS find a way of avoiding the u nfavourable
9 ttJge2 ttJf8 development of events.
11 . . . Si.d6?
Threatening 1 2 . . . ..txh2+ 1 3 �xh2 ttJg4+, but
Keres's main idea is to play 1 2 . . . ttJg6 and
1 3 . . . h6, forcing the exchange of bishop for
knig ht, and to recapture on f6 with the queen.
Then his pieces will be actively placed and
pressing on the opponent's kingside. Alas,
Botvi nnik refutes this idea .
The correct move order was 1 1 . . . ttJg6 (of
course, there are also other conti nuations,
for example 1 1 . . . ttJg4!?) 1 2 b4 ? ! Si.d6 (threat­
ening 1 3 . . . ..txh2+ or 1 3 . . . h6) 1 3 ttJf4 Sl.xf4 1 4
exf4 'ili'd6 1 5 .l:1fe 1 Si.d7 1 6 f5 ltJf4 , and Black
is excellently placed (Lazarev-Fedorenko ,
Why was Keres not in a hu rry t o play h i s Ternopol 1 964 ).
 Prophylactic Thinking l2J 43

White should remember about prophylaxis 1 5 i.. x e7 lhe7

and frustrate this plan by 1 2 i.. xf6! i.. xf6 1 3 1 6 t'Dg3
b4, retaining somewhat the better game. 'Wh ite does not h u rry to advance e3-e4 ,
12 'Oti>h1 ! t'Dg6 remembering Tarrasch's sayi ng that the
After 12 . . . h6 1 3 i.. h 4 (or 1 3 i..f4) the knight threat is stronger than its execution . But it is
can no longer go to g6. not put off for long , only to a time when Black
will not have any active replies . ' As you see,
while converting his advantage Botvinnik is
th i n king prophylactically.
16 . . . t'Df6
1 7 'ii'f2 i.. e 6
1 8 l'Df5 i.. xf5
Otherwise 1 9 g4.
19 i.. xf5 'ii' b 6
20 e4 dxe4
21 fxe4 .l:td8
22 e5

13 f3 ! !
This move practically wins the game, since it
frustrates Black's plans. If now 1 3 . . . h6, then
14 .bf6 ( 1 4 i.. x h6 gxh6 1 5 i.. x g6 fxg6 1 6
'lxg6+ �h8 1 7 'ii'x h6+ t'Dh7 is u nconvi ncing)
14 . . 'i'xf6 1 5 e4 with the terrible th reat of 1 6
.

e5 - it transpires that all the black pieces are

badly placed and are very vul nerable.
13 . . . i.. e 7
An admission of the faultiness of Black's
preceding strategy. But what else could he
do? - the th reat of e3-e4 was too serious.
1 4 l:l.be1 22 . . . t'Dd5
It is probable that Botvinnik avoided the The prophylactic move 22 . . . t'De8 was more
immediate 1 4 e4 because of 1 4 . . . dxe4 1 5 tenacious, preventing the manoeuvre of the
fxe4 tt:':lg4 . Now Black should have played wh ite knight to d6.
14 . . h6, provoking his opponent i nto unclear
23 t'De4
.

compl ications after 1 5 i.. x h6 gxh6 1 6 i.. x g6

It should be said that, against Yusupov,
fxg6 17 'i!lxg6+ �h8 or 1 5 i.. xf6 i.. xf6 1 6
lvanchuk managed to obtain roughly the
.bg6 fxg6 1 7 'it'xg6 l:txe3 1 8 t'Df4 i.. x d4 1 9
same position far more q uickly.
J:Ue3 i..x e3 20 .l:te 1 d4. Keres opts for
passive tactics and comes under terrible 23 . . . t'Df8
positional pressure. 24 t'Dd6 'ii'c 7
14 . . . t'Dd7 25 i.. e 41
44 � Prophylactic Thinking

White prepares to exchange the opponent's M iddlegame

only decently-placed piece - the knight on
I n order to improve i n a certai n aspect of
d5, and simultaneously he vacates the f5-
chess, it is useful to study the games of
sq uare for his own knight.
players who are masters of this particular
25 . . . lt'le6 field . Therefore I a m going to show you two
26 'i!Vh4 g6 extracts from the play of Anatoly Karpov, one
26 . . . h6 27 lLlf5 .l:red7 28 lt'lxh6+! gxh6 29 of the 'classics' in the field of prophylaxis.
'it'xh6.
27 .11.. x d5 cxd5 Ka rpov - Bag i rov
28 l:!.c1 ! 38th U S S R Championship, Riga 1 970
It is useful to wrest control of the c-file and at
the same time prevent the exchange sacri­
fice on d6.
28 . . . 'ii'd 7
29 .:c3 .l:rf8
30 lLlf5 !
Of course, Wh ite does not allow the freeing
advance 30 . . . f5! . Now 30 . . . gxf5 3 1 �g3+
lt'lg7 32 'i!Vf6 1eads to mate, while if 30 .. J:tee8
the simplest is 31 lLlh6+ 'it>h8 32 'ji'f6+ ltJg7
33 lt'lxf7+.
30 . . . .l:i.fe8
31 lLlh6+! 'it>f8
32 'iVf6 lt'lg7 Wh ite has an undisputed positional advan­
33 .l:i.cf3 tage. He controls more space, and the knight
The f7-point can not be defended . 34 'i!Vxf7+ ! on b7 has nowhere to go. But if it were Black
i s th reatened . to move he would play 27 . . . a 5 ! , and if 28 b5
33 . . . .l:i.c8 a4 , obtaining the c5-sq uare for his knig ht.
34 lLlxf7 �e6 Wh ite also has to reckon with . . . f7-f6 .
35 'iVg5 lLlf5 After 27 .l1.. b 1 a5?! 28 'iVd4! 'iVxd4 29 lLlxd4
axb4 30 axb4 Wh ite's advantage increases.
36 lt'lh6 l1Vg7
However, the opponent can play more
37 g4
strongly: 27 . . . .l:i.fc8! ? (followed by a possible
Black resig ned . . . . .l:i.c4 ) , or 27 .. .f6 ! ? .
I was impressed by the strength of the 27 llfd2 looks q u ite good , b u t then too Black
seemingly modest move 1 3 f3 ! , which has the reply 27 .. .f6, giving him some
refuted Keres's strategy. It should be men­ cou nterpl ay.
tioned that in such positions play i n the
Karpov found the opti mal solution .
centre with e3-e4 is the best response to the
plan of . . . .l1.. d 6, . . . lt'lg6 and . . . h7-h6. Wh ite's 2 7 'ji'g4!
knight could have been on f3 - then f2-f3 is Wh ite gains firm control of the central sq uare
no longer possible, of course, but there is d4 and at the same time he creates the th reat
.l:i.ae1 (or .l:i.fe1 ) followed by e3-e4 . of 28 .11.. x h7+ 'it>xh? 29 �h4+ 'it>g8 30 'ir'xe? .
 P rophylactic Thinking tLJ 45

Now 27 . . . a5 no longer achieves its aim i n 36 ..tf1 !

view o f 2 8 'ifd4 , or even simply 28 b5, when The bishop withdraws beforehand from a
29 . . a4 is not possible (here 28 ..txh7+?!
.
tempo-g ain in g attack (after . . . tt'lxe5 or
'Ot>xh7 29 'ii'h 4+ ..ti>g8 30 1Vxe7 is less strong: . . . tt'lf4).
30 . axb4 3 1 'ilfxb4 'ii'x b4 32 axb4 tt'ld8;l; ).
. .
36 . . . tt'lf7
After 27 . . .f6 there fol lows 28 'ii' h 4 tt'lg6 37 tt'lc7 l:td8
(28 . . . h6 29 l:tfe 1 ) 29 ..txg6 hxg6 30 'ii'g 3,
Or 37 . . . .l:.b8 38 b6 axb6 39 a6! tt'ld8 40 a?
while if 27 ... tt'lg6 28 ..txg6 - in both cases
l:ta8 ! ? 41 tt'lxa8 l:txa7 42 l:.c8 ! and wins.
with an obvious advantage for White .
38 llc6
27 . . . f5
Now the point of the knight manoeuvre to c7
28 'ii'd4
becomes clear. The e6-pawn is attacked ,
28 exf6 Uxf6 29 'it'g3 J:.bf8 is not so and therefore the knight is forced to defend
convincing. it, instead of captu ring the e5-pawn . Mean­
28 . . . l2'ld8 while, White is ready to create a passed
Black is hoping to place his knight at c6 . pawn on the q ueenside.
29 b5! 38 . . . l2'lf8
Karpov has achieved his favou rite domina­ 39 b6 axb6
tion - the opponent's pieces are deprived of 40 a61 tt'lh6
any active possibilities. White is free to carry 40 . . . d4 41 a? (4 1 ..tc4 ! ? ) 41 . . . tt'lxe5 does not
out a pawn offensive on the q ueenside. help i n view of 42 a8'it' (42 .l:r.xe5? l:txc7 ; 42
29 . . . g5 tt'lxe6?! l;lxa7 43 'ii' b 1 tt'lxe6 44 l;lxe6 tt'lg6)
The opponent is hoping to attack the e5- 42 . . . l:.xa8 43 'ii'x a8 tt'lxc6 44 tt'lxe6 .
pawn , but he is not able to create any real 41 :ec1
counterplay. 41 'ii'c 1 !? .
30 a4 tt'lg6 41 . . . tt'lg4
31 'i!i'a 1 ! 42 a7 tt'lxe5
From here the q ueen defends the e5-pawn 42 .. .'ii' x e5 43 'it'xe5 tt'lxe5 44 l:lxb6 was
and supports the offensive on the q ueenside. eq ually hopeless.
Therefore on the previous move it would 43 .l:!.6c2 tt'lc4
have m ade sense for Black to exchange the 44 a8'fi l::t x a8
queens: 30 .. .'ii' x d4 31 tt'lxd4 tt'lg6, although
45 tt'lxa8 b5
after 3 2 .Ufe 1 his position would have
remained difficult. In tu rn , Wh ite also could 46 .U.a2
have retreated his q ueen a move earlier. Black resig ned .
31 . . . 'ili'b7
32 l:tfe 1 1i'g7
Karpov easily forestalls the threat of 33 . . . tt'lf7
by attacking the weak e6-pawn .
33 tt'lc5 lif7
34 a5 .l:te7
Again 35 . . . tt'lf7 has to be parried .
35 tt'la6! l:ta8
46 � Prophylactic Thinking
Karpov - Hort
Moscow 1 97 1
I n contrast to the previous example, here the
situation is very tense. Both kings are stuck
in the centre. Black's main threat is 22 . . . \!Vh4 ! ,
a n d i t was for t h i s reason that Karpov
rejected the natu ral move 22 'it>c2 .
Black's activity can be prevented by 22
�g5!?. Karpov was concerned about 22 . . .
'ii'b6 2 3 .i.e3 'ii'c7 (the h2-pawn i s attacked ,
and Black wants to castle queenside, i nclud­
ing his queen's rook in the game). But why
retreat the bishop? - there is the simple 24
.l:r.g2 !?, while Kasparov recommended the
energetic 23 'it>c2 ! .i:!.xh2 24 l:l. h 1 'tlff2 25
.l:!.xh2 "i!i'xh2 26 .l::!.f 1 .
However, the cou rse chosen by Karpov is
also very strong.
22 .l:r.g4! !
A multi-purpose move . The rook takes
control of the important h4- and f4-sq uares,
prevents the black queen from going to h4,
and prepares the advance h2-h4 . I n add ition
it can be switched to the queenside along the
4th rank.
22 . . . "i!i'f6 !
After 22 . . . .i.xh2 23 'it>c2 Black's position is
difficult. He loses a piece after 23 . . . t'Lle5 24
.:!.g2! (24 . . . 'tlt'h4 25 .l:.xh2 'i!i'xh2 26 �b5+),
while if 23 .. .'it'd7, then 24 Itf1 or 24 .l:.e4+ is
strong .
23 h4!
Black was hoping for 23 <it>c2? t'Lle7! 24 .i.e3
l:.xh2 or 24 .i.g5 'ii'f5+ . But now neither
23 . . . 0-0-0 nor 23 . . . t'Llxh4 is possible because
of 24 �g5.
23 . . . 'iff5
24 .l:!.b4! �f6
Again castling is impossible (24 . . . 0-0-0?? 25
�g4), but 24 . . . t'Lle7 was more tenacious, or
24 . . . .i:!.g8 25 �d3 (25 .i:!.xb7? t'Llf4) 25 . . . 'ii' h 3!,
not conced ing the important f4-square to the
opponent ( Kasparov).
25 h5 t'Lle7
25 . . . t'Lle5? 26 l:tf4 .
26 llf4 'ife5
The culmi nating moment of the battle! As
Kasparov remarked , playing the rook to f2 , f1
or even a4 would have retai ned a g reat
advantage, whereas the move in the game is
sign ificantly weaker. However, Vlastimil Hort
failed to exploit his opponent's mistake.
27 l:.f3?
For the first time Black has gained an
opportunity to castle, and he should defi n ite­
ly have made use of it. After 27 . . . 0-0-0 ! the
outcome would have remai ned unclear. How
should Wh ite reply? 28 'it>c2? is a mistake
because of 28 . . . l:.xh6 29 'i¥xh6 'ii'x e2 + . If 28
l:.d3 there follows 28 . . J�dg8 29 .i.f4 'ii'f5
with the un pleasant threat of 30 . . . .i:!.g2. 28
 Prophylactic Thinking
i.e3 is unconvi ncing: 28 . . . lZJxd5 29 .i.d4
"le6.
28 i.f4 is tempti ng, hoping for 28 . . . '�xd5?
29 l:td3 'ii'h 1 + 30 'itc2 ! Wxa 1 31 .i.g4+ (31
J:Ixd6? 'ii'g 1 ! ) 3 1 ... '1tb8 32 llxd6 llxd6
33 'i'xd6+ 'ita8 34 'ii'xf6 and wins. In the
event of 28 . . . 'ife4?! Wh ite does not play
29 l::t d 3 on account of the pretty reply
29 . . . ltJg6! ! , poi nted out by Utut Ad ianto , but
29 'i'd3 ! , retaining the advantage. But it is
not evident how to cal l i nto q uestion the
simple 28 . . . 'ii' x h5 ! .
27 . . . lZJxd5?
27. . .'ii'x h5? would have lost to 28 llxf6 'ii' h 1 +
29 .if1 ltJg8 30 'ii'e 1 + , but the captu re of the
central pawn is l ittle better.
28 lld3 llxh6
28 . . . tt:\e7 29 .i.f4 .
29 l:txd5!
29 'i'xh6? �g5 and 30 . . . lZJe3+
29 . . . 'ike4
30 .l:i.d3 !
The rook's manoeuvres have d isru pted the
oppon ent's defences.
30 . . . 'ii' h 1 +?
Kasparov's suggestion 30 . . . 'ii h 7 was much
more tenacious - here Black would still have
retai n ed chances of a successful outcome.
31 'it>c2 'i!Vxa1
32 'ii'x h6 �e5
33 1i'g5
Black lost on time.
Endgame
The fol lowi ng ending is taken from N imzo­
witsch's My System. Old books give nu mer­
ous positions in which one of the players,
much weaker than the other, fails to put up a
worthy resistance. I n the com mentary all the
attention is usually focused on the play of the
w i n n e r, and defensive possibil ities are not
lZJ 47
even mentioned . As a result, such ga mes
receive a one-sided coverage and a non­
objective assessment. It is probable that at
some stage of chess study this even has a
defi n ite pedagogical point. But when 'at a
matu re age' you again turn to them and look
with q u ite d ifferent eyes, you easily notice a
certain naivety of such examples and their
book interpretation .
Gottschall - N i mzowitsch
Hannover 1 926
It is Black to move . How should the position
be assessed? N i mzowitsch 's chances are
certainly better thanks to his superior pawn
structu re - his one pawn on a4 is holding
back two of Wh ite's. To j udge by the
grandmaster's com ments, his win was the
logical outcome. In fact, with correct play the
game shou l d , of course, end in a draw. After
a l l , material is equal with opposite-colour
bishops, and in addition Wh ite controls the
only open file.
Every player is obl iged at ti mes to try and
'squeeze' a microscopic advantage, or on
the contrary, defend in a slightly i nferior
end ing. Therefore it is instructive to follow
the actions of N imzowitsch , who completely
outplayed his opponent, and to understand
why this happened .
48 � P rophylactic Thinking

How can Black strengthen his position? It

would not be bad to play his king to f5.
However, if 28 . . . �g6 Wh ite has 29 g4! hxg4
30 hxg4 f1h8 31 �g3. Generally speaki ng,
g3-g4 is a move which Black has to reckon
with . Is it not possible to prevent it?
28 . . . .l:th8!
An excellent prophylactic move. Now Black
is th reatening to play 29 . . . �g6, and if 30 g4 -
30 . . . hxg4 3 1 hxg4 l:th2+ or (after 29 l:ld2 or
29 .l:!.b4) - 3 1 . . . .l:!.h3. Regarding this N imzow­
itsch writes:
'To demand of a piece only direct attacking
activity is the stamp of the mere "wood­ N i mzowitsch was planning . . . e6-e5. Howev­
sh ifter" . The keener chess mind qu ite rightly er, if 3 1 . . . e5 there follows 32 fxe5 fxe5 33
demands of the pieces that they also g4+ hxg4 34 hxg4+ and 35 l:ld6+, driving
undertake preventive action. The fol lowing back the black king . Therefore the g rand­
situation is typica l : a freeing action (usually a master does not hu rry.
pawn advance) plan ned by our opponent 31 . . . .l:!.f8 !
would in the result give us an open file. This Another 'mysterious' move. The rook pre­
potential file, to open which does not lie in pares to operate along the f-file, which will be
our power, we nevertheless seize, and in opened in the near future. In any case, it is
advance, with the idea of g iving our oppo­ usefu l fi rst to strengthen the position as
nent a distaste for the freeing action . The much as possible, and only then change the
"mysterious" rook move is an ind isputable pawn structure.
ingredient of a rational strategy . . . I will take
Let us again take White's side and th ink how
the liberty of asserti ng that the prevention of
to combat the opponent's plan .
freeing moves by the opponent is far more
important than considerations about whether He can remove h i s rook from a possible
the rook is fu nctioning at a g iven moment or attack, by playing it to b4. Then if 32 . . . e5 he
is occupyi ng a passive position . ' has a pleasant choice between 33 fxe5 fxe5
34 h4, 33 fxe5 fxe5 34 g4+ , and fi nally, the
B u t now let's reason for Wh ite . Black i s
tactical stroke 33 g4+ ! . However, the move
obviously preparing to play h i s king t o f5 .
32 l:l.b4 has a serious drawback - the rook
How can this be prevented? Very simply - by
abandons the open d-file, and Black immedi­
the prophylactic move 29 l:l.d6! . The king is
ately exploits this factor: 32 . . . l:td8! 33 i.e3
now tied to the e6-pawn , and in the event of
l:td 1 or 33 �e3 l:!.d5 34 l:tc4 i.b5.
29 . . . i.d5 the rook occupies the 7th rank. It is
Let us move the king off the f-file. For
not apparent how Black can strengthen his
example, 32 �g 1 (after 32 'it>e2 Wh ite has to
position.
reckon with 32 . . . i.g2, although this is proba­
29 l:td 1 ? ! <;tJg6 bly not dangerous) 32 . . . e5 33 fxe5 (33 g4+ !?)
30 l:td4 �f5 33 . . . fxe5 34 g4+ hxg4 35 hxg4+, and Black
31 i.d2 does not achieve anyth i n g .
3 2 ii. e 1 ? ! e5
(see diagram) 32 . . . g5!? was evidently more accu rate,
 Prophylactic Thinking
since here Wh ite had the now familiar
tactical stroke 33 g4+ ! .
33 fxe5? fxe5
34 llh4? g5!
The move 3 1 .. J:If8! has its effect - Wh ite
can not play 35 Ilxh5? �g6+ .
35 llb4 �e6+
36 �e2 e4
36 . . Jif3? 37 .:txa4 .
37 .1f2 .l:tf3
38 .l:tb6
Nimzowitsch has achieved much , and now it
is indeed not easy to defend. 'The passed
pawn, the penetration of the rook into the
enemy position, and a certain weakness in
White 's c5-pawn slowly wrought the destruc­
tion of White 's game. '
How can Black strengthen his position? It is
unlikely that he will be able to prepare . . . e4-
e3 in circumstances such that all possible
replies - llxc6 , .1xe3 and .1e1 will prove
unsatisfactory. He would like to place his
king on d5, but what next? I n some cases
. . . h5-h4 makes sense, althoug h in principle
the exchange of all the kingside pawns is to
White's advantage .
Let us suppose that with the king on d 5 and
the rook on b4 it is Wh ite to move. If .l:td4+ ,
then simply . . . �xc5 , not fea ring a discov­
ered check. And if llb6 Black can reply . . . h5-
tZJ 49
h4, when after the captu re by the bishop on
h4 he can eliminate the c5-pawn with gain of
tempo.
Of cou rse, the win of a pawn when there are
opposite-colour bishops by no means guar­
antees a wi n . But all the same Black has
noth ing better, and after captu ring on c5 he
creates the u npleasant th reat of a bishop
check on b5.
38 . . . �e5!
39 ilb4 �d5
A zugzwang position is created .
40 h4 gxh4
41 gxh4 .l:th3?!
New prospects a re opened for the black
rook. However, now the wh ite rook also
breaks free , and yet it could have been kept
at b4 by making the prophylactic move
41 . . . �e5 ! . The rook will i nevitably reach h3 a
l ittle later, whereas Wh ite can not play 42
.l:.d4? because of 42 . . J 1xf2 + .
4 2 .:td4+ �e5
43 .l:td8 .1d5
44 .l:te8+ .1e6
Threatening 45 . . . llb3 . The concluding stage
of the game has been analysed in detail by
g randmaster Robert H u bner. I will briefly
acquaint you with his main conclusions.
45 l:td8 �f4
The noose a round the wh ite king is d rawn
ever tighter. Perhaps it would have been
slightly more accu rate to fi rst d rive the king
away from the e2-sq uare: 45 . . . .1g4+ ! ? 46
�d2 �f4 47 llf8+ .1f5 48 'it>e2 llh2.
However, in H ubner's opinion , here too with
accu rate play Wh ite would have gai ned a
d raw.
46 llf8+ .1f5
47 l:r.f7 l:.h2
Not i m mediately 47 . . . e3 48 .1g 1 .
48 l:l.e7?
If 48 �e 1 ? , then 48 . . . e3 49 .1xe3+ 'it>xe3 50
.l:r.xf5 llh 1 + 51 .:tf1 llxf1 + (51 . . . .l:!.xh4) 52
50 � Prophylactic Thinking

�xf1 �d2 ! and wins (but not 52 . . . �f3? 53 Fischer - Donner

�e 1 �g3 54 �d2 'it>xh4 55 b4 ! axb3 56 a4). Olympiad , Varna 1 962
However, by playing 48 �1 ! , Wh ite would
have retai ned quite good drawing chances,
for example, after 48 . . . 'it>g4 49 �g7+ 'it>f3 50
.tf.g3+ 'it>f4 51 l:tc3.
48 . . . ..tg4+
49 'it>e1 'it>f3
In Hubner's opinion , 49 . . . � h 1 +! 50 'it>d2
.tf.d 1 + and 5 1 . . . I:::!.d 3 was stronger.
50 k!.f7+ 'it>g2
51 ;t>d2?
Now the game concludes quickly, whereas
after 51 ..td4! �h 1 + 52 'it>d2 �d 1 + 53 �e3 it
is possible that Black might not have been
able to wi n .
51 . . . 'it>f1 ! The exchange of q ueens on a7 or e3 (24
52 'it>e3 �f3 tt::l d 6 'ifxe3 25 fxe3 Ji.xd6 26 �xd6 f6 27 b3)
53 �g3 l:Ixb2 leads to a roughly equal end i n g . What
54 �d6 �b3+ alternative does Wh ite have? Only the knight
sacrifice on h6. Let's try to calculate its
55 'it>d4 'it>f2
consequences .
56 l:tg7 e3
24 tt::l x h6+ ! ? gxh6
57 Ji.g3+ �f1
There is no direct way to g ive mate, but the
58 .l:!.f7 e2
rook can be incl uded in the attack via d4 .
59 l:te7 ..tc6
Before calculating variations, let's ask our­
White resig ned . selves how Black will defend . Obviously it is
Take note: at the board (in contrast to his very important for h i m to advance his f-pawn .
comments in the book) N imzowitsch thought to i nclude his q ueen in the defence along the
prophylactically - he took account of the 7th rank.
opponent's resou rces and endeavoured to Alas, Bobby Fischer underestimated this
forestal l them. This factor ensured the factor and played 25 l:Id4? , when after
grandmaster an enormous playing superiori­ 25 .. .f5! Black parried the attack. The game
ty over his opponent, who did not even think continued: 26 .l:Ifd 1 tt::l c 5 27 �d8 'iff7 28
about prophylaxis and as a result allowed .l:txe8 'ifxe8 29 �d4 tt::l e4 30 f3 e5! 31 fxe4
Black to carry out his plans. (31 Ji.b6 l:txb6) 31 . . . exd4 32 'ii' g 3+ �g7 33
exf5. Now 33 . . . c5! would most simply have
C o m b i nation decided matters, but the q ueen exchange
33 ... 'ii'e 3+ also proved sufficient for a wi n .
Here, it wou ld seem, there is altogether no
The key to the success o f Wh ite's attack l ies
place for prophylactic thinking, and what
in prophylaxis - he must prevent . . . f7-f5.
proves decisive is imagination and specific
calculation. But take a look at the following 25 'ii'g 3+ 'it>h7
example. 26 h5! !
 Prophylactic Thinking
Now 26 . . . f5? is no longer possible because
of 27 'ii'g 6 mate. The black pieces are
huddled together on the q ueen side, whereas
White is intending 27 l:td4 followed by l'::tfd 1
or I!.g4 . Bad , for example, is 26 . . . lt:'lc5(b6) 27
l:td4 lt:'ld7? 28 'ii'd 3+. If 26 . . . c5, then 27 lld3
'le7 28 f4! (again prophylaxis - it is
important not to allow 28 .. .'il'g5) 28 . . . lld8 29
l:.ff3 o r 29 l:ta3 , preparing 'iig 4 and .Ug3 .
26 . . . 'ii'e 7
27 1::t d4 tt'lc5
28 l:tf4!
Not 28 l:tg4? f6. 28 l:tfd 1 ? ! l:ta7 is also
inaccurate, si nce if 29 .l:.f4 Black now has
29. . .tt::ld 7! 30 �d6 'ii'g 5.
28 . . . .Ua71
28. . .tt::\d 7 is hopeless: 29 �d6! 'iig 5 (29 . .
l'xd6 30 �+ �h8 3 1 'ikg6) 30 :Xf7+ �g8
3 1 'i'xg5+ hxg5 32 .l:.xd 7 .
2 9 �f6 'ifd6
30 'ii'g 4!
(see diagram)
An attacking and simulta neously prophylac­
tic move! After parrying the th reat of 30 . . . e5
(31 'ii'f5+ ), Wh ite prepares l:td 1 or l:.f4-f3-
g3. If 30 . . . lt:'ld7 he has the decisive 3 1 .ltd4 ! .
White's attack looks i rresistible, a n d here, I
have to admit, my i n itial analysis came to an
end . But in fact the battle continues.
ltJ 51
- position after 30 'ifg4! -
30 . . . 'ifd3 !
A n excellent prophylactic move, pointed out
by John N u n n . The immediate th reats are
parried : 3 1 l:!.f3? 'ii'e4! 32 l:tf4 'ii'c2 , or 3 1
l:td 1 ? 'ii'c2 3 2 �e5 f6 . However, Wh ite's play
can be improved .
3 1 l:l.e1 !
Aga i n Black's position seems hopeless. 32
l:te3 or 32 l:.f3 is th reatened , and it is bad to
play 3 1 . . . l:td7? 32 l:tf3 'ii'd 2 33 l:tg3! or
3 1 . . . 'ifd2? 32 l:te3 l:td7 33 l:tg3 'ild 1 + 34
�h2 'ii'x g4 35 l:tfxg4, when the only way of
preventing the deadly 36 l'::t g 8 is to return the
piece by 35 . . . lt:'le4 .
. It seemed t o me that after 31 . . .e 5 ! 32 l:te3 !
Wh ite's attack should also ach ieve its goa l .
However, t h e German analysts Klaus Dieter
Mayer and Karsten Mu ller thoroughly chec­
ked this position with a computer and
established that with accu rate defence Black
can hope to save the game. I will g ive the
i n itial moves of their analysis: 32 ... 'ili'b1 +! 33
�h2 lt:'le6! (33 . . . exf4? 34 .l:txe8 lt:'ld7 35
'ii'x d7! l:txd7 36 l:txf8) 34 l:tfe4! ? (34 .:tg3
tt'lg5) 34 . . . .ltg7! 35 'iff5+ �h8 36 .:tg3 l:tg8
(but not 36 . . . tt'lg5? 37 l:txg5! hxg5 38 h6).
Their variations extend for a fu rther dozen
moves, but we will d raw the line here.
52 � Prophylactic Thinking
The exami nation of interesting examples on
the theme of prophylaxis could have been
continued . But for the mastery of prophylac­
tic thinking (as also, in genera l , any practical
skill) theory alone is i nsufficient - i ndepend­
ent training is required . I i nvite you to solve
Exercises
1. Wh ite to move
3. Wh ite to move
several exercises of d ifferent types, some
easy and some more d ifficult. They are
u nited by just one factor: everywhere the key
to the solution is one and the same q uestion:
'What does my opponent want, and what
would he play if it were h i m to move?'
2. Black to move
4. White to move
 P rophylactic Thinking ttJ 53

5. White to move 6. Black to move

7. White to move 8. Black to move

54 �
Sol utions
1 . Kholmov-Geller (Vilnius 1 957).
White needs an escape square for his king ,
and in addition it is important to prevent the
exchange of queens 21 . . . iff5! 22 'ii'xf5 gxf5
which would favour the opponent.
2 1 g4!
This move solves both problems and reta ins
somewhat the better chances for Wh ite .
21 .. .'ii' b4 22 'iti>g2 "ike7
Now Wh ite has to reckon with 23 . . . ii.g5(g7)
and 24 .. .f5. Therefore Ratmir Khol mov opens
the centre.
23 d5! exd5? (23 . . . cxd5 24 cxd5 b6 was
essential) 24 'ii'x e7 lbe7 25 g5! (25 ii.xa7?
was weaker: 25 . . . l:la8 26 ii.c5 l:te4 with
equal ity) 25 ... ii.e5 26 ii.xa7 ii.c7 27 cxd5
l:ted7 28 h4, and Wh ite gai ned the advan­
tage.
2. Ti mma n-La rsen (Mar del Plata 1 982).
23 ... ii.xc5? 24 lt::l e4 lt::l xe4 25 ii.xe4 is
unfavourable for Black. If Wh ite should
occupy the e4-square u n h indered , by play­
ing 24 lt::le4 , his chances will be much better.
Simple prophylaxis comes to Black's aid .
23 . . . 'iti> h 8 !
Now i f 2 4 lt::l e4? ! t h e pin 2 4 . . . ii.f5 i s
un pleasant. After 2 4 e3 .l:tbc8 25 lt::l e 2 'ii'f7
Black seized the initiative .
I should mention that his attack on the
kingside could also have been begun imme­
diately, without resorting to prophylaxis:
23 . . . iff7! 24 lt::le4 lt::l g 4 followed by . . . 'ii' h 5.
3. Hort-Mestel (London 1 982).
I n choosing a way of defending his f4-pawn ,
White has to reckon with the freeing advance
. . . e6-e 5 ! .
25 .l:.f1 ? is wrong in view o f 25 . . . e5. 25 g3!?
suggests itself, but in this case too the
Prophylactic Thinking
opponent repl ies 25 . . . e5!? 26 fxe5 fxe5 27
l:txe5 l:!f8! (of course, not 27 . . . 'ii'f7? 28
'ii'x c6 ! , and not 27 . . . .l::t x c5? 28 l:te8+ 'iti>f7 29
l:tf3+ 'iti>g6 30 .l:f.e6+ or 30 'ika6+ ). Black is
threatening both 28 . . . .l::t x c5 , and 28 . . . 'ii'f7
with dangerous pressure on the f-file.
Wh ite probably sti ll has the right to go in for
this variation , if he finds a set-up enabling
him to parry the opponent's immediate
th reats : 28 'it'b5! 1i'f7 29 'ii'e 2 .l:tf6 30 .l::t b 2.
But even here Black retains defi n ite counter­
chances, by conti nuing 30 . . . ifg6 3 1 'iti>g2 (31
'i!i'g2? 'ii'd 3 ! ; 3 1 l:td2 ! ? ; 3 1 l:te7 ! ? ) 3 1 . . . h5!?
32 'ii'x h5?! (32 h4) 32 . . . ifd3.
After 25 g3!? e5!? an i nteresting idea was
suggested by grandmaster Matthew Sadler:
26 l:!.be3! ? exd4 (in the event of 26 . . . e4 27
l:tb3 the position favou rs Wh ite) 27 l:.e7 'i!Vb8
(27 . . . ifd8 28 �e8+) 28 'it'd 1 ! (th reatening 29
'i!Vg4 or 29 ifh5) 28 . . . 'iti>h8 29 'ii' h 5 .l::tg 8 30
.l:l.xg 7 ! ! . Let us conti nue the variation:
30 . . . .l::tx g7 31 .l::t e 8+ l:.g8 32 l:txb8 .l:txb8 33
'ii'x d5 l:tbc8 34 'ii'x d4 'iti>g7. Wh ite has a clear
advantage, but is it sufficient for a win?
Vlastimil Hart preferred a prophylactic move,
enabling h i m to avoid complications.
25 .:tf3 ! ?
From the practical point o f view this decision
is very sensible - White maintains a position­
al advantage, without the risk of miscalcu lat­
ing in complicated variations. For example, if
25 . . .'ii' b 8! (with the idea of 26 . . . e5!) he can
reply 26 'ii' b 3!? 'ikxb3 27 :Xb3 with the better
endgame (27 . . . .l:l.a6 28 .l:tb7 or 27 . . . .l:f.8c7 28
a4!?).
After 25 . . .'ii' b 7?! 26 'ii'd 1 ! (with the idea of
.l:l.fe3 and 'ii'e 2 ) Black should have defended
against the threatened breakthrough on the
e-file by 26 . . .f5 ! .
2 6 . . Jla6? 2 7 l:tfe3 1i'd7
After 27 . . . l:txa2 28 l::t x e6 'ii' b 2 White would
 Prophylactic Thinking
have won by 29 l:16e2 ! 1Wa3 30 .:te7! 'ii' b 2 3 1
'fg 4 'i'xd4+ 3 2 'it> h 1 f5 33 1Wxf5 .:tf8 3 4 .:tea
'ff6 35 'i'xd5+ 'it>h8 36 1Wxa2.
28 'i'e2 lia4 (28 .. .'ii' a 4 29 ::txe6 'ilfxd4+ 30
�h1 ; 28 .. J:tcc6 29 f5) 29 .l:!.xe6 .l:!.xd4 30 c6!
(31 J:te7 'i'f5 32 g4 ! was another way to the
goal) 30 . . .'i'f7 31 .Ue8+ Uxe8 32 'ii'x e8+ 'ilff8
33 'i'xf8+
Black resigned in view of 33 . . . 'it>xf8 34 :c1 .
4. Tai-Ribli (European Team Champion­
ship, Skara 1 980).
23 'ii' b 5!
But not 23 'ii'a4?! i..f8 .
23 . . . a6
23 .if8 24 llxf6! 'ii'xf6 25 'ii'd 5+ and 26
...
'fxa8.
24 'i'd5 'ii'x d5
25 cxd5
By preventing the development of the knight
on b8, White achieves a winning position .
23 'i'd 1 ? ! is far weaker in view of 23 . . . lbc6
24 l::!. d7 l::!. d 8! . The conti nuation in the game
was also unsuccessfu l : 23 lbd2? ! lbc6 24
tt:le4? (24 'ii'a4 'ii'e 8 25 lbe4 f5 26 lbg5 was
stronger) 24 . . . lba5 25 'ii' b 5 lbxc4 (th ree
moves earlier could the knight have dreamed
of such a fate? ! ) 26 .l:tc6 (26 l:r.d7? lbxe3!)
2 6 . . l:l.xc6 27 'ii'x c6 .U d 8 , and Black equal­
.
ised .
5. Makarychev-Be l l i n (Hastings 1 979/80).
Black wants to play 1 8 . . . lbd8 and 1 9 . . . f6 ,
driving away the menacing knight on g5.
White also has to reckon with the manoeuvre
. . . tt:ld7-f6-h7 and with 1 8 . . . 'ii'f6 . 1 8 .l:tf1 lbd8
19 f4 f6 20 lbh3 (with the th reat of 21 'ii'x h5!)
is not bad , but in this case Black can
successfully defend : 20 . . . exd4 21 cxd4 f5 . It
would be desirable to find a more active way
of playing, after which the opponent does not
succeed in implementing his plan ned defen­
sive constructio n . But noth ing is given by
lZJ 55
1 8 l:e3 (with the threat of 1 9 lbxh5+) in view
of 1 a . . . :ha.
18 f4! exf4
1 8 . . . lbd8 1 9 f5 .
1 9 lbe2 lbd8
Little better was 1 9 .. .f3 ! ? 20 gxf3 lbf6 21 lbf4
lbh7 22 lbd5 'ii'd 8 23 'ii'd 2 with advantage to
White.
20 lbxf4
By taking control of the e6-sq uare at just the
right time, Sergey Makarychev has prevent­
ed the important defensive move . . . f7-f6.
Now the attack develops of its own accord .
20 . . . c6 21 'iid 3 .l:.h8 (2 1 . . . lbf6 22 e5! ) 22
l:te2 ! f6 23 lbge6+ lbxe6 24 lbxe6+ 'it>h7 25
.l:tf1 lbf8 26 e5! dxe5 27 dxe5 lbxe6 28 exf6
'ii'c 5+ 29 'it> h 1 ltJfB 30 .l:te7+ 'itr>h6 31 'ii'e4
lbd7 32 f7 .l:tf8 33 �c2 Black resig ned .
Of cou rse, it would be strange to call the
attacking move 1 8 f4 ! prophylactic. But at
any event it was found with the aid of
prophylactic thinking , suggesti ng the need to
prevent the opponent from strengthening his
position .
6. Psa khis-Speelman (Hasti ngs 1 987/88).
Black is the exchange u p , but his knight is in
danger. Wh ite is th reatening �b7-a6-d3. If
29 . . . lld2 30 �xd2 l:!.xd2, then 31 'it>e1 .l:1d7 is
u nclear, but 3 1 1Ic1 ! is very strong.
29 . . . g5! !
30 �a6 f4
3 1 i.. d 3
The waiting move 31 i.. b 5 (or 3 1 �c4) came
into consideration .
31 . . . .l:lxd3
32 exd3 llxd3
3 3 'it>e2 .l:td5!
It becomes clear why the kingside pawns
were advanced : if 34 l:.xd 1 ? Black now has
34 . . .f3+ . 34 i.. d 2? also does not work in view
56 <;i? Prophylactic Thinking
of 34 . . . tt::J xf2 ! . I ncidentally, in the event of
33 . . . .l:!.d4? (instead of 33 . . . l::t d 5) Wh ite would
have won material by 34 gxf4 gxf4 35 .id2 !
tt::Jxf2 36 .ic3 l::t d 3 37 .ie1 .
Now White should probably have played 34
'iitf3. If 34 . . . .l:ld4 he retreats his king with 35
'iot>e2 , creating the th reat of 36 gxf4 gxf4 37
i.. d 2!. And 34 . . . tLlc3 35 i.. x c3 .l:f.d3+ 36 'iot>g4
.l:f.xc3 37 �xg5 leads to an unclear rook
ending.
Lev Psakhis chose 34 gxf4 gxf4 35 .l:tc1 ?
(here too 35 'iot>f3 was necessary). After
35 . . . tLlb2 36 .ic3 tiJd3 Black went on to win .
7. Stei n-Keres (Moscow 1 967).
It would appear that there is noth ing to th ink
about here, and that Wh ite must defend his
e5-pawn with 19 f4 . But let's ask ourselves
what Black will do then . And we establish
that he is plan ning to defend his d5-pawn by
19 .. Jlad8 and then play 20 . . . c4! , including
his knight via c5 and hoping at some point to
advance . . . d5-d4.
Leonid Stein finds a way of forestalling the
opponent's idea .
1 9 a4! !
Now after 1 9 . . . .l:tad8 Wh ite has the strong
reply 20 axb5 axb5 21 .Ua6. If 1 9 .. .'ii' x e5 ,
then 20 I1e 1 ! 'ifd6 (20 . . . .ic7 2 1 i.. g 1 ) 2 1
axb5 axb5 2 2 l::tx a8 l::tx a8 23 i.. xf5 , a n d the
opening of the position is clearly to the
advantage of White, who has the two
bishops. It is no better to play 1 9 . . .f4 20 .if2
'ii'x e5 2 1 axb5 axb5 22 'i!kd 3 ! .
The game continued : 1 9 . . .tLl a 5 2 0 i.. f2 ! 'it>h8
(20 .. .f4 21 b4 ; 20 . . .'i1Vxe5 21 l:.e 1 ! 'ii'd 6 22
axb5 axb5 23 b4 ! ; 20 . . . tt::J c4 21 b3! tt::J x e5 22
axb5 axb5 23 l::tx a8 llxa8 24 Iie1 or 24
i..xf5!?) 2 1 .f:.e1 .l::[ a 7 22 'ii'e 2! b4 23 cxb4
cxb4 24 i.. x b6 'ii'x b6 25 l:1ad1 'i¥c5?
(25 . . . b3 26 .ib1 'iie 6 was more tenacious)
26 .id3 ! , and soon Wh ite converted his
advantage.
8 . Kozu i-Ma rja novic (Yugoslav Champion­
ship, Novi Sad 1 985, variation from the
game).
It is clear that Black has to play for
zugzwang. White will have to sacrifice his
knight on f5, since if it moves anywhere else
the rook will invade on the 2nd rank.
After 64 tt::Jxf5 .llxf5 Wh ite has two possibili­
ties: 65 e4 and 65 g4. Before turning to
calculation , let's see whether it is possible to
prevent at least one of them, i n order to deny
the opponent a choice.
63 . . . �g 1 ! !
I n the event of 6 3 . . . .l:f.a5? 64 tt::J xf5 Itxf5 65
g4 ! hxg4+ 66 �xg4 Wh ite would have easily
gained a d raw: 66 .. J:lf8 67 f5 'ii?e 2 68 �f4!
'ii? d 3 69 'ii? e 5! 'ii? x e3 70 f6.
64 tt::Jxf5 .l:f.xf5
And 65 g4 is not possible because of
65 . . . h4! 66 gxf5 h 3 .
Now let us analyse the conseq uences of 65
e4 .
65 e4 .l:f.a5 66 f5 .l:f.a3+ 67 'ii?t4 'iiff2 68 f6
Other continuations also do not help:
68 'ii? g 5 'ii? e 3 (68 . . . 'ii?x g3 is also good ) 69 e5
'ii?e4 70 e6 .Uxg3+ 7 1 'ii? h 4 l:.g4+! 72 �xh5
'ii?xf5;
68 e5 .l:f.f3+ 69 �e4 (69 'ii? g 5 'ii? e 3 70 e6 �e4
7 1 f6 .l:f.f5+ ! , transposing into the main
variation 68 f6) 69 . . . �xg3 (69 . . . l::t x g3 70 f6
.l:f.f3! is also possible; if 70 e6 Black has both
70 . . . h4 7 1 f6 h3 72 f7 .l:f.f3 73 e7 h2 and
74 . . . h 1 'ii' , and 70 . . . l::tf3 71 �e5 h4 72 e7 h3)
70 e6 .l:.f1 ! 71 'ii? d 4 .l:.e 1 72 'ii? d 5 'ii?f4 (72 . . . h4)
73 f6 l:le5+ ! 74 'ii? d 6 'ii? f5.
68 . . . I:If3+ 69 �g5 'ii?e 3!
Not 69 .. .'�xg3? 70 e5! (70 'ii? x h5? �f4 ! )
70 . . . h 4 7 1 e 6 h3 72 e7 h2 73 e8W h 1 'it' 74
We5+ with a d raw.
70 e5 �e4 71 e6 l:tf5+! 72 'ii? g 6 �e5 73 e7
l:.xf6+ 74 'ii? g 7 .l:!.e6 75 'ii?f7 'ii?tS , and Black
wins.
 tZJ 57

Max D l u g y

A Novelty i s born

W arrives in the mail, I really get excited !

hen the most recent Chess Informant Gaj i c - Dl ugy
New York Open 1 985
Now I can look at some of my favou rite Caro-Kann Defence
openi n g variations and see what novelties
1 e4 c6
the world came u p with , i n the last few
months. Now I ' l l be both armed with new 2 d4 d5
excitin g ideas and forewarned against un­ 3 ltJd2 dxe4
pleasan t surprises in my next tou rnament. 4 ltJxe4 ltJ d7
I can look in the Queen's Gambit Accepted 5 ltJf3 ltJgf6
section and see that the line I started playing 6 ltJxf6+ ltJ xf6
with Black in 1 985 is all the rage now, but that 7 ..ie2 g6
I am still safe there. Skipping back to the
8 0-0 ..ig7
Richter-Rauzer Sicilian B66 variation I 've
been playing since 1 983, I find that Serper's 9 c3
novelty in the previous Informant has been 9 c4 is the topical line assessed as ';!;' in
virtually refuted by Ralf Lau's excellent 1 4 EGO. But more on that later.
Wb1 ! . Good thing I did some work o n that line 9 . . . 0-0
before and that's not going to be a problem 1 0 ltJe5
either. If I only knew how to handle that
GrOnfeld Defence!
Novelty, innovation, new idea , improvement
- what are we talking about here? How is a
novelty born?
This is a story about the evolution of one
seemingly unimportant improvement i n a
.. . tt:'ld7 variation of the Caro-Ka n n . It's only
been played once on a GM level, but it was
enough to put Wh ite's set-up out of com mis­
sion .
The story starts with a game I played in the
1 985 New York Open agai nst loran Gajic, a
player I felt I had to beat to have a shot at the
top prizes i n that first class Swiss. A well known and popular set-up back in
1 985. After the usual 1 O . ..ie6 11 l:te1 White
. .

maintains the centre and although Black

doesn't have particu lar d ifficulties, it's ex-
58 � A Novelty is born

tremely hard to complicate the game. After 32 'i!i'e3+ 'iti>h51

some thought I came u pon an i nteresting, if 33 ..tg3 ..td7
somewhat artificial knight manoeuvre in the 34 ..tc7 'iti>g6
attempt to mix it up.
35 ..td8 l:.a8
10 . . . ttJe8 1
36 'ii'g 3+ 'iti>f7
1 1 ..tf4 ttJd6
And Black, armed with an extra piece, won
1 2 .l:.e1 ..te6 easily.
1 3 ttJg4?! So the risky play paid off ( I even managed to
My opponent is already confused by the new tie for fi rst in the tou rnament), and I
set-up and he begins an artificial knight remembered the . . . ttJe8-d6 manoeuvre as a
manoeuvre of his own in the hope of distinct possibil ity in this variation .
attacking my king. Along comes the following:
13 . . . 'ilfb6
1 4 b4 a5
Gertler - Dl ugy
1 5 tLlh6+ 'iti>h8
Long Island Open 1 985
1 6 a3 'ili'd8
Caro-Kann Defence
1 7 ..td3 tLlb5
1 e4 c6
18 'ii' d 2 axb4
2 d4 d5
1 9 axb4 .l:ixa1 dxe4
3 tLld2
20 .l:.xa 1 'it'd5 4 ttJxe4 ttJd7
21 I:.e1 tLld6 5 tLlf3 ttJgf6
22 h 3 l:ta8 6 ttJxf6+ ttJxf6
23 ttJg4 l:!.a2 7 ttJe5 ..te6
24 'ii'c 1 1t'b3 8 ..te2 g6
25 ..tg5 tLlf5 9 0-0 ..tg7
26 ..txf5 gxf5 1 0 c4 0-0
27 tLlh6 1 1 ..te3
The knight comes back for more but this time
he doesn't come home.
27 . . . f6
28 i.. h 4?
After 28 ..tf4! ..tf8 29 'i!fe3 ..td7 30 'i!i'g3!
Wh ite could have put Black's 'win at all costs'
strategy to the test.
28 . . . ..tf8
29 'ili'e3 'iti>g7!
The start of an unlikely king manoeuvre
which is especially hard to stop in view of
Wh ite's mounting time pressu re.
30 'ii'f4? 'iti>g6!
31 'ii' g 3+ 'iti>xh6 This, of course, is the main line favou red by
 A Novelty is born ltJ 59

ECO. I stopped to think. Once again I need to I n the ending only Black will have win n i ng
complicate the game and make an effort to chances.
solve my open ing problems at the same 1 7 ii.e3 'ii'x d4
time. EGO g ives 1 1 . . . lt:'ld7 , 1 1 .. .'ii' c 7 and 1 8 ii.xd4 e6!
11 . 'i'c8 as the possible moves , evaluating
The simplest way - Black completes his
. .

the position as somewhat better for Wh ite in

development.
all lines. Remembering my experiment from
earlier in the year I started contemplating 19 dxe6 �xe6
11 . ttJe8. And the more I looked - the more I
. .
20 .U.fd 1 l:tfdB
liked it! The knight on d6 will be threatening 21 b3 f5 !
the c4-pawn , the d4-pawn and e3-bishop via 22 gxf5 lt:Jxf5
f5, while the open ing of the long diagonal will 23 �b2 .U.acB?!
make the break . . . c6-c5 all the more
23 . . . lt:lh4 with an advantage was stronger.
plausible. I n fact, I didn't really see what
White could do to get any kind of play going . 24 .i.f3 b6
And so I played . . . 25 �b7 l::. b B
11 . . . li::l e B! 26 �e4 li::l d 6
1 2 li::lf3 li::l d 6 27 �d5 �xd5
1 3 b3 lt:'lf5 28 l:txd5 li::l f7
14 'ikd2 lt:Jxe3 29 l:.xdB+ l:txdB
Black is already equal and eventually he won 30 li::l x f7 �xf7?
a messy and somewhat lucky game. Black's last chance to play for a win was
In October of the same year I played i n a GM 30 . . . �xb2 3 1 li::l x d8 �xa 1 .
tournament and once again encountered the 31 .i.xg7 �xg7
same variation, this time against a tougher 32 llc1 .l:td7
opponent. 33 a4 �f6
34 a5! bxa5
Gruenfeld - D l u gy 35 l:tc5 l:tb7
Manhatta n Chess C l u b l nternational 1 985 36 l:txa5 l:txb3
Caro-Kann Defence
Draw.
(from previous diagram) Is this story over? Is the novelty born? In the
11 . . . lt:JeB ! same tou rnament Yehuda Gruenfeld put my
novelty to a real test with an in novation of his
1 2 �f4
own .
White prepares to strike i n the centre, but the
attem pt backfires as Black is extremely solid.
G ruenfeld - Lei n
12 . . . li::l d 6
New York 1 985
13 d5 cxd5
Caro-Kann Defence
1 4 cxd5 �f5 ( from previous diagram)
15 g4 �cB 11 . . . lt:JeB
White has gai ned some time at the cost of 1 2 'ii' b 3!
seriously compromising his pawn structu re . Wh ite targets the b7-pawn , tying Black's
1 6 'ii' d4 'ii' b 6! knight to its defence.
60 w A Novelty is born

12 . . . ltJd6 1 3 dxe5 ltJd2 ! 14 .ixd2 "ii'x d2 with full

1 3 .l:!.ad 1 ! equal ity, and all other moves transpose to
White successfully defends his centre and 1 1 . . . ltJe8 lines as Black safely reaches the
keeps a comfortable edge. d6-square with his knig ht.
13 . . . 'ii'c B The novelty is born !
1 4 'ii'c 3 l:.d8 Over a year later someone else caught up
with me. It seems the g reat practitioner of
1 5 h3 f6? !
prophylactics did his homework.
1 6 ltJf3 'ii'd 7
Behold !
1 7 l:.fe1 lieS
1 8 .if1 .if7
A.Sokolov - Karpov
1 9 .ic1 a5
Candidates Match 1 987
20 a4 b6 Caro-Kann Defence
21 b3 c5 ( from previous diagram)
22 .if4 'ii'c 6 11 . . . ltJe4!
23 g3 .l:!.adB 1 2 'ili'c2 ltJd6
24 .ig2 'ii'c B 1 3 b3 c5!
25 dxc5 'ii'x c5 1 4 liad 1 ltJf5
26 .ie3 'ii'c 7 1 5 d5 .ixe5
27 ltJd4 f5 1 6 dxe6 'ii'c 7
28 'ii'c 1 f4 1 7 exf7+ .l:lxf7
29 gxf4 1 8 g3 .l:laf8
And Wh ite eventually converted his extra 1 9 .ig4 ltJxe3
pawn i nto a full point. 20 fxe3 l:txf1 +
I mmediately after this game I started looking 2 1 .l:!.xf1
for improvements for Black after 1 2 'ii' b 3, but Draw.
the longer I looked the more convinced 1
You hardly see this once fashionable set-up
became that Wh ite had the upper hand . The
for Wh ite against the . . . ltJd7 Caro-Kan n . It's
tournament was over and I was off to
been replaced by 5 ltJg5 and 5 .ic4 . Yet
Montpellier to watch the Candidates Tou rna­
when the new revised edition of EGO comes
ment, to play in an Open there and to take
out I am sure the assessment of the line will
lessons from Mark Dvoretsky.
revert from ';!;' to ' = ' . It takes q uite a bit to
Mark's approach to teaching chess is d iffer­ erase a l ittle plus sig n , doesn't it?
ent from all other trai ners; he teaches
P. S . Grandmaster Yu ri Razuvaev reports
methods of thinking over anyth ing else. We
having played 1 1 . . . ltJe4! in nu merous blitz
concentrated on prophylactics - or the
games over 20 years ago against Karpov. He
prevention of your opponent's ideas, and
was su rprised to see Karpov remember this
when I came back to New York, I found the
move for his match against Sokolov. It
answer I was looking for. I n stead of reacting
seems there is noth ing new under the s u n , or
to 12 'ii'b 3, Black has to prevent it! So I need
is there?
the knight to go to d6, while preventing 1 2
'ii'b 3. The answer - 1 1 . . . ltJe4 ! ! . The un iversal
move. Now if 1 2 'ii' b 3 Black plays 1 2 . . . .ixe5!

Mark Dvoretsky
Pos ition a l Exercises
Yted with various aspects of positional
ou have to solve ten exercises, con nec­
play (manoeuvres, excha nges, prophylaxis
etc.). Time for the taking of the decisions will
be limited - from 5 to 1 5 minutes.
In my view, this is q u ite sufficient - after a l l ,
you don't have t o calculate any lengthy a n d
complicated variations. You merely have to
approach the position correctly: try to gain a
feeling for it, recogn ise the main problem
facing you , and q u ickly point out the promis­
ing possibil ities for you and you r opponent.
When you solve a study or try to find a
forcing combinatio n , on verification it is not
difficult to ascertain (if, of cou rse, the
exercise is correct) that the solution devised
is the only correct one, and that other
continuations are sign ificantly weaker. With
1 . White to move (1 0 mins.)
61
positional examples things are more compli­
cated - here the situation sometimes al lows
d ifferent approaches. In comparing their
strength , we base this not on precise
variations, but on general evaluatio n , about
w h ich arg u ments are possible. Even so , I
hope that in the exercises offered to you
there will be no particular arg u ments - they
have been carefu lly checked and already
solved earlier by many of my pupils.
For each correct reply you receive two
poi nts, and for an i ncorrect one you score
zero . If the reply is only partially correct, and
you fail to g ive some important variations,
then you receive one point. I assume that in
some cases I will have to use i ntermediate
values: a half or one and a half points.
2 . Black to move (5 m i ns.)
62 � Positional Exercises

3. Wh ite to move ( 1 0 mins.) 4. Black to move (1 0 m i ns.)

5. White to move (1 0 mins.) 6. Wh ite to move (5 m i ns.)

7. Wh ite to move (1 5 mins.) 8. Black to move (1 5 m i ns.)

 Positional Exercises ctJ 63

9. Wh ite to move (1 0 mins.) 1 0. Black to move (1 0 mins.)

Sol utions

1 . Knaak-Geller (Moscow 1 982).

White has good attacki ng chances, but for
the moment some of his pieces are not
taking part in the activity on the kingside. He
must first consolidate his position .
22 ..tc2 !
An excellent reg rouping . The bishop goes to
b3, intensifying the pressure on the d5-point
and simultaneously covering its own pawn
on b2. Then the rook at b 1 will defend the d4-
pawn from d 1 , after which the knight will
occupy the very i mportant f4-point, once
again attacki ng d5 and at the same time
approach ing the enemy king . Black is u nable Wh ite has a decisive advantage. The con­
to oppose this plan of action. For example, in cluding moves were made in severe time­
the event of 22 . . . llc4 there is the strong reply trouble.
23 l:.h4! followed by 24 ..tb3.
28 h6 (28 lDcxd5 was simpler) 28 ... g6 29
22 ... i.f7 23 ..tb3 .l:.ce8 24 .l::!. b d1 lDg5 lDcxdS (29 lDxb5! ? ) 29 ... a5 (29 . . . 'iti>h7!?) 30
In the event of 24 . . . lle3 25 lDf4 'ifd6 Wh ite h7+ lDgxh7? (30 . . . 'iti>h8! was essential) 31
wins by 26 lDfxd 5 ! l:ie2+ 27 lDxe2 l::t x e2+ 28 lDxg6 Black lost on time. His position is
lt>xe2 'i!i'xg3 29 lDe7+ 'iti>h8 30 ..txf7. hopeless, as is apparent from the variation
25 t2Jf4 'ii'd 6 (25 ... l:te3 26 lDfxd5) 26 l:td3! 31 . . . lDxg6 32 lDe7+ l:texe7 33 1i'xg6+ 'iti>f8
(intending 'ii'g 4 followed by lDcxd5 ) 26 ... b5 34 'ii' h 6+ 'iti>e8 35 'ii'x h7 'iti>d8 36 ..txf7 l:.xf7
27 'i'g4 l:td7 37 'ii'f5.
64 �
2. Zil berman-Taimanov (Moscow 1 979).
White is th reatening to gain an advantage by
advancing his c-pawn . This must be prevent­
ed .
17 . . . �a6!
A good prophylactic move - it was suggested
by nearly all the participants in the competi­
tion ( 1 8 c5? is not possible in view of 1 8 . . . bxc5
1 9 dxc5 lllxc5!). But Black had to reckon with
the following reply by the opponent: anyone
who did not see it receives only half a point.
1 8 'i!t'a4
Now the plausible 1 8 . . . b5? is i ncorrect i n
view of 1 9 'it'b4 ! with a b i g advantage for
White. Those who wanted to play this each
lost a point.
Vova Baklan earned one and a half points:
he saw the refutation of . . . b6-b5, but he did
not manage to fi nd the correct cou rse, which ,
unfortunately, was suggested only by Sasha
Chernosvitov.
18 . . . i.b7!
The captu re of the a7-pawn is dangerous ­
the queen risks becoming trapped in the
enemy position . Wh ite should reconcile
himself to the roughly equal position arising
after 1 9 'ii' b4 'Wic7 20 d5 (20 tt:'!f4? �xg2 2 1
'it>xg2 'ili'b7+ 2 2 'it>g 1 e5! ) 20 . . . tt:'!c5 2 1 l:.d 1 .
I n the game there followed 1 9 'i¥xa7? �xg2
20 �xg2 ii'c6+ 21 d5 (21 'it>g1 l:ta8)
21 . . . exd5 22 tt:'!d4 'ii'd 6.
Positional Exercises
White is in serious d ifficulties, for example:
23 'iia 4 tt:'!e5! 24 cxd5 Wxd5+ 25 f3 (25 'it>g 1
'ii'x d4) 25 . . . b5. He should probably have
sought salvation i n the variation 23 tt:'!b5
'ii'e 6! 24 l:te 1 'ili'c6 25 tt:'!d4 'ili'xc4 26 tt:'!e6!
(26 1i'xd7 'i!Vxd4 27 l:tad 1 'i!kc5 is less good)
26 . . . fxe6 27 'i!Vxd7 .
23 l:td1 ? tt:'!c5! (th reatening . . . l:t b 7 or . . J:ta8;
it does not help to play 24 tt:'!b5 'ii'g 6 25 'ilie7
dxc4) 24 tt:'!f5 'i!i'e5! 25 tt:'!e7+ 'ith8 26 Ite1
(26 tt:'!c6 'ii'e 4+ 27 �g 1 l:tb7) 26 .. .'ii' d 6 27
tt:'!f5 'ii'f6 Wh ite resig ned .
3. Pinter-Adorjan (Prague 1 985).
17 tt:'Jxe4! dxe4
1 8 i.xb4!
1 8 i.xe4? �xe4 1 9 l:txe4 is much weaker;
Black can choose between 1 9 . . . 'ii'd 5 20 'ilie2
tt:'!xa2 and 1 9 . . . tt:'!d3 20 l:tc3 tt:'Jxf2 21 'it>xf2
il.xd4+ .
18 . . . exf3
1 8 . . . cxb4 1 9 i.xe4 leads to the loss of a
pawn without any compensation .
The game went 1 8 . . . e3 1 9 .l:.xe3 i.xf3 20
'ii'c 2! g6 2 1 dxc5 �g5 22 .l:.d3 'ii'c 8 23
'ii'c4+ 'itg7 24 'ii'c 3+ 'ith6 (24 . . . i.f6 25 'ii'd 2)
25 h4 Black resig ned .
1 9 dxc5 'i!i'c8
1 9 . . . bxc5 20 il.xc5.
20 c6
 Positional Exercises ltJ 65

This variation had to be calculated by Wh ite But, u nfortu nately, most of you chose a
when he embarked on his operation. His resou rceful , but not very successful way of
position is won , for example: 20 ... .ltxc6 parrying the opponent's main idea - 1 8 . . . b5?!
(20 . .'i'h3 21 'ii'xf3) 2 1 'ii'c 2 'ii' h 3 22 'ii'x h7+
. (for it only half a point is awarded ). After 1 9
'fxh7 23 ..txh7+ <it>xh7 24 ..txf8 . cxb5! followed by 2 0 lt:Jc4 Black does not
Jozsef Pinter found a purely concrete way of have sufficient compensation for the sacri­
achieving an advantage. Only by a great ficed pawn .
stretch of the imagination can this example Less convincing is 1 9 'iix b5? ! , hoping for
be called 'positional' (Wh ite carried out a 1 9 . . . .l::ta b8 20 'iig 5 'ii'x g5 2 1 fxg5 l:txb2 22
series of favourable exchanges). But equal­ gxf6 l:txd2 23 fxe7 .l:txe7 (or 23 . . . .l::tx a2 24
ly, White's solution can not be called combi­ .l::t a 1 ) 24 .l::tf2 , when in the rook ending the
native - after all, noth ing was sacrificed . It l im it of Black's dreams is a draw. It is
would perhaps be more correct to cal l it stronger to i nterpose 1 9 . . . lt:Jg4 ! , and only
tactical . after 20 h3 - 20 .. Jlab8 (but not 20 . . . l:teb8?
Tactics a re an immeasu rably broader con­ 21 'iig 5 'ii'x g5 22 fxg5 .l:r.xb2 23 tt:Jxe4 ).
ception than combinations. When we say
that Emanuel Lasker was a great tacticia n ,
we don't mean that he was constantly
sacrificing something. No, simply the world
champion was excellent at fi nding the strong­
est resources for both sides - accurate
moves , precise variations.
Tactical skil l plays an enormous role i n
chess , and b y no means o n l y in sharp
combinative situations. With its help a player
can tenaciously hold difficult positions, con­
stantly erecting new barriers i n the oppo­
nent's path , or on the contrary, he can find
the quickest way to convert an advantage .
Even the solving o f purely strategic problems
in quiet positions can not be done without The reply 21 �a4? suggests itself, but it is
tactical elements - after a l l , our plans can prettily refuted by 2 1 . . .lt:Jf5! 22 hxg4 lt:Jh6!
only be carried out by means of specific with irresistible threats. Wh ite's position also
moves, which have to be seen and, if looks anxious after 21 'ii' a 5?! :xb2 22 hxg4
necessary, calculated . 'ii'x g4 23 :f2 lt:Jf5! 24 lt:Jxe4 l:tb6 followed by
25 . . . l:th6. He is forced to play 21 'ii'g 5! 'i!i'xg5
22 fxg5 l:txb2 23 tt:Jxe4 ! lt:Jg6 24 lt:Jc5 lt:Jh4 25
4. Lisitsyn-Tolush (Len ingrad 1 938).
hxg4 or 24 . . .lt:Jxe3 25 .l:r.f2 with roughly equal
What does Wh ite want? Without exceptio n , chances.
a l l the participants in t h e competition correct­
The strongest is a cool-headed prophylactic
ly decided that Black should be thinking not
move.
about the defence of his c7-pawn , but about
parrying the th reat of the q ueen exchange 18 . . . h6!
('fg5). It is incorrect to play 1 8 . . . c6? (or After 19 'ii'x c7?! lt:Jf5 there is no satisfactory
18 . b6?) 1 9 'ii'g 5! lt:Jg6 ( 1 9 .. .'i!i'xg5 20 fxg5
. . defence against the threat of 20 . . . lt:Jg4 . For
lt:ld7 2 1 tt:Jxe4) 20 'ii'x h4 lt:Jxh4 2 1 f5! . example: 20 l:!.e2 lt:Jg4 21 g3 lt:Jxg3 (2 1 . . . 'ikh3
66 � Positional Exercises

also looks tempting, intending either to

captu re the e3-pawn with the knig ht, or to
play . . . h6-h5-h4, or at an appropriate
moment to nevertheless sacrifice the knight
on g3) 22 hxg3 'ifxg3+ 23 'it>h 1 'ifh3+ 24
'it>g2 tt:Jxe3 25 .l:tff2 l:tac8 with advantage to
Black.
After 20 h3 tt:Jg3 Georgy Lisitsyn had to
accept the loss of the exchange, since if 2 1
l:.f2 there follows 2 1 . . . tt:Jg4! . 2 0 . . . 'i!r'g3 ! ? ,
attacking the e3-pawn, also came into
consideration.
Remember: a queen together with a knight
(and the more so, with two knights) in the - position after 23 f5 -

vicinity of the enemy king constitutes a

powerful force! 23 . . . f6 24 l:.d6 .tea 25 .l:txd8 (25 l:txf6?! is
weaker in view of 25 . . . tt:Jd7 and 26 . . . tt:Je5)
5. B el i avs ky-H erzog (Mexico 1 977). 25 . . . l:txd8 26 e5 fxe5 27 'ilfxe5 'ilff7 28 tt:Je4
White has an excellent position. It can be l2Jd7 29 'i!r'c3 gxf5 30 tt:Jg5 'i!r'd5 31 tt:Je6 tt:Jf6
strengthened either by the q uiet move 18 g3, 32 'i!r'xf6 Black resigned .
or by 18 tt:Je2 or 18 'iot>h2(h 1 ), prepa ring g2- Why did this example prove so difficult? The
g4. This is all not bad , but rather slow. The sharp transformation of the position carried
position lends itself to more energetic meas­ out by Alexander Beliavsky is somehow not
ures. in keeping with our usual approach to such
The only one to suggest the correct way was favou rable situations - i n them we prefer to
l lya Makariev. manoeuvre q u ietly, g rad ually improving the
tt:Jf4 placing of the pieces. And sometimes we will
1 8 g4!
miss favourable concrete possibilities. Gen­
19 tt:Jxe5 .txe5
erally speaking, the transformation of an
20 .txe5 tt:Jxh3+ advantage is a psychologically d ifficult pro­
21 .txh3 ced u re, demanding at the same time both
21 'i!r'xh3 .l:f.xe5 22 f4 l:tee8 23 f5 (or 23 e5!? dynamic th inking , and subtle positional eval­
h5 24 'i!r'h4 ! ) was eq ually good . uation .
21 . . . l:.xe5 Later the American player Maurice Ash ley
22 f4 .:t eeS suggested another, also very concrete way
of playing for Wh ite : 1 8 .tc4 ! ? . The pressure
23 f5
on the f7 -point is rather u npleasant, and also
(see diagram) the strateg ic threat 1 9 g4 tt:Jf4 20 tt:Jxe5
remains in force. The critical reply is 1 8 . . . ..te6.
The exchanging operation has enabled Then follows the unexpected 1 9 ..txf8 ! .l:i.xd 1
Wh ite to beg in a very da ngerous pawn 20 .l:txd 1 ..txc4 (20 . . . ..txf8 2 1 tt:Jxe5 is no
storm . The f5-pawn has restricted the mobil­ better) 2 1 ..txg7 tt:Jxg7 22 tt:Jxe5! :xe5 23
ity of the opponent's bishop and knight. 'ii'd 4. However, this clever idea can be cal led
Things are bad for Black - Wh ite is threaten­ into question by 20 . . .'it>xf8 ! 2 1 ..txe6 l:.xe6 22
ing 24 f6 or 24 e5 followed by tt:Je4 . tt:Jg5 (22 Il.d8+ 'iit e 7! , but not 22 . . . I!.e8? 23
 Positional Exercises
'i!i'd2 ) 22 . . . lle8 23 ltJxh7+ 'it'g8 24 ltJgS .i.h6,
when the position remains unclear.
6. Hort-Karpov (Amsterdam 1 98 1 ).
Black obviously wants to play . . . b6-bS ,
supporting his c4-pawn and preparing . . . ltJe4
or . . . ltJb6 . The same reply 1 4 . . . bS! follows
both after the attempt to undermine the pawn
chain by 14 b3, and after the tem pti ng 1 4
tt:'le5.
1 4 a4!
An important prophylactic move. Now 1 S b3
and 1 5 ltJeS are dangerous positional th reats
for White.
Anatoly Karpov played badly and was soon
in serious d ifficulties.
1 4 . . i.c6?! (in order to answer 1 S b3 with
.
15 . . . b5) 1 5 ltJe5 \!Vc7 1 6 ltJxc6 'i!i'xc6 1 7 i.f3
After the exchange of Black's bishop his dS­
pawn has become weak. 1 8 ltJxdS is
threatened . After 1 7 . . . .l:tfe8 Wh ite has the
strong reply 1 8 i.xf6 ltJxf6 1 9 e4 , while if
1 7 . . .l::!. a e8 , then 1 8 b3! bS 1 9 axbS axbS 20
.
'i'f5!.
1 7 ... i.b4? 1 8 ltJxd5! ltJxdS 19 �fS (but not
1 9 li'xc4? �xc4 20 l:txc4 in view of 20 . . . bS
followed by . . . ltJ7b6) 1 9 . . .'il¥xa4 20 .i.xdS
.l:!.ac8 21 b3! cxb3 22 l:lxc8 llxc8 23 'i!i'xf7+
'.t>h8 24 .i.xb3 'ii' b S 25 i.e6 l:.f8 26 i.xd7!
(not 26 'ii' x d7?! 'i!VhS ! ) Black resigned.
lD s1
The solution that you are looking for does not
necessarily lead to i m med iate success. lf the
opponent also rises to the occasion and finds
the best response, the outcome often re­
mains u n clear. That is the case here:
although 14 a4 ! was u ndoubted ly correct, if
Black had replied 1 4 . . . ltJe8! he would have
retai ned a defensible position. I ncidentally,
1 4 . . . ltJe4?! was weaker in view of 1 S i.xe7
'i!i'xe7 1 6 ltJxe4 dxe4 1 7 ltJd2 bS 1 8 b3!.
7 . Geller-Fischer (Curac;ao Cand idates
1 962).
Wh ite has to reckon with the threat of
29 . . .'it'xaS. He doesn't want to put his roo k
on a 1 - this is too passive. Noth ing is
ach ieved by 29 'ii' b 6 'i¥xb6 30 l:.xb6 (30 axb6
'ili>f8 ) 30 .. .'.ti>f8! (of course, not 30 . . . .l::!. x aS? 3 1
l:td6) 3 1 d6 .l::!.x aS 32 h3 .l:!.cS , a n d he is also
u nsuccessful with 29 l:td 1 'i!i'xaS 30 \!VxaS
l:!.xaS 31 d6 i.. d 7 32 l:tb 1 bS.
But if he were able to play his rook to b6 . . .
This would solve the problem of the aS­
pawn , and Black's central blockade would
prove insecure.
29 'ii'a 4! i.d7
29 . . . 'it'f8 30 l:tb6.
30 \!Val !
31 .l:.xb7 is th reatened, and if 30 . . . i.c8 there
follows 3 1 l:tb6 with a decisive positional
advantage. Black is forced to captu re the aS­
pawn with his rook, al lowing the cou nter­
stroke on b7. 30 . . . 'ii'x aS 3 1 'ii'x aS .l:!.xaS 32
.l:!.xb7 is totally bad for h i m .
30 . . . .l:!.xaS
31 .l:!.xb7!
Diana Darchiya and Sergey Movsesian
suggested 3 1 "ii' e 7? ! , which is much wea ker
- because of this they each lost half a point.
The completely correct solution was found
by Vova Baklan and Vad i m Zviag intsev.
31 . . . 'i!i'xb7
31 . . . .l::!.x a3 is hopeless: 32 .l::!. x c7 .:a 1 + 33 .i.f1
i.fS 34 g4! (34 f3 hS 3S 'ittf2 l:ta2+ is fa r less
68 � Positional Exercises
convincing) 34 . . . i.xg4 35 h3!? (35 'it>g2)
35 . . . .txh3 36 'it>h2 - after the exchange of
bishops, the con nected passed pawns in the
centre decide the outcome.
32 'i\Vxa5
White has achieved a decisive positional
superiority.
32 . . . g6
33 h3 'iVb1 +
34 �h2?
As Garry Kasparov pointed out, 34 iLf1 ! was
correct: 34 . . . 'iVb7 35 1!Vd8+ 'it>g7 36 "iile 7 or
34 . . . ..l1Lf5 35 d6 i.d3 36 d7. The move in the
game allows Black a savi ng chance.
34 ... i.f5?
34 .. .'it'c2 35 'iVd8+ �g7 36 't\Vxd7 't\Vxe2 was
necessary. The variation by Braslav Rabar:
37 't\Vc7 a5 38 f4! a4 (38 . . . exf4 39 11Vxf4) 39
fxe5 a3 40 e6 a2 41 'iVxf7+ 'it>h6 42 'iVf6
contains many weak points. For example,
instead of 37 . . . a5 Black ca n consider
37 . . .'iVe4!? 38 g3 1!Vd4 39 '.tg 1 e4 , and on
the next move 37 . . . 11Ve3!? g ives a d raw.
Finally, instead of the 'cooperative' 40 . . . a2?
there is 40 . . . 11Vf2 ! 4 1 e7 (4 1 'iVe5+ f6! )
4 1 . . . "iile 3, and Black saves the game,
35 "ifc3!
White has consolidated and his central
pawns have become a powerful force.
35 . . .'i!Ve4 36 i.f3 �d4?! (36 . . . �d3 was more
tenacious) 37 'ii'x d4 exd4 38 g4! ii.c8
38 . . . a5 39 gxf5 a4 40 d6 '.tf8 41 c5 or
38 . . . i.. c2 39 c5 d3 40 c6 i.a4 41 d6 was no
better.
39 c5 aS 40 c6 �f8 41 d6
Here the game was adjourned , and Fischer
resigned : 41 . . . 'it>e8 42 i.d 1 , 41 . . . d3 42 d7
i.xd7 43 cxd7 'it>e7 44 i.c6 d2 45 i.a4 , or
4 1 . . . . a4 42 c7 a3 43 i.c6 a2 44 d7 i.xd7 45
i.xd7 a 1 'iV 46 c8'ii' + .
8. Karpov-Lerner (Moscow 1 983).
Black is a pawn down . Should he exchange
rooks? At first sig ht it may seem that his
hopes of cou nterplay i nvolve creating an
attack against the wh ite king , which is stuck
in the middle of the board , which means that
the rooks should not be exchanged . That is
how Konstantin Lerner reasoned , in choos­
ing 41 . . . .l:te8? .
But after 42 .l:tg2! (42 .l:th2?! f6! was less
accu rate) there was the terrible th reat of
exchanging the q ueens: 43 'iVg4 ! . I n the rook
ending White's extra pawn and centralised
king would give h i m an easy wi n .
B u t i f the exchange i s avoided , he is the first
to beg in an attack on the enemy king:
42 . . .'ii' d 7 43 h5 �dB 44 hxg6 'ii'd 4+ 45 '.tf3
'iVd 1 + (45 . . . iVd5+ 46 'it>g4 '.tg7 was more
tenacious) 46 l:i.e2 'it'f1 47 '.te3 f5 48 l:i.e1
'iVbS 49 'i!Vh3 'ii'c 5+ 50 '.tf3 Black resigned .
A similar picture arises after 4 1 . . . lta8? 42
l:th2! (better than 42 l:lg2 lta1 ! ) . If 42 . . . lta1 ,
then 43 h5, but otherwise Wh ite offers the
advantageous exchange of q ueens: 42 .. .'ilt'h5
43 �f3 or 42 . . . lta2 43 11Vg5 .
It turns o u t that t h e rooks should have been
exchanged .
41 . . . ltxd2!
4 1 ... ltd5? is much weaker in view of 42 ltxd5
cxd5 43 'ii'g 2! d4+ 44 'it>xd4 'ifxf4+ 45 'ii'e4 ,
and Wh ite should win the q ueen ending. All
his pawns are securely defended by the
 Positional Exercises
queen, whereas the black b7-pawn is weak
and will soon be won .
42 'iti>xd2 'ii'e4
43 'ii'e 3
43 h5 'it'd4+ 44 'iti>e2 'ii'e4+ .
43 . . . 'ii' h 1
After 44 'iff2 'ii'e4 ( o r 4 4 . . .'iti> h 7 ) t h e material
advantage is not felt i n view of the activity of
the black queen and the vulnerabil ity of the
white pawns. Black retains excellent d rawing
chances .
9.Korchnoi-Geller (Candidates Match , 7th
Game, Moscow 1 97 1 ).
Black is more actively placed . I n particular
White has to reckon with a pawn offensive on
the queenside: . . . b7-b5-b4. How can this be
opposed? The correct solution was found by
Maxim Boguslavsky and Vad i m Zviagintsev.
1 7 a3 ! !
A subtle prophylactic move. I f 1 7 . . . b 5 Wh ite
intends 1 8 lt:Ja2 ! and then lt:Jb4, exploiting
the weakness of the c6-sq uare. At the same
time he plans to evict the powerful knight
from c4 by 'ii'a 2, tt:Jce2 and b2-b3.
Yefim Geller was unable to devise an
effective cou nter-pl a n , and as a result White
soon completed the development of his
pieces and seized the in itiative. Black should
have opted for a change of the pawn
ttJ 69
structure, by advancing his e-pawn: 1 7 . . . e6!?
1 8 dxe6 fxe6 (recommended by Alexander
Shabalov), or 1 7 . . . b5 1 8 tt:Ja2 e5! 1 9 dxe6
fxe6 .
1 7 . . . 'ii'c 5 1 8 'ii' a 2 .l:tc7 1 9 tt:Jce2 ! l:l.bc8
( 1 9 . . . lt:Jxd5? 20 lt:Jb3) 20 b3 lt:Je5!? (20 . . . lt:Jb6
21 e4 with an obvious advantage to White)
21 ..ltd2 ! ?
With t h e opponent's pieces s o active, White
has to be extremely carefu l . I n the event of
2 1 f4? ! 'ifb6(a7) the immediate captu re of
the knight is clearly bad - 22 fxe5? dxe5
(with the th reat of the rook invasion at c2 ),
while after 22 l1b 1 (or 22 'ili'b 1 ) 22 . . . ..te8 it
also looks risky.
But 2 1 e4! ? followed by 22 ..lte3 deserved
serious consideration .
2 1 . . . ..lte8 (2 1 . . .tt:Jxd5? 22 b4) 22 l:tac1 'iib6
23 .l::!.x c7 .l::tx c7 24 .tel ..tb5 25 'iid 2 ..txe2
26 ..lta5! 'ii'a 7 27 tt:Jxe2 .l::!. c 8 28 .l:!.c1 'ii b 8 29
.l:!.c2 (29 .l:!.xc8+!? 'ifxc8 30 'ii'c 1 ) 29 ... lt:Jed7
30 'ii' c 1 .l:!.c5 31 .l:!.xc5 tt:Jxc5 32 'ii'c 2 11i'e8 33
lt:Jd4, and Wh ite stands clearly better - he
has the two bishops and more space.
An example typical of Victor Korch noi's play.
He aimed for positions with a spatial advan­
tage, even if this involved a certain risk. By
cool-headed actions he usually managed to
exti nguish the opponent's activity and, by
exploiting his strategic trumps, seize the
in itiative.
70 � Positional Exercises

1 0. Gavri kov-Vitolins (Severodonetsk After 23 fxe4 .l:txg2+ Black retains a danger­

1 982). ous attack. It is impossible to calculate all its
Black is a pawn down , and his opponent is consequences in advance, but it is clear that
intending f2-f3 . He must act with the utmost there is practically no risk of losi ng, whereas
energy. By resou rceful play Alvis Vitolins a win may well be found (although it also may
succeeds in emphasising the insecure posi­ not). I , incidental ly, do not see one. Here is a
tion of the enemy king . possible variation : 24 'ii'x g2 �xe4 (24 . . .
l:lxg2+!? 2 5 'Ot>xg2 'ii'g 5+) 25 lDf3 (25 llf2 ;
17 . . . g5!
25 'ii'g 3!?) 25 . . Jixg2+ 26 <iii>x g2 'it'g5+
18 'it'h3 (26 . . . 'i!i'h4 27 h3) 27 <iii> f2 'iff5 28 <itg3 ! .
1 8 'ii'h 6 is completely bad : 1 8 . . . g4 with the The game concluded 23 g4? .ll 5 g6 (th reat­
threat of 1 9 . . . l:lh5 20 'iWf4 e5. ening 24 . . . ltJg5) 24 'ii' h 5?! (if 24 'ii'g 2 Black
18 . . . g41 would have repl ied 24 . . . 'ii' h 4 or 24 . . . f5)
1 9 �xg4 ltJxg4 24 . . .ttJf6 25 'it'b5 ttJxg4 26 fxg4 'ifh4, and
White resigned in view of 27 'ii'x b7 l:lxg4+ 28
20 'i¥xg4+ l1g5
� h 1 'ifxh2 + ! .
At the cost of another pawn , Black's bishop
and rook have establ ished coord ination - Let us sum up. T h e competition proved
they are both attacking the g2-point. But the sign ificantly more difficult for you than I
calculation of the variation should probably expected . Only Vad i m Zviagintsev correctly
be continued for a couple more moves . solved more than half of the exercises. By a
21 'i!i'h3 'Ot>h8 large marg i n he took first place (with twelve
22 f3 and a half points). In second place was Petya
llfg8!
Kirjakov with th ree points less, while Vova
Baklan finished th ird .
As you see, taking a positionally correct
decision in a l i m ited time is no easier than
finding a combination or accu rately calculat­
ing a long variation . Probably because here
you simulta neously have to display both a
correct evaluation of the resulting position,
and a clear vision of the tactical resou rces.
The slig htest mistake i n either of these
factors makes the search much more diffi­
cult, demands additional expenditure of time,
and i n general takes you far away from the
correct course.
 71

PART I I

Ways of Looking for

Positional Solution s

Artur Yusu pov

Ma noeuvri n g

O positional
ne of the most complicated elements of thought from N imzowitsch's book g ives an
play is the sh ifting of the excellent descri ption both of the events in
struggle from one part of the board to this game, and of the essence of manoeu­
another. The point is that a game is rarely vring against weaknesses:
won by breaking through the opponent's 'The process of manoeuvring against two
defences at one place. Normally one has to weaknesses can roughly be characterised
seek roundabout ways and try to give the as follows: two weaknesses, in themselves
opponent new weaknesses , in order to then quite defendable, are in turn put under fire,
began manoeuvring against them. This the attacker relying mainly on his te"itorial
procedure is closely linked with the principle superiority - his superior lines of communi­
of two weaknesses , which we have men­ cation. The game is lost because at some
tioned many times. This is one of the most moment the defender is unable to keep pace
important components in the technique of with the opponent in speed of regrouping. '
converting an advantage.
If the following game had been played by Anand - Kamsky
Aaron Ni mzowitsch , it would certainly have
PCA Candidates Match , 9th Game, Las
been incl uded i n the chapter 'Manoeuvring
Palmas 1 995
against enemy weaknesses when possess­
ing a spatial advantage' from his book Chess Ruy Lopez
Praxis. 1 e4 e5
It is usefu l to follow how Anand constantly 2 ltJf3 ltJc6
changed the direction of the attack, creating 3 i.. b 5 a6
one problem after another for the opponent, 4 i.. a 4 ltJf6
and how subtly he combi ned offensive and 5 0-0 i.. e 7
prophylactic actions. The superficial impres­
6 .l::.e 1 b5
sion, that Gata Kamsky lost the game
submissively, proves deceptive : simply he 7 .ltb3 d6
was confronted with insoluble problems, 8 c3 0-0
defend ing on different fronts. The following 9 h3 .ltb7
72 � Manoeuvring
10 d4 Ue8
In this match Kamsky pin ned his hopes on
the Zaitsev Variation of the Ruy Lopez. In my
view, this game practically decided the
outcome of the enti re duel. Victory in it gave
Anand additional confidence. In tu rn , Kams­
ky's failure forced him to change his open ing
and to take a risk in the 1 1 th game, by
playing the Sicilian Defence, which proved
disastrous for him.
�f8
1 1 lt:Jbd2
1 2 a4 h6
1 3 �c2 exd4
14 cxd4 lt:Jb4
1 5 ..tb1 'ii'd 7
Kamsky constantly varied his handling of the
Zaitsev Variation. The move made by h i m ,
apart from t h e abstract aim o f con necting the
rooks, also has a somewhat more camou­
flaged aim: if 16 axb5 Black can recapture with
the queen , creating the threat of . . . lt:Jd3. The
standard 1 6 e5 dxe5 1 7 dxe5 lt:Jh5 1 8 axb5
led after 1 8 . . . 'iVxb5 1 9 'iVb3 .l:.ad8 to double­
edged play in the game Beliavsky-Smejkal
(Sarajevo 1 982). Theory recommends 1 6
.l:ta3, but in this case, by continuing 1 6 . . . bxa4 ,
Black gains a tempo compared with the
variation 1 5 . . . bxa4 16 .l:txa4 a5 1 7 .l:l.a3 it'd? .
1 6 b3 ! ?
Anand deviates from the well-studied paths
and employs a rare , but solid and logical
conti nuation. Wh ite does not hu rry to deter­
mine the situation either in the centre , or on
the queenside, preferri ng fi rst to complete
his development.
16 . . . g6
16 . . . c5? ! is premature, since after 1 7 �b2
the wh ite bishop beg ins to threaten the
knight on f6 . True, in the variation 1 7 . . . cxd4
1 8 �xd4 'iYd8 (Anand) Black can hold on: 1 9
axb5 axb5 20 :xa8 ..txa8 2 1 'ife2 ..tc6 , but
the simple 18 lt:Jxd4 sets him more un pleas­
ant problems.
17 ..tb2
This natu ral move would appear to be a
novelty. I n the only game known to me in
which this position occu rred , Van der Wiei­
Karpov (Amsterdam 1 99 1 ), Wh ite conti nued
1 9 axb5 (on moves 1 1 - 1 2 the players
employed a repetition of moves, typical of
this variation - 1 1 lt:Jg5 l:!.f8 1 2 lt:Jf3 lle8 , and
the move order was also d ifferent: instead of
1 3 . . . exd4 the game went 1 5 . . . 'ili'd7 1 6 b3
exd4 1 7 cxd4 lt:Jb4 1 8 � b 1 g6) 1 9 . . . 'ii'x b5!?
20 d5, but after 20 . . . c6 21 ..tb2 lt:Jh5 Black
created cou nterplay in the centre.
17 . . . ..tg7
1 8 'i!i'c1 !
This subtle prophylactic move is sign ificantly
stronger than 1 8 d5, when there could have
followed 1 8 . . . c6 , suggested by Va n der Wiel .
White defends his bishop and prepares a
reg rouping on the long d iagona l : �c3 and
'i!Vb2 . Simultaneously he takes control of the
f4-sq uare and defends against the possible
movement there of the black knight (now if
1 8 . . . lt:Jh5 there fol lows the simple 1 9 lt:Jf1 ).
The wh ite pieces beg i n as though to by-pass
the knight on b4 , which is g radually trans­
formed i nto a detached observer.
18 . . . l:tac8? !
Black prepares the . . . c7-c5 advance, but
now in many variations Wh ite acq u i res
additional possibil ities on the a-file. The
immediate 1 8 . . . c5 looks stronger, since if 1 9
 Manoeuvring
dxc5 Black has 1 9 . . . llac8 ! ( 1 9 . . . dxc5 is
weaker because of Anand's suggestion 20
'lxc5 tLlxe4 21 .i.xe4 .i.xb2 22 .i.xb7 ! ) .
However, then he would have t o reckon with
1 9 e5!?.
19 .i.c3 c5
20 d5 'ii'e 7?!
Anand gives a n interesti ng assessment of
the position : 'It is not hard to see what
White's advantage comprises. He has two
"bad pieces" - bishop at b 1 and rook at a 1 ,
but both conta i n considerable potential ener­
gy. It will be easy for me to open the a-file "at
my leisure", while the bishop is restra i n i ng
the f7-f5 break. But Black's knight at b4 and
bishop at b7 are simply bad - and this is the
main distinction . '
I n view of the constant threat o f a n i nvasion
on the a-file it is extremely d ifficult for Black
to carry out a blockade on the dark sq uares,
which is practically his only pla n : . . . "iie 7 and
. . lL'ld7 . Therefore he should have consid­
.
ered the immed iate 20 . . . tLlh5 21 .i.xg7 'it>xg7
(2 1 . . tt::l x g7 22 lLlf1 with advantage) 22 'ii' b 2+
.
f6, although this position looks i n favou r of
White . I prefer another defensive ma noeu­
vre: 22 . . . 'ii'd 8!?, retaining the possibil ity of
covering the q ueen side with the q ueen . After
23 tt::lf1 Black can choose between 23 . . Jic7
followed by . . . .i.c8 , and 23 . . . ttJd7 24 .i.xg7
'lt>xg7.
21 tLlf1
The knight has no futu re on d2, and so Wh ite
transfers it to the kingside. The routine 2 1
'1b2 would merely have eased the oppo­
nent's defence, but now in the event of
Kamsky's plan ned exchange of dark-square
bishops the wh ite q ueen will exert un pleas­
ant pressu re along the c1 -h6 diagona l .
21 . . . tt::l h 7
The knight on h7 proves to be out of play.
2 1 . tt::l h 5 is also du bious: 22 .i.xg7 �xg7 23
. .
t2Je3 followed by tLlg4 (but not Ludek
Pachman's recommendation 23 g4? in view
of 23 . . . 'ikf6 ! ) . It looks better to play 2 1 . . . tLld7
ttJ 73
22 .i.xg7 �xg7 23 axb5 axb5 24 J:.a7 l:tb8
25 tLle3 h5, although in this case too Wh ite
holds the i n itiative.
22 .i.xg7 �xg7
23 tLle3
This knight causes Black serious anxiety.
The th reat is 24 tLlg4. He is forced to weaken
somewhat his castled position.
23 . . . h5
Anand also analyses other possibilities for
Black:
1 ) 23 .. .'it'f6 24 tLlg4! 'ii'x a 1 25 'ii'x h6+ �g8 26
e5! .l:.xe5 (26 .. .'ii' c3 is weaker: 27 .i.xg6 fxg6
28 lLlf6+ tLlxf6 29 'ii'x g6+ �f8 30 1i'xf6+ �g8
31 .l:te4 ! ) 27 tLlgxe5! dxe5 28 .i.xg6 1i'xe 1 +
29 ttJxe 1 fxg6 30 'iix g6+ with advantage;
2 ) 23 . . . tLlg5 24 ttJxg5 hxg5 (or 24 .. .'ii' xg5 25
axb5 axb5 26 f4 ! 'with strong prospects on
both wings' - Anand) 25 axb5 axb5 26 l:r.a5!
'ilkc7 27 lla7 .l:ta8 (if 27 . . . 'ii' b 6, then 28 tLlf5+
gxf5 29 1i'xg5+ 'it>f8 30 'ikh6+ 'it>e7 31 e5!) 28
tLlf5+! gxf5 (28 . . . 'iitf6 29 .l:.xb7 'i!Vxb7 30
ttJxd6) 29 'ii'x g5+ 'iitf8 30 'ili'h6+ 'it>e7 31 e5!
with a very strong attack.
24 'ii'd 2
One of many su btle positional moves in this
game. Wh ite method ically strengthens his
position, exploiti ng the fact that the opponent
has no active counterplay. Now Black has to
reckon with a possible tLld4 .
24 . . . 'it>g8
24 . . . tLlf6 is dangerous in view of 25 lLlh4 ,
while if 24 . . . 'ii'f6 Wh ite has 25 l:ta3. Possibly,
if Black had foreseen the following manoeu­
vre by his opponent, he would have chosen
the lesser evil - 24 . . . bxa4 ! ? , although i n this
case the wh ite knight would have gai ned an
excellent post at c4 .
25 axb5!
Wh ite has waited for a long time before
opening the a-file, but now, when the
opponent has focused his attention on the
kingside, Anand suddenly switches to the
queenside.
74 � Manoeuvring

25 . . . axb5 arise n , one that is typical of certai n variations

of the Benoni Defence.
29 . . . l2Jxb5
30 ..txb5 .l:ted8
31 ..tc4
After this seemingly modest move there is
the strong threat of a breakth rough in the
centre.
31 . . .
I n the event of 3 1 . . .f6 the line recommended
by Pachman is i nteresting: 32 e5 fxe5 33
l::tx e5 'Wf6 (33 . . . dxe5 34 d6+ 'it>g7 35 dxe7
Jb:d2 36 l2Jxd2 ..tc6 37 .l:r.a7 .tea 38 l2Je4) 34
l:.ee 1 l:.e8 35 l:ta7.

26 lDd1 ! !
An excellent and timely manoeuvre, consoli­
dating Wh ite's advantage. He prevents the
defence-relieving move . . . l2Jg5, and the
opponent now has to reckon with the central
breakthrough e4-e5. But, above a l l , it tran­
spires that in the enemy position there is a
serious weakness - the b5-pawn . The same
manoeuvre, but without the preliminary
exchange on b5, would have been weaker in
view of 25 . . . bxa4 ( cf. the note to Black's 24th
move).
The following stage of the game can serve as
a textbook illustration of how to manoeuvre
agai nst enemy weaknesses. 32 'ii' h 6!
26 . . . l2Ja6 Wh ite makes use of all the space on the
Black is forced to defend passively. If board and all the resou rces of the position to
26 . . . .Ua8 Wh ite has the u npleasant 27 l:.xa8 d isrupt the coordination of the opponent's
.:.Xa8 28 l2Jc3 'ii'd 7 29 e5. forces. Now the threat is 33 e5 dxe5 34 d6
27 l2Jc3 b4 and 35 'ii'x g6+ .
28 lDb5 32 . . . 'ii'f8
28 l2Ja4 with the idea of lDb6-c4 also looks The only move. The su icidal 32 . . . l2Jxe4? is
good . meet by the simple 33 .Ua2 with the
irresistible threat of 34 l:tae2 .
28 . . . l2Jc7
33 'ii'g 5
29 .i.d3
33 1i'f4 is also good .
Even the exchange of knights does not bring
Black any rel ief. The light-square bishop 33 . . . 'ii'g 7
moves with gain of tempo to an active If 33 . . . l2Jh7 34 1i'f4 , and the black pieces are
position. A structure favourable for White has even more badly placed . 33 . . . 'ife7 is no
 Manoeuvring ltJ 75

better in view of 34 J:la7 ! . but his opponent is vigilant to the end.

34 .l::t a 7 44 .l::t b 1
For nine moves Wh ite has patiently been Anand chooses the safest way, althoug h he
awaiting the most conven ient moment to could also have won with the direct 44 tt::lg 5
invade with his rook. After the white queen's b2 45 J:lb1 .l:l.a7 46 �h2! (but not 46 %:txb2
diversionary raid on the kingside, the tu rn 'it'xb2 47 'i!fd8+ tt::l e 8! 48 'ii'x e8+ 'it>g7)
has come for active play on the q ueenside. 46 . . . Iia 1 47 l::r.x b2 'ii'x b2 48 'iVd8+ �g7 49
34 . . . l:tc7 'ii'f6+ �h6 50 tLlxf7+ �h7 5 1 'ii' h 8 mate.
34 . . . l:!. d7 is strongly met by 35 e5! tt::l e 8 44 . . . b2
(35 . . . tt:Jxd5 36 e6 fxe6 37 lixe6 , or 35 . . . dxe5 45 'iVc5 ! l1b3
36 tt:\xe 5 .l:.dc7 37 d6 and wins) 36 e6 l:[dc7 46 'ii'd 4 'ii' b4
37 exf7 + 'ii'xf7 38 .l:!.e6 (Anand).
47 tLlg5
35 ..ta6 .Ub8
The weakness of the f7-sq uare qu ickly
35 . . .l:i.dd7 also loses to Anand's suggestion
.
decides the outcome.
36 .bb7 .l:!.xb7 37 .Ua8+! �h7 38 'ilkf4 tLlg8
47 . . . .U.c3
39 e5 dxe5 40 J:lxe5 with complete domina­
48 'ii'f4
tion.
The final stroke, showi ng how defence and
36 e5!
attack should be combined .
Now, when the opponent's forces are tied
48 . . . f6
down both on the queenside, and on the
kingside , the breakthrough i n the centre 49 exf6 tLld5
finally destroys his defences. 50 f7+
36 . . . tt::l e 8 Black resigned .
36 . . .dxe5 is bad because of 37 d6 .l:!.d7 38
tiJxe5. Here is a nother example from the same
37 J:lxb7 J:.cxb7 match , splendidly illustrating the importance
38 ..ltxb7 .l:.xb7 of a 'second front' .
39 'ii' d 8 'ii'f8
40 lla1 ! Anand - Kamsky
The cleanest solution , securing Wh ite a 3rd Match Game, Las Palmas 1 995
decisive gain of material. The plausible 40 e6
fxe6 41 :Xe6 is less clear in view of
41 . . .tt'lg7, when Black acq u i res some saving
chances.
40 . . . tt::l c 7
40 . . . 'i!i'e7 1oses to 41 J:la8.
41 'ii'd 7 'ii b 8
41 . . . 'i'e8 is also hopeless because of 42
l'c6 ! .
4 2 'ii'x d6 c4! ?
43 bxc4 b3
Kamsky finds the best saving opportun ity,
76 <;t>
Black's castled position is weakened , and
Anand immediately exploits this factor. He
begins a crafty knight manoeuvre, the aim of
which is to establish control over f5 .
20 l2J3h2!
Threatening 2 1 'ii' h 5 followed by l2Jg4 .
20 . . . 'iff&
21 l2Jg4 'it'g7
22 l2Jge3 l2Jxe3
23 l2Jxe3 l:.f4
Although White has made prog ress in
carrying out his plans, as yet he has not
managed to gain complete control of the f5-
square. Little is promised by 24 l2Jf5 'it'f6 , 24
g3 Iif7 (24 . . . .l:.f6!?) or 24 i.f5 l:.f8 . In tu rn ,
Black wants to set up pressure on the f-fi le.
White wou ld very much like to exchange a
pair of rooks and at the same time bri ng his
queen's rook into play. A 'second front' is
needed !
24 a4! !
This move secu res a clear advantage. Since
Black could not be satisfied with either
24 . . . b4 25 g3, or the more natu ral 24 . . . bxa4
25 llxa4 .l:!.af8 26 .uxf4 Iixf4 (no better is
26 . . . exf4? 27 l2Jf5 'ii'f6 28 .ll e 6 or 26 . . . gxf4 27
l2Jf5 1i'f6 28 'ii' h 5) 27 i.f5! (variations by
Anand), he decides on a pawn sacrifice, but
he does not gain sufficient compensation .
24 . . . .l:!.af8
Manoeuvring
25 axb5 aS
25 . . . axb5 is completely bad i n view of 26
l:!.a7.
26 l:lf1 i.c8
27 g3 .l:l.4f7
It is curious that in the event of 27 . . . l2Jf6 28
gxf4 gxf4+ 29 �h2 fxe3 30 fxe3! (30 l:tg 1 ?
l2Jg4+ ! ) 3 0 . . . 'i!i'g5 Black's counterplay is
suppressed by the inclusion of the rook in the
defence along the 4th rank: 3 1 l:la4! (Anand).
28 b4
This emphasises Wh ite's advantage, al­
though it is possible that Anand's suggestion
of 28 i.e4! l2Jc5 29 i.g2 e4 30 l2Jc4 would
have been more accu rate. S ubseq uently,
despite Kamsky's resou rceful play, White
converted his adva ntage into a wi n .
2 8 . . . e4! 2 9 i.xe4 l2Je5 3 0 i.g2 axb4 3 1
cxb4 l2Jf3+ 3 2 i.xf3 .l:.xf3 33 lla8 i.xh3 34
'iWxf3 l:.xa8 35 l:!.c1 ltf8 36 'i¥e2 i.d7 37 .l:!.c7
I1f7 38 .l:!.b7 'ii'a 1 + 39 l2Jf1 'i.t>g7 40 .l:!.xb6
'it'd4 41 .l:tb8 'ii'x b4 42 l2Je3 h5 43 b6! h4 44
g4 i.b5 45 'ii'd 1 'ii' b 2 46 l2Jf5+ l:.xf5 47 gxf5
i.e2 48 'ilfa4 (48 'ii'e 1 ! ) 48 . . . i.f3 ! 49 'i!Vd7+
(49 f6+! ) 49 . . . 'i.t>h6 50 'ili'e6+ �h5 51 'ii'e 8+!
'it>g4 52 1i'e1 i.xd5 53 .l:!.e8 i.f3 54 f6 ! <ithS
55 f7 'ii'd 4 56 .l:!.e4! 'ili'f6 57 b7 i.xe4 58
ii'xe4 Black resigned .
Anand's play makes a g reat impression and
provokes strong associations with the games
of Alexander Alekh ine - an unsu rpassed
master of changing the direction of attack.
Typical of Alekh ine was his splendid vision of
the enti re chess boa rd and his abil ity to fi nd
latent tactical resources, supporti ng his
strategic ideas. The followi ng th ree classic
Alekhine examples illustrate best of all his
style of play.
 Manoeuvring l2J 77

Alekhine - Bogolj u bow

Tri berg 1 92 1

After Wh ite's operation in the centre has

d iverted the opponent's forces away from the
kingside, there follows a swift and sudden
Regaining the pawn leads merely to simpl ifi­ attack.
cation after 1 2 'i!Vxc4 tt:Ja5. Alekh ine finds a
1 6 i.e4 ! !
far from obvious way of developing his
initiative, based on the strength of his two As shown b y Alekh ine, Black i s not saved by
bishops. either 1 6 . . . h6 1 7 i.xh6 f5 1 8 'li'g5 'i'c7 1 9
.i.xg7 'i'xg7 20 "it'xg7+ 'it>xg7 2 1 �d7+
1 2 .l:i.d 1 ! 'li'c8
followed by i.xb7, or 1 6 . . . g6 1 7 �f6 tt:Jxd5
If 1 2 . . . 'i'e7 Alekh ine g ives 1 3 i.g5 h6 1 4 1 8 .i.xd 5 . The move i n the game leads to
ixf6 'i'xf6 1 5 "it'xf6 gxf6 1 6 �d7 with loss of material and a q u ick finish.
advantage. (Here and later Alekhine's analy­
16 . . . f5
ses are taken from his book My Best Games
of Chess 1 908-23. ) And in the event of 1 2 . . . 1 7 .i.xf5 �xf5
"i'e8 Dvoretsky's suggestion i s very strong: 1 8 .l:i.d8+ �xd8
13 .ih6!! gxh6 14 'it'xf6 followed by tt:Je4 . 1 9 i.xd8 .l:!.c8
1 3 i.g5! lLJd5 20 .l:i.d1 .l:!.f7
14 tt:Jxd5 exd5 21 �g4 lLJd3
1 5 l:lxd5 ! 22 exd3 .l:!.xd8
A n accu rately calculated operation - the 23 dxc4 .l:i.df8
rook is switched to the kingside. In this 24 f4 .l:r.e7
chapter we will see fu rther examples of a 25 'it>f2 h6
rook being included in an offensive along a
26 J:te1 �c8
rank. It is often qu ite d ifficult to activate a
rook because of its own pawn chains, but if it 27 �f3 .l:i.ef7
does manage to take part in an attack, it 28 'ilt'd5 g5
normally causes the opponent a mass of 29 .l:!.e7 gxf4
problems. 30 gxf4
15 . . . l2Jb4 Black resig ned .
1 5 . . 'i'e6! was much more tenacious (Dvoret­
.

sky).
78 <;i( Manoeuvring
Alekh ine - Sterk
Budapest 1 92 1
The critical position of the game. Black has
developed strong pressure in the centre, but
Alekhine fi nds a way of retaining approxi­
mate equal ity.
1 6 Ji.d3
1 6 e5 was weaker in view of 1 6 . . . lt:Jg4 , as
was 16 .l:!.ac1 ..txc3 1 7 i.d3 lt:Jc5 1 8 l:txc3
i.xe4 ! 1 9 Ji.xf6 Ji.xd3 (Aiekhine ). 1 6 1Ifc1 !
i.xc3 1 7 Ji.d3 would have transposed .
16 . . . i.xc3
1 7 l:!.fc1 lt:Jxe4?
Correct, as suggested by Alekhine, was
1 7 . . . lt:Jc5 1 8 .l::t xc3 i.xe4 1 9 i.xf6 Ji.xd3 20
'ii'e 3! (now it is clear why it was the king's
rook that had to go to c1 ) 20 . . . gxf6 21 b4
Ji.g6 22 bxc5 bxc5 23 .l:!.xc5 'ii'd 7 24 h4 with
good compensation for the sacrificed pawn .
With the move in the game Black also wins a
pawn , but he comes under an un pleasant pin
on the c-file.
18 i.xe4 �xe4
1 9 'ii'xe4 tt:Jcs
20 'i!Ve2
Alekhine lures the black pieces onto poor
squares on the queenside. However, he may
have been wrong to avoid 20 'i¥b 1 ! i.b4 2 1
a3, since i f 2 1 . . . 'i!i'b7 Wh ite does not play 22
axb4 lt:Jb3, and not 22 'ii'c 2 'ii'd 5 ! , but wins a
piece with the simple 22 'ii'a 2! lt:Jd3 (22 . . . 'i!i'd5
23 axb4 lt:Jb3 24 l:.d 1 ) 23 l:.d 1 .
20 . . . i.a5
21 .l:tab1 'ii'a 6
22 .l:!.c4 lt:Ja4
If 22 . . . l:tac8 there could have followed 23 b4
lt:Ja4 24 b5 or 23 . . . lt:Jd7 24 .l:r.e4 and wins.
Black parries this th reat (if 23 b4 he has
23 . . . lt:Jc3), but now, when the opponent's
pieces are tied up on the q ueenside,
Alekh ine unexpectedly switches the play to
the kingside.
23 i.f6! !
Creating the highly u n pleasant th reat of 24
.l:tg4 . Thus, after 23 . . . h 5 24 .l::tg 4! 'i!i'xe2 25
l:xg7+ 'it;lh8 26 lt:Jg5 there is no defence
against 27 l::t h 7+ and 28 1Ih8 mate (Aie­
khine ). 23 . . . h6 also fails to save Black in view
of 24 lt:Je5 with the threat of 25 'i!Vg4 .
23 . . . .l:r.fc8!
24 'i!i'e5!
The point of Wh ite's idea. Alekh i ne's analy­
sis shows that he wins in all variations:
1 ) 24 . . . 'i!Vxc4 2 5 'i¥g5 <it>f8 26 'i!i'xg7+ 'it;le8 27
'ii'g 8+ 'it;ld7 28 lt:Je5+ <it>c7 29 1i'xf7+ and 30
lt:Jxc4;
2 ) 24 . . . .l::txc4 2 5 'ii'g 5 .l:.g4 26 1i'xg4 g6 27
'i!i'xa4 ;
 Manoeuvring ttJ 79

3) 24 . . . gxf6 25 l:i.g4+ and mate in two moves. 22 . . . l::i. d 7

24 . . . .l:!.c5 23 g 3 ! 'i!t'b8
In the hope of 25 l::i. x c5 gxf6 , but Wh ite In the event of 23 . . . 'i!Vd6 i n Aiekh ine's opinion
decides the game more simply. Wh ite had two win n i n g lines:
25 'ii' g 3 ! g6 1 ) 24 l:i.fd 1 'ii"x d 1 + (24 .. ."it'xc6 25 �xc6
26 l"::i.x a4 'iWd3 �xd 1 + 26 l:i.xd 1 l::i.e ? 27 �d8 } 25 l:i.xd 1
27 l"::i.f1 'iWf5 l:i.xd 1 + 26 'it>g2 �d? 27 'ii"x b6 �xa4 28 'ii"x a6
�d? 29 tt'lg5 Wg8 30 'iWe2 ;
28 'iWf4 'it'c2
2 ) 2 4 'it'c4 'it>g8 (24 . . . 'iWe7 25 tt'le5 l"::i.d 6 26
29 'iWh6
�g6 ! ) 25 .1i.c6 l"::i. c 7 26 l"::i.fd 1 'ii"e ? 27 �d3 ! .
Black resigned .
I n fact the second way is unconvincing in
view of 26 . . . l::i.xc6! (instead of 26 . . . ii'e7?) 27
Alekh ine - Rubi nstein �xc6 'ii"xc6 28 l:txc6 ii.b? 29 �c3 i£.xb4
Carlsbad 1 923 (John N u n n ).
24 tt'lg5!
Wh ite does not al low his opponent a mo­
ment's respite . The th reat is 25 tt'lxf?+.
24 . . . l::i. e d8

There are two vul nerable poi nts in Black's

defensive lines: 1 ) with the loss of the h 7-
pawn his castled position is weakened ; 2 )
there i s a lack o f harmony in t h e placing of
his queenside pieces, and they may come
25 i£.g6! !
under attack. With energetic action against
With gain of tempo Wh ite clears the way for
the queenside Alekh ine forces the opponent
his q ueen to the kingside , and the black
onto the defensive there, after which he
pieces , tied down to the defence of the
unexpectedly switches to an attack on the
opposite wing , simply do not have time to
king.
come to the aid of their king . Thus if 25 . . . fxg6
21 b4! �f8
(25 . . . �b7 26 'ii"c4 ) there follows 26 'ii"e4
The only reply, since other bishop moves �xb4 27 'ii" h 4+ 'it>g8 28 'ii" h 7+ 'it>f8 29 'ii" h 8+
would have allowed the decisive 22 'iWxc8 . 'it>e? 30 'iWxg?+ 'it>e8 31 'ii"g 8+ ii.f8 32
22 'iWc6 ii'xg6+ We? 33 'ii'x e6 mate (Aiekh ine).
In view of the double attack on the rook at e8 To avoid being mated i m mediately, Black
and the b6-pawn , Black's reply is forced . g ives up material.
80 � Manoeuvring

25 . . . �e5 6 . . . tt:lc6 7 c3 'i!Vh5 came i nto consideration,

26 tt:lxf7+ �xf7 as in the game Log inov-Karpeshov (Volgo­
27 .txf7 'i*'f5 donsk 1 983).
28 l:!.fd 1 .l:i.xd 1 + 7 "t!Vxd2 'ii'x d2+
29 .Uxd 1 "ifxf7 8 'iit> x d2
30 'it'xc8 Wh7 An endgame has a risen i n which Wh ite has a
purely symbolic advantage. At the given
31 "it'xa6 'i!Vf3
moment he is slightly ahead in development,
32 "t!Vd3+ but there are no weaknesses in Black's
Black resigned . position . Wh ite's first objective is to trans­
form his slight i n itiative i nto someth ing more
I n the following game Wh ite was able to win concrete, by creati ng a target in the oppo­
a roughly equal end i n g , by consistently nent's position .
employi ng the principle of two weaknesses. 8 . . . .tf5
After creating targets to attack on opposite 9 e3
wings, White began manoeuvri n g . In this If 9 .tg3!? Black does not reply 9 . . . tt:ld7?
example the process of creati ng weakness­ because of 1 0 tt:lh4, but simply 9 . . . c6 10 e3
es in the opponent's position was perhaps e6 or 1 O . . . tt:ld7 with a solid position .
more interesting that the manoeuvring against
9 . . . e6
them.
1 0 .te2 .td6
If 1 0 . . . .te7 1 1 ii.xe7 'iit> x e7 1 2 tt:lh4 and
Yusu pov - Wirthensohn
Wh ite exchanges knight for bishop (an
Hamburg 1 99 1 achievement, if only a slight one). 1 O . h6 . .

Torre Attack came i nto consideration . In this case I would

1 d4 tt:Jf6 have continued 1 1 �hc1 followed by c2-c4.
2 tt:lf3 d5 1 1 .tg3 h6
3 .tg5 tt:Je4 The threat was 1 2 tt:lh4 with the exchange of
A good move , casti ng dou bts on the early the bishop. But what should Wh ite do now?
bishop sortie. Natural play, involving c2-c4 , promises little ,
4 .th4 "t!Vd6 since after the exchange on c4 the black
bishop will be splendidly placed to defend
Another tempting plan involves the standard
the q ueenside from e4 . After a long think I
advance . . . c7-c5 followed by the develop­
found a way of maintaining my in itiative .
ment of the queen on b6.
1 2 .txd6 cxd6
5 tt:lbd2 "ifh6
If I am not mistaken this manoeuvre was first (see diagram)
employed by Vlastimil Hort. Black has
created the un pleasant threat of . . . g7-g 5 . 1 3 a4!
With the queen on h6 the retreat o f the In this way Wh ite becomes active on the
bishop to g3 looks extremely unappealing, q ueenside: his rook obtains the promising
since after the captu re on g3 White is forced route a 1 -a3-b3 . For the moment Wh ite is
to spoil his pawn structure and recapture accu mu lating barely perceptible pluses : he
with the f-pawn . has the more flexible pawn structu re and the
6 "t!Vc1 ! ? tt:Jxd2 potentially better bishop.
 Manoeuvring lD a1

20 llca 1 !
20 axb6 was weaker in view of 20 . . l:.xb6 2 1
.

b 3 li:Jb8 ! , and White has merely a m i n i mal

advantage.
20 . . . b5
21 a6
Thus, Wh ite has finally succeeded in creat­
ing a first real target - the b-pawn .
21 . . . b4
22 l:ta5 .l:tc6
23 f3

- position after 1 2 . . . cxd6 -

13 . . . r3ile7
14 aS li:Jd7
If 14 . . .:!.c8! ? 1 5 l:thc1 lt:Jc6 1 6 l:ta4 b6 1 7
.

axb6 axb6 1 8 .:!.ca 1 with a minimal i n itiative

for White.
15 .l:.hc1 l:thc8?!
1 5. . .ttlf6 was more logica l , forcing the retreat
of th e king to e 1 .
1 6 lt:Je1 !
Active prophylaxis! The knight had no pros­
pects on f3 , and therefore it aims for b4 . I n
addition, now i f 1 6 . . . lt:Jf6 there will b e the It is not normally possible to win a game by
good reply 1 7 f3 . attacking one single weakness. Therefore
Wh ite's next objective is to i n itiate play on
16 . . . .l:.c7
the kingside, with the aim of exploiting the
17 li:Jd3 .i.xd3 power of his long-range bishop and the
This exchange is practically forced - the mobility of his rooks.
knight would have been too u npleasantly 23 . . . .l:tcb6
placed at b4 . But now, although for the
24 h4 li:Jf6
moment the position is a closed one, the
white bishop is stronger than the opponent's 25 g4 l:th8
knight, since it will be able to take an active Black real ises the danger and takes precau­
part in the play on both wings. tionary measures on the kingside. Therefore
18 ..txd3 .l:1b8?! I exchange a pair of rooks, after which the
weakness of the b-pawn will be more keenly
19 l:ta3 b6
felt.
This attempt to play actively, which was
planned on the previous move, in fact leads 26 .l:tb5 .l::!. h b8
merely to the creation of weaknesses on the 27 .l:txb6 .l:txb6
queens ide. 28 g5
82 � Manoeuvri ng

Acting in accordance with the 'two weak­ [Black incorrectly restricts himself to passive
nesses' principle, White changes the d i rec­ defence. He should have tried 34 . . . e5!,
tion of the attack, shifting the emphasis to the intending the manoeuvre . ltJf�6 - Dvoret­
. .

kingside. sky.]
28 . . . hxg5 34 'it>b4 'it>c7
29 hxg5 ltJd7 35 'it>a5 ltJb6
29 . . . ltJe8 30 Ith 1 ltJc7 is bad because of 3 1 If 35 . . . ltJb8 , then 36 .i.b5 ltJc6+ 37 i.xc6
.l:!.h7! 'it>f8 3 2 .l:.h8+ 'it>e7 3 3 'it>c1 ! , a n d if 'it>xc6 38 b4 is strong , with a won rook
33 . . . ltJxa6 Wh ite has 34 l:r.a8 . ending.
30 .:th1 l:ib8 36 .l:lh2
31 .l:th7 .l:lg8 After the immediate 36 b3 I did not l i ke the
reply 36 . . . ltJc8 with the idea of . . . ltJe7-c6 .
Better defensive chances were offered by
31 . . . 'it>f8 . 36 . . . l:tcB
37 b3 'it>d7
38 lth7 .l:lg8
39 f4
A useful move, since now Black has to
reckon with a possible f4-f5 .
39 . . . 'it>e7
40 .i.b5
The tempti ng 40 'it>b5 does not promise any
immediate gains, in view of the u nexpected
resou rce 40 . . . ltJa8 ! , and if 41 'it>c6??, then
after 4 1 . . . .l:tb8 Wh ite is mated.
40 . . . ffi
Fearing the manoeuvre of the bishop to b7,
Black tries to i n itiative cou nterplay on the
32 c31 kingside. But this attempt is parried by White,
Now, when the black rook has taken u p a who reverts to his i n itial plan - the advance
passive position on the kingside, Wh ite of his king i nto the opponent's position via
again changes the d i rection of the play and b5.
activates his king . I n the process the weak b­ 41 .i.e2 fxg5
pawn is exchanged , but this is not so 42 fxg5 ltJd7?!
important, since Black remains with a real
It is possible that the best defence was
weakness on the queen side - he has to take
42 . . . ltJa8 ! ? , trying to prevent the wh ite king's
measures against the breakthrough of the
manoeuvre.
king to the a7-pawn . 32 'it>c1 (with the idea of
playing the king to b3) was less accu rate , 43 'it>b5
since Black would have succeeded in gain­ After the activation of the king Black's
ing counterplay after 32 . . . ltJb6 33 'it>b1 defences begi n to creak.
ltJc4 . 43 . . . ltJf8
32 . . . bxc3+ 44 .l:lh2 ltJd7
33 'it>xc3 'it> dB 45 'it>c6 .l::r. c 8+
 Manoeuvring t2J 83

46 'it>b7 .l:r.b8+ with d rawing chances after both 5 1 a7?!

47 'it>xa7 l:txb3 :Xd 1 52 aS 'if l:tb 1 + 53 'it>c6 l:tc 1 + , and 51
l:tg8 l:txd 1 52 .l:r.xg7+ 'it>e8 53 a7 l:tb 1 + (or
53 . . . l:ta 1 54 a8'iV l:txa8 55 'it>xa8 e5) 54 'it>c6
.l:.a 1 . [However; 51 .i.a4! is very strong -

Dvoretsky.]
I n stead of 50 'it>b7 the more cun n i ng move
50 'it>b6! is possible, and if 50 . . . l:ta3 51 'it>b5
followed by 52 .i.a4.
50 �c2 .l:.c4
52 �d3 .l:.b4
After gaming an important tempo, Wh ite
again changes the direction of the offensive,
strengthening his position on the kingside.
52 g6 e5
If 52 .. J1b3 there would have followed 53
Black appears to have gained cou nterplay, �c2 lhe3 54 'it>b7 .l:te2 55 l:txf8 ! and wins.
but White has a nother attacking resou rce, The move i n the game also fails to ease
which he had to foresee i n advance . Black's position.
48 .l:th8! 53 .i.f5! exd4
With my small army I create th reats to the 54 exd4 l:txd4
opponent's king! It transpires that 48 .. J::txe3
55 'it>b6 l:ta4
is bad because of 49 .i.h5 ltJf8 50 l:tg8 g6 5 1
Or 55 . . . l:tb4+ 56 'it>a5 .l:tb2 57 l:th4 and the a­
l:.g7+ 'it>d8 5 2 �xg6. Black i s torn i n two: he
pawn decides matters.
has to keep watch over the dangerous
passed a-pawn and at the same time repel 56 l:tg8!
the invasion of the white pieces on the The simplest and most thematic solutio n ,
kingsi de . illustrating t h e strength o f p l a y against two
48 . . . ltJf8 weaknesses.
49 .i.d1 l:tb4 56 . . . ltJe6
49 J:txe3 was more tenacious. If 50 'it>b7
. .
57 a7
Black has an u nexpected defence: 50 . . . .U.e 1 ! Black resigned .
 Alexey Kosikov

How to d raw u p a P l a n

W opponent is able to parry them without

hen we create simple threats , the
particular difficulty. The secret of success
often lies in the abil ity to create several
th reats simultaneously. I will illustrate this
idea with a study by Richard Reti which will
be well known to all of you .
ing an advantage, you will have met the
'principle of two weaknesses'. I think that this
principle is also usefu l for the compiling of a
strategic plan in the middlegame. I n the
broad sense of the word , a weakness is a
long-term strategic threat. That is, not only a
vulnerable enemy pawn , but also the threat
of promoti ng you r own pawn , the th reat of
i nvad ing along an open file, a mating net,
and so o n .

Botv i n n i k - Zagoryansky
Sverd lovsk 1 943
Reti Opening
1 l"Llf3 d5
2 c4 e6
3 b3 l"Llf6
4 .1i.b2 .1i.e7
5 e3 0-0
6 t"Llc3 c5
After 1 'it>g7 h4 2 'it>f6 'it>b6 (2 . . . h3 3 'it>e7 ) 3 7 cxd5 t"Llxd5
'it>e5! Wh ite saves the game only because 8 t"Llxd5 exd5
he simultaneously has two threats: to stop
don't wish to dwell on the opening
the enemy pawn by 4 'it>f4 and to support his
subtleties. I will merely remark that it would
own pawn by 4 'it>d6. Black can easily parry
have been more promising for Black to
either of these threats, but not both of them .
capture on d5 with his quee n , and if with the
The idea of simultaneously creating two pawn, then a move earlier.
threats will serve as our starting point for
9 d4 cxd4
understanding the process of compiling a
plan. 1 0 ii'xd4 .1i.f6
1 1 ii'd2 l"Llc6
The plans which we make pursue the aim of 1 2 .1i.e2 .1i.e6
creating th reats , but not simple, tactical Black plays too passively. I would have
threats, as in the example we have just preferred to develop the bishop at g4.
analysed , but long-term strategic threats. 13 0-0 .1i.xb2
In trying to master the technique of convert- 1 4 ii'xb2 ifas
 How to d raw up a Plan ctJ 85

22 e4 was th reatened .
22 �b2 .l::. c 8
23 'i!i'e5 l:tcd8
By repeati ng moves Wh ite has gai ned time
on the clock.
24 l:!.d4 aS

Up to here it would have been premature to

talk about drawing u p a pla n , but now the
position is ripe for th is. Of cou rse, to foresee
to the end everything that will happen is not
possible, but nor is it necessary. All you need
to do is correctly imagine to you rself the
character of the forthcoming play and decide
what should be done now.
White obviously stands better. He has a The first part of the plan - the attack on the
target to attack - the weak d5-pawn, the enemy weakness - has been completed . All
black bishop is passive, and the opponent the wh ite pieces are ideally placed , whereas
has no cou nterplay. the opponent's pieces are passive and
restricted i n their movements . But even so
In the first instance Mikhail Botvinnik con­
Wh ite has not managed to win the d5-pawn ,
centrates his forces against the enemy
which is not surprising - an attack on only
weakness.
one weakness rarely proves successfu l . A
1 5 l:tfd 1 l:tfd8 new target is needed , and Botvi n n i k finds it
1 6 .t1d2 .l:i.d7 on the kingside.
1 7 .t1ad 1 .l:i.ad8 25 g4!
18 h3 h6 'At the cost of weakening the position of his
19 tDe5! own king (which is possible only thanks to
The black knight was controlling the blockad­ the fact that the enemy pieces are tied
ing d4-square and restricting the white d own ), Wh ite opens up the play on the
queen, so Botvin n i k happily excha nges it. At kingside. It soon transpires that Black is
the same time the f3-square is vacated for unable simu lta neously to defend his king's
the bishop, which will join the attack on the fortress and the ill-starred d5-pawn . ' -
d5-pawn . Botvinnik.
19 . . . tDxe5 25 . . . �c6
1 9 . l:td6 20 tbc4 ! .
. . 26 g5 hxg5
20 �xe5 iic5 27 'i!Vxg5 f6
21 �f3 b6 I n principle, in such cases it is better to keep
86 � How to draw up a Plan

the pawns on their initial squares. But in view 37 l:Ig5 .l:.f7

of the threatened attack on the h-file (.l:.h4 38 'i¥h5 'it'a1 +
and 'i¥h5) all the same Black can not avoid 39 'iit g 2 g6
pawn moves.
40 'i¥xg6 .fl.. h 7
28 'ii'g 6 .fl..f1
41 'ii'd 6+ �be7
29 'ii'g 3 f5
42 'ii'd 8+
'An impulsive move, assisti ng the develop­
Here the game was adjou rned . Black re­
ment of Wh ite's in itiative , although the
sig ned without resu ming.
inevitable transference of the rook from d 1 to
g1 would in any case have set Black Let us remind ourselves how events devel­
insoluble problems . ' - Botvinnik. oped . After creating a weak d5-pawn in the
enemy position , White then assailed it with
Evgeny Zagoryansky essentially sig ned his
all his forces. However, his aim was not to
positional capitulation . Now all the dark
win this pawn (the opponent is always able to
squares in his position are hopelessly weak.
defend one weakness), but merely to tie the
30 'ii'g 5 'ii'e 6 enemy pieces to its defence. Then Wh ite
31 'iit h 1 'i¥e5 located a second target and successfully
32 l:tg1 .l:.f8 attacked it. The principle of two weaknesses
33 'it'h6 llb8 i n action!
Both 34 .l:h4 and 34 'it'xb6 were th reatened .
If 33 . . . .l:td6 34 .l:xg7+, while after 33 . . . g6 Of course, i n such positions Botvi n n i k was
Wh ite decides matters with h3-h4-h5. not the first to employ plans based on the
34 l':th4 'iii f8 principle of two weaknesses. Before h i m too,
35 'i¥h8+ i.g8 lead ing players consciously or sub-con­
sciously made use of the same weapo n . We
will now examine another classic game, and
follow how the g reat strategist Akiba Rubin­
stein provokes the creation of weaknesses in
the opponent's position and then exploits
them.

Rubi nstein - Takacs

Budapest 1 926
Queen 's Gambit
1 c4 lLlf6
2 d4 e6
3 tLlc3 d5
4 .fl.. g 5 lLl bd7
36 .l:If4
5 e3 .fl.. e 7
As the attack has developed , new weak­
nesses have appeared in Black's position . 6 lLlf3 0-0
Wh ite now turns his attention to one of these 7 .U.c1 c6
weaknesses - the f5-pawn . 8 'i¥c2 a6
36 . . . .l:.bb7 9 cxd5 exd5
 How to d raw up a Plan l2J 87

1 0 .ltd3 lle8 1 8 'ii' b 3! 'ii' b 6

11 0-0 lll f8 Black really doesn't want to play 1 8 . . . l:r.a7.
1 2 l:!.fe1 .ltg4? ! 19 tt::la 4 'ifxb3
1 3 tt::l d 2 tt::l 6 d7?! Maxim Boguslavsky suggested a good de­
Black's desire to simplify the position is q u ite fence: 1 9 . . . 'i¥b5 ! ? . Wh ite would probably
understandable, but the move in the game have replied 20 tt::l c 5, but then an exchange
does not achieve its a i m . of minor pieces eases Black's defence:
20 . . . tt::lx c5 21 lbc5 'ii'x b3 22 tt::l x b3 tt::le 6 and
1 4 .ltf4 .ltg5
23 . . . a5 .
1 5 h3! .lth5
20 tt::l x b3 tt::l e 6?
The bishop is forced to retreat, since after
Now Black's position becomes difficult. What
1 5 . . . � xf4? 1 6 exf4 it is trapped . Sandor
does White want to play next move? That's
Takacs apparently overlooked this tactical
right, 21 llla 5 ! . This defi n itely should have
subtlety. As a result Wh ite has retained his
been prevented : 20 . . . a5! 2 1 tt::l b c5 tt::lx c5 22
important dark-square bishop, and the oppo­
tt::lx c5 l:ta7.
nent's last few moves have merely led to a
loss of time and to his pieces being badly 2 1 tt::la 5! l:ta7
placed . 22 'it>f1 1
1 6 .lth2 .ltg6 A good prophylactic move. Wh ite prepares to
tra nsfer his rook to b3 (the immediate 22
17 .ltxg6 hxg6
.Uc3? is not possible because of 22 . . . tt::lx d4 ).
The opponent prevents this manoeuvre.
22 . . . .ltd8
23 b4 f5
24 lll b 2
At d3 the knight will be better placed than at
a4 .
24 . . . g5
25 tt::l d 3 'it>f7
26 .lic2 ..tb6
27 .ltd6! tt::l d 8?!
Wishing to free his rook on a?, Black
weakens his control of c5, and Rubinstein
i mmed iately exploits this.
White stands better and it is now time for h i m
to decide on a p l a n o f further action . With the 28 lllc 5! tt::l x c5
light-sq uare bishops no longer on the board , 29 .txc5 .ltxc5
the routine minority pawn attack is u npromis­ 30 bxc5
ing, since in this case the important c4-point The weakness on b7 has been fixed . Now
is weakened , and the enemy knight will aim (as in the Botvinnik-Zagoryansky game) as
for there ( . . . b7-b5 and . . . tt::l b6-c4). many attacks as possible should be concen­
Let's select a target in the opponent's trated on it, in order to tie the opponent's
position, which it will be possible to attack. Of pieces to its defence.
course, this is the b7-pawn . 30 . . . c:l;e7
88 � How to draw up a Plan

31 l:r.b2 �d7 exchange the terrible rook on a7?

32 .l:teb1 �c8 41 . . . fxe3
33 �e2 l:te7 42 fxe3 �d7
34 �f3 .Ue4 43 .l:tg2 r!e8
44 l:!.xh4 lle7
45 .ll h 8 �c7
46 l:lgg8 l:!.d7
47 lDb3!
The knight has completed its work at a5 and
it is now switched to a more active position -
the e5-square.
47 . . . a5
48 lDc1 .l:ta8
49 lDd3 b5
Against passive defence Wh ite could, for
example, have placed his rook on e8 and his
knight on e5, and then played his king to f6.
White has strengthened his position to the 50 cxb6+ �xb6
maximum. Now, in accordance with the
51 lDc5 .Ud6
principle of two weaknesses, he opens a
52 a4!
'second front' on the kingside.
' Do not hurry!' Black is deprived of any
35 g4! g6
moves at all on the q ueenside, and in
36 l:!.g1 lDf7 addition his king finds itself i n a mati ng net.
37 h4! 52 . . . lieS
The open ing of lines emphasises the unfor­ 53 'it>g4!
tunate position of the rook on a7, which is Black resig ned . The march of the king to e5
taking practically no part in the play. is threatened , and if 53 . . . �c7 there follows
37 . . . gxh4 54 l:tg7+ �b6 (54 . . . �b8 55 .Uhh7) 55 .l:txd8
38 gxf5 gxf5 l:txd8 56 llb7 mate .
If 38 . . . g5, then 39 lDc4! is very strong . A wonderfu l positional game!
39 J:tg7
I will once again remind you i n what order our
Strategically the game is decided . Wh ite has strategy develops in such cases.
created a second weakness in the oppo­ 1 ) Creation of the first weakness. This is
nent's position - the g-file for an i nvasion , perhaps the most difficult thing - to be able
and h e confidently exploits it. somewhere to 'latch on' to the opponent.
39 . . . liJd8 2 ) Attack on the weakness (not with the aim
40 .l:tg8 f4 of destroying it, but in order to tie the
41 .l:th8 opponent's pieces to its defence).
I think that Rubinstein did not even seriously 3) Creation of a weakness on another part of
consider regaining the pawn by 4 1 lDxb7 the board .
:Xb7 42 llxd8+ �xd8 43 llxb7 - why 4) Break through the opponent's defences.
 How to draw up a Plan l2J 89

The games which we have exam ined were the other weakness should be the e6-pawn . '
played rather a long time ago. Of cou rse, (Shirov)
modern players have assi milated the les­ 32 . . . f5?
sons of the past and successfu lly make use This makes things easier for the opponent.
of the same strategy. 32 . . . fxe5 33 dxe5 Wf8 was stronger, but here
too after 34 .l:!.g2 Black's position remains
S h i rov - Kinsman d ifficult. For example: 34 . . . .l:!.e7 35 �d 1 ! ..tg8
Paris 1 990 (35 . . . l:lcd7 36 l:td6) 36 l:lgd2 (note that White
immed iately switches to exploiting the new
weakness which has a risen - the d-file)
36 . . .We8 37 Wf4 followed by Wg5 and f3-f4-
f5 with a n easy wi n . Or 34 . . . ..tg8 35 .l':Ibg 1 ,
intending h4-h5-h6, and the weakness of
the e6-pawn does not allow Black to play his
bishop to f5.
33 l:tg2 g6
34 l:tbg1 l:tc8
35 .llx g6 f4+
36 Wd3 licd8
37 l:tf6 I!.xd4+
38 �c3 lld1
It is Wh ite to move . He has an obvious 39 l:tg7 �c1 +
advantage. How best to exploit it? 40 Wb3 .l:i.b1 +
The first part of the standard plan has 41 '.t>c2
already been completed . In this commentary Black resig ned .
on the game Alexey S h i rov writes:
'One weakness (the b7-pawn) is securely I n positions without cou nterplay for the
fixed . It is also important that the black rooks opponent, such as those we have examined ,
are tied down . . . But back i n my child hood I each of us would feel very comfortable. But
was taught that to win you need at least one i n practice th ings are usually much more
more weakness. And it turns out to be - the complicated and it is not often that the
g7-pawn . ' principle of two weaknesses can be put i nto
3 0 h4 gxh4 practice in such pure form . To d raw up a plan
In the event of 30 . . . .ltf7 31 hxg5 fxg5 Wh ite in sharper situations one also has to be
will at some point play f3-f4 (but, of cou rse, g uided by other pri nciples. One of these
not immed iately 32 f4? gxf4+ 33 gxf4 e5!), pri nciples, wh ich , i ncidentally, is by no
when the pattern of the game remains means well known , will now be described .
roughly the same.
31 gxh4 Ji..f7 Ka l i kshte i n - Vysoc h i n
32 e5! C I S J u n ior Championship, J u rmala 1 992
The point of Wh ite's idea. When he begins Slav Defence
attacking the g7-point, the b7-pawn will no 1 c4 tt:lf6
longer need to be defended ; this means that 2 tt:lc3 c6
90 <;t> How to draw up a Plan

3 d4 d5
4 li'lf3 dxc4
5 e3 b5
6 a4 b4
7 li'lb1
The knight more often retreats to a2, in order
after the capture of the c4-pawn to conti nue
.td2 and li'lc1 -b3. The player with Wh ite is
obviously not aiming for an opening advan­
tage and is intending to transfer the entire
weight of the struggle to the midd legame.
7 . . . .ta6
8 li'lbd2 e6?! - position after 1 7 . . . �b7 -
Theory recommends 8 . . . c3 with equal ity, but
the game continuation is also q u ite possible. pawn on f2 it is always possible to place a
9 li'lxc4 .te7 barrier in the path of the opponent's bishop
1 0 .i.d3 0-0 by f2-f3 , but now this becomes impossible.
1 1 0-0 li'lbd7 The e4-point is weakened .
1 2 b3? ! The c 1 -h6 diagonal is blocked , and the
A superficial move! 1 2 li'lce5 is more logica l , bishop on b2 is now altogether without
a n d after 1 2 . . .'ii'c8 Wh ite's position remains prospects .
slightly preferable. I n positions with an isolated d4-pawn the
12 . . . c5 move f2-f4 is sometimes made, but only
when there is a hope of playing f4-f5. But
1 3 ..ib2 l:tc8
here Black immediately prevents the further
14 .l:tc1 cxd4
advance of the pawn .
1 5 exd4?
18 . . . g6!
There was no need for Wh ite to g ive h i mself
19 .U.f2
an isolated pawn . However, Black would
Wh ite's active moves have come to an end
also have been excellently placed after 1 5
and he beg ins marking time, whereas the
li'lxd4 li'lc5 followed by . . . li'ld5. White feels
opponent consistently strengthens his posi­
the weakness of his c3-square, which he
tio n .
incautiously weakened with his 1 2th move.
19 . . . li'le8
15 . . . li'ld5
20 l:tcf1 li'ld6
16 'ii'd 2 li'l7f6
21 'iit h 1 li'lf5
1 7 li'lfe5 .tb7
2 2 ..ixf5?
(see diagram)
The decisive strategic mistake. Now there is
1 8 f4? nothing with which to oppose the bishop on
Another positional mistake , which puts White b7.
in an extremely difficult position. What are 22 . . . exf5
the defects of the pawn advance? 23 .l:tc1
The h 1 -a8 diagonal is weakened . With the What would you have now played for Black?
 How to draw up a Plan l2J 91

Let's first see what happened in the game, with an u n usual material balance. That is
and then return again to this position . also the case here: 29 tt'lxb4! .ll x e 1 + 30
23 . . . tt'lf6 .llx e 1 deserved very serious consideratio n .
Of course, a very natural move - the e4- F o r t h e q ueen Wh ite has rook, knight a n d
square itself is as though inviti ng the knight pawn - al most a sufficient equivalent. I f
to go there. 30 . . . i.xb4 3 1 �xb4 it'xd4, t h e n 3 2 i.d6 ! . I
would like to play �e5 and then exploit the d­
24 'i*'e3 tt'le4
file for an attack (�d2, .Ued 1 ). Of cou rse,
25 .Ufc2 'it'd5 White has to reckon with the cou nter-stroke
26 tt'ld3 l::!. fe8 32 . . . �xc4 , but it leads only to a draw: 33
27 �e1 �e8+ 'it>g7 34 �f8+ 'it>f6 35 �e7+ 'it>g7 (if
Black seems to have played well , but his 35 . . . 'it>e6?, then 36 i.c5+! is strong) 36
actions have not been systematic. Whereas �f8+.
White, who j ust now was losing, has sudden­ [I do not agree with this evaluation. After 32
ly gained cou nter-chances. The b4-pawn is i.d6 f6! Black does not allow the bishop to
attacked , and 27 . . . a5 28 tt'lb6 is bad for go to e5. The e-file can always be blocked by
Black. . . . i.e4, and switching to the d-file requires
27 . . . tt'lc3 too much time. Meanwhile, the b3- and g2-
points are vulnerable. I do not see what there
Practically forced .
is to prevent Black from converting his
28 i.xc3 i.f8 material advantage. - John N u n n . ]
Black prepares to play his bishop to g7. I n t h e g a m e Wh ite missed his chance a n d
lost without a fig ht.
29 tt'lde5? bxc3
30 'ilt'xc3 i.h6
Despite Wh ite's extra pawn , his position is
d ifficult. Black has two powerfu l bishops, and
his pieces control the entire board .
31 tt'ld3
If 31 .l:!.f1 , then 31 . . . i.xf4! 32 .Uxf4 .l:!.xe5! . [32
tt'lxg6! is White's best chance in this variation
- N u n n .]
31 . . . .Ued8
32 tt'lc5?! �a8
33 b4 �xf4
A very interesting moment. Here it is 34 .l:!.d1 .l:!.e8
appropriate to remember a procedure which 35 'ir'd3 .l:!.cd8
is constantly employed by Mark Dvoretsky. 36 l:!.e2 .l:!.xe2
In unfavourable situations he recommends
37 'it'xe2 'it> g 7
looking for a way of rad ically changing the
character of the struggle. Sometimes it is Black prepa res 38 . . . l:Ie8.
possible to complicate the play by means of 38 tt'lb2? Ite8
a positional sacrifice (for example, of a pawn 39 'ir'f1 .tc7
or the exchange) or by going i nto a position 40 b5
92 � How to draw up a Plan
This loses immed iately. 40 tZ'lc4 was more
tenacious.
40 . . . 'ii'd 6
41 11'g1 .i::t e2
42 tZ'lc4 'i¥d5
White resig ned .
Let's now return to the situation after Wh ite's
23rd move and consider whether it wasn't
possible to convert Black's advantage in a
more convi ncing way, without allowing the
opponent any counter-chances.
For myself I have formulated a rule, which I
call 'the principle of the worst piece ' . I n chess
literatu re you will not find such a formu latio n ,
although , o f course, lead ing players make
use of this principle.
I magine that you are a designer, and you
have to improve a machine which consists of
ten parts. Nearly all of them are operati ng at
90% of their power, but one is only at 1 0% . It
is clear that if you can find a way of raising
the efficiency of the 'backward' part, this will
be the best way of improving the operation of
the mach ine as a whole. In order to extract
the maximum from your position (wh ich is a
kind of chess machine), you must first of all
raise the efficiency of all the pieces which for
the moment are taking no or hardly any part
in the play.
In positions with strategic manoeuvring
(when the time factor is not of decisive
importance) look for the piece which is
worse placed than all the others. The
activation of this piece is often the most
reliable way of improving your position
as a whole.
Let's analyse the placing of the black pieces.
The knight at d5 and bishop at b7 are
operati ng, if not at 1 00 % , then at least 90%.
The rook at c8 also stands well , and in one
move the other rook will be able to occupy an
excellent sq uare at e8 or d8. The q ueen is a
mobile piece , and it is not hard to move it to
where it is required . But for the moment the
bishop on e7 is not taking part in the play.
How can it be included in the attack? By
. . . �f8-h6 (after the preparatory 23 . . . l:te8!).
At the same time the way for the q ueen to h4
is vacated .
I n cidentally, in choosing such a plan we are
also acting in accordance with a principle
discussed earlier - the principle of two
weaknesses. The fi rst weakness , the d4-
pawn , has already been fixed , and Black
beg ins looki ng for new targets - the white
king and the f4-pawn . S ubseq uently (with his
bishop on h6 and q ueen on h4) he will be
able to consider playing his knight from d5 to
g4 or e4 to strengthen the attack on the king.
The fol lowi ng game was played i n the same
tournament.
M i rumian - Baklan
CIS J u n ior Championsh i p , J u rmala 1 992
French Defence
1 e4 e6
2 d4 d5
3 tZ'lc3 .i.b4
4 e5 tZ'l e7
5 .i.d2 b6
A comparatively rare move (5 . . . c5 is usual).
 How to draw up a Plan
6 'iig4 lt:Jf5
If now 7 i.d3, then Black exploits the
vulnerabil ity of the d4-pawn to force the
advantageous exchange of q ueens by 7 . . . h5
8 'lf4 g5! (the game Leko-Panno, Argentina
1 994, went 8 . . .it'h4 9 'i!Vxh4 lt:Jxh4 1 0 g 3 lt:Jf5
1 1 lt'lb5 i.xd2+ 1 2 �xd2 tt:Ja6 1 3 tt:Je2 with
the better chances for Wh ite) 9 'ii'x g5 'i!Vxg 5
1 0 .bg5 lt:Jxd4 (Martin Gonzalez-Psakhis,
Benasque 1 995).
7 lt:Jge2
A thematic move, even though it looks rather
awkward . Wh ite defends his d4-pawn and
prevents the opponent's plan ned exchange
of the lig ht-sq uare bishops after . . . i.a6. The
exchange of the bishop for the knight on e2 is
not so advantageous for Black.
7... h5
8 'it'f4
If I had been playing Black, without th i n king
for long I would nevertheless have played
B .ia6 - after all, this was the reason for the
. . .
move . . . b7-b6. The rapid development of
the pieces is a basic principle of opening
play. But our Vova Baklan is not a dogmatist,
he does not often remember about general
principles, and he always seeks his own way.
8 . . . i.e7 ! ?
This second move with an already devel­
oped piece would simply not have occu rred
ttJ 93
to me, because I am in fact a dog matist. I n
o n e instance out of ten , such violations of the
chess laws by Vova prove successfu l , in a
fu rther five he gets away with it, but in three
or fou r cases out of ten he is punished .
However, on this occasion Black's u n usual
decision is justified . The threat of 9 . . . g5 is
very u n pleasant, and it provoked the oppo­
nent into making a sharp reply, which proved
unsuccessfu l . I would recommend 9 0-0-0 gS
1 0 'iif3 g4 1 1 Wf4 , and if 1 1 . . . l:tgB 1 2 h3.
9 h4?! tt:Jxh4
1 0 0-0-0 tt:Jc6
Black h u rries to bring his pieces into play.
Generally speaking his position is rather
dangerous - after a l l , the opponent has a
lead in development and the open h-file for
an attack.
11 g4 lt:Jg6
1 2 We3? !
A n over-optim istic decision. I n an analysis of
the game we decided that it would have been
better to retreat the q ueen to g3, in order to
provoke . . . h5-h4 , and only then play 'iie 3
with the strateg ic threat of f2-f4-f5 .
12 . . . hxg4
1 3 .l:.xh8+ lt:Jxh8
14 lt:Jf4 i.g5
By p i n n i ng the knig ht, for the moment Black
does not allow it to go to h5.
15 i.d3 ..ib7
1 6 l:th1
The evaluation of the position depends
l iterally on one tempo. I magine that the
queen were already on d7 - then Black
would castle and the opponent would have
no compensation for the two missing pawns.
But Black does not have this tempo and the
situation remains rather tense.
16 . . . i.h6
Now someth ing must be done, as otherwise
Black will prepare to castle.
(see diagram)
94 <t> How to draw up a Plan

25 'iff5
What would you play now?

17 .l:.xh 6 ! ? gxh6
1 8 lDfxd5!
If 1 8 lDh5 there was the good reply 18 ... 'ili'h4 .
Let us employ the principle of the worst
18 . . . exd5 piece . It is q u ite obvious that Black's worst
1 9 'ifxh6 piece is his king . If it were to be removed
Wh ite's attacking looks menacing, but don't from the centre and placed at b8, the
forget that for the sake of it he has sacrificed opponent would have to resig n . Therefore
a whole rook. In such cases the opponent 25 . . . 'it>c8! followed by 26 . . . 'it>b8 suggests
usually has an opportunity to buy his itself.
opponent off, by g iving u p part of his extra National master Telman (the trainer of
material . Seryozha Ovseevich , who was playing in the
19 . . . lDxd4! same tou rnament) suggested another, tacti­
cal solution : 25 . . . l:th4 26 f3 d4! 27 .ixd4
20 'ii'x h8+
�xf3 , achieving simpl ification adva ntageous
Nothing was given by 20 lbb5 lbxb5 2 1 to Black. Also not bad , although the march of
�xb5+ c6 2 2 �xc6+ i.xc6 2 3 'ii' x c6+ 'it>f8 24 the king away from the centre appeals to me
'ii'h 6+ 'iti>g8 25 �g5 'ii'f8 . more.
[Instead of 22 �xc6+ ? White maintains the It was a pity that, as a consequence of
balance by 22 �g5!, for example: 22. . 'ii'xg5+.
tiredness and approach ing time-trouble (of
23 'ili'xg5 cxb5 24 'ili'f6! 'it'd7 25 "iVd6+. The course, this is not a j ustification , but merely
same move 20 �g5! was also not at all bad an explanation ) , Vova did not fi nd the correct
immediately, instead of capturing the knight path and m issed an almost certain win .
- Dvoretsky.]
25 . . . 'it>e7?
20 . . . 'it>d7
26 lDe2 �c8?
21 'ii' h 5 'ili'g8
A clear waste of time - the bishop stood
22 �e3 lbe6 better at b 7 .
It is apparent that the worst for Black is over, 27 lDg3 'i!kg6
and the attack should be parried . 28 'ii'f3 �b7?
23 i.f5 'ii'g 7 [ The logical continuation of Black's preced­
24 �xg4 .l:th8 ing moves would have been 28 . lbg5! 29 . .
 How to draw up a Plan L'iJ 95

lf6+ (29 'ikf4 �xg4 30 'ikxg4 l'De4) 29 . . . 'Wixf6 A rather passive pla n . Theory recommends
30 exf6+ �xf6 3 1 ii.d4+ cJ;g6 32 �xhB �xg4 11 . . 'ika5 .
.

with an extra pawn and definite winning 1 2 IIc1 e6

chances, despite the opposite-colour bish­ 1 3 dxc6
ops - Dvoretsky.]
In the event of 1 3 f4 Black could have
29 c4 i nitiated compl ications not unfavou rable for
Black's position has become dangerous. h i m by 1 3 . . . l'Dxc4 1 4 ii.. xc4 exd5. Here in
After th inking for al most all his remaining reply to 1 3 . . . bxc6? the move 14 f4 would now
time, Baklan took the correct decision - to be much stronger.
force a draw. 13 . . . l'Dxc6
29 . . . d4 1 4 ii'b3 'ii'e 7
30 l'Df5+ 'it>d8 1 5 �fd 1
31 'it'xb7 'Wixg4 Wh ite sees a weakness in the opponent's
32 'Wia8+ 'it>d7 position - the d6-pawn , and he concentrates
33 'it'd5+ 'it>e8 his attacks on it.
34 'Yi'c6+ �d8 15 . . . �ed8
35 'ir'a8+ 1 6 �d2 .l:i.d7
Draw agreed . 1 7 l::tc d1 1Iad8
Rememberi ng the turning point of this game, 18 'ir'a3 .ltf8?!
we see that any of our pieces can be the For the moment the d6-pawn was adequate­
'worst' , including the king . ly defended , so was it worth retreating the
bishop to a passive position?
Now I will illustrate the principle of the worst 19 .ltg5
piece in a game not between two young With the bishop on f8 this pin is especially
candidate masters, but between two experi­ u npleasant.
enced grandmasters. 19 . . . h6
20 ii.. h 4 g5
Dorfman - Balashov Otherwise 21 f4.
Tal l i n n 1 983 21 �g3 a6
King 's Indian Defence In ' H edgehog'-type positions this is a normal
1 l'Df3 l'Df6 move, taking away the b5-square from the
2 c4 g6 white knig ht. But i n the g iven instance it is not
too successfu l and l ossif Dorfman subtly
3 l'Dc3 �g7
exploits its d rawback - the weakening of the
4 e4 d6
b6-square. 21 . . . b6 was preferable.
5 d4 0-0
6 �e2 � g4
7 .lte3 l'Dc6
(see diagram)
8 d5 .ltxf3
9 .ltxf3 l'De5
10 ii.. e 2 c6
11 0-0 �e8
96 � How to d raw up a Plan

25 'ii'e 3 'ii'f6
26 tt:lb6 �c7
It's done - Wh ite has switched his q ueen and
knight to more active positions. Now he
clarifies the situation i n the centre (if it were
the opponent to move, he would happily play
27 . . . ..tf4).
27 ..txe5 tt:lxe5
If 27 . . . 'i!Vxe5, then 28 c5 with the th reat of 29
tt:lc4 . And if 27 . . . dxe5 there fol lows 28 �xdB
tt:lxd8 29 tt:ld7 and 30 tt:lxe5 .

Let's approach this position from the stand­

point of the 'worst piece' principle. White's
rooks and bishops are excellently placed .
For the moment his knight is not very
strongly involved - it would not be bad to
transfer it to b6. But fi rst the e4-pawn must
be defended . How? I n the event of 22 f3 the
opponent acquires not only the . . . tt:lh5-f4
manoeuvre, but also 22 . . . d5, which , it turns
out, Yuri Balashov had prepared by playing
2 1 . . . a6. It could not be played immediately:
2 1 . . . d5? 22 'i!Vxe7 tt:lxe7 23 cxd5 exd5 24
..lli. b 5! .
28 g3!
But what can b e said about the wh ite queen?
It has already fulfilled its function on a3 - 'This and especially the next move came to
lured the enemy bishop to the passive me with great difficulty. After investigating
square f8 . But now there is noth ing for it to do the position, I realised that the exchange of
here: to attack a well-defended pawn is a queens (after 28 . Wif4 or 29 . ."ikg5) would be
. . .

rather pitiful role for Wh ite's strongest piece. advantageous to the opponent, since the
Where would he like to play it to? Of cou rse, knight on b6 would become vulnerable. ' -

to e3 - from here the queen defends the e4- Dorfman.

pawn, releasing the knight, and has the The problem of which pieces to exchange is
possibility of taking part in an attack against an i n separable component of chess strate­
the opponent's kingside, which has been gy!
weakened by . . . g7-g5. 28 . . . g4
22 'i!Vb3 ! 29 h4!
Wh ile intend ing the manoeuvre 'i!Vb3-b6-e3, After 29 . . . gxh3 30 f4 Wh ite's adva ntage is
White simu ltaneously prevents . . . d6-d 5 . obvious. 29 h3?! is weaker: 29 . . . 'ii'g 5! 30
22 . . . ..tg7 'ii'x g5 hxg5 3 1 hxg4 �c6 , and Black regains
23 'ii' b6 tt:le8 his pawn .
24 tt:la4 ..te5 29 . . . �b8
 How to d raw up a Plan ltJ 97

What should Wh ite do now? We a re not able 35 c5!

to intensify the pressure on the weak d6-
pawn . Accord ing to the principle of two
weaknesses, our attention should be switched
to new targets. Black's kingside pawns are
weakened and demand attention . His king is
unlikely to be mated , but it will feel uncom­
fortable.
But also on the q ueenside Wh ite can make
progress. It makes sense to place the pawns
on a5 and b4 , after which the opponent will
have to reckon with the breakth roughs b4-
b5 or c4-c5.
30 b4 �g7
31 a4 aS
Balashov does not want to al low a4-a5 , with
a bind on Black's q ueenside.
Wh ite's pieces are very active, whereas the
opponent's forces are disun ited - with the
exception of the knight on e5, he has no
good pieces. In such cases it makes sense
to q u ickly open the position, and come into
d i rect contact with the opponent, before he
has coord inated his forces.
35 . . .
3 5 . . . dxc5 36 'ili'xc5 was also bad for Black.
36 �g2 dxc5
37 l:.xd8 'ilfxd8
38 Wxc5
[38 a5!, intending 39 l:.xc5, was more
accurate, since after the move in the game
Black could still have put up a tough defence
by playing 38 . 4Jc6! - Dvoretsky.]
. .

32 bxa5 .l:.c5
38 . . . 'iid 6
33 llb1
39 'ili'c3
Of cou rse, not 33 l:txd6? tt:Jxd6 34 'ili'xc5
39 'ili'e3! ? .
4Jxe4.
39 . . . l:txb6
33 . . . lba5
40 llxb6 'ili'xb6
34 llb5 .l:ta6
41 'iix e5+ f6
In the event of 34 . . . l:txb5 35 cxb5! Wh ite
acquires a menacing passed pawn on the a­ 42 'ii h 5
file. Black resig ned .

Of cou rse, chess strategy does not red uce

merely to the two principles we have
considered . But even so, I hope that, by
using them, you will find it easier to plan your
actions at the board .
98 �
Al exey Kosi kov

Sens i n g the Tem po

Tby th ree main factors : material, placing

he evaluation of a position is determi ned Like other practical chess skills, sensing the
tempo can be developed by solving specially
of the pieces and pawns (king safety, the selected exercises with a tactical content,
existence of weaknesses etc. ) and time. which demand sharp combinative vision. I
The role of time i n chess is not clearly recommend that, when a nalysing games,
defined . In tactical positions every tempo is both you r own and those of other players,
of decisive importance, whereas in situations focus you r attention on this problem . And, of
of strategic manoeuvring the time factor course, of g reat help here is a general g rowth
someti mes fades into the background. in you r erud ition , you r chess culture.
I n many types of sport the concept of a Today we will analyse various situations, in
'sense of rhythm' exists. I n its best years the which the time factor plays a g reater or
Dynamo Kiev football team played at an lesser role.
irregular tempo - fi rst quietly, d u l l i ng the
vigilance of the opponents, and then sudden­ Korneev - Kos i kov
ly with a sharp increase in speed . This ski lful
Smolensk 1 99 1
organisation of the play (of cou rse, i n
combination with h i g h individual skills) helped
the team to become one of the strongest in
Europe.
Another example. A long-distance ru nner
has to control his speed . He supposed ly
should ru n as qu ickly as possible, but
nevertheless not too q u ickly - otherwise he
won't have enough strength for the entire
distance. It is very important for h i m to seek
his optimal running rhythm .
I n chess too we can talk about a similar
feeling. It may relate to expend itu re of time in
the direct meaning of the word (abil ity to
decide when moves can be made q u ickly,
and where one should stop and have a good Black's position i s won , of cou rse. However,
think). This is a sepa rate , very importa nt the simple variation 56 . . . � c3 57 b5 'ii'x a3 58
topic, but today we are talking about some­ b6 'ii' b 2 59 b 7 did not completely satisfy me ­
thing else - about 'sensing the tempo ' . This the far-adva nced pawn on b7 makes the win
is what I call the ability to sense how rather difficult. The question is, doesn't Black
important the time factor is at a g iven have a more reliable way to win?
moment, and at what tempo one should be The wh ite queen is overloaded , si nce it
acting: qu ietly, even wa iting, or very specifi­ simultaneously has to defend the f2- and a3-
cally and energetically. pawns. There immed iately comes to mind
 Sensing the Tempo ttJ 99

the idea of zugzwang - a device with the help 6 cxd5 exd5

of which nu merous endings a re won . 7 .te2 a6
56 . . . ..td4! 8 0-0 c4
If now 57 'iif4, then 57 . . . ..tc3 58 'ii'f3 'ii'x a3 The Tarrasch Defence to the Queen's Gam­
59 b5 i.b2 (or 59 . . . ..tb4), and the pawn does bit has been played . But the ideas of different
not succeed in advancing to b7. openings sometimes i ntersect. Tel l me, does
57 'it>g1 .tc3 this position remind you of anyth ing? That's
57 'i!t'c3 would also have won easily.
. . .
rig ht, we've reached the Panov Attack
against the Caro-Kan n Defence with colours
58 b5 .td4!
reversed and an extra tempo for Wh ite .
The b5-pawn is attacked , and if 59 a4 there
There the plan i nvolving c4-c5 is not
follows 59 . . ."ii' a 1 + and 60 . . . Wxa4 . This is
considered very favou rable for Wh ite. And
why Black lured the king onto the 1 st rank! this applies even more here - with a tempo
White has only one reply. less, it is hardly good to advance the c-pawn .
59 'ii'f4 .te5 9 lLle5 'ii'c 7
60 'ifc4 Now 1 0 f4 , supporting the knight on e5,
Otherwise 60 . . . 'i!kxb5. suggests itself. But this move has a position­
60 . . . .td4! al defect - it weakens the e4-poi nt. Black
The bishop has moved to this sq uare three completes his development with 1 O . . . .tb4,
times, and each time with increasing effect. and later at some point he will captu re on c3
61 'ii'f1 'iVxa3 and invade at e4 .
62 'it>g2 1 0 lZ:lxc6 'it'xc6
'ii' b 2
63 'i!tf3 'it>fB 1 1 b3

White resigned .
Thus a zugzwang position is one of the
situations where the time factor tells, and i n a
highly d istinctive form .

In the opening stage your forces must be

quickly developed , and here any loss of
tempo may prove fatal . N u merous games
have been lost merely because one of the
players neglected his development. I will
restrict myself to one example on this theme.

Petrosian - Sueti n
27th USSR Championsh i p , Len ingrad 1 960
The danger of Black's position begins to be
Queen 's Gambit
felt, and he must be extremely carefu l . His
1 c4 c5
king is still i n the centre and the opponent is
2 lLlf3 lZ:lf6
already taking the in itiative and creating
3 lLlc3 lZ:lc6 concrete threats. He should complete his
4 e3 e6 development as q u ickly as possible by
5 d4 d5 playing 1 1 . . . .tb4 (with gain of tempo ! ) and at
1 00 � Sensing the Tempo

the fi rst convenient opportun ity - castle. But Soloviov - Kosi kov
that which I cal l 'sensing the tempo' appar­ Smolensk 1 99 1
ently betrayed Alexey Sueti n . French Defence
11 . . . b5? 1 e4 e6
1 2 bxc4 bxc4
2 d4 d5
1 3 e4!
3 e5 c5
If Wh ite wants to pun ish the opponent for
4 c3 tt:'lc6
neglecting his development, he must delay
it, come into direct contact with h i m , and 5 tt:'lf3 'ili'b6
sharpen the play. 6 a3 �d7
13 . . . dxe4 7 �e2
The lesser evil was 1 3 . . . tt:Jxe4 (exchanges The combination of 6 a3 and 7 �e2 is hardly
usually favour the defending side), but even good ; the latest word i n fashion in this
here after 14 tt:Jxe4 dxe4 1 5 d5 'ii'g 6 1 6 Ji.h5! variation is 7 b4 . However, when playing
(a recommendation by Yuri Razuvaev) Wh ite one can take some liberties without
16 . . . 'ii'f5 1 7 lle 1 Wh ite has the advantage. being pun ished .
1 4 .ll g 5 �f5 7 . . . tt:'lh6
The middlegame is in full swing, but Black is 8 b4 cxd4
still not in a hu rry to develop his kingside. 9 cxd4 tt:'lf5
Good or bad , he had to play 14 . . . �e7. 1 0 �b2 �e7
15 d5 "ilc7 1 1 0-0 0-0
1 6 �xf6 gxf6 White has problems with the completion of
1 7 �g4! his development - he can not bring out his
The last accu rate move , exchanging the knight i n view of the loss of the d4-pawn . The
opponent's only developed piece - the move b4-b5 is anti-positional , since it seri­
bishop which is defending the e4-pawn (if ously weakens the q ueenside.
1 7 . . . �g6 Wh ite wins by 1 8 tt:Jxe4 �xe4 1 9 12 'ii'd 2?
.l:!.e 1 'ii'e 5 20 'il'a4+ with mate). The outcome
1 2 Ji.d3 and then �xf5 was necessary, with
is now not in doubt.
roughly equal chances.
17 . . . �xg4
What would you now have played as Black?
1 8 'ifxg4 'iie 5
1 9 tt:Jxe4 f5
20 'ii' h 5 0-0-0
21 tt:'ld2 c3
22 tt:'lc4 'il'd4
23 'ii'xf5+ .Ud7
24 tt:Je5
Black resig ned .

It is not only neglect of development that can

be the cause of failure. The same conse­
quences can follow from a superficia l , ill­
judged handling of the opening.
 Sensing the Tempo ltJ 101

You suggest playing for the exchange o f the

light-square bishops? It is too slow. With his
last move the opponent began carrying out
what is general a sensible idea : he is
planning 1 3 l:td 1 and 1 4 ltJc3. To frustrate his
plan, you must play more energetically. Let
us check specifically: 1 2 . . . ltJb8? 1 3 ltJc3 or
12 . a6?! 1 3 .l:.d 1 ltJa7 1 4 ltJc3 , and Wh ite is
. .

excellently placed .
Another suggestion is 1 2 . . . g5. You know, I
like playing . . . g7--g5 in the French Defence,
but here this is really too sharp !
There is a nother possibil ity: 1 2 . . .f6 . B u t , i n
suggesti ng i t , did you m i s s 1 3 g4 ltJ h 6 1 4 The i n itiative is on my side, but the position is
exf6 followed by the fork g4--g 5 ? You missed one where every tempo is important. If Wh ite
it? But even so, that's what I played . should succeed in playing ltJd2-f3--e5 or
12 . . . f6! ltJd2-b3-c5, I will have to forget about any
The tactical idea associated with this move is hopes of an advantage. Sensing the tempo
wel l known - it is analysed i n Aaron suggests that Black should urgently 'latch on'
Nimzowitsch's famous book My System. to the opponent, using concrete threats to
divert him from the completion of his devel­
1 3 g4 ltJh6
opment.
14 exf6 .l:.xf6
22 . . . a5!
1 5 g5 .:txf3
23 bxa5 l:lxa5
16 ..txf3
But not 23 . . . ltJxa5 because of 24 ltJc3. The
Black also has the advantage after 1 6 gxh6. black rook comes i nto play, and Wh ite still
16 . . . ltJf5 cannot move his knight.
1 7 l:id1 24 a4
The positional exchange sacrifice has g iven Here I stopped to th i n k and I found what I
Black a splendid position . It is amusing that believe is a good solutio n . I was helped by
at this point my opponent offered a d raw. the ' principle of the worst piece'. The bishop
17 . . . on d7 is not taking any part in the play. The
'ii'd 8!
standard route for the bishop is via e8 to h5,
The g5-pawn is lost. The next few moves are
but from there it will be fi ring i nto empty
forced .
space. It is most probably better to attack the
1 8 ..tg4 ..txg5 a4-pawn with it, i . e . move the knig ht.
1 9 f4 ..tf6 24 . . . ltJd8!
20 ..txf5 exf5 The knight goes either to e6, or via f7 to d6. If
21 'ilr'f2 Wb6 Black should captu re the a4-pawn , he will
Wh ite wanted to develop his knight, but I already have two pawns for the exchange.
prevent this by tying h i m to the defence of the 25 ltJc3 ltJe6
d4-pawn . The d4-pawn is again under fi re.
22 l:td3 26 ltJd1
1 02 � Sensing the Tempo
Here the knight is not much better placed
than on b 1 . Now I could have simply
captured with my rook on a4 with an
advantage, but in such cases it is important
not to sell yourself too cheaply. You should
check whether or not there is someth ing
stronger. And indeed , if you see Black's next
two moves , it immediately becomes clear
that this is what should be played .
26 . . . 'it'd&
The f4-pawn is attacked .
27 �c1 b5!
The pawn is won in a version that is more
advantageous to Black. His passed pawn will
be much more dangerous on the a-file than
on the b-file.
28 �e3 bxa4
29 l2'lc3
A new question : how would you characterise
the resulting situation (from the standpoint of
our topic - 'sensing the tempo') and what
wou ld you suggest playing?
I thought that I had already gai ned a decisive
advantage (two pawns for the exchange,
dangerous passed a-pawn , weaknesses on
d4 and f4) and I spent some time looking for
concrete ways to break through the oppo­
nent's defences. But gradually I realised that
the defensive resou rces were qu ite consid­
erable, and that the position was not yet ripe
for decisive action - it was not tempo­
dependent. For a time Black needs to
maintain it, to manoeuvre. Such manoeu­
vring, on the one hand, enables the position
to be strengthened to the maxi m u m , by
making all moves that will be usefu l in the
futu re, and on the other hand, it allows the
most appropriate moment to be chosen for
switching to positive action , when the oppo­
nent goes wrong and makes the task easier.
30 . . . h6!
Now the king will feel more secu re, and
. . . g7-g5 is also a possibil ity.
31 .:d2 'i!i'b4
32 l2'la2 'ii'd 6
33 lt'lc3 'it>h8
As you see, in non-tempo positions the
principle 'do not hu rry' comes to the fore. I
very much did not want to play . . . g7-g5 (the
opponent is condemned to passivity, so why
sharpen the play? ), and yet now Wh ite has to
reckon with this move. You see, after 34 . . . g5
35 fxg5 hxg5 36 "ii'xf5 I have the reply
36 . . . lt'lxd4, which would not have been the
case with the king on g8. And in general, the
king stands slightly better at h8 than at g8,
even if only marginal ly.
Note that my last few actions ( . . . h7-h6,
. . . 'i!i'b4 , . . . 'it>h8) have not involved the
slightest risk. But the opponent has to be
constantly on the alert, since any move of his
may tu rn out to be a serious mistake. It is
very d ifficult to defend in such situations.
34 .Ub2 'ii'c 7
Advancing the pawn to a3 is prematu re - this
move should be held over Wh ite l i ke a sword
of Damocles and made only with decisive
effect. But for the moment Black should
prolong the manoeuvring, trying with minor
th reats to d isrupt the coordination of the
wh ite pieces.
35 'iid 2 'iia 7
36 .l:td1
36 lt'le2 is bad in view of 36 . . . i.b5, when the
 Sensing the Tempo l2J 1 03

bishop comes very strongly i nto play. But I n the previous game no particular imagina­
after the rook has moved , the advance of the tion was demanded of Black - self-control
a-pawn will crack the opponent's defences. and patience were more necessary. But for
We see the principle of two weaknesses in seizing and retaining the in itiative, accurate
action - Wh ite is u nable simultaneously to and resou rceful play is requ i red , and the
defend the d4-pawn and to combat the value of each move is usually extremely
passed a-pawn . high.
36 . . . a3
37 l:ta2 �a4
After both a move of the roo k and the Roma n i s h i n - Farago
excha nge o n a4, the d4-pawn is lost. The E u ropean Tea m C h a mpionsh i p ,
game is practically decided . Skara 1 980
In such situations the opponent usually 'goes R e ti Opening
berserk' , trying at any cost to create some 1 liJf3 ltJf6
counter-chances, and here one should be 2 g3 d5
especially carefu l . 3 i.g2 c6
3 8 l:lxa3 i.xd 1 4 0-0 �g4
39 ltJb5 5 c4 e6
A clever reply. But since Black's previous 6 d4 liJbd7
strategy was correct, the tactical compl ica­
7 ltJe5 i.. f5
tions should tu rn out to his advantage.
8 ltJc3 i.. d 6
39 . . . l:txa3
9 �f4 'iWb8
40 ltJxa7 l:txe3
Black is slig htly slow in castling. Oleg
41 ltJc6
Romanishin tries to exploit this factor by
41 'i'xd 1 �xd4 was totally bad . sharpening the play.
41 . . . Iie4 1 0 ltJxd7 ltJxd7
42 'i¥xd 1 ltJxf4 11 cxd5 exd5
Of cou rse, 42 . . . ltJxd4 would also have won , 1 1 . . . i.xf4? 1 2 dxc6 .
but why exchange the wh ite knight when i t is
1 2 e4! dxe4
shut out of the game?
1 3 i.. x e4 i.. xe4
43 'iVa4
1 4 �xd6 "ii'x d6
43 'iff3 .l:!.e 1 + 44 'it>f2 �h4+ .
1 5 ltJxe4 'ii'g 6
43 . . . �h4
Here is the position , for the sake of which we
White resigned .
are a nalysing this game. What would you
suggest?
We have seen what sort of approach a player
should adopt i n non-tempo positions (the
principle 'do not hu rry' etc . ) . The fol lowi ng
examples will be devoted to the problem of
the initiative . (see diagram)
1 04 � Sensing the Tempo
If Wh ite does have any advantage, it is clear
that after the slig htest inaccu racy it will
instantly evaporate. After all, there are no
weaknesses in Black's position, and the only
straw at which one can clutch is the tempo
which the opponent will have to spend on
castl ing. But how to exploit it?
1 6 lDc5? is incorrect in view of 1 6 . . . 0-0-0,
and Black already stands better.
Grandmaster Igor Za itsev pointed out the
strongest conti nuation : 1 6 'i¥b3 ! . Now 1 6 . . .
lDb6 1 7 lDc5 0-0-0 1 8 a4 i s clearly advanta­
geous to White , which means that kingside
and queenside castl ing must be examined .
I t i s good i f after 1 6 . . . . 0-0-0 you instantly saw
the tactical motif 'ifxf7 ! . But the immed iate
1 7 'iixf7? does not work because of
1 7 . . . 'iixe4 . 1 7 llfe 1 suggests itself, but Black
replies 1 7 . . . I1he8, when 1 8 'ifxf7? .:Xe4 1 9
'ii'x g6 .l:txe 1 + is not possible. Therefore 1 7
.l:tae1 ! must be tried . Now after 1 7 . . . .l:the8 1 8
'i!i'xf7 Wh ite wins a pawn . If 1 7 . . . lDb6, then
1 8 lDc5 , intending 1 9 a4 or 1 9 .U.e7 , for
example: 1 8 . . . .l:txd4 1 9 .l:l.e7 'ifd6 20 'ifxf7
'ii'xc5 2 1 'i!Vxg7! and wins, or 1 8 . . . 'iff6 1 9
'ii'a 3! 'it>b8 20 .l::!. e 7! . Finally, if 1 7 . . . 'it>b8 Wh ite
can reply 1 8 .l:te3 ( 1 8 a4? ! l:lhe8 is less
accu rate) with the better chances.
In the 16 . . . 0-0 variation White can fight for an
advantage, only by fi nding after 1 7 11¥xb7
'ii'xe4 1 8 'iVxd7 l:.fd8 ( 1 8 . . . .l::!. a d8 1 9 Vxa7)
the i ntermediate move 1 9 .l:!.ae 1 ! . How
should Black defend? He loses his q ueen
after 1 9 . . . 'it'xd4? 20 .l:!.e8 + ! , while 1 9 . . . 'ii'xe 1
20 'ifxd8+ .l:txd8 2 1 .l:txe 1 leaves Wh ite a
sound pawn to the good . Wh ite also has the
advantage after 1 9 . . .'ii' f3 20 'i!i'b7 followed by
l:te3 or l:te7. It remains to check 1 9 . . .'ili'c2 20
'ii' b 7 l:tdb8 (20 . . . .l:tab8 21 'i!i'xa7 'iVxb2 22
.l:r.e4) 21 11¥a6 'ii'x b2 (2 1 . . . .l:tb6 22 'ili'e2;
2 1 . . . lhb2 22 l:tc1 'ilfe2 23 'i!i'xc6 or 22 . . . 'ii'e4
23 .l:!.xc6, retaining the extra pawn) 22 'ii'xc6
'il'xd4 23 .:r.e7. Although material equality
has been restored , the i n itiative undou btedly
remains with Wh ite (analysis by Zaitsev and
Dvoretsky).
Romanishin is a player with an original,
dynamic style, and even so he missed this
opportun ity. However, the move made by
him looks very natu ra l .
1 6 d5? ! 0-0 !
1 7 l:te1 ? !
1 7 dxc6 ! 'ii'xc6 1 8 l:te1 followed b y l:t c 1 was
stronger, when Wh ite still retains some
in itiative, although not as promising as after
1 6 'ii' b 3! . I n chess, you know, mistakes often
come in series. Romanishin was apparently
oppressed by the subconscious feeling that
somewhere he had not exploited all the
advantages of his position. In trying to
mai ntain his fading advantage, he lost his
objectivity and soon encou ntered serious
difficulties.
17 . . . .l:tad8
1 8 d6?
Here too he should have captu red on c6,
although now after 1 8 . . . lDe5! (with the idea
of 1 9 . . . lDxc6 ) the in itiative would have
passed to Black.
18 . . . f5!
Apparently Roma nishin overlooked or un­
derestimated this move. The pawn on d6
becomes a real weakness and will soon be
lost.
19 'ii' b 3+ 'it>h8
 Sensing the Tempo 1 05

20 tbc3 tbc5 28 .Uxd 1 "it'xc3

21 'i!Vc4 'ii'x d6 29 'ii'xf5 'it'f6?
22 .l:!.ad1 tbd3 Only by retaining the q ueens (29 . . . 'it'a3 ! )
23 l':!.e2 could Black h ave contin ued playing for a win .
23 .Ue3? lbxb2 . 3 0 'ir'xf6 gxf6
23 . . . b5! 3 1 .l:!.d7
24 �h4 Now W h ite is no worse , i n view of the activity
of h i s rook and the u nfortunate placing of the
opponent's pieces .
31 . . . c5 32 �xa7 �c8 33 �b7 (33 \it.Jf1 !?)
33 . . . b4 34 'it>g2 \it.Jgs 35 h4 h5 36 'lt13 �as
37 �b5 �a3+ 38 \it.Je4 .l:f.xa2 39 l:!.xc5 .l:i.xf2
40 l:tb5 'it>g7 Draw.
The game we h ave analysed , in w h ich from
the very start W h ite's in itiative h u ng by a
thread, shows h ow d ifficult it can be in such
situations to find the only conti nuations,
someti mes far from obvious, which do not
allow the flame of the i n itiative to be
extinguished . Sensing the tempo should
help you to decide on that moment, when the
maximum concentration is required in search­
Black has an obvious advantage. However,
ing for the correct solution .
because of the pin on the knight at d3 there is
still some tension in the position, making it
tempo-dependent, and forcing h i m to look for l l i nchenko - Kosi kov
a clear, concrete solutio n . After saying ' a ' , h e Kiev 1 99 1
should h ave said 'b' - advanced h i s b5-pawn English Opening
further, disru pting the coordination of the 1 d4 tbf6
enemy pieces. After 24 . . . b4 ! 25 tba4 'i!t'd5 26
2 tbf3 c5
J::l.e3 (26 �ed2 i¥f3 with the threat of
3 c4 cxd4
27 . tbe 1 !) 26 . . . iib5 (but not 26 . . . 'ir'xa2? 27
. .

ILleS) Wh ite's position would h ave become 4 tbxd4 e5

extremely dangerous. For example: 27 b3 f4 ! 5 lbb5 d5
or 27 'ir'e7!? .l:!.g8! 28 b3 f4 29 gxf4 l:!.d5 6 cxd5 .iLc5
(analysis by Dvoretsky). The 'Vaganian Gambit' . For the pawn Black
Ivan Farago found another way, also very obtains an end u ring in itiative , which is not at
concrete - a forcibly simpl ifying operation, all easy to extinguish .
which , h owever, seems to me to be less 7 tb 1 c3?
convincing. A poor move. The c3-sq uare should h ave
24 . . . iic5?! been occu pied by the other knig ht.
25 �ed2 l:td4 7 . . . 0-0
26 'ii' h 5 lbxb2 ! 8 e3 a6
27 .l:i.xd4 tbxd 1 9 tba3 b5
1 06 � Sensing the Tempo

1 0 lDc2 .if5
Although Black is a pawn down , for the
moment he can play without particu lar
thought - so natu ral is the development of
his pieces . He is not th inking about regaining
the material , but is planning . . . lbbd 7 , . . . e5-
e4 and . . . lbe5 . From e5 the knight will exert
pressure on both wings.
11 �d3
My opponent was apparently afraid of me
regaining the pawn by 1 1 . . . .ixc2 and
12 . . . lbxd5, which did not come into my plans
at all. However, after 1 1 . . . lbbd7 and 1 2 . . . .Uc8
the threat of the exchange on c2 would
indeed have become real .
11 . . . e4
1 2 �e2 �g6
This is not yet a tempo-position , and so I
decided to spend time on a prophylactic
move. After the immed iate 1 2 . . . lbbd7 1 would
have had to reckon with 1 3 g4! ? .
1 3 b4? !
The start of an over-sharp pla n , which in the
end led to my opponent's defeat.
13 . . . .id6
1 4 a4
A continuation of the same faulty strategy.
With the wh ite king still stuck in the centre,
the opening of lines on the queenside is to
Black's advantage. Wh ite would have done
better to 'let sleeping dogs lie'.
14 . . . bxa4
1 5 l:.xa4 'ii'c8
I did not like 1 5 . . . lDbd7 because of the
manoeuvre lDc2-d4-c6 .
1 6 .ib2 lbbd7
If now 1 7 lDd4 , then 1 7 . . . lDb6 followed by
1 8 . . . lbbxd5, 1 8 . . . lbc4 or 1 8 . . . �xb4.
1 7 b5 lDc5
1 8 lta2
How should I now contin ue?
If the opponent should succeed in removing
his king from the centre, he will obtain an
acceptable position . Moreover, his slight
delay in castling will then acq u i re a logical
basis. In fact, if Wh ite had castled a few
moves earlier, then by placing my knight on
e5 I would have beg u n an attack on the
kingside. But now, when the open ing of the
play on another part of the board has
diverted my pieces, the king will feel com­
pletely safe on the kingside.
Thus, I am obl iged to undertake something.
1 8 . . . lDd3+ suggests itself, but how should
one assess the position arising after 1 9
�xd3 exd3 2 0 lbd4 axb5 2 1 .l::tx a 8 'ii'xa8 22
lbdxb5 (or 22 lbcxb5)? Of cou rse , Black
retains the i n itiative , and this cannot be bad
for him. But nevertheless it is pity that at the
end of the variation the opponent has a
choice - he can captu re on b5 with either
knight, 22 lbdxb5 apparently being the stron­
ger option. And I found a transposition of
moves, which denied Wh ite this possibil ity.
18 . . . axb5
1 9 llxa8 'ii'x a8
20 lbxb5
[20 0-0 was better - Dvoretsky.]
20 . . . lbd3+
21 �xd3 exd3
22 lbcd4
 Sensing the Tempo lZJ 1 07

'Da7 l:.xc3 27 'Dxb5 .Uc1 mate .

26 . . . 'Dxd5
26 . . . 'De4 was also strong . Here Black was
able to terminate his calculations, since it
was now obvious that the opponent's de­
fences were collapsin g .
27 'Da7 'ii'x b4
28 'Dxc8
28 'iVxb4 Itc1 + 29 'it>d2 I1c2+ .
28 . . . 'ii' b 1 +
29 'it'd 1 d2+!
30 'iii> x d2 'iVb4+
In aiming for this position , I did not see a Wh ite resig ned .
concrete solution , but I sensed that some­
thing shou ld be fou n d . It is now possible to We have talked about the development and
latch on to the wh ite king , by checking with use of the i n itiative, but this is only part of the
the queen at a5 or the bishop at b4 , and the story. Before that you have to be able to
d3-pawn is very dangerous. seize it. The fight for possession of the
I d e l ve d into the position , and i n the end I in itiative (as well as subsequent actions to
found and calculated a forced way to win . exploit it) is usually of a tempo-dependent
nature, demanding accu racy and resou rce­
22 . . . �b4+
fu l ness.
23 i.c3
23 �1 was more tenacious, but then Black
llyin-Genevsky - Ragozin
simply captu res on d5 with an overwhel ming
advantage. [ There is a quicker way to win: 3rd Match Game, Len i n g rad 1 929/30
23 fia2! 24 �c3 tDe4 - Dvoretsky.]
. . .
French Defence
23 . . . 'iVa5 1 e4 e6
24 "ii' d 2 2 d4 d5
If 24 ..ltxb4 Wxb4+ 25 'ii'd 2 I had prepared 3 'Dc3 �b4
25 'i'b1 + 26 'i!Vd 1 d2+! 27 'it>xd2 'De4+ 28
. . . 4 �d3 c5
lt>e 1 (28 'iii> e 2 'it'b2+) 28 .. .'ii' x d 1 + 29 'iii> x d 1 4 . . . dxe4 5 'Dxe4 'Df6 is also good .
liJxf2+ and 30 . . . 'Dxh 1 , winning the rook. 5 a3
24 . . . l:.c8! Many years later the Soviet master Pavel
The key move, which , of cou rse , had to be Kondratiev i ntroduced an interesti ng gambit
seen beforehand . If now 25 �xb4 , then line: 5 exd5 'ilfxd5 6 �d2 �xc3 7 ..txc3 cxd4
25 .'i!i'a 1 + 26 'ii'd 1 l:.c1 27 'Dc3 l:r.xd 1 + 28
.. 8 �xd4 'iix g2 9 1i'f3 Wxf3 1 0 'Dxf3 f6 1 1 l:.g 1
liJxd 1 tt:lxd5 29 �d2 Wa2 - Black's material 'it>f7 1 2 0-0-0 . For the sacrificed pawn White
and positional advantage ensure h i m a n has a lead in development and the two
easy win . bishops.
25 'Dc6 'iVxb5 5. . . �xc3+
26 �xb4 6 bxc3 c4
26 tt'lxb4 'De4 was bad for Wh ite , as was 26 7 �e2
1 08 � Sensing the Tempo

7 ..tf1 ! was stronger, in order to answer arranged his pawns on light sq uares, to give
7 . . . dxe4 with 8 'iig 4. scope to the dark-sq uare bishop, wh ich has
7. . . dxe4 no opponent. But the opposite has happened
8 ..txc4 "ilc7 - the pawns are fixed on dark squares and
the bishop at b2 has been transformed i nto a
9 "ile2 tLld7
'large pawn ' .
1 0 ..tb2 tLlgf6
14 . . . tLld5
11 f3 0-0
1 5 tLlf3
1 2 ..tb3
1 5 1i'xb5? ttJxc3 (the g2-pawn is under
Give some thought to the situation that has
attack) 1 6 'iff1 'ii'a 5 is clearly bad for White.
arisen.
15 . . . ttJxc3
Black has restored material equality and
seized the i nitiative. This happened because
Ragozin correctly recognised the moment
when it was necessary to delve deeply into
the position and find the correct solution
( 1 2 . . . b5! ) .
1 6 "ild3

As yet the evaluation of the position has not

been determined - now is the time when it
will be decided which of the two players will
seize the in itiative. White is threatening not
only the captu re 1 3 fxe4 , but also 1 3 c4,
activating his dark-square bishop.
12 . . . b5!
An excellent solution . If 1 3 'ii'x b5 there
follows 1 3 . . . a5! (intending 1 4 . . . ..ta6 or 1 4 . . . a4!
1 5 ..txa4 .l:!.b8) 14 'i!fc4 'ii' b 6, creating
dangerous th reats with the wh ite king caught
in the centre.
Incidentally, it would have been very good to
play this a move earlier: in the 5th game
of the match Vyacheslav Ragozin chose
1 1 . . . b5! .
1 3 fxe4 ..tb7
1 4 e5
With the two bishops White should have
16 . . . ..txf3 !
The routi ne 1 6 . . . .l:!.ac8? would have allowed
Wh ite to launch a counterattack by 1 7 tt:\g5!
g6 1 8 0-0 with the th reat of 1 9 'ii' h 3, and if
1 8 . . . h6 there follows 1 9 ttJxe6 . The in itiative
is a fragile th ing - it does not last long, and it
ca n be lost after the very first i naccu racy.
1 7 "ilxf3 l:!.ac8
1 8 0-0 tLl b6
Black is excellently placed . Now he is
intending to advance his a-pawn .
 Sensing the Tempo ltJ 1 09

1 9 l:r.ae1 aS soon come to an end.

20 'ilkh5 [ This note contains a flaw. After 25 ... 'iit h 8 26
20 'i'g4 !? came into consideration , althoug h cxb3 f4 White can play 27 .l:!.g6. In order to
after 20 . . . lt:'lbd5! 2 1 lif3 a4 22 ..txd5 lt:'Jxd5 23 prevent a deadly sacrifice on h6 (which wins
l:ig3 g 6 24 c3 "ike? Black would have against 27 . . . I!xc3, for example) Black must
retained the advantage. defend by 27 . . . .l::t f7. After the continuation 28
20 . . . h3! ltJd5 29 .i.d2! White undoubtedly has
'ii'd 7!
some advantage, since Black is a pawn
But not 20 ... a4? 2 1 .Uf3 .
down. In view of the lavish praise bestowed
21 l:tf4 a4 on Black's moves throughout the game, it
22 .l:th4 h6 would appear that a reassessment is in
23 .l::t e 3 order. - John N u n n .]
23 l:tg4 axb3 24 'ii'x h6 did not work because [It seems to me that the entire game does not
of 24 . 'ii'x d4+ ! .
. . need a reassessment, but only the recom­
23 . . . axb3 mendation 25. . . 'iit h 8: cf. the note to White 's
24 l:tg3 f5 2 7th move - Dvoretsky.]

25 ..txc3 Alas, the concl uding stage of the game was

far from fau ltless . Black won , but not without
How does Black best exploit his advantage?
the help of his opponent.
25 . . . 1J.xc3
26 .l::t x c3 b2
27 'ifd 1 ?
27 l::t b 3 was necessary, since now 27 . . . b4!
would have i mmediately decided the out­
come: 28 axb4 l:.a8 or 28 l:.b3 bxa3 29 l:ixa3
lt:'Jc4 .
[However, after 27 l::. b 3 Black would also
have retained a dangerous initiative, by
continuing 27 . . . ltJa4! 28 'ifd1 .l::t c8, intending
. . . .l::t c4 and . . . 'ifc6, or in some cases . . . lt:'Jc3.
Here is a possible variation: 29 1J.hh3 .l:tc4 30
.l::thd3 'ii c 7! (stronger than 30. . ."ii c 6 31 "ikf3!)
There is no time for 25 . . . bxc2? in view of the 3 1 1J.d2 r1c3! and wins - Dvoretsky.)
threat of 26 "ifxh6, and 25 . . . f4 26 'ii'x h6! fxg3 27 . . . lt:'Jc4?
27 'i'h8+ �f7 28 .l::t f4+ is also bad for Black. 28 l:tb3 "ike7
[In fact this leads to a draw after 28 .. . <�g6! 29 29 llhh3?
l:l.g4+ rtJf7, and if 30 .l::txg 7+ �e8 31 'ii' h 5+,
Appa rently Wh ite could have saved the
then 3 1 . . . 'iif7! 32 'i:!xf7 'i:!xf7 - Dvoretsky.]
game with 29 .l:lf4 ! .l:ta8 30 "ikb 1 followed by
The prophylactic move 25 . . .'ith 8 ! would ltxb2.
have won : if 26 cxb3 f4 , severing the link
29 . . . 'ii'g 5!
between the rook on h4 and the d4-pawn and
preparing 27 . . . l:txc3 on the next move, while Th reatening 30 . . . 'ii'c 1 .
after 2 6 l:tg6 l:txc3 (26 . . . .l:!.f7) 27 l:txh6+ gxh6 30 .l:.hf3 lt:'ld2
28 'i'xh6+ 'iit g 8 29 'ii' h 8+ 'iit f7 the checks If now 31 .l:.g3, then 31 . . . lt:'Jxb3! 32 .l::txg 5 lt:'Jc1 .
1 10 � Sensing the Tempo

31 h4 tt:'Jxf3+ 1 1 �b3 d6
32 l:l.xf3 �c1 1 2 l:l.fd 1 �e7
33 .Uf1 �e3+ 1 3 .l:i.ac1 l:l.ac8
34 .l:i.f2 b4! 14 e3 eS
35 a4 .l:i.d8 1 5 lL'le1
36 'it'b1 .l:i.xd4 A thematic, but com mitting move . White
Wh ite resig ned . wants to attack the b4-pawn with his knight,
in order to resolve the situation on the
q ueenside. But as a result of the exchange of
The last game that we will look at today is
bishops, the wh ite king's position is weak·
perhaps the most important one for our topic.
ened somewhat. However, there are hardly
Its leitmotif is 'change of rhyth m ' .
any black pieces there, so that this factor
Many players remark that i t is most d ifficult
seems purely academic.
for them to fi nd their beari ngs when there is a
sharp change of scene: the transition from 15 . . . �xg2
attack to defence or from defence to attack, 1 6 Wxg2 �b7+
from a position played in accordance with the 1 7 �g1 e4
principle 'do not hurry' , to tempo play and 1 8 lbc2
vice versa , and so o n . Even for lead ing
grandmasters this is sometimes a serious
problem . To make progress i n this field you
must consciously develop and train you r
'sensing the tempo ' .

Tu kmakov - Vitolins
Yerevan 1 980
Bogo-lndian Defence
1 d4 lL'lf6
2 c4 e6
3 lL'lf3 J.b4+
4 �d2 cS
S �xb4 cxb4 Let us dwell on this position a l ittle. Vladimir
Tukmakov has carried out his pla n , and after
6 g3 b6
1 8 . . . bxa3 1 9 bxa3 followed by .l:i. b 1 he will be
7 �g2 �b7 the fi rst to 'latch on' to the opponent, by
8 0-0 exploiting the defects in his pawn structure.
A slight inaccuracy! If Wh ite had played 8 a3 Alvis Vitolins was an u nusual player, excep·
immed iately, the opponent would not have tionally inventive and always seeking the
been able to obtain the position that occurred in itiative. He had a subtle feeling fo r the
in the game. rhythm of a game, and he knew how to
8. . . aS change it. Here too he emerged with honour
9 a3 tt:'Ja6 from a d ifficult position .
10 lL'lbd2 0-0 18 . . . 'it'd7 !
 Sensing the Tempo
By giving u p a pawn , Black essentially burns
his boats behind him. If he does not g ive
mate, he will of cou rse lose. Yet for mate
there would appear to be i nsufficient force.
After . . . ii'h3 and . . . tt::l g 4 there follows tt::lf 1 ,
and what then?
19 axb4
It is not possible to take away the h3-sq uare
from the q ueen - if 1 9 'it>g2 there fol lows
1 9 a4 .
. . .
19 . . . 'i!Vh3
20 bxa5 tt::l g 4
21 tt::l f1 tt::l c 7!
This is the point! Black transfers his isolated
knight to the kingside, where it will take part
in the attack. (The 'principle of the worst
piece' in action - the correct strategy is to
improve the position of the piece that is least
well placed . )
I t i s quite obvious that if the knight should
reach g5, Black's threats will become ex­
tremely dangerous, and probably even irre­
sistible. White must forestall his opponent and
undertake something before the appearance
of the knight on g5. He should have played
22 tt:le1 ! tt::le 6 23 f3 . Then 23 . . . tt::l x h2 24
I/Jxh2 'iNxg3+ 25 'it>h 1 tt::l g 5 (which Tu kmakov
feared, apparently) does not work i n view of
26 l:tc2 exf3 27 tt::l exf3 ! tt::lxf3 28 e4 , when the
pin a long the 3rd ra nk proves decisive. And
CtJ 111
i n the event of 23 . . . exf3 24 tt::l xf3 bxa5 White
remains with an extra pawn , for which the
opponent has insufficient compensation .
2 2 d5?!
I n mechan ically preventi ng the manoeuvre
. . . tt::l c7-e6--g5 , White weakens the i mportant
e5-sq uare.
22 . . . tt::l e 8!
The knight changes cou rse and aims for g4
(after the other knight, by moving to e5, has
made way for it).
23 tt::l d 4 tt::l ef6
Now it was essential to prevent . . . tt::l e 5, by
playing 24 f4! exf3 25 tt::l xf3 . However, after
25 . . . bxa5 the position would remain u nclear
- v u l nerable poi nts appear in Wh ite's posi­
tio n , giving Black q u ite good compensation
for the pawn .
24 axb6?
Tukmakov clearly underestimates the dan­
ger of his position.
24 . . . tt::l e 5
Threatening 25 . . . tt::l f3+ 26 'iii> h 1 tt::l g 4 (or
26 . . . tt::lx h2 27 tt::l x h2 tt::l g 4 ).
25 f4 exf3
26 e4?? 'i!Vg2 mate
Wh ite's last move shows that he was
completely unprepared for the sharp change
in the situation on the board . However, his
position had already become difficult. To 26
l:td2 there is the strong reply 26 . . . tt::l e4 , while
if 26 l:tc2 , then 26 . . . l:!.xc4 27 l:!.f2 (27 b 7 l:!.xc2
28 'i!Vxc2 tt::lfg4) 27 . . . tt::le4 28 tt::l xf3 tt::lxf2 29
tt::l x e5 dxe5 30 'it>xf2 'iNf5+ followed by
31 . . . l:!.c2 .
[In fact in the final position it would appear
that White could have parried the attack and
e ven claimed an advantage. Here is a
sample variation: 26 l:!.c2! l:!.xc4 27 b 7! l:txc2
28 'iNxc2 tt:lfg4 (28. . . ttJe4 29 l:!.a 1 ; 2B. . . 'Wid7
29 'iNb3 l:!.bB 30 l:!.a 1) 29 l:ta 1! f2+ 30 'iii> h 1
tt:lxh2 3 1 tt:lxh2 tt:lg4 32 tt:lf3 f1'Wi+ 33 l:!.xf1
ttJxe3 34 l:!.a 1 ! ttJxc2 35 l:!.aB and wins.
1 12 � Sensing the Tempo

I think that in this game Black attacked really much stronger 1 8 . b5! 1 9 cxb5 1kxb5 20
. .

too recklessly, too riskily. At some point he
should have captured the a5-pawn (for
example, on the 22nd move). And instead of
the clever, but objectively not fully correct
queen manoeuvre ( 1 8 'iid 7?!) he had the
. . .
axb4 axb4. Black stands better - all the
opponent's pieces are cramped, and in
addition he constantly has to reckon with the
switching of the queen to the kingside by
.. ."ilb5-h5 - Dvoretsky.]

Mark Dvoretsky
Tra nsformation of a Position
W problem, which in one form or another
e w i l l now discuss a complicated
we invariably encounter in every game - the
transformation of a position.
What is understood by transformation is a
sharp change in the position , lead ing to a
change in the character of the play, which
usually occu rs with a n exchange of pieces
(often several pieces) and/or a significant
change in the pawn structu re.
During the cou rse of a game the position
changes repeated ly, but by no means every
change is perceived as a transformation .
Sometimes such changes are beyond our
control and depend completely on the will of
the opponent, but at other times they are so
natural , so self-evident, that they are not
perceived as any specific chess tech nique.
In my analysis of various instances of trans­
formation of a position, I should like to beg i n
with a classic example.
Fischer - Petros ian
Final Candidates Match , 7th Game,
Buenos Aires 1 97 1
l2J 113
Wh ite's positional superiority i s obvious. He
has n umerous tempting conti nuations. He
can place his quean's rook on the open file,
bring his king towards the centre by 22 �2 .
or play 22 g4 , creating a th reat against the
d5-pawn .
But if we employ prophylactic th inking , and
ask ourselves about the opponent's inten­
tions (and this is how we should th ink i n such
situations), Wh ite's choice is immed iately
narrowed . It is clear that if it were Black to
move he would play 22 . . . i.b5. It is not hard
to prevent the exchange of bishops by 22 a4 ,
but then there follows 22 . . . i.c6 , th reatening
to exchange knights by 23 . . . lt:'Jd7 . After this
exchange Black gains the opportun ity to
attack the b4-pawn with his rooks.
Robert Fischer found a wonderfu l solution to
the position , which came as a complete
su rprise to the experts gathered in the press
centre.
22 lt:'Jxd7+ ! ! l:r.xd7
23 .Uc1
Commenting on this episode, grandmaster
Lev Polugayevsky remarks:
'Very typical of the present-day Fischer. He
often resorts to the possibility of transforming
one type of advantage into another. '
Why did Wh ite g ive up his fi nely-placed
knight for the passive black bishop? First of
a l l , he eliminates the opponent's hopes of
counterplay. His rooks control the open files
and the enemy rooks do not have anyth ing to
attack. The b4-pawn is invul nerable: . . . .Ub7
is always met by a2-a3; if . . . a6-a5, then b4-
b5. The wh ite bishop remaining on the board
is much stronger than the black knig ht. Both
1 14 � Transformation of a Position

24 i.xa6 and 24 :c6 are th reatened , and if Charush i n - F ra n ke

Black defends the 6th rank with his rook, the European Correspondence Championship
wh ite rook will i nvade on the 7th rank. All 1 979/83
these advantages are q uite sufficient for a
win .
23 . . . .l::t d 6
24 l:tc7 lZJd7
25 l:te2
If the knight moves from d7 there is the
un pleasant reply .l:tee7. There is l iterally
noth ing that Black can move.
25 . . . g6
26 'iiii2 h5
27 f4 h4
27 . . . lZJb6 28 l:.ee7 !tf6 was more tenacious.
28 �f3 ! f5
29 �g4 was threatened . White's position is d ifficult. Both 39 .. .l:tg4
29 �e3 d4+ and 39 . . . tt:Jxg3 40 hxg3 J:txd4 are th reatened.
30 �d2 If 39 lZJf4+? Black does not reply 39 . . .<iti>g5?!
in view of 40 tt:Jxe4+! .l:Ixe4 41 tt:Jxh3+ 'itg4
Zugzwang! Any move will worsen Black's
42 �g2 with a probable d raw, but 39 . . . i.xf4!,
position.
tra nsposing i nto a favou rable rook ending.
30 . . . lZJb6
For example, 40 l:txf4 tt:Jxg3 41 hxg3 .l::txf4 42
31 .l::te e7 tt:Jd5 gxf4 �5 followed by 43 . . . �e4 , and the
32 .l:tf7+ �eB i nvasion of the king is bound to win . Or 40
33 l:tb7 tt:Jxb4 exf4!? l:tg4 (if 40 . . . � there is the strong
34 i.c4 reply 4 1 � h 1 ! ) 4 1 <ifi>f1 tt:Jxg3+ 42 l:txg3 .l:txg3
43 Iixe7 l::tg 2 44 l1xb7 l:txh2 . Here the
Black resig ned .
outcome is still not altogether certa i n , but it is
clear that Wh ite is in serious danger.
Transformation is one of the methods for
Such a transformation of the position did not
converting an advantage. The example we
satisfy the player with Wh ite, and he found a
have examined (like some of those that
more favou rable series of exchanges.
follow) illustrates this particular case. But
one also has to resort to transformation in 39 tt:Jf5! !:tg4+
other situations, for example, in the defence 40 �h1 !
of difficult positions. Of course, not 40 �f1 ? l:.f7 .
40 . . . .l:r.f7
41 tt:Jxd6 .l:.xf3
42 tt:Je5+ �h5
43 tt:Jxf3 tt:Jxd6
44 lZJg1 !
This move certa i n ly had to be foreseen when
 Transformation of a Position ttJ 115

the exchanging operation was beg u n - 61 <it>d6 d4

otherwise the opponent could have played 62 <it>xc6 d3
44 .. Jie4 and 45 . . . t:Df5. However, 44 t:De5! ? 63 'it'b7 d2
l:te4 4 5 'it'g 1 w a s nevertheless possible
64 c6 d1 'ii'
(45 . . . t:Df5 46 'it'f2). Black would probably
have changed the pattern of the play by 65 c7 'ifd5+
45 . . . t:Dc4!? 46 t:Dxc4 l:tg4+! 47 'it'f1 dxc4 66 'it'a7 'it'c5+
(now White has to reckon with 48 . . . .l:tg2) 48 67 'it'b7 'ii'x a5
e4 \ot>g5! , retaining some chances of success 68 c8'i¥
in a sharp rook ending. Draw.
But now it is awkward for Black to defend the
h3-paw n : 44 .. .<it> h4 45 t:Df3+ or 44 . . . l:th4 45
In each of the endings exam ined it was not
l:if2 lbe4 46 l:tf3 'it'g4 47 l::tf4+ 'it'h5 48 l:tf3 .
easy for Wh ite to take the correct decision,
44 . . . .l:.e4 but these d ifficu lties were of a d ifferent
45 t:Dxh3 t:Df5 natu re. In the Fischer game the problem was
46 t:Df2 1 to overcome a psychological prejudice ('good'
White would have lost after 46 t:Df4+? 'it'g4 knig ht, 'bad' bishop) and to assess the
47 l::tg 2+ 'it'f3 48 l:.g5 t:Dxe3 (but not advisabil ity of the u nexpected exchange. I n
48 . . . .l:!.xe3? 49 t:Dg2 ! ) . the second example an accu rate calculation
of the variations is required , and the evalua­
46 . . . .Uxe3
tion of the final positions is made difficult by
47 l:lxe3 t:Dxe3 the fact that it is not easy to picture them in
48 <it>g1 t:Dc2 your mind - they d iffer so much from the
49 t:Dd3 i n itial one. (However, in this particular in­
The knight ending is d rawn. stance this did not cause any particular
problem , since the game was played by
49 . . . 'it'g4
correspondence and it was possible to move
50 'it'f2 t:Dxd4
the pieces).
51 t:Dc5 'it'f4
From this discussion it is apparent how
52 t:Dxb7 t:De6! diverse are the qualities (both chess and
After 52 . . . t:Dc2 53 'it'e2 t:Dxb4 54 t:Dc5 'it'e5 55 psychologica l ) wh ich you have to develop, i n
'it>d2 it is now Black who has to exercise order t o be a b l e t o transform a position to
caution: 55 . . . 'it'd4? loses to 56 h4! . your adva ntage at the appropriate moment.
5 3 'it'e2 'it'e4 Now let us examine a n example of an
54 t:Dc5+ ! ? unsuccessful transformatio n .
Yet another transformation !
54 . . . t:Dxc5
55 bxc5 <it>d4
56 h4 'it'e4
57 <it>d2 'it'f4
58 'it'd3 'it'g4
59 'it'd4 'it'xh4
60 'it'e5 'it'g5
1 16 <;;t> Transformation of a Position

Kupch i k - Marshall leads t o a d ifficult rook e n d i n g : 4 1 .l::tx h3

New York 1 9 1 5 .l::tx g6 . It is the same after 40 l:.g 1 (g5)? l:r.g8
or 40 lig4? lih8! 41 .l:tg 1 l:.g8! .
40 . . . l:th8
Noth ing was ach ieved by 40 . . . .l::tg 8 4 1 ..th5
(or 4 1 i.d3 .l:tg2 42 �c3) 4 1 . . . l:.g2 42 ..tf3.
41 .l::te 1 ! h2
42 .l::t h 1 .l:.g8
4 3 i.h5!
At d3 the bishop is less well placed : 43 .id3?
l:.g2+ 44 �c3 �c7 , and the march of the
black king to g3 is th reatened .
43 . . . .l:.g2+

Black is a sound pawn to the good and he

has excellent winning chances after the
simple 35 . . . .l:tf8 or 35 . . . 'ii'e 5 36 l:.f6 'ife2+ ! .
Frank Marshall took a different decision - he
carried out a complicated combination in­
volving a piece sacrifice, by which he hoped
to obtain a won ending.
35 . . . 'ii'x h2?!
36 ..te4! 'ii'g 3!
After 36 . . .'ti'g 1 37 l:tf8 ! Black again has to
give a check on h2, since after 37 . . .'i!Vxd4?
38 l:.xd8+ �c7 39 cxd4 �xd8 40 ..txg6 he Marshall was aiming for this position when
remains a piece down . 36 .. .'ifd6 37 .l::tf6 also he began his combination . He obviously
leads to a repetition of moves. considered it to be won . This evaluation
37 .l::t g 2 'ir'f4! really is too bol d , especially if one considers
38 ..txg6 that it had to be made nine moves earlier,
The variation 38 lhg6 'ti'xe4+ 39 "ii'xe4 dxe4 when the material balance and the arrange­
40 l:.h6! (but not 40 l:lg4? .l:th8) was also ment of the pieces were completely d ifferent!
sufficient to casts doubts on Marshall's idea . Now 44 �c3 suggests itself, followed by the
I n my view Wh ite has significant drawing removal of the q ueenside pawns from the
chances here, and in any case better than second rank. I n his commentary Marshall
after 35 . . . l:!f8 or 35 . . . 'ii'e 5. g ives this variation : 44 . . . '1t>c7 45 a4 �d6 46
b4 'i!t e6 47 'i!t d3 'iiif5 48 �e3 .l:!.g3+ 49 'it>f2
38 . . . 'ifxd4
l:th3 50 ..td 1 �f4 5 1 �g2 l:l.h8 52 l:txh2
39 cxd4 h3
.l::tx h2+ 53 �xh2 �e3 54 �g2 �xd4 55 �f3
40 l:l.e2 ! 'iit c4 , when Black, i n his opinion , should win.
Only this move (as wel l as 40 .l::t d 2 ! ) enables I am not convinced that this is so (56 b5!), but
White to retain his extra piece. 40 .l::t h 2? i:.g8 ! there is no need to study the concluding
 Transformation of a Position
position , since the entire variation is uncon­
vincing. Instead of 49 �f2 Wh ite plays 49
if3! .l::th 3 50 .ti.b 1 and he is out of danger.
Abraham Kupch ik was apparently too frig ht­
ened by the powerful pawn on h2 and he
hurried across to it with his king , abandon ing
his queenside pawns to their fate .
44 �d3? .ti.xb2
45 ..te2 .t!.xa2
46 .l::!. x h2
Here the th ree pawns are stronger than the
bishop, and this means that the transforma­
tion carried out by Wh ite has worsened his
position . His defence is difficult - he has to
try and halt the advance of the opponent's
pawns and at the same time not forget about
the defence of his weak d4-pawn. I don't
know whether this problem can be solved ,
but at any event Kupchik failed to cope with
it.
46 .ti.a3+ 47 �c2 a5 48 �b2 .ti.e3 49 �c2
...
b5 50 i..d 3 �c7 51 l:!.h8 a4 52 .t!.a8 .tl.g3 53
J::!.a 7+ �b6 54 .tl.a8 'it>b7 55 .tl.f8 b4 56 l:!.f7+
'it>b6 57 .U.f8 .ti.g2+ 58 �c1 a3 59 .tl.b8+ �c7
60 J::!.a8 (60 .ti.xb4 .l::tg 1 + 61 �c2 a2) 60 . . . l:!.g4
61 'it>c2 .l::t x d4 62 ..te2 .tl.e4 63 ..td3 .tl.e3 64
J::!.a 4 c5 65 '.td2 l:!.g3 66 l:!.a5 c4 67 ..tf5 �d6
68 � c8 a2 69 ..tb7 b3 Wh ite resig ned .
Marshall's clever although insufficiently justi­
fied decision is explai ned to some extent by
his romantic natu re - he simply could not
resist the temptation to put into effect a deep
and pretty idea .
Very often players commit similar mistakes
due to lack of composure. They aim to pick
the fru it as soon as possible, not sensing that
it is not yet fully ripe. Such haste once cost
me very dearly.
117
Tsesh kovsky - Dvoretsky
42nd U S S R Championsh i p ,
Len ingrad 1 974
This game was played in the last round and a
win would make me bronze prize-winner i n
t h e championship o f t h e cou ntry. I had
managed to outplay my opponent and gain
an appreciable positional advantage. In
add itio n , Vitaly Tseshkovsky was acutely
short of time - for 1 7 moves he had less than
20 m i nutes left on his clock.
First of all I considered 23 . . . ..txd4 . In the
event of 24 ..txd4 'ifxd4 25 'ifxa7 'ife4+
Wh ite comes out a pawn down . Playing for a
pin does not work: 24 .l'lad 1 ? e5 25 e3 .l'lb6 ! ,
a n d after the q ueen moves Black gives a
check with his q ueen from b7.
I was concerned about the variation 24 .l::i.fd 1 !
e5 25 ..txd4 exd4 26 �xa7. By conti nuing
26 . . . 'i!Vxa7 27 .l:.xa7 c3! Black retains an
advantage, but whether it is sufficient for a
win is not clear.
For an opponent who is in time-trouble the
most u npleasant strategy is defi n itely not
forcing play, which allows h i m q u ickly to
make several obligatory moves. It is better to
force h i m to search, all the time maintaining
the tension and posing one problem after
a nother. From this point of view I should
have simply played 23 . . . .l'lc7 ! .
1 18 <t> Transformation of a Position

When short of time it is al most impossible to occu rred in the game.

find the tactical variation 24 ..ta5!? 'i¥d5+ 27 .tel !
(with the idea of playing the rook to d7) 25 I overlooked this move, or more precisely,
e4! ! 'ifxe4+ 26 f3 . At any event, several not the move itself, but the fact that after it I
precious minutes would have to be spent on do not win a pawn , since if 27 . . . l:txe2 there
it, whereas Black would have the right to follows 28 'ito>f3! and 29 llxe7.
revert to the idea of captu ri ng the d4-pawn
27 . . . e6
after 24 .. J:tcc8 25 .i.c3 .
28 l:!.fa 1 ! .:tbb8
Wh ite would probably have repl ied 24 .l:tfb 1 ,
but after 24 . . . g5! things would not have been 29 .:tc1 .:tc6
easy for him, especially in time-trouble. 25 30 .l:tc3 h5
b6 'ikc6+ leads to the loss of a paw n , and 3 1 .l:ta4 l:.bc8
how otherwise can he parry the terrible 32 f4!
threat of . . . l:.b8-b6-h6 ? I n the event of 25
Wh ite prevents . . . e6-e5 and intends �f2,
'ii'a4 the idea of exchanging pawns by b5-b6
<it>f3 and e2-e4 . The i n itiative is now on his
is no longer possible, and Black can strength­
side, and I have to act carefu lly, to avoid
en his position with 25 . . . 'i!Ve6 or 25 . . . h5
myself ending up in a difficult position.
followed by 26 . . . h4.
32 . . . .:t6c7
I saw the correct pla n , but u nfortu nately here
33 h3 f6 ! ?
I lacked composure and patience. I wanted
immed iately to transform my positional ad­ 34 g4 hxg4
vantage into someth ing more tangible. Alas, 35 hxg4 e5
a mistake crept into my calculations and my 36 f5 gxf5
winning chances promptly evaporated . 37 gxf5 l:te7 ! ?
23 . . . .:r.xb5? ! 38 <it>f3 .:teeS
24 'ii'x a7 'i!Vxa7? 39 llxc4 exd4
If Black had avoided the exchange of queens 40 l:txd4 .l:.e5
by 24 . . . l:!.b7!? 25 'ii'a 4 'ii'd 5+ 26 'ito>g 1 h5, he
41 l:tf4
would still have reta ined the i n itiative.
24 . . . 'iie 6! was even stronger, and if 25 An amusing situation : Wh ite can not streng­
'it'a6? , wh ich I was afraid of (25 e3 is better), then his position , but I do not have a single
then not 25 . . . 'i!Ve4+? 26 f3! , but 25 . . . l:tb6! 26 harmless waiting move - each one involves
'ir'a5 ..txd4! 27 ..txd4 'ii'e 4+ 28 <it>g 1 'i!i'xd4 some concessio n . After 4 1 . . . .:t8e 7 there
with a decisive advantage. However, I follows 42 ..tc5 l:tc7 43 .i.d4, while if
continued playing in accordance with my 4 1 . . . ..th6 42 .l:l.g4+.
plan . 41 . . . �h7!
25 .l:txa7 l:lb3 The lesser evi l !
26 .i.d2! 4 2 <it>f2
Black was hoping for 26 .l:!.c1 ? �xc3! 27 l:!.xc3 This leads to an i m mediate draw, but a lso
..txd4 with an extra pawn in the rook end ing. after 42 lih4+ <it>g8 43 l:th5 .:t5e7 White
26 . . . l:tb2 can not undertake anyth ing.
I could hardly have hoped to win after 42 . . . .i.h6!
26 . . . ..txd4 27 l:.xe7 , but here at least my 43 .l:!.c7+
bishop would not have remained inactive, as The variations are easily calculated : 43 l:!.h4
 Transformation of a Position
�e3 44 l:txe3 l:txe3 45 l:txh6+ �xh6 46
�xe3 'itg5 4 7 �e4 �g4 or 43 l:r.f3 i.xe3+ 44
lkxe3 'iil h 6! with a d raw.
43 . . . �h8!
43 . . . .l:.8e7 was worse : 44 l::t x e7+ :Xe7 45
l:tf3.
44 .l:!.h4 .l:!.xf5+
Draw.
In all the examples we have exam ined the
question to be decided was whether to
change sharply the character of the play, or
maintain the existing situation. But some­
times it is possible for a player to transform a
position in several ways . Grandmaster Boris
Gulko once said to me that he considered
such problems to be the most d ifficult i n
chess, making t h e most severe demands on
a player's mastery, his calculating tech nique
and his depth of positional evaluation .
I will show how Gulko h i mself copes with
such problems. In the followi ng game I
remember how one of his decisions made a
strong impression on me.
G u l ko - Dvoretsky
Vi lnius 1 978
Slav Defence
1 c4 c6
2 tt:'Jc3 d5
3 cxd5 cxd5
4 d4 tt:'Jf6
5 .if4 'ii' b 6
In my preparations for the game I glanced i n
the Encyclopaedia o f Chess Openings ( i n
the first edition t h e corresponding section
was written by g randmaster Alexey S ueti n )
and saw there a recommendation that after
the move order chosen by Wh ite, Black
should respond with 5 . . . 'ii' b 6. Generally
speaking, it is dangerous to trust S ueti n's
assessments - too much hack-work was
produced by his pen . But I considered it
ttJ 1 19
u n l i kely that my opponent would play this
particular system , and so I did not bother to
check the book variations, but simply accept­
ed them.
6 l:!.c1 !
And i m mediately I ra n into a novelty, found
by Gulko at the board . The opening g u ide
only considered strange variations such as 6
'ii'c2? ! lL'lc6 7 e3 i.f5! or 6 1i'b3 1i'xd4. The
rook move to c1 is logical - Wh ite develops a
piece, defends his knight in advance in the
event of the capture of the b2-pawn , and,
i ncidentally, prevents the immediate 6 . . .
'ii'x b2? because of 7 tt:'Ja4 'ii' b 4+ 8 .id2 .
6 . . . lL'lc6
7 e3 iVxb2?!
Consistent: if Black doesn't take the pawn , it
is not clear why his q ueen was developed at
b6. Even so, 7 . . . i.f5 or 7 . . . i.g4 8 f3 i.f5
would have been more cautious.
8 i.d3 i.g4
There is no longer time for qu iet develop­
ment: 8 . . . e6? 9 lL'lb5 i.b4+ 1 0 �f1 0-0 1 1
.l:!.c2 .
9 ltJge2 .ixe2
1 0 i.xe2 ! e5
The transformation of the position caused by
this move proves clearly advantageous to
Wh ite. 1 O . . . e6 ! ? came into consideration , for
example: 1 1 lL'lb5 'iib 4+ 1 2 �f1 .l::t c8! 1 3
tt:'Jc7+ ( 1 3 .l:t b 1 'ifa5 1 4 ltJc7+ .l:!.xc7 1 5 .l:!.b5
iVxa2 1 6 i.xc7 tt:'Je4) 1 3 . . . �d8 14 .l:!.b1 'it'e7
1 5 .l:!.xb7 g 5 ! , or 1 1 0-0 i.e? 1 2 lL'lb5 0-0 1 3
a4 (th reatening 1 4 .l:tb1 'ii'a 2 1 5 lL'lc3 'ii'a 3 1 6
l:tb3) 1 3 . . .'it'b4 ( 1 3 . . . i.b4 ! ? ) 1 4 i.e??! ( 1 4
ltJc7 followed b y l:.b 1 i s stronger) 1 4 . . . ltJe4
1 5 l:t b 1 'ii'd 2! (but not 1 5 . . . tt:'Jc3? 1 6 tt:'Jxc3
1!i'xc3 1 7 .l::t b 3). However, in these variations
Black's position looks uneasy, and I recom­
mend the readers to look for an improvement
i n Wh ite's play - I would not be at all
su rprised if one should be found .
1 1 dxe5 i.b4
1 2 0-0 ! i.xc3
1 20 � Transformation of a Position
1 3 exf6
1 3 l:!.c2 1\Yb4 1 4 exf6 i.xf6 1 5 'il'xd5 is also
not bad . The game Rashkovsky-Arnason
(Sochi 1 980) contin ued 1 5 . . . 0-0 ( 1 5 . . . l:td8?
16 'Wf5! 0-0? 1 7 i.d3) 1 6 i.f3! (nothing is
given by 1 6 i.d6 l:tfd8) 1 6 . . . l:tad8 1 7 11Yf5 ! ,
a n d Black experienced serious difficulties.
13 . . . i.xf6
Wh ite's superiority has become obvious. I n
a n open position h e has two strong bishops,
and the black king is still i n the centre.
Several attractive possibilities immediately
suggest themselves. For example, i.d6 is
tempting, or l:tb1 followed by .l:lxb7. It is also
not bad to capture the d5-pawn with the
queen (after the preparatory 1 4 l:tc2), or with
the bishop after 14 i.f3 .
Most probably there is not just one way for
White to maintain his superiority, but how
does he best exploit the adva ntages of his
position? Gulko had a long think, and d u ring
this time I tried to find an acceptable defence
agai nst the opponent's various attacking
attempts. I decided that to 14 i.d6 I would
reply 14 . . . i.. e 7, although after 1 5 i.xe7
tbxe7 1 6 .Ub1 or 1 6 'il'a4+ Black's position is
unattractive. But 14 . . . .Ud8?! 1 5 "i!Vxd5 ap­
pealed to me even less - later I learned that
the game Matsula-Filipenko ( Krasnodar
1 978) went 1 5 . . . i.e7 1 6 l:.xc6 0-0 1 7 i.d3!
with a decisive advantage for Wh ite.
1 4 i.f3 ! l:td8
The alternative is 14 . . . 0-0 1 5 i.xd5 . Here is
one of the possible variations: 1 5 . . . l:tad8 1 6
e4! ( 1 6 l:. b 1 Wc3 1 7 l:txb7?! i s weaker:
1 7 . . . CiJb4 1 8 e4 CiJxd5 1 9 exd5 'i'c4)
1 6 . . . CiJb4? ! 1 7 .I:I. b 1 'ifa3 1 8 i.c1 'ii'c 3! 1 9
i.d2 CiJxd5 2 0 ..txc3 CiJxc3 2 1 'it'c2 CiJxb1 22
.l:.xb 1 , and White will most probably succeed
in converting his material advantage.
1 5 i.xd5 CiJe7
I was expecting 1 6 'ilka4+ b5 1 7 1\Y'e4 0-0
( 1 7 . . . l:txd5? 1 8 .l:f.c8+) 1 8 i.b3 a5! , when it is
not easy for Wh ite to maintai n his advantage.
Here is a possible variation : 1 9 i.e? (the aS­
pawn is under attack) 1 9 . . . :c8 20 l:tc2 'i'a3
2 1 'iikd 3 (th reatening not only the b5-pawn ,
but also the queen) 2 1 . . . b4 22 it'd? a4! 23
1\Yxa4 Wxa4 24 ..txa4 CiJd5 with cou nterplay.
To my surprise, my opponent allowed me to
exchange his strong lig ht-square bishop,
and i n add ition he sacrificed his a2-pawn.
16 e4! CiJxd5
1 6 . . . 0-0 1 7 i.e? ! .
1 7 exd5 0-0
1 8 d6
1 8 i.e? .U.d7 1 9 d6 also looks good .
18 . . . 'it'xa2
If I am going to suffer, let it at least be for a
pawn !
1 9 d7
 Transformation of a Position tLJ 121

Gulko had ai med for this position, when he 34 .l:i.1 d4 11fc3

took his decision on the 1 4th move. He
subtly judged that, by restricting the black
pieces, the far-advanced passed d-pawn
would more than compensate for the lost
material and prom ise him enormous winning
chances. Only a player of the highest class is
capable of taking such a decisio n !
The conclusion o f t h e g a m e confirmed that
White's choice was correct - I don't know
where Black's defence could be seriously
improve d .
19 . . . 'it'e6
20 !i.e? was threatened .
20 �c7 .1l.. e 5
If 20 . . . a5, then 21 .l:i.e 1 it'a6 22 .1l.. d 6 or
2 1 ."i'b6 22 'i'd5 (22 'it'd6).
..
35 l:.d6 �h7
36 �g2 'ii' b 3
37 .l:i.4d5 'it>g7
38 .l:i.g5
Black resigned .
Attempts to rehabilitate the entire variation
were made by the Soviet master Alexander
Filipenko. He found new resou rces for Black
and several times successful ly upheld his
position i n practice. Other players, using his
analyses, also began employing this system .
M a n y years later, G ulko, w h o knew noth ing
about these new investigations, again ob­
tained the same position with White, this time
against a well-prepared opponent.

21 .1l.. x e5!
G u l ko - Shcherbakov
Of course, not 21 .l:i.e 1 ? ! .1l.. xf4!? (2 1 .. .f6 is
Helsinki 1 992
also possible) 22 .l:i.xe6 .1l.. x c7 , and the
position becomes u nclear. Slav Defence

21 . . . 1 d4 d5 2 c4 c6 3 cxd5 cxd5 4 .1l.. f4 ifb6

'i�Nxe5
5 lDc3 lDf6 6 .l:i.c1
22 .l:i.xb7 a5
23 g3 h5
24 h4 g6
25 l:i.e1 it'c5
26 it'a4 '.t>h7
27 .l:tb5 'i!Vc3
28 .l:i.bb1
White gradually reg rou ps his pieces, movi ng
them closer to the kingside and preparing a
decisive attack on the king . At the same time
he keeps a carefu l eye on the a5-pawn - he
does not al low it to advance and, of cou rse,
he is ready to captu re it at the first convenient
opportu nity.
28 . . . M.b8 6. . . .1l.. d 7 ! ?
29 "ii'e4 l:i.bd8 A recommendation o f Filipenko, w h o consid­
ered this move to be more necessary for the
30 .l:i.bd1 li'b4
defence than my move 6 . . . lbc6 . Wh ite must
31 .l:td4 'i!Vc3 sacrifice a pawn , since after 7 'ikd2 e6 (with
32 .l:i.ed1 "ii' b 3 the threat of 8 . . . lbe4) he has noth ing.
33 �d5 '.t>g8 7 e3 li'xb2
1 22 � Transformation of a Position
8 �d3 e6
9 ltJg e2
9 ltJf3? �b4 .
9. . . 'ii'a 3!
1 0 0-0
Noth ing is given by either 1 0 ltJb5 'ii'a 5+ 1 1
'it>f1 ltJa6, or 1 0 .l:tb 1 �c6 1 1 ltJb5 'ifa5+.
10 . . . a6
This is the idea of the defensive plan
suggested by Filipenko.
I n the game Gleizerov-Filipenko (Kursk
1 987) Wh ite did not even try to solve the
problem of the position, but chose the
passive knight retreat 1 1 ltJb 1 ? , and after
1 1 . . . 'ii'b4 ! 1 2 a3 'ii' b 6 1 3 ltJec3 �d6 he did
not gain any compensation for the sacrificed
pawn .
If 1 1 l:lb1 Black repl ies 1 1 . . . b5 1 2 �c7 �c6!
(of course, not 1 2 . . . i.e7?? 1 3 .l:.b3) 1 3 lib3
'ille 7.
To exploit his lead in development, it is
advantageous for Wh ite to open l i nes.
However, after the immediate 1 1 e4 dxe4 1 2
ltJxe4 ltJd5! he does not ach ieve anyth ing.
As in the previous game, after weig hing u p
the various possibil ities, Gulko chose the
most promising one. He found a way of
advancing e3-e4, without conced ing the
central squares to his opponent.
11 i.e5 ! ! i.e7
Other replies also do not bring any relief:
1 1 . . . i.. c6 1 2 i.. xf6 gxf6 1 3 e4 dxe4 1 4 i..x e4;
1 1 . . . ltJc6 1 2 i.. xf6 gxf6 1 3 e4 dxe4 1 4 i.xe4
with the th reat of 1 5 d 5 .
1 2 e4!
Now after 1 2 . . . dxe4 1 3 ltJxe4 the knight
can not go to d5 - the g7-pawn is under
attack. Wh ite also retains a dangerous
i n itiative after 1 2 . . . ltJc6 1 3 �xf6 �xf6 1 4
exd5 ltJxd4 (or 1 4 . . . exd5 1 5 ltJxd5) 1 5 ltJe4
ltJxe2+ 1 6 'ii'x e2 �e7 1 7 l::t c 7.
12 . . . �c6
1 3 exd5 exd5
A timely and skilful transformation has
enabled Gulko to obta i n an attractive posi­
tion . But now he is again at the cross-roads.
Which is better, to direct the knight via g3 to
f5, or, after exchanging on f6 , attack the d5-
pawn by ltJf4 ? When he showed me the
game, Gulko suggested that the second way
would have been stronger. I ndeed : after 1 4
�xf6!? �xf6 1 5 ltJf4 0-0 1 6 ltJcxd5 (weaker
is 1 6 ltJfxd5?! ltJd7 ! ) Black ca nnot play
1 6 . . . ltJd7? in view of 1 7 l::t c 3 'ii'd 6 1 8 i.. x h7+!
'it>xh7 1 9 .l:.h3+ 'it>g8 20 'ii' h 5 . 1 6 . . . �xd5? is
also bad : 1 7 ltJxd5 'i!i'd6 1 8 'ili'f3 ltJd7 1 9 'ii'f5!
g6 20 'ii'x d7 . There only remains 1 6 . . . 'ii'd6
1 7 ltJxf6+ 'ifxf6 1 8 1i'd2 . It is clea r that Wh ite
has the advantage, but how g reat is it?
 Transformation of a Position tLJ 1 23

In my view, the conti nuation chosen by the 23 t"Llc3 i.c6

grandmaster is no weaker. 24 t"Lle2 i.b5?!
1 4 t"Llg3 ! ? t"Llbd7! 24 . . . i.a4 is more tenacious, but here too
It is not possible to cover the fS-square: Wh ite gains an advantage by playing 25 t"Lld4
1 4 . . g6? 1 5 t"Llxd5 ! . If 14 ... 0-0? Wh ite de­
. .l:r.b8 26 t"Llc6 .
cides matters with 1 5 t"Llf5 l:te8 1 6 t"Llb1 'ii' b4 25 t"Llxb5 axb5
1 7 a3 'i¥b6 1 8 t"Llxe 7 + l:txe 7 1 9 i.xf6 gxf6 20 26 i.xb5 t"Llc5
'fg4+ .
26 . . . :xa2 27 t"Llc3 .l:ta7 28 t"Llxd5 .
1 5 t"Llf5 g6
1 5 . . 0-0? 1 6 t"Llb1 is bad for Black, while if
.

1 5 ..1ib4 there follows 1 6 t"Llxg7+ �8 1 7

. . .

ltlh5.
1 6 t"Llb1 'i¥a4
Now Wh ite restores material eq ual ity and
transposes i nto a favourable end ing. G u l ko
considered the strongest reply to be 1 6 . . . 'ifa5
and after 1 7 t"Llxe7 <it>xe 7 he was intending to
play 18 i.f4 h6! 1 9 t"Lld2 <it>f8 20 t"Llf3 �g7 2 1
ltJe5. Wh ite certainly reta ins excellent com­
pensation for the pawn , but the outcome still
remains u nclear - this is why Gulko had
doubts about his choice on the 1 4th move .
Later grandmaster Viorel Bologan suggest­ 27 f5!
ed strengthening the attack by 1 8 'ii'f3! , for The attack on the king is mai ntained even in
example: 1 8 . . . .l:r.he8 (in the hope after 1 9 the endgame. A sample variation goes
l:tfe 1 ?! of ru n n i ng away with the king by 27 . . . i.gS 28 f6+ <it>h6 29 l:.c3 i.d2 30 .l:r.h3+
1 9 . . �f8 ! ) 1 9 'ii'f4 ! , or 1 8 . . . t"Llxe5 1 9 dxeS
.

�gS 31 l:Ig3+ <it>h6 32 t"Lld4 t"Lle4 33 l:Ih3+

ltld7 20 .l:tfe 1 followed by 'ii'f4 , and Black's <it>gS 34 t"Llf3+.
position is very d ifficult.
27 . . . gxf5
1 7 t"Lld6+! <it>f8
28 t"Llg3 l:Ib8
If 1 7 . . . i.xd6 Wh ite would have interposed 1 8
29 t"Llxf5+ <it>f8
i.c2 ! , and only then played 1 9 i.xd6.
30 t"Llxe7 .l:.xb5
18 'ikxa4 i.xa4
3 1 t"Llxd5 <it>g7
1 9 t"Llxb7 t"Llxe5
31 . . . t"Lld3 would not have helped : 32 .:l.c8+
20 dxe5 t"Lld7
<it>g7 33 t"Llf6 t"LlxeS 34 l:l.g8+ <it>h6 35 h4! .
Black should probably have tried 20 . . . t"Llg4! ?
32 l:lc4 l:.a4
2 1 .l:i.fe 1 (2 1 f4 t"Lle3 22 .l:r.f3 d 4 ) 2 1 . . . i.g5 22
J:tc5 i.f4 . 33 t"Lle3 ! lixa2
21 f4 .l:r.a7 34 t"Llf5+
Black avoided 21 . . . �g7 because of 22 t"Llc3 After playing 34 . . . 'it>g6 in this hopeless
�c6 23 t"Lla5, but here not all is clear after position , Black lost on time.
23 . . . i.b5.
22 t"Lld6 �g7 I should like to show you one more example
1 24 � Transformation of a Position

from the games of Boris Gulko, connected

with the topic u nder discussion.

G u l ko - Kupreich i k
5 2 n d USSR Championsh i p , R i g a 1 98 5
King 's Indian Defence
1 d4 lt:Jf6
2 c4 g6
3 lL'lc3 ..tg7
4 e4 d6
5 f3 a6
6 ..ie3 0-0
- position after 1 3 . . . e5 -
7 'ii'd 2 lt:Jc6
8 lt:Jge2 l:.b8
d6-pawn are attacked . But take note: the
9 l:tb1 position has been opened u p , and White is
A rare pla n . Theory considers the strongest behind in development. In such positions
reply to be 9 . . . b5 1 0 cxb5 axb5 1 1 b4 e5! 1 2 one must be extremely cautious, especially
d5 lt:Je7. when sitting opposite you is such a resource­
9 0 . . ..id7 ful tactician as Kupreich ik. He will most
10 b4 'ii'c 8 probably play 1 S . . . l:te8 ! , i ntending to meet 1 6
If now 1 O . . . b5 1 1 cxb5 axb5 1 2 dS lt:JeS 1 3 ..txd6 o r 1 6 'ii'x aS with 1 6 . . . lt:Jxe4 ! . It is
lt:Jd4 with the better chances for White . possible to defer winning material, by rein­
forcing the e4-point with 1 6 lt:Jf2 . The
1 1 b5 lt:Jas
position after 1 6 . . . lt:Jxc4 1 7 ..txc4 is advanta­
1 2 lt:Jf4! c6 ! ? geous to Wh ite. But the opponent finds a
After 1 2 . . . b 6 1 3 eS!? or 1 3 l:.c1 ! ? c6 1 4 bxc6 bri l l ia nt cou nterattack: 1 6 . . . d S ! ! 1 7 'i'xa5
Black would have stood worse. Therefore dxe4 , or 1 7 ..txb8 'ii'x b8 ( 1 7 . . . dxe4 1 8 i.e5
Viktor Kupreichik i nitiates risky play - he exf3 can also be considered ) 1 8 'i!i'xaS dxe4
abandons to its fate his knight on aS, which with dangerous threats.
now has no retreat square. As we will see , Let us now tu rn to the pawn exchange 1 4
this idea has a clever tactical basis and it is dxeS dxeS. Perhaps here White should
not at all easy to refute . retreat his knight to d 1 ? Let us check: 1 5
1 3 b6! lt:Jd 1 exf4 1 6 ..ixf4 l:te8 1 7 lL'lf2 ( 1 7 'ii'xa5
Stronger than 1 3 eS?! lt:Je8 1 4 b6 ..ie6 ! . lt:Jxe4 ! ) 1 7 . . . lt:Jxc4 1 8 ..ixc4 lla8 1 9 0-0 i.e6.
13 0 . . e5 The next move will be 20 . . . lt:Jd7 , and White
has nothing.
(see diagram)
Let us check 1 S lt:Ja4 (instead of 1 S ttJd 1 )
What position should Wh ite go in for? It is 1 S . . . exf4 1 6 ..txf4 . After 1 6 . . . l:e8?! 1 7 ttJc5
clear that he must attack the knight on aS, the wh ite knight is far more actively placed at
but in retu rn he will evidently have to part cS than at f2 . But on the other hand, the
with his knight on f4 - he does not want to opponent acq u i res an excellent tactical
retreat it to h3! 1 4 lt:Jd 1 exf4 1 S ..txf4 resou rce: 1 6 . . . cS! 1 7 'ifxaS ..txa4 1 8 ..txb8
suggests itself, since both the knight and the ( 1 8 'iix a4 lt:Jxe4 ! ! 1 9 fxe4 'ili'g4) 1 8 . . . lt:Jxe4!?
 Transformation of a Position ttJ 1 25

( 1 8 .'i'xb8 1 9 'it'xa4 l2Jh5!? is also possible,

. . 17 . . . i.. e 6
with fine compensation for the sacrificed 1 8 'ii'c 2 l:td8
exchange ) 1 9 fxe4 'it'g4 , and again Black is The attempt by Black to sharpen the play by
able to in itiate i rrational complications. 1 8 . . . l2Jd5 does not achieve anything: 1 9
It should also be mentioned that the modest exd5 cxd5 20 i.. d 3 'ii'xc2 2 1 .txc2 d4 22
15 <ilh3 allows the opponent to gain good i.. d 2 .txa2 23 .:tb2.
counterplay by 1 5 . . . i.. e 6 1 6 l:!b4 l:r.d8 1 7 'iff2 19 0-0 .txc4
c5! ( 1 7 . . J:td4 !?) 1 8 i.. x c5 l2Jc6 .
20 �xc4 l2Je8 !
It can happen this way: you analyse compli­
The knight heads via e8-d6-b5 to the central
cated variations, and at some point you
d4-point.
grasp the essence of the position and real ise
what you r reply should be. Thus here the 21 "ii' b 3! l2Jd6
solution is to 'deaden' the play- not allow 22 i.. g 5! .Ue8
Black to activate his pieces, especially his 22 . . . .l:r.d7? 23 lL'lc5.
bishop o n g7. 23 .:tbd1 lL'lb5
1 4 dxe5! dxe5 24 lL'lc5 lL'ld4
1 5 l2Jxg6! ! hxg6 25 'ii'c 4 i..f8
1 6 l2Ja4 l2Jxc4 26 i.. e 3 !
1 7 .txc4 Wh ite has retained a positional advantage
White has a spatial advantage, and his and subsequently he successfully converted
knight will reach c5 . it.

Exercises

1 . Wh ite t o play 2 . Wh ite to play

1 26 � Transformation of a Position

3. Wh ite to play 4. Wh ite to play

5. White to play 6. Wh ite to play

 Transformation of a Position
Sol utions
1. Marshaii-Ed.Lasker (New York 1 924 ).
lf the knight retreats, then Black, with his two
strong bishops, will gain excel lent play: 1 9
<ilc3 'i'c5 ! , or 1 9 ltJe3 'iff4.
1 9 e5! �xe5
19 . 'it'c5?? 20 b4 is bad for Black, while after
. .
1 9 . 'i'd7 or 1 9 . . . 'ii' b 8 he will no longer have
. .
sufficient compensation for the pawn .
20 'ifxe5 cxd5
20 'ii'x e5? 21 ltJe7+ .
. . .
21 Wxd6 .l:txd6
22 c5
After retu rn ing the pawn , in exchange Wh ite
has obtained some appreciable positional
gains: control of the d4-point and a pawn
majority on the q ueenside.
22 .l:.a6 23 a4! ? �d7 24 .l:.fd 1 ! .l:.xa4 25
...
J::l.xa4 ..txa4 26 .l:.a1 ..tc6 27 Iixa7 .ll e 8 28 b4
White has an undoubted positional advan­
tage, although it is u nclear whether it is
sufficient for a win.
2. Medn is-Keene (Mannheim 1 975).
White's position is d ifficult, despite the equal
material. After some preparation the central
passed pawns are threatening to beg in
advancing. The best defensive possibil ity is
the sacrifice of a knight for two pawns.
41 .l:.d1 ! ! ltJxb3
42 ltJxe6+ ! .l:.xe6
43 ltJxd5+ ltJxd5
44 .l:.xd5
There is no forced d raw here - Wh ite does
not succeed i n exchanging the opponent's
last pawn . But the ending with rook and g­
and h-pawns agai nst rook, knight and h6-
pawn turns out to be d rawn . This conclusion
is worth remembering - there may be a time
when you are able to save a d ifficult position,
lLJ 1 27
by taking play into such an ending .
44 . . . 'it>c6 45 .l:.h5 ltJd4 46 g4 .l:lg6 47 'it>t2 !
ltJe6 48 'it>g3 'it>d6 49 h4 ri;e7 50 l:tf5 (Wh ite
would have lost after 50 g5? ltJg7! ) 50 . . . ltJg7
51 .l:.a5 l:td6 52 .l:.a7+ ri;JS 53 l:ta8+ 'it>f7 54
l:.a7+ 'it>g6 55 .l:.a3 ltJe6 56 .Uc3 ltJd4 57 .l:!.a3
l:.b6 (th reatening 58 . . . .l:.b3+ ) 58 l:t.a5! ltJe6
59 l:a3 ltJc5 60 l:ta5 .l:.b3+ 61 'it>g2 l:tc3 62
.l:!.b5 llc4 63 l:lb6+ 'it>g7 64 'it>h3 ltJd7 65
.l:.b5 .l:.c5 66 .l:t.b7 lld5 67 l:l.a7 .l:.d6 68 .l:.a5
.l:.d3+ 69 'i;g2 lle3 70 l;la7 l:Ie7 71 'it>g3 rJ;;fT
72 l:ta6 l:te6 73 .l:.a7 l:r.d6 74 l:la5 'it>g6 75
'it>h3! (Wh ite does not want to allow the
knight to go to f6 , and for this he must
prevent it from giving a check in reply to g4-
g5) 75 . . . l:.c6 76 .l:.d5 ltJc5 77 l:ld4 lle6 78
:c4 ltJd3 79 l:ta4 l:te3+ 80 'iti>g2 ltJc5 81 l1c4
ltJe6 82 .l:.a4 l:id3 83 l:ta6 'it>f7 84 .l:.a7+ 'it>g8
85 .ll a 8+ 'it>f7 86 l:ta7+ 'it>f6 87 g5+ hxg5 88
hxg5+ 'it>xg5 89 l:te7 'it>f5 Draw.
3. Geller-P. Littlewood (European Team
Championsh ip, Plovdiv 1 983).
21 ltJb6! ltJxb6
If 21 . . . l:lbd8, then 22 'ikf3 is a satisfactory
reply.
22 ..txe6 fxe6
23 axb6 .l:.xb6
24 ltJe4
With his pawn sacrifice Yefim Geller has
obtai ned a nu mber of important positional
gains: he has left the opponent with a ' bad'
bishop and created pawn weaknesses in his
position . Black faces a d ifficult defence.
24 ... ltJb8 25 'ikg4 ltJd7 26 ..th6! g6 (26 . . . ..tf8
27 l:te3 'iti>h8 28 .ll g 3 or 28 .l:.f3 gxh6? 29
l:.g3) 27 cxb4 cxb4 28 l:.ac1 'ii'd 8?
28 . . . 'i'b7 ! 29 'ikf3 .if8 30 .ie3 l:ib5 was
better, although after 31 l:tc6 ! Black still
experien ces serious d ifficulties.
1 28 <;t> Transformation of a Position
29 i.e3 .l:tb5 30 'ifxe6+ 'it>h8 31 'ifxa6
It is now Wh ite who is a pawn up, and in the
subsequent play he successfully converted
it.
4. Miles-Romanishin (Tilburg 1 985).
White's best saving chance is to transpose
into a heavy piece ending.
36 ltJxe4! 'ifxe4
37 i.xg7 'it>xg7
38 'iff6+ 'it?g8(h7)
39 h3
White's queen is well placed on f6 - it
defends the b2-pawn and ties the rook to the
defence of the f7-pawn , making it hard for
Black to create an attack on the king . And in
the event of the queens being exchanged ,
the famous formula may apply: ' Rook end­
ings are always drawn . '
The tempting 3 6 'ii'f4? chosen b y Tony Miles
was weaker because of the reply 36 . . .'it'd6! .
Now 37 ltJxe4? 'iVxf4 loses immediately, and
37 'ii'xe4?! i.xf6 is also bad for Wh ite, since
after 38 'ife8+ 'it>g7 39 ltJe4 Black has the
decisive 39 . . . 'ii'd 3! 40 ltJg3 h4. There only
remains 37 'ifxd6 ltJxd6 38 i.xg7 'it>xg7 , but
this endgame is much more difficult than the
one with heavy pieces, si nce Black can
improve without hindrance the placing of his
pieces (in particu lar, h is king ).
I n the game there followed 39 .l:.d1 l:!.e7 40
'iitg 1 (40 g4!? ltJc4 41 gxh5 g5!?) 40 . . . lLlf5 41
.l:ic1 'it>f6 42 b4 .Ue5 43 'it>f2 'it>e6 44 g3 ltJd4,
and Black won .
5 . Psakhis-Romanishin (49th USSR Cham­
pionship, Fru nze 1 98 1 ).
It is not easy for White to exploit his
exchange advantage. His king is exposed ,
his pieces are poorly coord inated , and the
opponent's position is solid. For a long time
Lev Psakhis was unable to find a plan which
promised chances of success. And suddenly
it dawned on h i m - he saw a way of retu rning
the extra material and forcibly transposing
i nto an ending with a g reat positional
advantage.
44 'ii' g 2! 'ii'c 7 !
45 'i!i'g3 ! ! 'ii'c 1 +
45 . . . 1!i'xg3 46 fxg3 i.b7 47 l:ta4 is hopeless
for Black.
46 lle1 'ii'x e1 +
47 'it>xe1 i.xg3
48 fxg3
48 . . . i.b7
lf 48 . . . i.d7, then 49 l:ta7 is strong . This move
was also not bad now, but Psakhis decided
to exchange the rooks, since he correctly
judged the bishop ending to be won .
4 9 l:txg8+ 'it>xg8 5 0 a 4 'it>f7 51 a 5 bxa5 52
bxa5 'it>e7 53 'it>d2 'it>d6 54 'it'c3 'iitc6 55
'iti>b4 i.c8 56 'it>a4! (zugzwang) 56 . . . i.b7 57
a6 Black resigned .
6. Grau-Eiiskases (Olympiad , Buenos
Aires 1 939).
Of cou rse, 1 8 dxc5?? 'ii'x d2 is not possible,
while 1 8 lLlb3 cxd4 leads to an eq ual game.
The only way of fig hting for an advantage is
the critical move 1 8 lLle4 , the consequences
of which must be accu rately calculated and
correctly assessed .
 Transformation of a Position l2J 1 29

1 8 tt::l e 4! cxd4 22 l:!.d3! l::t c2 23 i.. xf7 + ! , obtaining a position

1 9 tt::l x d6 dxc3 with the same advantageous material bal­
20 tt::l x b7 l:td7 ance as in the previous variation.
In the opinion of Alexander Alekhine, by Wh ite made a weaker move, and the
attacking the knight Black refutes the oppo­ subsequent play also saw some serious
nent's raid. The world champion g ives the mistakes.
following variation: 21 tt::l a 5 tt::l d 5 ! 22 f3 �e6 18 tt::l b 3?! cxd4 19 tt::l x d4 tt::lf5?
23 e4 tt::lf4 24 .U.xc3 i.. xc4 25 .l:txc4 .l:.d2 with
This loses a pawn . The balance would have
the better chances for Black.
been maintai ned by 1 9 . . . a6 or 1 9 . . . tt::l c 6 .
But White can play more strongly.
20 tt::l xf5 �xf5 21 'ii' a 5! i.. e4 22 'ii'x a7 (22
21 tt::l c 5! .l:tc7
f3? b6! ) 22 .. .'ifg6 23 i.. f1 l:.d2 24 'ii'a 5 lld5
22 l:txc3 .l:txc5 (24 . . . l:!.fd8 25 l:tc8 ! ) 25 'ii' b4 l:tfd8 26 l:.c4??
Little is changed by 22 . . ..l:.fc8 23 l:tac1 l:txc5 A blunder, which changes the evaluation of
24 i.xf7+ <it'xf7 25 l:txc5 l:.xc5 26 l:.xc5. the position to diametrically opposite. After
23 i.. x f7+ <it'xf7 26 a4 or 26 g3 followed by �g2 Wh ite would
24 l:l.xc5 have retai ned his extra pawn .
In such endings a rook and two pawns are 26 . . . i.. x g2! 27 .l:!.g4 (27 �xg2 l:.d 1 +) 27 .. .'iif6
stronger than two minor pieces. Wh ite can 28 l:.b1 i..f3 29 .l:Ig3 l:td 1 30 .l:.xd 1 l:txd1
hope for success. (threatening 31 . . . Wa6) 31 'ilfc4 'ilfc6 32 'ilfxc6
Apart from 20 . . . l:td7, 20 . . . l:td2 21 l:.xc3 llc8 i.. x c6 33 f4 l:ta 1 34 f5 <it'f8 35 f6 gxf6 36 l:th3
must also be considered . Then there follows i.. b 5 37 .l:tf3 <it'e7 38 e4 <it'e6 Wh ite resig ned .
1 30 \t>
PART I l l
Typical Position s
Mark Dvoretsky
Oppos ite-Colou r B ishops i n the Mi d d legame
S u nderstanding of some typical position
uppose that you want t o try a n d gain a n
from the midd legame or endgame. The most
desirable is to have in you r possession a
good book or article, in which it is all clearly
explai ned . But, u nfortunately, there is not a
great deal of material of this sort, and the
standard of much of it is rather low.
The alternative is independent study. Com­
pile as many examples as you can on the
topic that interests you , desirably with good
an notations. Study these examples, analyse
them and try to pick out typical ideas and
make general ising concl usions, which then
may be verified with new examples.
I n practice you usually have to combine both
methods of worki ng. Thus for me a stimulus
for the study of midd legame positions with
opposite-colour bishops was provided by an
article on this topic by grandmaster Vlad imir
Simag i n , publ ished in 1 962 in the magazine
Shakhmaty v SSSR. The grand master's
assessments interested me, although by no
means all the examples seemed convi ncing .
I compiled an extensive amount of material ,
and in particular I used the games of Simagin
himself, who played skilfu l ly with opposite­
colour bishops. As a result I was able to gain
an u nderstanding of the given problem.
The laws which operate with opposite-colour
bishops i n the middlegame and the endgame
are d ifferent, and someti mes completely
opposed . In endings the presence of
opposite-colour bishops usually increas·
es the weaker side's chances of saving
the game ('drawing tendencies'). In the
middlegame, opposite-colour bishops
strengthen an attack and increase its
chances of success, and this also means
the chances of a decisive outcome to the
game.
We will now draw u p a few fu rther general
principles, which will help in the understand­
ing of midd legame positions with opposite­
colour bishops, and analyse some examples
illustrating them.
The Initiative
Play as energetically as possible, and
endeavour at all costs to seiz e the initia·
tive. With opposite- co/our bishops, the
possession of the initiative is a serious
advantage.
I n my book School of Chess Excellence 1 :
Endgame Analysis (p.64) I g ive an ending
from the 4th game of the Alexand ria­
Liti nskaya Candidates match . There , apart
 Opposite-Colour Bishops in the M idd legame
from opposite-colour bishops, each side had
only a rook. Nevertheless , at the basis of the
paradoxical solution found by Wh ite was the
same idea - the striving to avoid passive
defence and at any cost to create cou nter­
threats. If you are not familiar with this end­
ing , I recommend that you take a look at it.
But now we will examine some examples
from the middlegame.
Si magi n-Chistyakov
Moscow Championship 1 946
It is White to move. He would appear to stand
better, si nce his bishop has some prospects ,
whereas in the immed iate future it is dou btfu l
whether the enemy bishop will be able to
take part in the play. However, the wh ite king
does not feel too comfortable, and the
opponent retains counter-attacking possibili­
ties on both wings. Therefore Wh ite must
play accu rately and energetically.
The di rect attempt to exploit the h-file by
playing 26 'it>g3 is ineffective in view of the
reply 26 . . . 'ii'g 6. Perhaps Wh ite should sim­
ply captu re the c4-pawn? Simagin does not
even analyse this move - a general assess­
ment is sufficient for h i m :
The position is s o sharp, that the black pawn
is a 'trifle ', and time should not be wasted on
capturing it.
lZJ 1 31
26 'ii'f2 !
The q ueen is switched to h4. Here it will
safeguard its own king and th reaten the
enemy king from f6. In the event of the
exchange of q ueens on f6, the d iagonal of
the b2-bishop will be extended .
26 . . . c3!?
Black creates scope for his rook. The
i n itiative is more important than a pawn !
I n endings with opposite-colour bishops an
outward ly very similar principle appl ies:
' Even seemingly imperceptible nuances are
often far more important than pawns.' But
there pawns are sacrificed with a completely
d ifferent aim - for the sake of constructing or
destroying a fortress.
It perhaps made sense to activate not the
rook, but the q ueen by 26 . . . 'i!i'g6!? 27 'i!Vh4
fxg4+ 28 fxg4 'i!Vd3+. This would have led to
a sharp situation , d ifficult to evaluate .
27 �xc3 .l:txa4
The f4-pawn is under attack, but Simagin is
not at all concerned about this.
28 'ifh4!
Wh ite prepares a check at f6 or 29 'it>g3.
Black's position is difficult. In Simagin's
opinion , he should have fanned the flame of
a cou nter-in itiative , by harassing the enemy
pieces. He recommends 28 . . . l:ta3 . If the
bishop moves, Black captu res on g4, while in
the event of 29 l:tac1 he can consider . . . b7-
b5-b4 . However, by playing 29 'ii'f6+ .l::t g 7 30
�b4! (weaker is 30 l:tad 1 fxg4+ 31 'it>h4 �d7
32 �b4 .l:txa2) 30 . . . fxg4+ 31 'it>h4 White
creates decisive threats.
28 . . . Itxf4
Now Simagin launches an attack on the king ,
and i n it a very important role is played by his
bishop, which has no opponent. After all, the
offensive is conducted on the dark squares,
which are i naccessible to the opponent's
bishop.
29 'i¥f6+ l:tg7
1 32 � Opposite-Colour Bishops in the Middlegame

30 <t>g3 l:tc4 An instructive exa m ple! Both players fought

3 1 l:tad1 ! energetically for the i n itiative, willingly sacri­
An important tempo! ficing pawns. White's actions were more
vigorous and pu rposefu l , and it was this that
31 . . . i.d7
brought h i m victory.
32 i.d2 !
After the disappearance of the f4-pawn , the
Yaku bovich-5 i magin
c1 -h6 diagonal has been opened , and Wh ite
skilfully exploits this factor. Moscow 1 936
32 . . . ct>g8
32 . . .'i!Vg6 was a tougher defence.
33 i.h6 l:tg6

Who stands better? Black, you say, since the

position of the wh ite king is insecu re? B u t if I
play i.f4 , intending l:txa7, l:tg 1 and i.e5 -
which of the kings will be in danger? The
34 l:!.xd7! bishop at c8 is out of play, White will assail
As one of the classics said : 'A combination i n the g7-point, and the h5-pawn will help in the
such positions is as natu ral as a baby's attack.
smile.' Who is it to move? This is the q uestion which
34 . . . l:txf6 should have been asked at the start. If it were
35 l:tg7+ 'it>h8 White to move, the advantage would be on
his side, but in fact it is Black to move and he
36 exf6 'i!fb8+
immed iately opens lines in the vicinity of the
37 f4 l:tc3+
enemy king .
38 <t>h4 'ii'f8
34 . . . f4!
There was no satisfactory defence against
Of course, this pawn sacrifice suggests itself
the threat of 39 f7 .
even 'on general g rounds' . But Simagin
39 l:txh7+ 'it>xh7 l i n ked it with a combinatio n , lead ing to a
40 i.xf8 ct>g8 forced win .
41 i.e7 fxg4 35 i.xf4 g5! !
42 l:tg1 l:th3+ It is important to obta i n the g6-square for the
43 ct>xg4 rook.
Black resigned . 36 hxg6 .:txf4+!
 Opposite-Colour Bishops in the Midd legame
37 'it>xf4 llf6+
38 'it>g3
If 38 �g5 Black would have decided matters
with 38 . . . �b6 39 'it'c2 �e6 (or 39 . . . l:l.f3
followed by 40 . . . it'f6+) 40 .U.xa? lif5+ 4 1
Wh6 l:th5+! 4 2 Wxh5 'it'g4+, a n d mate next
move.
38 . . . l:txg6+
39 'lt>f3 ii.g4+
40 �g3 ii. e2+
41 'it>h4 'it'b7
42 'it'x e2 'ii'e 7+ !
Wh ite resigned . An attractive attack!
With these first two examples I pay tribute to
Vladimir Pavlovich Simag i n . Many reckon
that there's Kasparov, Karpov and perhaps a
few other g reats, but that all other players are
weaker and therefore u n i nteresti ng. I n deed ,
among the g randmasters of the second and
third echelon there a re some who skilfully
move the pieces, but do not d isplay any
striking creative individuality. However, among
them one also encou nters true artists with
orig inal ideas and deep conceptions, and
one can learn as much from their games as
from the games of the champions. Simagin
was one of these a rtists. I insistently recom­
mend that you make a study of his games.
For a time a book of his selected games was
one that I constantly referred to .
Attack
The correct strategy with opposite- colour
bishops is an attack on the king. M aterial
or positional gains are worth little if your
king is in danger. A ny opportunity to play
for an attack should be exploited.
In the episodes which we have already
exami ned , the game was decided by a direct
attack. I should now like to show you two
more examples from Simagin's games. I n
each o f them the opponent was enticed by
the possibility of winning a pawn on the
ltJ 1 33
queenside, underesti mating the threats to
his own king .
U usi-Simag i n
Gorky 1 954
Black's position is preferable. The mig hty
bishop on d5 is attacki ng the a2-pawn ,
putting pressu re on the kingside and also,
fi nal ly, blocking the d4-pawn . Even so, for
the moment Wh ite is not in any real danger.
For example, he can play 2 1 �a3, defending
the a2-pawn and threatening 22 l2Je4. If
21 . . . 'iff4 , then 22 'ikd2 or 22 'ike5, aiming for
simpl ificatio n .
21 .l:i.a3? ! a5
2 2 �c3?!
Assessing the plan chosen by his opponent,
Simagin writes: 'White decided to win a pawn
on the queenside. Black happily gives up this
pawn and, for a minimal material cost, he
creates a strong attack on the kingside. '
22 . . . e5!
Exploiting a convenient moment, the pawn
adva nces to e4 . Up till now it was difficult for
Black to develop an offensive , since the
opponent was always able to neutral ise the
powerfu l bishop at d5 by f2-f3 . But now the
pawn on e4 will cramp Wh ite on the kingside
and the attacking possibil ities are increased.
Of cou rse, it is hard to establish beforehand
1 34 � Opposite-Colour Bishops in the M iddlegame
whether the attack will be successfu l . But i n
principle this is sound strategy, a n d this i s
how o n e should act when there are opposite­
colour bishops.
23 'Yid2
White would have lost after 23 .l:ixa5? exd4 ,
but it would have been more accu rate to play
the queen to e3 (or d1 ).
23 . . . e4
Apart from moves that are part of a pla n , you
should always be on the lookout for chance
tactical resou rces such as 23 . . . ltJe4 ! ? . After
24 'ife3 ttJxc3 25 'ifxc3 'it'g6 Black's advan­
tage is obvious.
24 h3
White has to prevent 24 . . . ltJg4 . If 24 'iVg5
Simagin was intending 24 . . . e3! 25 fxe3 h6 26
'ifg3 'Yixg3 27 hxg3 .l:ixe3 or 26 'Yif4 li'xf4 27
exf4 lle2 , retaining an advantage in the
ending.
24 . . . lbh5
Since the knight has not been allowed to go
to g4, it aims for f4 , from where together with
the bishop it will threaten the g2-point.
25 l:i.xa5 .l:ixa5
26 �xa5 lLlf4
Threatening 27 . . . ttJxg2 ! .
2 7 'i!Ve3
Wh ite has won a pawn and he has parried
the i m mediate th reats. The knight on f4 is
dangerous, of course, but for the moment the
other pieces are not able to help it: the queen
can not go to g6, and the bishop is ru nning up
against the e4-pawn . Has Black's strategy
proved to be a fiasco?
With opposite-colour bishops you must be­
l ieve in the attacking possibil ities of a
position! Black's pressure on the kingside is,
after a l l , more important than the opponent's
material gains. I n order to i ncrease it, he
needs to drive the queen from the blockading
e3-square, and for this the pawns have to be
i ncluded in the attack.
27 . . . f5!
With the impending th reat of 28 . . .<�'Jxg2! 29
Wxg2 f4. A 'Tal-like' piece sacrifice also came
i nto consideration : 27 . . . ltJxg2!? 28 'it>xg2 f5
with dangerous threats .
28 lbb7?!
An interesting reply. I n the event of 28 ... .txb7?
29 �c7 and 30 �xf4 the attack passes to
White. However, it is not essential to take the
knight.
Available to Gunnar Uusi was a nother clever
resource: 28 �c7! 'Yixc7 29 lbd3. However,
after 29 . . . ttJxg2!? 30 �xc7 ltJxe3 31 lLlf4 (31
lbb4 ! ? ) Black would have retai ned the better
chances in the ending, by continuing 31 .tf7 . . .
32 fxe3 g5.
28 . . . 'it'h6!
29 �d2
29 �c7? lbe2 + ! .
29 . . . g5?
Black could have won by 29 .. .'it'g 6 ! ! 30 �xf4
e3 ( N u n n ) .
30 ttJcs
30 �c7? 'it'b6.
30 . . . 'ii'g 7
31 l::te 1
Wh ite overlooks the knight sacrifice, which
has been in the offing for a long time. But
what was he to do? After 31 'lt>h2 h6
 Opposite-Colour Bishops in the Middlegame ttJ 1 35

(31 . . . l2Jxg2? 32 'ii'x g5) 32 g3 l2Jg6 his Simag i n-Saigin

position would have remai ned anxious. And Vil n i u s 1 958
if 31 h4 the opponent would have had a
pleasant choice between 3 1 . . . gxh4 32 'ii'xf4
e3 and 3 1 . . . l2Jxg2 32 'ii'x g5 l2Jxh4 .

The position looks roughly equal. Black is

hoping to g radually prepare . . . c6-c5. Had
his pawn been not at h6, but at h7, he would
indeed have been q u ite alright.
31 . . . l2Jxg2 ! ?
20 .i.d3!
Here too the q u iet 3 1 . . . h 6 ! ? deserved
Simagin observes that the only defect in the
serious consideration , but Simagin is unable
opponent's set-up is the weak b 1 -h7 diago­
to resist the temptation and he forces events.
nal, on which he can set up a dangerous
32 <i&;>xg2 f4 battery with q ueen and bishop.
33 'ii'c 3 e3+ 20 . . . l2Je7
34 f3? 21 �b1 c5
If 34 �h2, then 34 . . . g4! 35 fxe3 g3+ 36 'it>g 1 22 dxc5 bxc5
f3 (Simag i n ) . However, John N u n n has 23 bxc5 .l:Ixc5
shown that this variation is unconvincing -
24 lDe2 .Uxc1
White saves h imself by 37 e4! f2+ 38 'it>g2
fxe 1 'ii' 39 �xe 1 .ltxe4+ 40 l2Jxe4 l:.xe4 4 1 25 l:txc1 .l::t b 8
'i!fb3+ . Regarding this move, Simagin writes:
34 . . . g4! 'It is possible that Black could still have held
the position, if he had realised that he stood
35 lig1 �xf3+
worse. For example, 25. . . e5 came into
36 'it>f1 g3
consideration, in order to answer 26 "i!id3
37 �c1 g2+ with 26 . . . e4. But he serenely tries to win a
White resigned . pawn, thinking that his position is very good. '
26 'ii'd 3 �b2
27 l:i.d1 'ii'x a3
28 'ii' h 7+ 'it>f7
29 lDf4
Threatening 30 �g6+.
1 36 � Opposite-Colour Bishops in the Middlegame

29 . . . i.f6 ii.g7 35 fxg5 'ii'a 4! ) 33 . . . ii.xg5 34 'ii' h 5+ 'itf6

35 'ii'e 8.
32 i.g6+ cJi;f6
After 32 . . . ltJxg6 33 'ii'x g6+ the e6-pawn is
under attack - this is the idea of the g2-g4-
g5 advance.
33 ltJh5+ rJi;e5
34 'ii'x g7+ rJi;d6
35 'ii'f8
White has broken through , but after 35 . . .
rJi;c7 !? the outcome would still have been
u nclear (pointed out by Artu r Yusupov).
However, the opponent was already in
severe time-trouble.
35 . . . 'i!i'a4
The situation resembles the previous game.
It may seem that Wh ite's offensive has come 36 .:tc1 'ii' b 5?
to a standstil l , because he achieves noth ing The decisive mistake. 36 . . . l:!.c4 ! was essen­
with 30 i.g6+? ltJxg6 3 1 'ii'x g6+ cJi;e7. I n fact, tia l .
as usual, the attack compensates for the 37 ii.e8 ! 'ii'd 3
sacrificed pawn, and there are even two 38 I1c6+ rJi;e5
different ways of strengthen ing it.
39 'ii'g 7+
30 g4!
Black resigned .
It transpires that after 31 g5 either captu re of
the pawn will weaken the defence. Also good It is curious that many years later a similar
was the manoeuvre 30 ii.d3!? followed by 3 1 strategic situation occu rred in the 4th game
ii.e2 a n d 3 2 ii.h5+. Even so , against of the second match for the world champion­
accurate defence by the opponent White can ship between Anatoly Karpov and Garry
hardly hope to win - there are too few Kasparov.
attacking pieces left on the board .
30 . . . Karpov-Kasparov
.:tb4
30 . . . 'i!Vd6! came into consideration, for ex­ Moscow 1 985
ample: 3 1 h4! (renewing the threat of g4-g5)
3 1 ... i.xh4 32 ii.g6+ ltJxg6! (it is hopeless to
play 32 . . . rJi;f6? 33 ltJh5+ rJi;g5 34 'ii'x g7 , or
32 . . .<ifi>f8? 33 .l:!.c1 ! with the th reat of 34 .:tc7 ! )
3 3 'ii'x g6+ cifilg8 34 ltJxe6 "fie? 35 l::t x d5 'iff7 ,
and Black maintains the balance.
31 g5! ii.xg5
It is also not easy to defend after other
continuations: 31 . . . hxg5 32 'ii' h 5+ g6! 33
ii.xg6+ ltJxg6 34 'ii' h 7+ rJi;e8 35 ltJxg6 'ifa4
36 'iig 8+ rJi;d7 37 lDf8+, or 31 . . . .l:!xf4!? 32
exf4 hxg5 (32 . . .'iff3 33 l:.d3 'ii'g 4+ 34 l:.g3 ! )
3 3 fxg5 (weaker is 33 'ii' h 5+ g6 3 4 'ii' h 7+
 Opposite-Colour Bishops in the M iddlegame ttJ 1 37

Black could have equal ised by playing 24 'ii'd 3 'i!th8?!

20 . . . �xd4! 21 l:.xd4 l:tdc8 (with the threat of It is evident that Kasparov does not properly
22 . . . l:tc2 ) . For example: 22 .l:tfd 1 ? ! l:.c2 23 understand the position. After a l l , sooner or
l:!.4d2 �f5 , or 22 .l:!.d2 'ii' b4 23 l:.fd 1 (23 'ikd 1 ? later Wh ite will set u p the battery on the b 1 -
l:!.c1 ) 23 . . . 'ii'x d2! (23 . . . llc1 ! ? ) 24 .:Xd2 l:.c1 + , h7 diagon a l , a n d then the king will have to
or, finally, 2 2 �d3!? .l:lc5 23 h3 ( 2 3 f4 �d7 ! ) exit hastily from the corner.
23 . . . 'ii'c7 followed b y . . . .l:!.c1 . 25 l:lfd 1 a5
20 . . . :deS?! 26 b3! .l:.c3
2 1 ltJxe6! 27 'ii'e 2 lif8 ! ?
I n the event of 21 . . . 'ikxe6 ! ? the d 5-pawn is 2 8 �h5!
rather weak, while 21 .. .fxe6 leads to the At last the bishop switches to its 'lawful'
structure with which we a re already familiar. diagonal.
21 . . . fxe6 28 . . . b5
Kasparov probably did not know the Sim­ 29 .i.g6 il.d8
agin-Saigin game, and therefore he under­ 30 il.d3 b4
estimated the threatened attack on his king
31 'ikg4 'ili'e8
along the light sq uares and exaggerated the
32 e4! il.g5
role of his pressu re on the q ueenside.
To me, this move seems not altogether
Objectively speaking, Black can still hope for
logica l . It would appear that, by placing his
a draw. I n the previous game a significant
rook on f8 and bishop on dB, in reply to the
role in the offensive was played by the wh ite
inevitable e�4 breakth rough Kasparov
knight, but here there are no knights on the
was preparing to play . . . il.b6, to develop
board and therefore it is easier to defend . I n
counter-pressure on the f2-poi nt. But sud­
the event o f the d i rect 2 2 'it'g6? 'ii'a 5 Wh ite
denly the bishop occupies a completely
loses a pawn, without creating any serious
different diagonal! The grandmaster was
threats in retu rn . Karpov resorts to more
probably tempted by a simple trap: 33 .l:te2??
refined strategy.
l:.f4 .
'What is required of White is systematic
33 .l:.c2
play, the essence of which can be described
as follows: the consolidation of his position
on the queenside, the switching of his queen
to the kingside, the opening of the position
by e3-e4, and only then the mounting of an
attack on the light squares, making use of
the now open e-file. In the game Karpov
skilfully put all these ideas into practice, but,
of course, not without substantial 'help ' on
my part. ' (Kasparov)
22 .ltg4!
Weaker was 22 l:ifd 1 ? 'it'b4 , intending
2 3 . . . 'ii'x d2 .
22 . . . l:tc4
23 h3 'it'c6 33 . . . lbc2?
1 38 <;t> Opposite-Colour Bishops in the Midd legame
A serious positional mistake . By exchanging
his active rook, Black submissively con­
cedes the in itiative to the opponent. He
should have exchanged not the rooks, but
the queens. Since if 33 . . .'ilkf7 there is the
reply 34 l:te2 ! , he should have chosen 33 . . .
'ii'c6 o r 33 . . . 'ii'c8 . For example: 3 3 . . . ii'c8 ! ?
34 exd5 exd5 (34 . . . l:.xc2? 35 1Ve4) 35 'ii'xc8
l:tfxc8 36 l:te2 l:Ic1 37 l:l.xc1 :Xc1 + 38 'itt h 2
.l:tc8 39 .tg6 .tf6 with an i nferior, but
probably tenable ending .
34 �xc2 'ii'c 6
35 �e2 'ii'c 5
36 l:tf1 'ii'c 3
37 exd5 exd5
38 Si.b1 !
The triumph of Wh ite's strategy - the queen
inevitably reaches the b 1 -h7 diagona l . I will
give the rema ining part of the game with brief
notes.
38 ... 'ii'd 2 39 'ii'e 5 .ll d B?! (39 . . . �f6!? 40 'ii'f5
'it>g8) 40 'ii'f5 'it>gB 41 'ii'e6+ 'it>h8 (4 1 . . . 'ittf8
42 .tg6 'i¥f4 43 .l:.e 1 ) 42 'ii'g 6 'it>g8 43 1!i'e6+
'it>h8 44 Si.f5! (43 .:f.e 1 l:tf8 ! ) 44 .. .'ii'c 3 45
'iig 6 'it>gB 46 Si.e6+ 'it>h8 47 .tf5 'itt g 8 48 g3!
'it>f8 49 'it>g2 'ii'f6 50 'ii' h 7 'ii'f7 51 h4 ..td2
(otherwise 52 l:te1 ) 52 l:td1 .tc3 53 l:.d3 lld6
54 .l:.f3! (54 l:te3? g 5 ! ) 54 ... 'it>e7 (54 . . . .:tf6 55
.l:!.e3 l:txf5 56 'ii' h 8+ 'iig 8 57 lieS+) 55 'i¥h8
d4 56 'ii'c 8 l:tf6 57 'ii'c 5+ 'it>e8 58 :t4 'i!i'b7+
59 l:te4+ 'it>f7 (59 . . . .l:!.e6! ? would not have
helped in view of 60 'S'c4! lbe4 61 'ii'g 8+
'it>e7 62 'iixg7+) 60 'ii'c4+ 'it>f8 61 .th7! l:tf7
62 'iie6 'S'd7 63 'ii' e 5 Black resig ned .
Arrangement of the pawns
The stronger side should (as in the end­
game) arrange his pawns on squares of
the colour of the opponent's bishop. But
for the weaker side, the endgame recom­
mendation (to place the pawns on squares of
the colour of his own bishop) is no longer
suitable. There this was done in order to set
up a fortress. If his bishop and king are
defending all their pawns and blocking a
passed pawn of the opponent, he does not
have to fear the penetration by the e nemy
pieces on sq uares of the opposite colour.
Of course, in the m idd legame this logic does
not apply. The pawns should cover the
squares which are not controlled by the
bishop. It is clear that, if Kasparov's kingside
pawns had been standing on l ight squares
(g6 and h7), he wouldn't have had any
problems. I should also mention the typical
regrouping of pawns on dark sq uares (g2-g3
and h3-h4) carried out by Karpov.
And here is another usefu l observation . The
flexibility of the pawn structure and the
presence of a mobile pawn chain can deci­
sively influence the evaluation of a posi­
tion. A pawn storm gains significantly in
strength if it is supported by an active
bishop.
Botvi n n i k-Tal
World Championship Retu rn Match ,
3rd Game, Moscow 1 96 1
Is it worth captu ring the b7-pawn? We have
already gained sufficient experience with
opposite-colour bishops to decide immedi­
ately: not, it is not worth it. The captu re on b7
merely leads to a loss of time, wh ich the
opponent will exploit to create counterplay.
For example, he can choose 35 . . . d3!? 36
 Opposite-Colour Bishops in the M idd legame
�xd3 (36 'i!Ve3 is stronger) 36 . . . 'ii' b 5 37 .td5
c6, and Wh ite can save his piece only by
playing 38 a4 ! . Of cou rse, he has no reason
to go in for such compl ications.
35 .tc4!
'It only remains for White to place his bishop
at d3, after which his pawns will begin
adva n cing . The capture of the b 7-pawn
would merely have diverted him from this
plan. ' (Botvinnik). Concise and to the point!
The bishop looks to be wel l placed on d5, but
that is all. By playing it to d3 (from where it
will be eyeing h7) and advancing f2-f4 , e4-
e5 and g4-g5 , Wh ite will cramp the enemy
pieces and then switch to a direct attack on
the king , exploiting the open h-file.
35 . . . c5
36 b5 .tf6
37 f4 d3
It was for the sake of this pawn sacrifice that
Black's two preced ing moves were made.
Mikhail Tal wants to exchange a pai r of rooks
and establish his bishop on the secu re
square d4. However, this does not ease his
position . The bishop on d4 is fi ring into
space, whereas the white bishop is th reaten­
ing the kingside.
38 .U.xd3
Less good was 38 .txd3 .U.d4 , when the
black rook is active . One rook is all Botvi nnik
needs for his attack.
38 . . . .U.xd3
39 .txd3 .td4
40 e5 g6
41 .U.h1 'it>g7
42 'ii' e4 b6
43 .tc4
A check on b7 is th reatened ; if 43 . . . 'i!Vd7 the
simplest is 44 'i!Vc6 �xc6 45 bxc6 �c8 46 e6.
If instead 43 .. .'ii' e 7, then 44 g5 (with the idea
of 45 'it'c6 and 46 'ikf6+ ! ) 44 . . . .U.c8 45 f5 gxf5
46 l:.xh7+ ! �xh7 47 "it'h4+ and 48 'ii' h 6 mate.
Black resigned .
tZJ 1 39
Bad bishop
Here the concept of 'bad bishop' has a rather
d ifferent meaning than usua l . Botvinn ik's
bishop was good , because it was attacki ng
the kingside, whereas Tal 's bishop was bad ­
it was not creati ng any counter-th reats. Thus
the main thing for a bishop is the pros­
pect of taking part in an attack, and often
this factor decisively influences the eval­
uation of the position.
A pawn obstructing the action of its own
bishop can be a serious defect in a posi­
tion.
Spass ky-Simagin
28th USSR C h ampionsh i p , Moscow 1 96 1
How should the position b e assessed?
Simagin is playing Black, and when there are
opposite-colour bishops we are already
accustomed to the advantage always being
on his side. Firstly, he has mobile pawns - all
the time the opponent has to reckon with the
th rusts . . . h5-h4 and . . . e5-e4. There is also
a second, exceptionally important factor:
Wh ite has an obstructi ng pawn on d 5 , on a
sq uare of the colour of his own bishop. It
would be better if it were not there at all! It is
absolutely clear that the bishop on c6 has no
prospects. But replace it at g2 - and it is also
doing noth ing here.
How can Black strengthen his position?
1 40 � Opposite-Colour Bishops in the Middlegame

36 . . . h4? 37 g4 is prematu re. Simagin finds
an excellent plan: he switches his bishop to
c7, setting up a battery which will th reaten
the wh ite king . Then the pawn breakth roughs
will become more dangerous.
36 . . . j_d8 !
37 bxc5
White cannot get by without this exchange:
attack the weak a 3- and c4-pawns. And if the
wh ite rook goes to their defence, the
exchange sacrifice on f3 then becomes
possible.
47 .l:tb1 'it>h6
Already now it was possible to give up the
exchange: 47 . . Jbf3! 48 'i!Vxf3 (48 'it>xf3
'i!Vf5+ and 49 . . . 'i!Vxb 1 ; 48 .l:tb7+ l:l.f7) 48 e2
. . .

from c7 the bishop will be defending the e5- 49 .l:!.b 7 + 'it>h6 50 'ii'f8+ �g5 5 1 h4+ '.tg4 52
point, and the b4-pawn will be attacked . 'i!Vf3+ 'it>xh4 53 .Uh7+ 'it>g5 and wins. Howev­
37 . . . bxc5 er, this possibility will never ru n away from
Black.
38 �b1 ?! j_c7
48 .i::f. b 3 j_d2
39 j_a4
49 .l::t b 6
Boris Spassky tries to include his bishop in
the defence. Now, when the bishop is still on
its way to the kingside and the black pieces
have already taken up ideal attacking posi­
tions, it is the right time to break through the
opponent's defences.
39 . . . e4!
40 dxe4 fxe4
41 .Uxf7+ .l:txf7
42 j_d1
42 iVxe4 'i!Vxg3 43 iVg2 'i!Ve3! is also
hopeless for White - Black effectively has an
extra piece, since the enemy bishop is taking
no part in the play. For example: 44 .Ug 1 �h8
45 .Uf1 (defending against 45 .. J::tf2) 45 . . . .Uxf1 + 49 . . . .Uxf3 !
46 'i!Vxf1 'it>g7 4 7 j_d7 'i!Vd2 ! 48 'ii'g 1 'it'e2 ,
50 .Ue6
and Wh ite is completely helpless (variation
50 'it>xf3 'i!Vh5+; 50 'i!Vxf3 e2.
by Simagin).
50 . . . 'i!Vxe6
42 . . . e3
51 dxe6 .Uf2+
Th reatening 43 . . . .l:tf2 .
52 'it>xg3
43 �f3 h4!
52 'i!Vxf2 gxf2 53 e7 e2 or 53 \t>f1 'it>g7.
Black has a decisive attack. 44 g4 or 44 gxh4
is not possible because of 44 . . . .l:txf3 . 52 . . . l:l.xe2

hxg3 53 e7 .l::f. g 2+!

44 l:!.f1
'i!Ve5 54 '.txg2 e2
45 'i!Ve2
55 e8'i!V e 1 'i!V
46 'lt>g2 j_a5
5 6 'i!Vf8+ 'lt>h5
Spassky has somehow managed to set up a
barricade, but his position remains lost. After 57 'i!Vxc5+ j_g5
playing his bishop to d2, Black will then Wh ite resigned .
 Opposite-Colour Bishops in the M iddlegame
Glushniov-Sakharov
U S S R 1 96 1
Here it i s more d ifficult to evaluate the
position. (Simag i n was not playing either
White , or Black, so we are forced to proceed
from other considerations).
The black bishop has more of a futu re. I n the
event of a pawn exchange in the centre it will
be able to press either on c2 , or on g2. There
is also a resou rce such as . . . h7-h5-h4 . Of
course, on no account should f4-f5 be
allowed - then the opponent's bishop will be
activated .
But in principle it is hardly correct to try and
evaluate such a dynamic position on general
considerations alone - one must look specif­
ically to see what may result from it.
Now 22 exf5 is th reatened , and if 21 . . . 'i!i'f7
there follows 22 lt:lf3 , forcing the reply
22 . . h6.
.
I n the game a highly non-routine move was
made.
21 . . . 'i!i'f6!
In the event of the q ueen exchange on f6
Black acquires a nu mber of useful moves
which strengthen his position : . . . 'it>f7, . . . .ll b 8,
..Jig S , and . . . h7-h5-h4 .
22 e5 dxe5
23 fxe5 'ii'g 5
The situation has become sharper. Black is
ltJ 141
in danger of losing his weak c5-pawn , but he
is hoping to develop a n attack on the
opponent's king . White has a weak pawn on
e5, obstructing his own bishop, knight (after
24 lt:lf3) and rook. 'We already know that an
obstructing move on a square of the colour of
a bishop creates nothing but problems. '
(Simag i n )
24 lt:lf3 'ilkg4
25 b3
Wh ite rejected 25 'ili'xc5 because of 25 .. .f4
26 ..tf2 ..id5, but he was wrong to do so. As
g randmaster Sergey Shipov rightly pointed
out, after 27 'ii'c3 he would not have stood
worse - the opponent has no immediate
threats, and the e-pawn is now ready to
advance. For example, 27 . . . lt:lf8 28 e6!
lt:lxe6 29 .l:.e5, or 27 . . . .i:te8 28 h3! 'i!i'f5
(28 . . . 'ii' h 5 29 e6) 29 lt:ld4 ! ? 'ilkg5 30 lt:lf3 .
25 . . . h5
25 .. .f4 26 ..tf2 ..id5? does not work, because
the knight on f3 is defended and White has
27 e6. The bishop must blockade the pawn
for the present, and move to d5 only at the
appropriate moment.
26 lt:lh4?!
By tactical means Wh ite manages to ex­
change a couple of pieces, but this does not
bring any rel ief, since the black rook breaks
through to d2, strengthening the attack on
the king . Here too 26 'ii'x c5 was correct,
since 26 . . . h4? can be met by 27 lt:lxh4! lt:lxh4
28 .i:tf4 .
26 . . . lt:lxh4
27 l:tf4 'i!i'g5
28 .i:txh4 .i:td2 !
Th reatening 29 . . . ..id5.
29 l:tf4 .l::t a d8
30 'ili'xc5
Now this pawn-g rabbing involves a loss of
time. 30 h4 was safer.
30 . . . h4!
3 1 lhh4 .l:r.d1 !
1 42 � Opposite-Colour Bishops in the M iddlegame

31 . . . i.d5 suggests itself, but after 32 e6! the Attack on the long diagonal
wh ite pieces come al ive. Therefore the
We will begi n with a rather simple example.
bishop remains on its blockading square to
the end of the game. Thus the threat of
. . . i.d5 remained (according to the i ronic Perl is-Marshall
defi n ition of Bobby Fischer) 'an eventual Vien na 1 908
possibility' .
3 2 �b4?
Wh ite parries the threat of 32 . . J:txe 1 + 33
i.xe 1 �d 1 in the most unfortunate way.
Meanwhile, in the event of 32 �c3 or 32 �f1
the outcome would have remai ned u nclear.
32 . . . iVe3+
33 i.f2 �8d2 ! !
I n this hopeless position (the black king
hides from the check q ueens at g6) Wh ite
lost on time.

We have probably analysed all the most

general principles of playing midd legame Both players are pressing on the long
positions with opposite-colour bishops. After diagonals, but, of cou rse, Wh ite has a great
assimilating these ideas and getting a feel for advantage: his rook is active, and his pawns
the spirit of such positions, you will certainly on the c-file can be used to divert the enemy
be able to find you r way more confidently i n pieces.
them a n d successfu lly solve specific prob­ 37 c4!
lems facing you . In order to assimilate the
As usual when there are opposite-colour
topic better it makes sense to analyse a few
bishops, pawns do not cou nt. It is important,
more practical examples, to check whether
if only for a moment, to free the q ueen from
the laws already studied apply, and to seek
the defence of the g2-point.
new ones.
37 . . . �xc4
It is advisable also to pick out and examine
If 37 . . . i.xc4 , then 38 �d8 is immediately
the patterns which occu r most often with the
decisive.
given material. The chief of these are:
38 �f6 iVa2
1 ) attack on the g7- (g2) point, as in the last
example, or attack on the long diagonal in 39 i.b2!
the absence of the g7- (g2) pawn , as in Again th reatening 40 �d8 followed by 4 1
Simagin-Chistyakov, the fi rst game exam­ iVh8+ ! , a s well a s 4 0 c6 . Black can resist,
ined by us; only by pestering the opponent with counter­
2) attack on the f7- (f2 ) point; threats . The i n itiative, first and foremost -
neither side can delay!
3) King's I ndian structure.
39 . . . �c4
40 iVe5 i.e6
41 �d8! f6
42 ifc7+
 Opposite-Colour Bishops in the M iddlegame tLJ 1 43

42 'i!fxf6 is also good . After 42 . . .'ii' xf4+ note 33 l:.e3 !

the typical king manoeuvre, with which it T h e rook must b e i ncluded i n the attack o n
diverts the enemy q ueen from the necessary the g7-point. This can not b e done on the g­
trajectory and avoids perpetual check: 43 file (33 l:tg3 l:.g6), and so Tigran Petrosian
'it>g 1 ! 'ii'e 3+ 44 �h 1 ! 'i!fe 1 + 45 'it>h2. plans an i nvasion on the e-file (.i.d4, 'i!Ve2
42 . . . l:Ig7 and l:.e7).
After 42 . . . .i.f7 White wins most q u ickly by an 33 . . . .Ug6?
already familiar manoeuvre: 43 l1xg8 'it>xg8 Now the black bishop is completely shut out
44 'ifd8+ 'it>h7 45 'ii'xf6 'ilfxf4+ 46 'it>g 1 'ii'e 3+ of the game. The pawn sacrifice 33 . . . d4!
47 'it>h 1 'ilfe 1 + 48 'it>h2 'it>h6 49 h4. suggests itself, for example: 34 .i.xd4 l:tg6
43 .i.xf6 ! 35 f3 .i.d5.
Black resig ned . 34 .i.d4! �h7
35 �c2
Petrosian-Pol ugayevsky In the event of 35 'i!Ve2 Black still has the
4th Match Game, Moscow 1 970 defence 35 . . . �c7 ! , so therefore Petrosian
threatens to seize the c-file with 36 I:tc3 . If
35 . . . l:.c6 , then 36 ike2 ike? 37 l:!.e5 l:tc2 38
'i!Ve3 is now decisive . 35 . . . l:!.e6 36 l:!.xe6
'ii'x e6 37 ike? is also bad for Black.
35 . . . 'ii'd 7
36 �h2!
There is no reason to hu rry - the opponent
simply has noth ing he can move. Besides,
now the threat of 37 1i'e2 gains in strength ,
since if 37 . . ."ilc7 there follows 38 l:te5 , and
there is no check on c1 .
36 . . . .i.c8?!
37 .:tel !
The rook fi nally breaks through onto the 7th
29 .i.e5!
ran k (the opponent cannot reply 37 . . . .l::t c6). A
The ex-world champion readily al lows the
triumph of flexible manoeuvring, typical of
opponent to exchange one of his bishops,
Petrosian's play!
since there will be no way of cou nteri ng the
37 . . . .i.a6
pressure on the g7-point.
38 .l::tc 7 'ili'e6
29 . . . l':tc8
ltJxd3 39 g4!
30 'ii' b 2
With the opposite-colour bishops Black has It would hardly have been possible even to
almost no chance of saving the game. But contemplate this move, if Black had got rid of
things would hardly have been any better for his d-pawn at the right time.
him after 30 . . . ltJe6 (in the hope at some point 39 . . . .i.f1
of playing . . . d5-d4) 31 'i!i'b 1 ! . 40 1i'xf5 11i'xf5
3 1 l:txd3 l::t c 6 41 gxf5 llg2+
32 h3 h6 42 'iit h 1
1 44 � Opposite-Colour Bishops in the M iddlegame

Black resig ned , because there is no defence better. He has been able to provoke a2-a3,
against the th reats of 43 f6 and 43 l1c1 . and now the wh ite rook has to defend the a­
pawn . The knight has occupied the strong
I now wa nt to show you two of my games, e4-poi nt and will soon be supported by . . . f7-
played with one and the same opening f5 . The exchange on e4 is u nfavourable for
variation. And although the positions arising Wh ite , whereas after tLlf3-e5 he has to
were nearly identica l , the character of the reckon seriously with the exchange on e5,
play in them was diametrically opposite. since then Black can occu py the c5-square
Everything depended on which of the players with his knight. Which is what happened in
was able to seize the in itiative. the game.
1 2 lLle5? i.. x e5!
Vi kulov-Dvoretsky 1 3 dxe5 lLlxd2
Moscow Championship Semi-Final 1 97 1 14 'ilfxd2 lLlc5
Queen 's Indian Defence 1 5 �c2 dxc4
1 d4 lLJf6 1 6 bxc4 .l:tfd8
2 lLlf3 e6 1 7 �d4? !
3 c4 b6 I n the event of 1 7 'ii'e 2 ..ie4 the kn ight is
stronger than the passive bishop on b2. But
4 e3 �b7
now Black reaches a favourable position
5 �d3 i.. b4+
with opposite-colour bishops .
The idea of this check is to lure the knight to
17 . . . lLle4!
d2, so that it should not occupy the best
1 8 �xe4 �xe4
square c3. Theory recommends 5 . . . d5 or
5 . . . c5. 19 f3 �b7
6 lLlbd2 0-0 1 9 . . . �g6! ? followed by 20 . . . c5 also came
into consideration .
7 0-0 d5
20 'ilfc2
8 a3 ..id6?!
8 . . . �e7 is preferable. Where does this tell?
Firstly, after 9 'ilkc2 ! ? lLlbd7 9 e4 dxe4 1 0
lLlxe4 the bishop would be better placed at
e7 rather than d6. Secondly, after 9 b4 ! ? c5
1 0 cxd5 it is desirable to capture on d5 with
the queen.
9 'ii'e 2?
But now my open ing set-up proves com­
pletely justified .
9 . . . lLle4!
1 0 b3
1 0 b4!? c5.
10 . . . lLld7
1 1 �b2 'ilie7 Black effectively has an extra pawn on the
Black can be pleased with the outcome of q ueenside, but the d ifference in the placing
the opening - he already stands slig htly of the bishops is even more important. My
 Opposite-Colour Bishops in the Midd legame ltJ 1 45

bishop is pressing on the kingside, and it can 26 . . . Ii8d 7 ! . I delayed and played someth ing
also attack the c4-pawn , whereas the wh ite slightly weaker, but this did not change the
bishop is obstructed by its own pawn on e5 character of the play.
and has no prospects at a l l . 26 . . . j_a8 ? !
In the first instance Black must prevent c4- 27 h3 h5
c5 and gain control of the only open file. 28 'ii'c 2
20 . . . c5
21 .ltc3 lid7
22 a4
My opponent wants to get rid of his vulnera­
ble rook's pawn and g ive me a weakness on
b6. But we know that, when there are
opposite-colour bishops, play on the q ueen­
side is less effective than activity on the
opposite side of the board , which Black will
soon develop. I would have preferred 22
l:tfd1 , although after 22 . . . .U.ad8 23 .U.xd7
'ixd7 Black has a n obvious advantage.
22 . . . .l:tad8
23 a5 'ii'g 5!
Black's pressu re on t h e kingside has ena­
24 .l::!.a e1
bled him to tie down the opponent, but for the
24 f4 ? would have lost i mmed iately to moment there is no direct way to wi n . To his
24 . . . .l:!.d2 ! . aid comes the principle of two weaknesses.
24 . . . �d3 He must stretch the opponent's defences by
I make on that I am attacking the pawn . But, creating d iversionary th reats on the opposite
of course, this is not so: it is not possible to side of the board . The new target is the c4-
capture on e3 in view of f3-f4 or the pin on the pawn .
c1 -h6 diagonal . 28 . . . j_b7
2 5 axb6 axb6 29 .l:.e2 �8d7
If 26 j_d4 I was intending 26 . . . 1:.8xd4! 27 30 �h1
exd4 .l:.d2 . Of course, I overlooked the 30 'ii' b 2 j_a6.
unexpected i ntermed iate move 27 h4 ! , win­ 30 . . . j_a6
ning the exchange, but i n overwhelming
3 1 'ii'a 4
positions such oversights are not too danger­
If 31 j_b2 Black can finally capture the e3-
ous. After 27 . . . 'ii'x h4 28 exd4 'ii'x d4+ 29 l::.f2
pawn , which has been en prise for a long
ic6 ! ? (more accu rate than 29 . . . h6 30 'iWa4)
time.
White's position is unenviable.
31 . . . l:.xc3
26 'ii' b 2
32 'ifxa6
Now Black would have liked to include his h­
pawn in the attack, but after 26 . . . h5?! 27 32 'ii'x d7 j_xc4 was no better for Wh ite .
'fxb6 .l:!.xc3 28 'ii'x b7 l:tc2 (or 28 . . .'i!i'xe5 ) the 32 . . . 'ii'd 8!
opponent has the good defence 29 f4! . The The q ueen defends the b6-pawn and takes
most accu rate move was the prophylactic up the 'approved' position beh ind the rook on
1 46 � Opposite-Colour Bishops in the Middlegame

the open file. Black's position is won . If 33 Of cou rse, 12 .. .f5 suggests itself. H owever I,

'ith2 I was intend ing 33 . . . h4 followed by was not sure that the advance of the f-pawn
34 . . . l:!.d 1 (or 34 . . . .l:.d2). was appropriate in the positions arising after
33 "ii'a 1 ? ! llxc4 1 3 b4 c5 or 1 3 cxd5 exd5 1 4 Ji.a6.
3 4 l:ta2 'ili'g5 On the basis of this, it was logical to play
1 2 . . . a5! , preventing both of these possibili­
35 .l:!.a8+ 'ith7
ties for the opponent. But I made a less
36 'ili'b1 + 'iig 6
accu rate move, wh ich neutralises only the
37 g4 second of them. What operated , apparently,
37 'ili'xb6 l:ic2 38 l1g 1 lld 1 . were associations with the previous g a me - I
37 . . . hxg4 remembered that there my rooks had operat­
ed q u ite well on the d-file, and I hurried to
38 hxg4?!
occupy it.
Of course, there were also no saving
chances after 38 'ii'x g6+. 1 3 Ji.c2 a5
38 . . . l:.c2 Again Black defers . . . f7-f5 because of 14
b4.
39 �g1 lld2
1 4 lDe5 Ji.xe5?
Wh ite resig ned .
Another move made by analogy. But where­
as i n the game against Alexander Vikulov the
N isman-Dvoretsky
exchanges led to an advantage for Black,
Moscow 1 972 here the result is the opposite.
Queen 's Indian Defence 15 dxe5 ltJxd2
1 d4 ltJf6 2 c4 e6 3 lbf3 b6 4 e3 Ji.b7 5 Ji.d3 1 6 .:txd2 dxc4
ii.. b4+ 6 lbbd2 0-0 7 0-0 d5 8 a3 Ji.d6?1 9
1 7 bxc4
'i!Ve2? ltJe4! 1 0 b3 lDd7 1 1 Ji.b2 'ili'e7
1 7 'ifxc4? ttJxe5! .
17 . . . lDc5

The same open ing moves have been made

as in the previous game. But this time my
opponent was not in a hurry to play his knight assu med that my opponent would be
to e5. unable to avoid exchanges on the e4-square
12 .l:lfd 1 l:tad8?! or the d-file, which a re advantageous to me.
 Opposite-Colour Bishops in the Middlegame
But I simply overlooked the following strong
move by Wh ite .
1 8 .l:i.d4!
The rook takes control of the e4-sq uare, at
the same time creati ng the threat of 1 9
�xh7+! 'it>xh7 20 'ifh5+ 'iit g 8 2 1 l:th4 . B ut the
main th ing is that now the exchange of rooks
becomes practically impossible for Black,
since Wh ite will recaptu re on d4 with his
pawn, obtaining a mobile pawn chain in the
centre , and then will soon advance d4-d 5 .
Instead o f the plausible 1 7 . . . t"Llc5 i t made
sense for Black to play 1 7 . . . t"Llb8, with the
idea of attacking the rook on d4 with the
knight. However, as Vad i m Zviagintsev point­
ed out, after 1 8 Ild4! g6 ( 1 9 ii.xh7+! was
threatened ) 1 9 .l:!.g4 ! lid? 20 h4 lifd8 2 1 ii.c3
his position would have remained anxious.
For example: 21 . . . t"Llc6 22 h5 t"Llxe5!? 23
�xe5 (less accu rate is 23 hxg6 t"Llxg6!? 24
J::l.g3 'ifh4) 23 . . . .l::r.d 2 24 hxg6! hxg6 25 .l:!.xg6+!
fxg6 26 'i!Vg4 1::!. xc2 (26 . . .�f7 27 �f1 �c2 28
t'h4) 27 'it'xg6+ 'it>f8 28 'ifxc2 with advan­
tage to White .
18 . . . g6
19 'i!Vg4
1 9 a4 !?.
19 . . . a4!
The only counter-chance! By placing his
kn ight on b3 Black will most probably
provoke the advantageous exchange of the
dangerous bishop on c2 . The far-advanced
pawn on b3 will promise tactical cou nter­
chances, or for a certain time will at least
divert the opponent from his attack.
20 liad 1 t"Llb3
21 ii.xb3 axb3
22 h4
As is customary with opposite-colour bish­
ops, White attacks on the kingside. It is
important to note that his rook can not be
driven from the d4-square by . . . c7-c5 - it will
occupy an even more powerful position at
d6. Exchanging it there will be altogether
ctJ 1 47
impossible - then the terrible d iagonal for the
b2-bishop is opened . I n cidentally, precisely
such a structu re occu rred in the highly i nter­
esting game Taimanov-Averbakh , which I
insistently recommend that you look at - you
will find it in David Bronstein's famous book,
devoted to the 1 953 Candidates Tournament.
22 . . . h5
23 'ii'f4
What do you think, whose bishop is better? It
may seem that the comparison is in favour of
Black - after a l l , his bishop is pressing on g2,
whereas its opposite nu mber is obstructed
by the pawn on e5. But let us look a l ittle
more deeply. Not one of my pieces is
supporting the bishop, so that its activity is
purely superficial. Whereas the opponent
has chances of penetrati ng on the weakened
dark squares on the kingside (after the
opening of the long diagonal, or on the c1 -h6
diagonal ) , and then my king will be in trouble.
Black cannot passively mark time - the
opponent will play 111 d 3 , capture the b3-
pawn , and then prepare either e3-e4 , or g2-
g4. He must try to seize the in itiative , but
how? He had to decide on a very risky
operation.
23 . . . l:l.xd4! ?
24 exd4 b5!
24 . . . .l:!.d8 25 1::!. d 3 .
1 48 � Opposite-Colour Bishops in the M iddlegame

25 d5!
Boris N isman correctly senses the spirit of
the position and , not paying any attention to
pawns, endeavours to open the diagonal for
his bishop. The miserable 25 cxb5?! would
have allowed me to activate my forces by
25 . . . �d7 26 a4 .:ta8 27 .:ta 1 'ifd5 28 f3 c6! ?
(but not 2 8 . . ."ii' c4? 29 .:tc1 ) 29 b6 'iic4 .
25 . . . 'ili'c5!
Activity first and foremost! I n the event of
25 . . . bxc4? 26 d6 things are bad for Black,
since the exchange on d6 is suicidal, and
after 26 ... 'i!i'd7 27 'ii'xc4 �d5 28 'ii'f4! he has
no counterplay. How should Black defend? 29 . . . �g7? 30
26 'iff6!? i.h6+ and 29 . . . �g8? 30 i.h6 .:ta8 31 'i'f6
Wh ite is intending to switch his bishop to h6. are completely bad . Only two possibilities
For the sake of this he is ready to part with remai n : 29 . . Jig8 and 29 . . . b2 30 i.xb2 (30
his pawns and even his rook. A clever idea , �xf8 'ii'xf1 +) 30 . . . �g8 .
but, as we will see, Black has a defence. It is easy to make the only possible moves,
However, all the same I do not see a direct but far more d ifficult when there is a choice.
way for the opponent to win . In the event of The price of a mistake in such a sharp
26 dxe6 fxe6 27 "ii'g 3 Black has a choice situation is extremely h i g h , and therefore a
between 27 . . . �e4 and 27 . . . �g7 28 l:.d7+ very carefu l calculation is demanded . Alas, I
.l:i.f7 . If instead 26 d6, then 26 . . . 'i!i'xc4 . In the failed to display this.
endgame it is now White who would have to When checking the variation 29 . . . b2 30
find a way to save himself: 27 'ifxc4? bxc4 i.xb2 <ittg 8 I was not afraid of the return of
28 d7 lld8 followed by 29 . . . �d5, or 28 dxc7 the bishop to the c1 -h6 d iagonal: 31 .ic1
�d5 and 29 . . . .l:!.c8 . If 27 .l:td4! Black can reply 'ii'x e5 32 i.h6 .l::f.a 8, or 31 �d4 �e4 ! . But I
27 . . . 'ii'c6!? (not 27 .. ."W/c2? in view of 28 d7! was frightened by the combi nation 31 dxe6
i.d5 29 l:!.xd5 exd5 30 e6, but 27 . . . 'ii'e 2 'i!Vxb2 32 exf7+ :Xf7 33 'iie 8+ 'it>g7 34 e6
comes into consideration ) 28 'ifg5 'i¥d7 and and wins, for example: 34 . . . .l:.f4 35 'iid 7+ (or
29 . . . c5. Here Black's position is uneasy, but 35 'W/e7+ 'it>g8 36 �d8+ llf8 37 e7) 35 . 'it>h6 . .

nevertheless his queenside pawns guaran­ 36 e7 ll.g4 37 'ii'x g4 ! . In fact Black can save
tee him counter-chances. himself, by playing 32 . . . �g7! (instead of
26 . . . 'i!i'xc4 32 . . . .:txf7?). There is an even simpler draw
27 �c1 ! 'it>h7 by 3 1 . . . ..txg2! (instead of 3 1 . . ."�xb2 ).
Of course, not 27 . . . 'ii'c2? 28 i.h6! 'ifxd 1 + 29 29 . . . l:.g8?
�h2 with unavoidable mate. 30 ..tg5! b2
28 'ife7 'i¥e2 3 1 i.f6 'i!i'xf1 +
29 l:[f1 32 �h2
Black is a rook up, and his pawn is on the
(see diagram) threshold of q ueening. And yet there is no
satisfactory defence against the mating
threats, created by j ust two enemy pieces.
 Opposite-Colour Bishops in the M iddlegame ltJ 1 49

There you have it - the formidable strength T h e f7 (f2) point

of an attack with opposite-colour bishops.
Boleslavsky-Sterner
32 . . . 'i!kc1
Sweden-U S S R Match 1 954
32 J:1g7 33 WfB .
. .

33 'ii'x f7+ 'it>h6

At first sight it may seem that a draw is

i nevitable. Material is equal, and all the wh ite
34 �g5+? pawns are on one wing . Practically any
A vexing, over-hasty move just a step away endgame will be d rawn , for example the one
from victory. 34 Wxg8? 'ii'f4+ would have led with 'pure' opposite-colour bishops where
to a draw, but the modest move 34 g 3 ! ! , with Black has lost his c5- and f7-pawns.
the idea of 35 �g5+ Wxg5 36 hxg5+ 'it>xg5 I n fact Wh ite has a g reat a n d , most probably,
37 �f4 mate, would have forced i m mediate decisive advantage. Exploiting the powerful
capitulation . position of his bishop and the vulnerabil ity of
34 . . . 'ii'x g5 the f7-poi nt, he condemns his opponent to
passive defence. And with opposite-colour
35 hxg5+ 'it>xg5
bishops we already know that the unchal­
36 'ili'f6+ lenged possession of the in itiative is usually
There is no longer a win : 36 g 3 'it>h6 ! , 36 c;.t>g3 a very important factor.
h4+ ! (but not 36 . . . 'it>h6? 37 'ii'x g8 b 1 'ii' 38 First the pressu re on the f7 -pawn must be
'i'h8+ 'it>g5 39 f4+ 'it>f5 40 'iff6+ and 4 1 intensified , in order to tie the enemy pieces
'i'xg6+ ), or 3 6 f4+ 'it>g4! 37 Wf6 g 5 ! .
to its defence.
36 . . . 'it>h6 38 l:td1 �c7
37 'ii'f4+ 'it>h7 38 . . . l:lf8 was weaker i n view of 39 .l:td7 �c7
38 'ii' b4 �xd5 40 g3. Now the rook cannot leave the back
39 'ii'x b5 c5 ran k because of the check on b 1 .
40 'ii'x b2 c4 39 'ii'd 7 llf8
41 'it>g3 40 e5!
White adjourned the game, after sealing this A typical move. Remember: with opposite­
last move. On the invitation of his opponent a colour bishops the pawns should be placed
draw was ag reed without the game being on squares of the colour of the opponent's
resumed . bishop. The bishop on c7 is now restricted by
1 50 � Opposite-Colour Bishops in the M idd legame

the e5-pawn , which at a conven ient opportu­ attacks the g7-poi nt: 4 6 We7 l:tb8 (46 . .'i'b8 .

nity may also be able to advance further, to 47 h6 i.. c3 48 .l:.d7 i.. d 4+ 49 l1xd4) 47 h6
open up the black king's position. 'iWc3 48 .l:id7 .l:!.b 1 + (48 . . . 1Wc 1 + 49 iJ1 ) 49
40 . . . 'it'b6 'it>h2 11h 1 + 50 'it>xh 1 Wc1 + 5 1 'it>h2 'iixf4+ 52
A waste of a tempo. The bishop is doing 'it>g 1 'iic 1 + 53 .tf1 . It is better to give up the
nothing on c7 - its place is at d4, and Black exchange: 44 . . .'ii'xc4 45 e7 'ii'e 6 46 exf8'1+
should immed iately have begu n manoeu­ 'it>xf8 , in the hope after 47 'ii'xe6 fxe6 of
vri ng it there: 40 . . . 'iib 8 4 1 f4 .ta5. putting up a stiff resistance in the endgame.
B ut it is not essential to excha nge queens -
41 f4 'ii' b 8
stronger is 47 it'a?! i.. b 6 (47 . . . 'ili'b6 48 'i'a8+
'itt e 7 49 l:t a 1 ! ) 48 'ili'b8+ 'it>e7 49 l:r.b1 i.d8 50
'fia7+ and 5 1 'ifxc5.
44 l:.d61
The bishop was wanting to go to d4, blocking
the d-file, and so the rook h u rriedly advances
to an active position . The threat is 45 h6. If
44 . . . h6, then 45 'iif5 , intending 46 .Uxh6, 46
l:td7 or 46 i.. d 3.
44 . . . 'ii b 1 +
45 'it>h2 h6
46 'ifxf7+! l:txf7
47 J:td8+ 'it>h7
48 i.. x f7
42 h4! Black resigned .
A typical attacking resou rce in such situa­
tions! The pawn wants to advance to h6,
King's I n d i a n structu re
breaking up the enemy king's defences. If it
is met by ... h7-h6, the b 1 -h7 diagonal is Levenfish-Kan
weakened and White's bishop and q ueen Moscow 1 927
can switch to it. The pawn on h 5 will also
come in useful if the opponent plays . . . g7-
g6.
I ncidentally, in reply to 42 . . . g6 Isaak Bole­
slavsky suggests the spectacular break­
through 43 e6 fxe6 44 f5. But after 44 . . . gxf5
45 .txe6+ 'it>h8 46 i..xf5 Black has a
defence: 46 . . . i.. h 2+! 47 'it>h 1 'fie? . Therefore
Wh ite should prefer the simple 43 h5! or 43
e6 fxe6 44 h 5 ! .
42 . . . .ta5
43 h5 i.. c 3
43 . . .'ii' b4 should also be considered , after
which Boleslavsky was i ntending 44 e6 ! . I n
the event of 44 .. .f6 45 'ifxe6+ 'it>h8 White This a characteristic position from the King's
 Opposite-Colour Bishops in the M iddlegame l2J 1 s1

Indian Defence (with reversed colours). The

typical pawn sacrifice made by Grigory
Levenfish should be i n the arsenal of every
King's I ndian player.
22 fS! gxfS
The challenge has to be accepted : weaker is
22 . . . i.f7 23 fxg6 hxg6 24 i.h6 followed by 25
i.h3.
23 i.h6 lbg7
24 exfS lbexfS
25 lL'lxf5 i.. xfS
26 lL'lf3
By placing his knight on h4, Wh ite i ntends to - position after 33 .l:lxf5 -
seize control of the e4- and f5-sq uares. What
can be done to oppose this? Probably
advancing his g-pawn , whereas Black has
26 . . . i.g6 27 lLlh4 f5 should have been tried ,
no cou nterplay.
although after 28 'ilkd2 with the idea of 29
33 . . . ILg7
tt:lxg6 hxg6 30 'ilkg5 Wh ite would have
retained the i nitiative . 34 'i¥d2
26 . . . i.e6 The q ueen goes to h6, to take part i n the
attack.
27 .l:tf2 J:Lf7
34 . . . bxc4
28 lbh4 bS
35 bxc4 �b6
White would have had the advantage after
28 .. .f5 29 lbf3 e4 30 dxe4 fxe4 31 'it'xe4 i.f5 Of cou rse , it would be crazy to go chasing
32 'ifh4 (32 'i!Vd5 ! ? ) 32 . . . i.g6 33 lbe5 . the a3-pawn : 35 . . . It.b3 36 'ifh6 .U.xa3 37 .U.h5.
29 ILef1 i.e 7 36 'it'h6 i.. d 8
But now 29 . . .f5! was simply essential. l lya 37 a4
Kan decided to keep his pawn on f6 , With the opponent completely deprived of
restricting h imself on the kingside to passive cou nter-cha nces, Wh ite can even permit
defence. Hopeless strategy! h imself this rather abstract move. 37 l:txe5?
30 i.. x g7 'iit x g7 fxe5 38 l1f8+ l:tg8 39 lixg8+ 'it>xg8 40 i.d5+
did not work i n view of 40 . . . 'i!Vxd5+ 41 cxd5
31 lLlf5+ 'iit h 8
l:txh6.
32 i.e4 i.xfS
37 . . . aS
33 .l:txfS
38 i.dS
(see diagram) Th reatening 39 .l:i.xe5.
38 . . . 'it'e7
Levenfish has carried out his plan a n d , Now the thematic adva nce 39 g4! strongly
despite being a pawn down , he has gained suggests itself. 39 . . . �xg4 40 .U.xe5! is bad for
an overwhelming advantage. To real ise th is, Black. Apparently Levenfish did not find
it is sufficient to compare the positions of the a nyth ing convincing after 39 . . . .Ug6 40 i*'h5
two bishops. Wh ite can attack the h 7 -point or (40 'ilr'h3! is stronger, and if 40 . . . i*'g7 41 .U.h5)
undermine the opponent's pawn chain by 40 . . . 'it'g7 41 g5 i.e?. He decided not to
1 52 � Opposite-Colour Bishops in the Middlegame

hu rry, but to keep manoeuvring and await a defends them from b4 , Wh ite advances his
more convenient moment for the break­ pawn to h6, putting the opponent in zug­
throug h . zwang.
3 9 'ii' h 5 .l:tg6 This conclusion is not altogether accu rate ­
40 .l:.5f2 ifd7 in fact Black saves h i mself by sacrificing two
41 i.. e4 .l::t g 5 pawns and changing the roles of his pieces:
54 .. .f5! 55 i.. xf5 e4! 56 .txe4 i.. f6 57 'it>g2
42 'ii' h 6 'ii'g 7
cJitg7 58 �f3 �f7 59 i..f5 �e7 60 �e4 'it>d6
43 'iVh3 'ii'c 7 with a d raw. But, of cou rse, Wh ite can easily
44 'ii h 6 'ili'g7 gain the tempo that he lacks i n this variation
45 'iVh3 'ili'c7 by 52 'ii'x d7 .l::t x d7 53 i..f5! .l::t g 7 54 .l::tx h 7+ (or
46 .l::tf5 ! l:lxf5 first 54 �g2).
If 46 . . . .l::t g 7 there would probably now have 50 lixe5??
followed 4 7 g4 with the threat of 48 g5 l:txg5 I n his book of selected games and reminis­
49 l1xg5 fxg5 50 .l::tf8+ cJitg7 51 l:te8 .l:lh6 52 cences, Levenfish lamented a serious defi­
'iif5 'ii'd 6 (or 52 . . . i..f6) 53 i.. d 5. ciency in his play. After outplaying his
4 7 l:txf5 .l::t d 6 opponent and gaining a decisive advantage,
48 g4 I:f.d7 he would often make a serious error and ruin
the fru its of his preced ing work. This was
also the case here. It is hard even to explain
why Wh ite felt the need to exchange rooks.
Surely not for the sake of winning the e­
pawn? But when there are opposite-colour
bishops, pawns are of no sign ificance - you
should be thinking only of attack!
After 50 .l::tf8+ ! cJitg7 51 .l::te 8 Black would have
had to resig n , whereas the move in the game
leads only to a draw.
50 . . . 'ii'x e5 51 'iix d7 'ii'e 7 52 'ii'f5 i.. c7 53
�g2 i.. d 8 54 �f3 i.. c 7 55 h3 i.. d 8 56 'it>g4
�g8 57 i.. d 5+ �g7 58 i.. e 4 �g8 59 i.d5+
�h8 60 �f3?! 'ii'e 3+ 61 �g4 'ii'e 2+ 62 'i'f3
'ii'e 7 63 'ii'e4 h5+?! 64 �xh5 Vxe4 Draw.
49 g5! fxg5
Kan reckons that he could have held the We have mainly been studying the strategy
position by 49 . . . .l::tf7 50 g6 .l::t g 7. Romanovsky of play with opposite-colour bishops, but for
retorted that the ending arising after 51 l1h5 dessert I invite you to solve a few combina­
'ti'd7 52 l:.xh7+ l:txh7 53 'ii'x h7+ 'ii'x h7 54 tive exercises. Most of them (although not
gxh7 was hopeless for Black. The wh ite king a l l ) are elementary, but nevertheless useful,
invades the enemy position via e4 and goes since they demonstrate tactical ideas typical
across to the queen side pawns. If the bishop of positions with opposite-colour bishops.
 Opposite-Colour Bishops in the M iddlegame ltJ 1 53

Exercises

1 . Wh ite to move 2. Wh ite to move

3. Wh ite to move 4. White to move

1 54 � Opposite-Colour Bishops in the Middlegame

5. Wh ite to move 6. Black to move

7. Wh ite to move 8. Black to move

 Opposite-Colour Bishops in the M iddlegame lZJ 1 55

Sol utions

1 . A. Petrosian-Moldagaliev ( U S S R J u n ior the hope of winning the enemy q ueen after

Championship, Yerevan 1 969). giving a bishop check). But the cool-headed
30 l:!.h8+! �xh8 reply 40 . . . l:.d6 ! forces the exchange of rooks,
31 'ii' h 5+ after which Wh ite's success becomes prob­
lematic. With opposite-colour bishops, an
Black resig ned .
extra pawn by no means guarantees a wi n .
4 0 'i!fe3 !
2. Wachtei-M ichel ( 1 953).
A decisive double attack - White is threaten­
1 l:!.e5 ! 1 ing not only 41 'ii'x b6, but also 41 l:!.h8+!
Paradoxically, it i s the exchange of rooks that �xh8 42 'ii'x h6 mate.
leads to inevitable mate.
6. N N-Rossolimo (Paris 1 957).
3. Wade-Ku ij pers (Holland-England Match The b7-point against the f2-poi nt! Whose
1 972). attack is stronger?
29 .l:.xh6+ ! ..ixh6 1 . . . .l:td1 ! !
30 '1Wxe5+ 2 c4
Black resig ned in view of 30 . . . ..ig7 31 'fih2+ There is noth ing else: 2 I1bxd 1 'i!i'xb2 , 2
with mate. 'iix b5 (or 2 .l:r.fxd 1 ) 2 . . . ..ixf2+ 3 �h2 .l:h8
mate , or 2 ..ixb7+ 'it;>b8.
4. Hartston-Pen rose (London 1 963). Now the conseq uences of the fol lowing two
1 .l:!.xf7 ! �xf7 l i nes are not too clear: 2 . . . l:!.xf1 + 3 �xf1
2 ..ic4+ �fB l:txf2+ 4 'ii'xf2 'iWxb 1 + 5 �e 1 , or 2 . . . ..ixf2+ 3
3 l:if1 + ..if6 'ii'xf2 l:r.xf2 4 I1 bxd 1 .l:.xf1 + 5 .:.xf1 .
4 l:xf6+ gxf6 2 . . . l:txf2 !
5 'figS+ �e7 3 ..ixb7+ �b8
6 'iie 6+ �f8 4 1i'xb5 .l:tfxf1 +
7 'il'xf6+ 5 �h2 Uh1 mate
Alas, Wh ite did not fi nd this combination and
he offered a draw, which , of cou rse, was 7. B l umenthai-McG u n n igle (correspond­
accepted . ence 1 962).
The prosa ic 3 1 'iWf7 ! ? l:!.d7 32 tiJe7 is not
5. Karpov-H ubner (Montreal 1 979). bad , forcing the opponent to g ive up the
Karpov played 39 'ii'c 4? (apparently expect­ exchange. However, after 32 . . . l:!.xe7 33
ing 39 . . . 'ii' b 5? 40 llc7 �8 4 1 .l:l.xe7! �xe7 42 'ii'x e7 'ilfd8 Black is still capable of putting up
'i'c7+ �e8 43 ..if7+ �f8 44 ..ic4), but after a tough resista nce .
39 . Jlf6 1 40 llc7 'ii'd 6 Wh ite's advantage
.
31 'ii' h 6! !
proved i nsufficient for a wi n . With this pretty move Wh ite beg ins a deadly
39 llg8+! �h7 attack on the g7-point. The q ueen is taboo:
Now the showy 40 l:!.xg6 suggests itself (in 3 1 . . . gxh6 32 ..id4+ �g8 33 tiJxh6 mate.
1 56 � Opposite-Colour Bishops in the M iddlegame

31 . . . lld7 This rook sacrifice i s the most energetic

32 .id4
32 ltJxg7! .l:.xg7 33 .ib6! was stronger
(V.Bologan). Black could have excluded this
possibility, by transposing moves: 31 . . . 'ii'c7
32 .id4 (32 .ib6 'ii'e 5) 32 . . . .l:td7 .
32 . . . 'ii'c 7
33 :1f3
By doubling rooks on the f-file Wh ite will
create the th reat of ltJxg7 followed by .l:.f7 .
33 . . . ltJg6 34 l:tbf1 <itg8
Black would also have lost after 34 . . . ltJe5 35
.l:te3! gxh6 (35 . . . ltJc6 36 l:.fe 1 or 36 'ii'e 6) 36
.l:txe5 (S.Sh ipov).
35 'ifxg7+? !
A spectacular blow, althou g h , as soon
becomes clear, Wh ite inaccu rately calculat­
ed its consequences. He would have re­
tained a winning attack after 35 'ili'e3 !
(th reatening 3 6 lbh6+ gxh6 37 'ii'e 6+)
35 . . . 'ii'c6 36 ltJxg7! llxg7 37 .ixg7.
35 . . . l:txg7
36 ltJh6+ <ith8
37 .l:t� 'ifx�?
37 . . . ltJe5? 38 .l:txc7 also would not have
helped . As was pointed out by Taylor
Kingston, Black could have saved h imself by
playing 37 . . . 'ii'e 5! 38 .ixe5 ltJxe5 39 l:r.f8+
.l:tg8 40 ltJxg8 llxf8 41 I!xf8 <itg7, for
example, 42 .l:tf5 cxd3 43 cxd3 liJxd3 44 ltJe7
ltJxb2 45 ltJd5 with a double-edged end­
game.
38 .l:tx� .l:tg8
39 .l:td7 !
It is important to cut off the bishop's path to
e8 .
39 . . . .ixd7
40 lb� mate
8. Kholmov-Gel ler (USSR Spartakiad ,
Moscow 1 959).
29 . . . l:txh3 ! !
continuation of the attack. In the event of
29 . . . 'ifh4 30 'iWe5+ .ic7 31 'iWxd4 'ifxh3+ 32
<itf3 g2+ 33 <ite2 gxf1 'if+ 34 l1xf1 or
29 . . . 'ii' h 6 30 h4!? (weaker is 30 'ife5+ 'i!i>a8!
31 .l:.h 1 J:lf4) Wh ite would have retained
d rawing chances.
30 <itxh3
As was shown by g randmaster Shipov,
Wh ite would have had a d ifficult endgame
after 30 'ife5+ ! ? 'ifxe5 31 .l:txe5 .l:txg4! 32
'it'xh3 g2 33 l:tee 1 (33 .l:tfe 1 g 1 'ilf) 33 . . . gxf1 'i'+
34 llxf1 l.'tg7 .
30 . . . 'ii' h 6+
3 1 <itxg3
31 �g2 'ifh2+ 32 <itf3 g2 was altogether bad
for White .
31 . . . Ji.c7+
32 �f2
Now noth ing is given by 32 . . . 'ili'h2+? 33 cJi>e3
or 32 . . . 'iWf4+? 33 <itg 1 ! (33 'ilff3? 'ilfh2+ 34
<ite3 .ib6) 33 . . . 'ifg5 34 .ixe6 ! . Black could
have won by 32 . . . 'ifh3 ! 33 'ii'f3 (33 'i'd2
l:txg4 34 <ite2 l:tg2+ 35 <itd 1 .l:r.xd2+ 36 'i!i>xd2
Ji.a5+) 33 . . .l:tf4 (33 . . .'ii' h2+!? 34 <ite3 .ib6 is
also strong) 34 'ii' xf4 .ixf4 35 l:.g 1 (35 J:l.e4
Ji.g3+) 35 . . . Ji.g5 ! . Yefim Geller played some­
thing d ifferent and, apparently, less strong.
32 . . . 'ii'g 5?!
A pretty q u iet move, with the terrible th reat of
33 . . . l1f4+.
33 J:.b1 ?
Wh ite vacates the e 1 -sq uare for his king. But
it was much stronger to play 33 'ife3! 'ii'x g4,
and only now 34 .l:t b 1 ! .
33 . . . l:r.xg4
34 <ite1 l:g2
If now 35 .l:!.xf7 , then 35 . . . .l:.xe2+ 36 <itxe2 b5.
35 d4 'ii' h 4+ 36 'it'd1 .l:txe2 37 <itxe2 'ii'e4+
38 <itd2 .ia5+ 39 <itc1 'ili'e3+ 40 <itb2 'ii'd2+
41 <ita3 'ii' b4+ Wh ite resigned .

Mark Dvoretsky
You ca n 't get by without a Combi nation !
W Secrets of Grandmaster Play by John
hen studying the excellent book
Nunn and Peter Griffiths, my attention was
drawn to a position which occu rred in the
game N u n n-Van der Wiel (Wij k aan Zee
1 982 ).
Wh ite to move
It was suggested to pupils of our school as a
piece of homework (with the right to move
the pieces on the board ) after studying the
topic 'Opposite-colour bishops in the midd le­
game' .
'But where are the opposite-colour bishops
here?' you may ask. This will soon become
clear.
White has a clear advantage. 28 .l:!.dh 1
suggests itself, but Black has the reply
28 . . . �f6. Not finding a nyth ing convincing
here, John Nunn decided not to allow the
bishop to go to f6 .
2 9 e5!?
The start of a combi nation calculated 12 ( ! )
moves ahead .
ltJ 1 57
29 . . . llxh8!
The only defence. All other attempts are
easily refuted .
A) 29 . . . d5. Petya Kiryakov and Vova Baklan
pointed out the amusing variation 30 Ji.g5!?
�xg5 (30 . . . .l:ta8 31 'ifxa8) 3 1 'ifc5+ 'it>b8 32
l:.xg8 (there is also 32 'i!i'xc8+ ) 32 . . . .l:txg8 33
"iif8+ 'ii'e 8 34 'ifxe8+ l:txe8 35 l:th 1 . Not bad ,
but why go for a combi natio n , when there is a
simple solution : 30 l:td h 1 with the threat of 3 1
�g8 .l:txg8 3 2 .l::t h 8.
B) 29 ... l':txg7 30 l2Jxe6+. (The suggestion of
l lakha Kadymova is also strong: 30 exd6+!
..txd6, but then she continued 3 1 l':txc8+
'it>xc8 32 ..tf4 ! Ji.xf4 33 "ii'c 5+, which is
unconvinci ng: 33 .. .'�b8 34 'ii'f8+ .tea 35
'iVxg7 'ii'd 7. Meanwh ile, 31 lLlf5 ! ! wins
i m med iately. ) 30 . . . fxe6 3 1 exd6+ . (There is
also another way: 31 i.. b 6+ �c6 32 llxc8+
.txc8 33 'it'c7+ 'itb5 34 exd6 (34 b3; 34
Ji.e3 ) . ) 31 . . . Ji.xd6 32 "ii' b 6+ 'itb8 33 "i¥xd6+
.l:tc7 (33 . . . '>t>a8 34 'ii'c 5! 'iii> b 8 35 Ji.f4+) 34
l:txc8+ ..txc8 35 �a7+ and wins.
C ) 29 . . . dxe5 30 lLlb3 "ii' b 5 3 1 Ji.b6+ 'i+'xb6 32
lid?+ 'itc6 (32 . . . 'it>xd7 33 'ii'x b6 �d5 34
1\i'a7+) 33 l2Ja5+ 'ii'x a5 34 'ii'x b7+ '>t>c5 35
.l:.xg8 l:.xg8 36 l:ic7+. Many pupils of our
school found another, perhaps even more
spectacular way of attacki ng: 30 l2Jxe6+!?
fxe6 3 1 Ji.b6+ 'itc6 32 l:id7! <;i(xd7 (32 .. .'it>b5
33 b3) 33 'it'xb7+ 'itd6 34 'iVxc8 llxc8 35
l:lxc8 .
30 exd6+ Ji.xd6
3 1 gxhS"ii'
The obvious move , but not the only one. We
will retu rn later to this position .
31 . . . l:txh8
1 58 � You can't get by without a Combination!
32 lbxe6+ fxe6
33 'ifb6+ �c8
34 'iVxd6 'ii'c 6
The only defence against the n umerous
threats. Bad was 34 . . . .i.d5 35 .lli. f4 �b7 36
'i!Vc7+ �a8 37 .i.e3 .lli. b 7 38 .l:!.d8+.
Noth ing is given by 35 'it'd4 .l:!.e8 36 .tf4 (with
the th reat of 37 'iie 5) in view of 36 . . . 'ifd5.
Therefore 35 'iix b4 suggests itself, but after
35 . . . .l:td8! 36 �xd8+ �xd8 Black successful­
ly defends, for example: 37 .lli. g 5+ �c7 38
'ifa5+ �b8 39 'ifd8+ .lli. c8 , or 37 'iff8+ 'ife8 !
(weaker is 37 . . . �c7 3 8 .i.f4+).
I n a midd legame with opposite-colour bish­
ops the most important th ing, as we know, is
the in itiative. Even if there are comparatively
few pieces left on the board , in the first
instance you should th ink not about winning
material, but about creating threats to the
enemy king .
35 'ife5! !
Threatening 36 .lli.f4; i f 3 5 . . . .Uh5 there follows
36 .i.g5.
35 . . . .l:rd8
36 .l:rxd8+ ..t>xd8
37 .i.g5+ �d7
37 . . . �c8? 38 'it'h8+ , or 37 . . . ..t>e8? 38 'ifh8+
�f7 39 'ii'h 7+ �f8 40 'ii'e 7+ 'itg8 41 .i.f6 .
38 'ii'g 7+ �d6
39 'iff8+ �d5
39 . . . �e5 40 .lli. f4+ �d5 41 'iix b4 was even
worse for Black.
40 'ifxb4
Only now is this an appropriate moment to
captu re the b4-pawn .
This was the position for which N u n n aimed,
when he made his 29th move. He sensed
that, despite the material equal ity, things
would be bad for Black. A very deep
calculation and a completely correct assess­
ment!
Why is Wh ite's advantage so appreciable?
The entire blame l ies on the awkward
position of the black king in the centre of the
board , and , as usua l , the presence of
opposite-colour bishops g reatly strengthens
the attack. N u n n 's pieces a re domi nant on
the dark squares. The enemy bishop is
unable to help here in any way, and the
q ueen also is hardly participating in the
defence, since it is tied to its own bishop.
Thus the king remains alone against Wh ite's
superior forces .
I should also mention the good position of
the f3-pawn (as stipulated by the rules - on a
sq uare of the colour of the opponent's
bishop) - it secu res the e4-poi nt for the
queen and takes away this square from the
black quee n . The q ueenside pawns are also
 You can't get by without a Combination ! l2J 1 59

ready to join the attack: b2-b3 followed by 5 1 . . . �b6 52 f5 e3, then after 53 �d4+ �c6
c2-c4+ . Wh ite is also threatening the i m me­ 54 'iig 6+ 'iid 6 55 'ilfe8 it will be lost.
diate 41 c4+ �d4 (4 1 . . . �e5 42 Wc3+ �d6 51 . . . 'ii'd 7
43 'it'd4+ ) 42 'ii'c 3+ �c5 43 b4+ �b6 44 52 'ifh6+ �d5
'i'd4+ with inevitable mate . I n the event of
53 Wb6 ! 'iic 6
40 . . . �e5? he g ives mate by 4 1 �f4+ �f5(f6)
42 'ii'fB . If 40 . . . 'ii'd 7 there is the strong reply 54 'ii'd 8+ �e6
41 i.f4 ! with the threat of 42 c4+ �c6 43 55 'ikf6+ �d7
'l'a4+ . 56 'it'g7+ �e6
The concl uding stage of the game convinc­ 56 . . . �c8 57 f5 e3 58 f6 e2 59 'ikg4+.
ingly demonstrates how helpless Black is. 57 'ii'g 4+ �
We will only examine the main variations - a 58 f5 'ifh6
detailed analysis can be found in the afore­
59 f6
mentioned book.
Black resigned .
40 . . . e5!
41 'ii'e 4+ �d6
Accord ing to the comments in the book,
4 1 . . .'iti>c5 42 'ii'x e5+ 'ifd5 43 'ii'c 7+ 'ikc6 44 transposing into a position with opposite­
'i'e7+ �b5 45 �e3! followed by b2-b3 and colour bishops was the only correct solution
c2-c4+. for Wh ite . However, in their analysis only two
42 'ii'x g6+ 'iti>d5?! of our pupils opted for this cou rse (and on the
42 ... 'iti>c5 43 �e3+ �d5 was a much tougher 35th move, alas, they only considered 35
defence, but even then after 44 'ii'f7 + �d6 45 'ii'x b4? instead of 35 'ike5 ! ! ) . All the rest tried
c4! 'it'd? (45 . . . 'ii'c 7? 46 c5+ �c6 47 'ii'e 6+ to exploit i m mediately the unfortunate posi­
'it>b5 48 c6 ! ) 46 'ii'f8+ �e6 47 'ii' h 6+ �f7 48 tion of the black king and the power of the g7-
'lh7+ Wh ite would have retained a powerful pawn . To my su rprise they succeeded , and,
attack. what's more, i n various ways.
43 'iff7+ 'ifi>d4 Let us return to the position arising after 29
44 'ii' b 3! e4 e5 .U.xh8 30 exd6+ .i.xd6 . Apart from the
The only defence against mate by the bishop move i n the game, N u n n also examines 31
from e3. ttJxe6+ fxe6.
45 .i.e3+ �e5
46 f4+
Apart from a continuing attack, White has
also acquired a new 'trump' - a passed f­
pawn .
46 . . . �f6
47 'i!Vg8 'ii'd 5
48 �d4+ ! �e7
49 'ifg7+ �d6
50 b3! �c6
51 �e5
Threatening not only a terrible check at c7,
but also 52 f5. If Black advances his pawn :
1 60 � You can't get by without a Combination!
The variation he gives is 32 'ii' b 6+ 'Ot>b8 33
'ifxd6+ 'Ot>a8 , and White has no time to
captu re on h8, since his c2-pawn is attacked .
Kiryakov and Baklan found the excel lent
qu iet move 32 'ii'd 4! 1 . If 32 . . . l::l. h d8 Wh ite
wins by 33 g8'ii' ! . If instead 32 '0t>b8, then 33
• • •
b3! 'ii'c6 (33 . . . 1\i'ea 34 1\i'xd6+ 'Ot>a8 35
gxh81\i' 1\i'xh8 36 'ii' b 6) 34 gxh8'ii' 'ii'x c2+ 35
'Ot>a1 l:.xh8 36 1\i'xhS+ i.. c 8 37 l;lc1 and wins.
However, as N u n n pointed out in the 2nd
edition of his book Secrets of Grandmaster
Chess, Black can put up a tough defence by
32 . . . i..d 5! 33 gxh8'i!V l:.xh8 34 'iig 7+ 'ii d 7 35
'ii'x h8 i..xf3 or 34 'ili'xh8 b3.
The analysis by Kadymova was less good ,
unfortu nately: 32 i.. b 6+ 'Ot>c6 33 gxh8'iV l:.xh8
34 i.. d4 .Uc8 35 'ii' b 6+ 'Ot>d7 . She then
considered 36 'iWxb7+ l:!.c7 37 'ii'e4, but this
is unconvincing in view of 37 . . . b3! 38 cxb3
'ili'xb3, and Black's chances are not worse.
36 i.. c 5! wins. However, Black can defend
more accurately: 34 . . . 'ii' b 5 (instead of 34 . . .
.:c8?) 3 5 i.. x h8 �c5 3 6 'ii'x c5 i.. x c5, and the
most probable outcome is a draw. Another
way of making a draw was later suggested
by N u n n : 34 . . . l:tb8!? 35 1i'b6+ 'Ot>d7 36 i.. c 5
'ii'a 2+! 37 'Ot>xa2 i.. d 5+ 38 lixd5 l:!.xb6 .
Perhaps the most unexpected and spectacu­
lar idea was found by g randmaster Sergey
Dolmatov during a training game which
began from the orig inal position: 30 i.. g 5!?.
After checking the variations we concluded
that Black is helpless, for example: 30 . Ir.he8 . .
31 i.. x e 7 l:lxe7 32 exd6+ �xd6 33 tt'lb5+
'Ot>e5 (33 . . . '0t>c6 34 l:td6+ 'Ot>xb5 35 'iY'b6+) 34
'ii'd 4+ 'Ot>f5 35 t2J d6+ 'Ot>g5 36 l:tg 1 + 'it>h6 37
l:.h 1 + 'it>g5 38 'i!kh4 mate . N u n n agreed with
our opinion , adding the i nteresting variation
30 . . . l:.hg8 31 i.. x e7 dxe5 32 tbb3 (perpetual
check results from 32 t2Jxe6+ fxe6 33 .id6+
'Ot>d7 34 i.. c 5+ ) 32 . . . 'ii' b 5! 33 tbc5 (33 .id6+
'Ot>c6 34 'ii'e 3 is unclear) 33 . . . l:.a8 34 If.d7+
'Ot>c8 35 'iVxb7+ 'ii'x b7 36 l:txb7 l:.xg7 37 .id6
and wins (the refinements in brackets are
mine). However, here Black's defence can
be improved : 34 . . . '0t>c6! (instead of 34 . . . Wc8?).
Going into an ending by 35 'ifxb7+ is now
u npromising, which means that White is
obl iged to repeat moves : 35 :d6+ 'it>c7 36
l:td7 + '.t>c6 ! . So that, alas, the brill iant bishop
move to g 5 is not good enough to win .
Let us retu rn t o the position with which we
beg a n . Seryozha Movsesian analysed the
consequences of 29 tiJb3 ! ? 'i!kb5 (29 . 'i'c6 . .
30 tba5 'ii' b 5 3 1 l:txg8 .l:txg8 32 J:!.d4).
I ncidentally, i n the game the knight had only
just been at b3, and the q ueen at b5 - the last
moves were 28 tbd4 'iWa4 .
30 l:.xg8 .l:txg8 31 e5! d5
Black loses q u ickly after 31 . . . Vxe5 32 ii.b6+
'it>c8 33 ttJa5 or 31 . . . l:.xg7 32 exd6+ i.. x d6 33
 You can't get by without a Combination ! ttJ 161

'i'd4 1Ve5 34 Wb6+. I f 3 1 . . . dxe5 Movsesian

suggested 32 ..tb6+ ! ? 'ifxb6 (32 .. .'it>c8 33
ltJa5; 32 . . . <iit c6 33 l2Ja5+ Wxa5 34 ..txa5 .i.c5
35 .l:td8 or 35 .l:.d6+ ) 33 l:td7+, but the
situation arising after the win of the q ueen
(33 . . . <iitx d7 34 'ilxb6 .i.d5, or 33 . . . <iit c6 34
ltJa5+ <iit x d7 35 'i!i'xb6 .i.d5), is unclear, in
Nunn's opinion . Wh ite's attack can be
strengthened by 32 .i.c5 ! , for example,
32 ... .i.xc5 33 l2Jxc5 <iit c8 34 l2Jxb7 'iVxb7 35
'i'c5+ <iit b 8 (35 . . .'i!i'c7 36 'iVf8+) 36 .Ud6, and
this time the black q ueen is lost without any
compensation .
3 2 c4! ? - position after 29 . . . .i.f6! -
A spectacular breakthroug h . However, a s
Volodya Kramnik rightly pointed out, the - the rook wants to break through on c4 (after
simple 32 .l:th 1 is also sufficient. l2Jxe6+ or l2Jb3). But also available to Wh ite
32 . bxc3
. .
is a nother, perhaps more convincing way.
Or 32 . . . dxc4 33 ii.b6+! 'ili'xb6 (33 . . . <iit c 8 34 30 l2Jxe6+ ! fxe6 31 Wb6+
ltJa5) 34 l:.d7+ <iitx d7 35 'ii' x b6 cxb3 36 Noth ing is promised by 31 l:txg8 .U.xg8 32
'i'xb7+ <iit e 8 37 'ili'c6+ <iit d 8 38 'iVa8+, and 'ifb6+ <iit c 8 33 'ifxd6 (without check ! )
Black loses his rook. 3 3 . . .'ifd7.
33 .l:.c1 ii.b4 31 . . . <iit b 8 32 Wxd6+ <iit a 8
No better is 33 . . .'ifxb3 34 l:.xc3+ 'i!fxc3 35 Now there is no time for the exchange on g8,
bxc3 l:txg7 36 .i.c5! (but not 36 Wb6+? <iit c 8 since the c2-pawn is attacked .
37 i.c5 .i.d8 with chances for both sides). 33 'ifc5 ! !
34 .i.d4 This excellent stroke , found b y Vadim Zviag­
34 i.c5? .i.xc5 35 .l:!.xc3 is a mistake in view i ntsev and Maxim Boguslavsky, decides the
of 35 . . .'ilt'd3+! . game i n White's favour. If 33 . . . <iit b 8 there
34 J:txg7 (34 . . . 'ild3+ 35 �a 1 c2 36 Wb6+
..
follows 34 lbg8 1:1xg8 35 .U.h8 :Xh8 36
'it>b8 37 1Vxb4) 35 .i.xc3 .i.xc3 36 l:bc3+ gxh8'if+ ii.xh8 37 'ifd6+ <iit c8 38 .i.b6 with
'it>d7 (36 . . . <iit d 8 37 'i!fb8+) 37 l2Jc5+ , and unavoidable mate.
Black has no defence. 33 . . . !1xc5 34 .Uxg8+ .i.c8 35 .i.xc5 <iit b 7 36
l:.xc8 ii.xg7 37 �g8 , and, as it is not hard to
In conclusion , let us check 29 l:tdh1 (with the see , Wh ite wins.
threat of 30 l:txg8 l1xg8 31 l::t h 8). In N u n n ' s
opinion , i t is insufficient because o f t h e reply Thus the position could have been won in
29 . . . i.f6 ! . various ways , and the path chosen by Nunn
(see diagram) was by no means the shortest. But this does
not imply that the win is achieved 'as you
An interesting way of attacking was later please' - success was not possible without
suggested by Artur Yusupov: 30 e5!? .i.xe5 the discovery of fa r from obvious combina­
(30 dxe5 3 1 l2Jxe6+! fxe6 32 l:txg8 .l:!xg8 33
. . .
tive subtleties and a very accurate calcula­
'i'b6+ <iit b 8 34 'i¥d6+ <iit a 8 35 .l:lh8) 31 l:t 1 h4! tion of variations.
1 62 �
I gor Khenki n , Vlad i m ir Kram n i k

Modern I nterpretation of the Dutch Defence

I. Igor Khenkin

W of the 'Stonewall' set-up in the Dutch

e will introduce you to the main ideas

Defence, and also touch briefly on the

Len ingrad Variatio n .
I have t o admit that t h e move 1 . . . f5 h a s
always seemed t o m e t o b e positionally not
altogether correct. When playing against it, I
usually avoid the main open ing lines, and
prepare some rare set-up, which neverthe­
less has a defin ite strateg ic basis. Such a
set-up will usually serve me faithfully for a
year or two , then my opponents adapt to it, Here Black has many possible moves. The
after which it has to be repa ired or replaced . most usual are 5 . . . d5 or 5 . . . c6 , while in
I n both of the afore-mentioned variations I recent times the idea of . . . c7-c5 h a s
aim to develop my king's knight at h3. The f3- appeared .
square remains free for the other knight, and 5 . . . d5 al lows me clear play on the dark
sometimes it does indeed move there. sq uares. The moves 6 tt:'ld2, 7 tt:'lh3 and 8
Let us make the following initial moves: tt:'lf3 follow al most a utomatically. Then the
1 d4 f5 knight goes from h3 to f4 (occasionally it can
2 g3 tt:'lf6 also be placed on g5) and the pawn to h4,
3 �g2 g6 after which Wh ite plays either h4-h5, or (if
this is prevented ) tt:'ld3 and �f4 .
Here for many years I successfu lly employed
an idea of Viktor Gavri kov. [Kramnik: All is not s o simple - if the knight
is developed at h3, the opponent acquire s
4 c3
quite a good counter-plan: . . . e 7-e6, t:Lle4,
Wh ite brings out his q ueen to b3, to prevent
. . .

. . . tt:'lc6, . . . a 7-a5, . . . b 7-b6 and . . . �a6. I

castl ing . A very u n usual set-up! Often the
have played this a couple of times with Black
black player proves unprepared for it. This is
and, in my opinion, here White has no
not surprising - he always also has plenty of
advantage. In reply to 5 . . . d5 1 am much more
problems in the mai n , more popular varia­
concerned about the simple 6 tt:'lf3.]
tions.
A few yea rs ago Sergey Gorelov tried 5 . c6
4. . . �g7
. .

against me. The game developed as follows:

5 'ifb3 6 tt:'ld2 'it'b6 7 tt:'lc4 'it'c7? ! 8 tt:'l h 3 , and then
9 0-0 , when Wh ite was evidently better. Later
(see diagram) I came to the conclusion that Black should
 Modern I nterpretation of the Dutch Defence lZJ 1 63

have exchanged q ueens by 7 . . . 'ifxb3 8 axb3 1 0 e4

lt:Ja6 , with chances of equalising . Neverthe­ This central break is normal in such posi­
less, I am satisfied with the resulting pattern tions. It would probably have been better to
of play. concern myself with maintaining equal ity, but
In 1 990, in a rapid-play tournament in I did not yet sense the danger.
Belgorod , Igor Glek chose 5 . . . c5. I n reply I
should have captu red 6 dxc5 ! , and after
6 . . . lt:Ja6 played 7 'ii'c4 . But at first sight it was
not clear why I shouldn't captu re the b7-
pawn . And I captu red it, but this did not tu rn
out well .
6 .ixb7? c4 7 'iVb4 (7 'ifb5? w a s bad : 7 . . a6 8
'i'b4 �xb7 9 'it'xb7 t"Llc6 with the irresistible
threat of 1 0 .. J::t a 7) 7 . . . lt:Ja6 8 �xa6 �xa6 9
tt'ld2? .l:!.b8 1 0 'ii'a 4 �b5 1 1 'it'c2 0-0 1 2 lt:Jgf3
d6. Despite the extra pawn , my position is
worse - the knight on d2 has no prospects ,
and the light squares are weak. There
followed 1 3 b3 �c7 1 4 a4? ! ( 1 4 bxc4 really
was better) 14 . . . �a6 1 5 b4 i.. b 7 1 6 0-0 e5
10 . . . b5
with advantage to Black.
1 1 a4 bxa4!
Instead of 9 lt:Jd2? it would have been
I had not considered this at all - I had only
stronger to play 9 lt:Ja3, but after 9 . . . l:.b8 1 0
reckoned with the attempt to win the d5-
"i'a4 'it'c8 all the same Black has excellent
pawn after 11 . . . c4 and 1 2 . . .fxe4 .
compensation for the sacrificed pawn . It is
hard for Wh ite to fight without his light­ 1 2 .l:!.xa4 �d7
square bishop! 1 3 l:ta3
However, i n view of the possibil ity of 6 dxc5, I thought for a long time about where to
I was still prepared to try and uphold this retreat the rook to , but even so I did not
variatio n . But soon Stuart Conq uest, an guess right. It would have been better to
international master from England, devised a place it on a2 , so that su bseq uently the b2-
set-up which would appear to put an end to pawn should be defended .
the plan with c2-c3 and 'ii' b 3. 13 . . . aS
1 4 c4 lt:Ja6
Khenki n-Conquest Black's position is clearly better, and he went
Gausdal 1 99 1 on to win the game.
5. . . t"Lla61 The set-up with . . . t"Lla6, . . . c7-c5 , . . . d7-d6
A logical move - Black prepares . . . c7-c5. and . . . t"Llc7 seems to be an excellent
With the queen on d1 it is poi ntless - White antidote to the plan of c2-c3 and 'ii' b 3. This
can even reply b2-b4 . plan can not be employed u ntil an improve­
6 lLJd2 c5 ment is devised for White .
7 d5 t"Llc7 I h a d t o change my opening weapon . I have
8 lt:Jh3 d6 to admit that, when playing against the
9 lt:Jf4 0-0 Leningrad Variation , I don't like placing my
1 64 � Modern I nterpretation of the Dutch Defence
pawn on c4 , and so I again tried to manage
without this move.
1 d4 f5 2 g3 tt:\f6 3 j_g2 g6
4 tt:\h3
My favourite knight manoeuvre!
4... j_g7
For some reason Black very rarely plays
4 ... d6 here, after which Wh ite should proba­
bly reply 5 tt:\c3 ! ? . The endgame after 5 . . . e5
6 dxe5 dxe5 7 �xd8+ �xd8 8 e4 is clearly in
Wh ite's favour. 5 . . . c6 6 d5!? j_g7 7 tt:\f4
leads to roughly the same situation as after
4 . . . j_g7. The move 5 . . . d5 is of independent
sign ificance. Wh ite's plan of action in this
case is simple: 6 0-0, then j_g5, tt:\f4 , e2-e3,
tt:\ce2 and tt:\d3. I cannot guarantee an
advantage, but this kind of position appeals
to me.
Incidental ly, this move - 5 tt:\c3 - was played
in the game Korch noi-M .Gurevich (Rotter­
dam 1 990), in which Wh ite scored a fine win .
Here are the opening moves: 5 . . . j_g7 6 tt:\f4
c6 7 d5! e5! 8 dxe6 d5! (if 8 . . . �e7, then 9
e4 ! is very strong) 9 h4! 'ii'e 7 1 0 h5 g5 1 1 h6
j_f8 12 tt:\h5!? .l::i.g 8 13 tt:\xf6+ 'i!Vxf6 1 4
j_xd5!? j_e6? (Black should have accepted
the challenge) 1 5 j_xe6 'i!Vxe6 1 6 e4! j_xh6
17 'i!Vf3 ! fxe4 18 'i!Vxe4 .l::i. g 6 19 j_e3 tt:\d7 20
0-0-0 , with an appreciable advantage for
White.
5 tt:\f4 d6
After 5 ... 0-0 there is the unpleasant reply 6
h4. The game Savchenko--M alanyuk (Kher­
son 1 989) conti nued 6 . . . d6 7 c3 ! ? c6 8
'ikb3+ d5 (8 . . . 'it>h8 9 h5) 9 h5 g5 1 0 h6 with
the in itiative for Wh ite.
6 d5
It is also possible to transpose moves: 6 tt:\c3
followed by d4-d5.
6... c6
7 tt:\c3 0-0
If 7 . . . cxd5 the correct captu re is 8 tt:\fxd5 ! , in
view of the variation 8 ... e6 9 tt:\xf6+ j_xf6 9
e4! .
This position was reached i n the game
Khenki n-Wi lson (Gausdal 1 992 ). The theo­
retical continuation here is 8 e4 . But I had not
prepared for the game (in Swiss tourna­
ments you fi nd out only at the last minute
who you are playing against). My opponent's
rati ng was fa irly modest (around 2300), and I
decided that it was not essential to create
tension i m mediately - it would do no harm to
first consolidate the knight's position with
h2-h4 . But this once again confirmed how
dangerous it is to underestimate your oppo­
nent - he manoeuvred very confidently.
8 h4? !
Threatening h4-h5-h6.
8 . . . tt:\g4!
An excellent response! If 9 h 5 Black replies
9 . . .'iVb6 ! , and only after 1 0 e3 or 1 0 0-0 -
1 0 . . . g5.
9 0-0 tt:\e5
1 0 e4 tt:\a6
1 1 exf5
It is unfavourable to conti nue 1 1 h5 g5 1 2
tt:\e6 j_xe6 1 3 dxe6 g4 (or 1 3 . . . f4! ? ) .
11 . . . j_xf5
1 2 h5
I felt qu ite optimistic about this position,
expecting only 1 2 . . . g5 1 3 tt:\e6 j_xe6 1 4
dxe6 h 6 1 5 'i!Ve2 d5 1 6 f4 , or 1 2 . . . tt:\b4 1 3
hxg6 hxg5 1 4 tt:\e4 .
 Modern I nterpretation of the Dutch Defence l2J 1 65

12 . . . tt'lc71 a ) 18 'ii'd 4 i.e2 .

The knight is excellently placed here - it is
attacki ng d5 and defending the only weak­
ness in Black's position - the e6-square.
Black's position is probably already more
b) 1 8 tt'lxe7+ Wxe7 1 9 .:ixf3 tt'lxf3+ 20 ii'xf3
d5!.
c) [ The most tenacious is 1 8 tt'lef6+! exf6 1 9
11xf3 tt'lxf3+ 2 0 Wxf3 f5 2 1 il. f4 Dvoretsky.]
-

promising. d) 1 8 l:!.xf3 tt'lxf3+. In the event of 1 9 <iftg2

13 hxg6 hxg6 Black can simply retreat his knight: 1 9 . . . tt'ld4
14 tt'le4? 20 tt'lxc7 ii'xc7 2 1 c3 ii'c6 22 ii'g4 tt'lf5. But it
is probably even stronger to g ive a cu nning
A continuation of the same opti mistic tactics
check: 1 9 . . . tt'le 1 + ! 20 'ii'x e 1 (2 1 <iftf2 tt'lxc2!
- the knight aims for g5. It would have been
22 tt'lxc7 'ifxc7 23 'i!Vd 5+ e6! 24 'ii'x e6+
better to 'stand still ' - 1 4 'ii"e 2 or 1 4 lle 1 .
iff?+ ) 20 . . . tt'lxd 5 , and Black is a sound pawn
14 . . . cxd5! to the good . 1 9 'ii'xf3 tt'lxd5 20 tt'lg5 'ii' b 6+ 2 1
15 tt'lxd5 i.g4 <iftg2 tt'lf6 is no better for Wh ite.
1 6 f3 16 . . . tt'lxf3+?
Of cou rse, I didn't want to play 1 6 'ifd2 tt'lf3+, Fortu nately, my opponent did not see what
but the move i n the game is no better. Now was bad about this move, and so he did not
Black was faced with solving a fairly simple consider the capture with the rook. I nciden­
little problem i n the calculation of variations. tally, when playing i n Swiss tou rnaments I
have noticed that less skilled players usually
go wrong in a tense tactical battle. At
decisive moments I have often succeeded in
out-calculating them.
17 l:txf3 !
Here everything became clear to my oppo­
nent, but it was no longer possible to change
course.
17 . . . il.xf3
1 8 i.xf3 ltf5
1 9 tt'lg5
The two minor pieces are much stronger
than the rook and pawn , especially since the
black king comes under attack.
With what should he take on f3? The most 19 . . . i.xb2
natural captures lose, for example: 1 6 . . . i.xf3? This loses by force.
1 7 llxf3! tt'lxf3+ 1 8 i.xf3 l:.xf3 1 9 tt'lxc7 , or 20 tt'lxc7 i.xa 1
16 . . . tt'lxf3+? 1 7 .l:txf3! i.xf3 1 8 i.xf3 .l:!.xf3 1 9 21 i.d5+ <ifth8
liJxc7 . Correct was 1 6 . . . l:txf3 ! ! (I saw this
22 �g2 ! l:lxg5
possibil ity only after the game). The cun n i ng
23 i.xg5 'ii'x c7
point is that all the subsequent captu res on
f3 are made with gain of tempo - with check 24 1i'h 1 + <iftg7
or with an attack on the white quee n . For 25 'ii' h 6 mate
example, 1 7 i.xf3 i.xf3 (or 1 7 . . . tt'lxf3+ 1 8 Everything tu rned out wel l , although it could
:Xf3 i.xf3): have been the other way rou n d .
1 66 � Modern I nterpretation of the Dutch Defence
For the moment this variation su its me for
White, although , of cou rse , the next time I
won't play 8 h4? ! , but the right move 8 e4 ! . I n
the opening you should fight in the centre,
and not on the wing.
After 1 d4 f5 2 g3 lLlf6 3 .ltg2 Black
someti mes changes his move order i n the
opening and plays 3 ... d6. Now it is rid iculous
to place the knight on h3 because of the reply
4 . . . e5. Of course, one can simply play 4 lLlf3 ,
transposing into the main l i nes of the Dutch
Defence. But I don't like developing my
knight on f3 and, when I encou ntered this
problem a couple of years ago, at the board I
devised a new pla n .
Khenki n-Vasyu kov
Voskresensk 1 990
4 lL\c3 ! ?
If now 4 . . . e 5 , then 5 dxe5 dxe5 6 'ii'x d8+
'it>xd8 7 e4 with the better endgame for
White. The reply 4 . . . d5!? deserves serious
consideration. Wh ite continues 5 ltJf3 and 6
0-0 . This leads to a rather unusua l , compli­
cated position, with chances for both sides.
4... c6
5 e4 fxe4
6 lL\xe4 lL\xe4
7 .ltxe4 ..tf5
8 'it'f3
8 ..txf5 'ii'a 5+.
8. . . ..txe4
9 'iix e4 'ii'a 5+
1 0 c3 'ii'd 5
Here Black offered a d raw. Evgeny Vasyukov
is an active player and it was apparently
ted ious for h i m to play without the queens .
B u t i n t h e endgame I a m n o t i n any danger ­
at best Black can hope for equal ity.
1 1 'ii'x d5 cxd5
1 2 lLlh3
1 2 lLle2 is eq ually good . Wh ite then castled ,
placed his rook on e 1 and his knight on f4,
and retai ned a slight but stable i n itiative.
After a lengthy struggle I managed to win.
I like the situation arising after 4 lL\c3 ! ? , and I
am ready to play it aga i n . I n recent ti mes I
have come to real ise that it is not so
important whether you gain an opening
advantage with Wh ite - this is problematic in
any opening. The main thing is to obtain
'your' position, in which you feel more
confident than you r opponent.
It is now time to tu rn to the main topic of our
lectu re - the 'Stonewal l ' .
1 d4 f5
2 g3 lLlf6
3 ..tg2 e6
4 c4
Wh ite has also tried i m med iately developing
his knight at h3 with his pawn stil l on c2 . But
I think that after 4 lLlh3 the reply 4 . . . c5!? is
u npleasant (qu ite logica l : as soon as the
knight has moved onto the ri m , Black attacks
the centre).
4 . . . d5
For some reason this is the reply I have
faced , although it is considered more accu­
rate to beg i n the construction of the 'stone­
wall' with 4 . . . c6! ? , waiting to see where the
g 1 -kn ight goes. After 5 lLlh3 Black places his
 Modern Interpretation of the Dutch Defence
pawn not on d5, but on d6, preparing . . . e6-
e5. For Wh ite I can recommend that you
check 5 d 5 ! ? , which , I think, as yet no one
has played .
Of course, I must also mention the l lyin­
Genevsky Variation (4 . . . .i.e7 , 5 . . . 0-0 and
6 . . . d6). 4 . . . .i.b4+ has also been employed ,
with the idea after 5 .i.d2 of retreating
5 . . . .i.e7 - the bishop on d2 is not too well
placed . 5 lDd2 is stronger, when it is not clear
what the black bishop is doing on b4.
5 lDh3!
For Black this plan is the most dangerous.
When I made my first g randmaster norm , I
scored a very important win in this variation
against Vlad imir Tukmakov.
Khenkin - Tu kmakov
Metz 1 99 1
5. . . .i.e7
5 ... .i.d6 6 0-0 c6 has also been played . If 7
i.f4 Black should reply 7 . . . .i.e7 ! . I n the
game Bareev-Vaisser (Pula 1 988) after 8
'ii'b 3 0-0 9 lDa3 (if 9 lDc3 there is the
unpleasant reply 9 . . . 'ii' b 6) 9 . . . h6!? 1 0 l:lad 1
g5 1 1 i.d2 a5! 1 2 f3 b5! Black seized the
in itiative . But later Kozul improved Wh ite's
play against Bareev (Biel 1 99 1 ): 8 lDd2 0-0 9
'ii'c2 h6 (after 9 . . . lDbd7 1 0 cxd5 Black does
not have 1 0 . . . exd5, the essential capture in
l2J 1 67
such cases - his f5-pawn is 'hanging') 1 0
.i.xb8 ! l:.xb8 1 1 lDf4 Wh ite's chances are
.
better - he places his knig hts on d3 and f3,
and then he beg ins an attack on the
queenside, by advancing his b-pawn .
7 b3 fie? 8 .i.b2 0-0 9 lDd2 is another
promising set-up. Then the d2-kn ight goes to
e5, and the other knight is qu ite well placed
on h3, controlling the f4-point. (If it had come
out in the opening to f3 , to obtain a similar
construction Wh ite would have had to spend
a couple more tempi: lDf3-e5-d3 and lDd2-
f3-e5). One of the possible subsequent
plans is 'ifc2 , l:tad 1 , lDf4 , f2-f3 and e2-e4 .
6 0-0 0-0
7 b3
7 'ili'c2 is also possible, but I prefer fi rst to
develop my bishop - who knows, perhaps
the q ueen will also come in useful on d 1 .
7 . . . c6
Grand master N igel Short, who constantly
employs the Dutch Defence with Black,
plays 7 . . . lDc6 ! ? and then . . . a7-a5 i n such
positions. The knight presses on d4, thereby
preventing the exchange of the dark-sq uare
bishops (8 .i.a3? .i.xa3 9 tDxa3 dxc4 ). After 8
.i.b2 a5 Wh ite is obl iged to make the not very
useful move 9 e3 - otherwise he ca nnot
develop his q ueen's knight. However, even
in this case I prefer Wh ite's position.
8 i.b2
The exchange of bishops does not bring
Wh ite any particular benefits - after 8 i.a3
.i.xa3 9 lDxa3 the black q ueen obtains a
comfortable post at e7, whereas the knight at
a3 is badly placed , and it faces a lengthy
journey via c2 and e 1 to d3 or f3 . If the king's
knight were at f3 , controlling the e5-point,
such a plan would make sense, but here,
while the q uean's knight is on its way, Black
will surely have time to play . . . d5xc4 and
. . . e6-e5 .
8 . . . lDe4
Black has problems with the development of
1 68 � Modern I nterpretation of the Dutch Defence

his queen's knight: 8 . . . tt:Jbd7? 9 lt:Jf4 is bad 1 1 'iVc2

for him. For the same reason - the weakness For the moment 1 1 f3 is prematu re in view of
of the e6-point - the normal (with the knight 1 1 . . . exf3 1 2 exf3 c5! .
on f3) plan of . . . b7-b6 and . . . ..tb7 does not 11 . . . 'iVe7? !
work.
T h e threat was 1 2 .l:lad 1 and 1 3 f3 . 1 1 . . .'ii'c7
Tukmakov plans . . . ..tf6 and . . . c6-c5, not 1 2 f3 exf3 1 3 exf3 e5! was more tenacious,
fearing 9 f3?! lt:Jd6 (or 9 . . . 4:Jf6) . I was able to but after 1 4 dxe5 ..txe5 1 5 ..txe5 'ikxe5 1 6
find a strategic refutation of his idea. With his l:tfe 1 White has a n obvious advantage. I f is
other knight on c6 , . . . tt:Je4 would have been very d ifficult to find a good post for the c8-
more justified - Black would have retai ned bishop.
pressu re on the opponent's centre. 12 f3 c5
Theory mainly deals with 8 ... 'ike8 . It is 1 2 . . . exf3 1 3 exf3 c5 14 d5! exd5 1 5 cxd5.
recommended that 8 ... b5 should also be
13 fxe4!
considered . But I am sure that Black's
Here 1 3 d5 is no longer so good i n view of the
problems cannot be solved in this way -
reply 1 3 . . . e3! .
Wh ite can simply reply 9 lt:Jd2 or 9 c5.
13 . . . ..txd4+
[Kramnik In the 'stonewall' the . . . b7-b5
1 4 .Yi.xd4 cxd4
advance has always seemed dubious to me
- a whole complex of squares is immediately 1 5 exf5 exf5
weakened , and Black does not gain any real 1 6 lt:Jf4 tt:Jc6
cou nterplay.] 1 7 .l:r.ad 1 ..td7
9 lt:Jd2 ..tf6 Now White has to find an energetic plan . He
only needs to delay slig htly, by allowing his
opponent time to play . . . 'ifi>h8, . . . l':.ad8 and
. . . ..tcB , and his entire advantage will evapo­
rate. After a l l , my position has its weakness­
es on the e-file. Of cou rse, I need to play e2-
e3.
18 c5! 'it>h8
18 . . . .l:r.ad8 1 9 b4 ! .
1 9 lt:Jd5 'ii'e 5

1 0 tt:Jxe4!
I exchange the opponent's only active piece,
and then open l ines in the centre by f2-f3 , to
exploit my lead in development. It is impor­
tant that the pawn is stil l at e2 - after the
exchange on f3 Wh ite will recapture with this
pawn .
10 . . . dxe4
 Modern I nterpretation of the Dutch Defence lZJ 1 69

20 e3! dxe3 28 . . . .l:lxd5

I also had to reckon with 20 . . . i.e6 . After this 29 i.xd5 'ii'x d5
I was i ntending 21 l:tfe 1 !! i.xd5 22 exd4 �f6 30 'ifxa5 �xc5
(22 . . . tt:'Jxd4 does not help: 23 l:t.xe5 tt:'Jxc2 24 3 1 'i!fc3 ! Wxb5
l:tdxd5! f4 25 gxf4 .l:txf4 26 l:.d7) 23 i.xd5
3 2 'ii'e 5! a6
l0xd4 24 'iff2 ! tt:'lc6 25 l:te6 with a winning
positio n . 33 l:tb3!

21 .l:ife 1 f4? ! Wh ite forces the most favourable version of

the exchange.
Tukmakov is a n active player, and he tries to
avoid going totally onto the defensive. 33 . . . 'ii'x e5
However, after 21 .. J1ae8 22 l:txe3 'i!fb8 23 34 fxe5 l:tf7
l:!.de 1 l:.xe3 24 l:.xe3 Black's position would 35 l:t.eb1 ! .:te7
also have remai ned d ifficult - his bishop on 36 l:txb7 .Uxe5
d7 has no futu re, and the threatened pawn 37 .l:tbB+ .tea
advance b3-b4-b5 is extremely u n pleasant.
38 .UaB 'itgB
The ending arising after 24 .. J1e8 25 1i'c3
.l:.xe3 26 'ii'x e3 'i!fe5 27 Vxe5 tt:'Jxe5 28 tt:'le7! 39 l:tbbB 'it>f7
is obviously i n Wh ite's favou r. 40 l:tb7+! 'itf6
22 gxf4 'ii' h 5 41 .Uxa6+ �5
23 l:lxe3 l:tadB 42 .Uxg7 i.g6
24 l:.de1 ! £L g4 43 �2 �g4
25 Vc4 44 lla3 l:tf5+
Black would have compensation for the 45 �e2 .l:th5
pawn if he were able to drive the knight from 46 h3+!
d5, but it is not possible to ach ieve th is. Black resigned .
Initially I wanted to exchange someth ing, to
simpl ify the position , but then I realised that
this was incorrect strategy. The active
I I . Vlad i m i r Kra m n i k
placing of the wh ite pieces should be
exploited for an attack.

I 'stonewall' system with the wh ite knight on

25 . . . 'i!ff5 w i l l show y o u t h e m a i n branches o f the
26 b4 i.h5
f3 . First I will express my views on this
27 b5 tt:'Ja5
opening set-up, and then I will g ive a
No better was 27 . . . i.f7 28 bxc6 iLxd5 29 theoretical review and show you a few
..txd5 .lbd5 30 cxb7 .Uxc5 31 1t'b4 l:.b5 32 games.
'i'xf8+! 'ii'xf8 33 .:tea �g8 34 lt 1 e7! , or
The 'stonewall' is one of the few opening
28 . . . bxc6 29 tt:'le7! i.xc4 30 tt:'Jxf5 l:txf5 31
systems i n which Black immediately gains
l:1e8+ l:r.f8 3 2 i.xc6 i.xa2 3 3 .:.Xf8+ .l:1xf8 34
some advantage in space. But, of cou rse,
..td7.
there is a price to pay for this - the
28 Vc3 weakening of the dark sq uares i n his
It was possible to support the knight with 28 position . However, it is not easy for Wh ite to
'i'd4 followed by .l:!e5 , but I decided to force exploit the e5-point - th ings often end in a
matters. simple exchange of pieces on this square.
1 70 \i? Modern I nterpretation of the Dutch Defence
The central idea of Black's strategy is the
restriction of the g2-bishop. In my view, here
it is no stronger at all than the bishop on c8 .
I have played this opening for both sides and
as a result I have become aware that it is
even simpler to play for Black than for Wh ite .
At any event, it is usually more difficult for
White to choose a pla n . His actions must
often vary depending on the opponent's
plans, i.e. he has to adapt flexibly to Black's
play, which is never easy. I like to employ the
'stonewall' as Black against a player with an
attacking, combinative style, because here
Wh ite will not give mate , and the strategic
problems someti mes prove too difficult for
such players.
I n one order or another let us make the i n itial
moves:
1 d4 fS
2 c4 lZ'lf6
3 lZ'lf3 e6
4 g3 dS
5 ..1ll. g 2 c6
6 0-0 il.d6
After Black castles, he has to choose one of
two ways of developing his q uean's bishop.
Sometimes he directs it via d7 to e8,
vacating the d7 -square for the knight. Later
the bishop can be taken further, to h5, after
which it can be exchanged , or . . . c6-c5
prepared .
But Black more often plays . . . b7-b6 , . . . .ib7
and . . . lZ'lbd 7 , developing his queenside as in
the Queen's I ndian Defence or the Catalan
Opening. The main d ifference with these
openings is the position of the pawn on f5.
Wh ite's thematic e2-e4 advance is now
g reatly h indered (if Wh ite prepares it with f2-
f3, Black i m med iately cou nters with . . . c6-
c5). On the other hand, after Black carries
out his basic plan - completes his develop­
ment, places his rooks on c8 and d8, and
advances . . . c6-c5 - pawn exchanges occur
in the centre and the weakness of the e5-
square may become perceptible.
Wh ite usually chooses one of two continua­
tions: 7 il.f4 or 7 b3. 7 lZ'lbd2 (or 7 'i¥c2) has
no i ndependent sign ificance - all the same
he cannot manage without b2-b3. The move
7 lZ'lc3 is not very dangerous for Black, and in
general it seems to me that the knight is not
best placed at c3 - all the time Wh ite has to
reckon with . . . d5xc4 . However, this is some­
times played with the idea of developing the
bishop at g5 on the next move. 7 l2Je5 also
occasionally occu rs .
I think that the development of the bishop at
f4 is more logica l . With it on b2, both white
bishops lack prospects: one is obstructed by
its own pawn on d4, and the other by the
enemy pawn on d 5 . I n cidental ly, I agree with
Khenkin : it is more d ifficult for Black to
defend if the opponent's king's knight is
developed not at f3 , but at h3.
Plan with b2-b3
7 b3 'iie 7
If 7 . . . 0-0 there fol lows 8 ..ta3, and White's
position is better - he gains control of the e5-
sq uare. There is no point i n al lowing the
exchange which is adva ntageous to White.
8 ..lll. b2
8 lZ'lbd2 is also played .
 Modern I nterpretation of the Dutch Defence ltJ 1 71

First let's analyse the plan with . . . .i.c8-d7--e8 . hardly doing anything here. I magine that
8 . . . 0-0 Black has not managed to exchange it: the
9 ltJbd2 knight has gone from f3 to e5, and the e2-
.i.d7
pawn has moved to e3. When Black advanc­
10 ltJe5 i.e8
es . . . c�5. the g2-bishop i m med iately
1 1 lDdf3 beg ins operating on the long diagona l , and
White's moves are natu ra l , but if you ask there is nothing to oppose it.
what does he want, what is his plan, it is not 8 . . . b6
easy to obtain a sensible reply. Most This move is probably more accu rate than
probably he has to adapt to the actions of his 8 . . . 0-0 .
opponent.
9 'iVc1
11 . . . �h5
Wh ite insists on the idea of exchanging
12 'ifc2 ltJbd7 bishops. A completely harmless undertaking!
1 3 ltJd3 I!ac8 It simply wastes too many temp i . Once I
Black prepares . . . c6-c5. myself played this, but I i m mediately ended
14 .l:.ac1 u p i n a somewhat inferior position .
9 . . . �b7
1 0 i.. a 3 ltJbd7
11 �xd6 'ii'x d6

The game Schmidt-Haba (Prague 1 989)

developed as fol lows: 14 . . . ltJe4?! 1 5 ltJfe5
lt:lxe5? 1 6 dxe5 i.. c7 1 7 f3 lDg5 1 8 h4 ltJf7 1 9
lt:lf4 .i.g6 20 h5, and the battle was decided . 12 'ika3?!
A curious example, but it vividly demon­ 12 e3 may still maintain the balance, but in
strates how ru inous superficial decisions can several games Wh ite went i n for the ex­
be (such as 14 . . . ltJe4?!). change of queens. For example: Alburt­
14 ... c5! is correct, with good chances of Short (Subbotica 1 987) and Goldi n-Dol­
equalising, although accu racy is still requ i red matov (Kiaipeda 1 988).
- after all, Wh ite can maintain the tension i n 12 . . . 'ili'xa3
the centre b y playing, say, 1 5 'ifb1 ! ?. 1 3 ltJxa3 1;e7
In principle, this entire plan does not really In my view, Black's position is preferable. His
appeal to me. Of cou rse, at h5 the bishop is king is closer to the centre and he is ready to
better placed than at c8 , but even so it is play . . . c�5 .
1 72 \t> Modern I nterpretation of the Dutch Defence
I n cidentally, after 8 . . . 0-0 (instead of 8 . . . b6) 9
1Vc1 b6 (9 . . . b5!?) 1 0 ..ta3 ..tb7 1 1 ..txd6
1Vxd6 1 2 'iVa3 Black can play the same
endgame, although without his king o n e7 -
here there is nothing terrible for h i m . But he
is not obliged to go i nto the endgame -
1 2 . . . c5 is not bad .
If Wh ite really wants to exchange the dark­
square bishops on a3, he should do this by 8
a4 and 9 ..ta3.
8 a4 aS
8 . . . 0-0 9 ..ta3 ..txa3 1 0 tt:Jxa3 a5 is equally
good , but not 11 a5! with the better chances
for Wh ite, as in the game Kasparov-Short
(London 1 987).
9 i.a3 ..txa3
1 0 tt:Jxa3 0-0
Here are a few examples showing how this
position is handled :
1 1 tt:Jes b6 1 2 tt:'lc2 ..tb7 1 3 tt:'le1 tt:'la6 1 4
cxd5 exd5
I thi n k that 1 4 . . . cxd5 is also good enough to
maintain the balance, but when a2-a4 and
. . . a7-a5 have already been played , Black
usually captu res . . . e6xd5. He then plays his
knight to b4 and advances . . . c6-c5.
15 tt:'l1 d3 c5 1 6 e3 .l:.ac8 1 7 l::r. c 1 tt:'le4
Black is excellent placed ; his next moves will
be . . . tt:'lb4 and . . . ..ta6 .
Akopian-U iybin (Mamaia 1 99 1 ): 1 1 tt:Jes
b6 1 2 'ifc2 ..tb7 1 3 l:fc1 (th reatening 1 4 c5)
1 3 . . . tt:'la6 1 4 cxd5 cxd5
Forced : 1 4 . . . exd5 is n ow i mpossible, while if
14 . . . tt:'lb4 , then 1 5 d6! 'i!Vxd6 1 6 'ii' b 2 followed
by tt:'lac4 is strong.
The resulting situation seems ha rmless, but
in analysis I have n ot i n fact fou n d a clear-cut
way to equal ise. Wh ite has n oth ing special
here , of cou rse , but nevertheless Black
faces a lengthy and ted ious defence in a
slig htly inferior position . I will g ive a few more
moves: 1 5 tt:'lb5 l::tfc8 1 6 'ifd2 tt:'lb4 1 7 h3
tt:'le4 1 8 ..txe4 dxe4 1 9 lixc8+ .:!.xc8 20 l:tc1
lbc1 + 21 1Y'xc1 tt:'ld5 22 e3 g5?! 23 'i'd1 !
i.a6 24 'ii h 5, and Black faces d ifficulties.
It is u nderstandable, therefore, why in the
prepa rations for my game with Vladimir
Akopian i n the 1 99 1 USSR Championship I
did not go i n for this with Black, but chose a
d ifferent set-up - I decided to fight immedi­
ately for the e5-point.
Akopian-Kra m n i k
Moscow 1 99 1
1 1 tt:Jes tt:'lbd7
With this move Black succeeds i n driving
back the enemy knight, but on the other hand
his knight is less actively placed at d7 than at
a6. Which method of development is better
will be revealed by future games.
 Modern I nterpretation of the Dutch Defence ttJ 1 73

1 2 tt::l d 3 b6
1 3 'ii'c 2 ii.a6
1 4 l:tfc1 l:tac8
1 5 'i¥b2 tt::l e4

A l ittle exercise for you : should Black play

1 1 . . . tt::l b d7 ?
No, he should n 't! After 1 2 cxd5! cxd5 (forced )
1 3 tt::l d c4 ! Wh ite excha nges his knight for the
dark-square bishop and gains the advantage
After a series of natural moves we have (Tu kmakov-Dolmatov, Odessa 1 989). This
reached a position which did not g reatly is the point of the set-up with the knight on e5
appeal to me, although objectively it is close - to h inder the development of the knight at
to equal. Neither side appears to have any d 7 . ( I magine that instead of tt::l e 5 Wh ite has,
serious plan . Of course, Wh ite has to reckon say, made the move e2-e3 - Black plays
with . . . c6-c5 and also even a possible . . . g7- 1 1 . . . tt::l b d7 , soon advances . . . c6-c5, and he
g5, and Black with b3-b4 . Akopian decided is perfectly alrig ht. )
on immed iate activity, although the u n h u rried However, the knight also has a nother satis­
16 .U.c2 !? and 1 7 l:tac1 came i nto considera­ factory route.
tion. 11 . . . a5
16 b4 axb4 Black is intending 1 2 . . . tt::l a 6 and at some
17 tt::l x b4 ..tb7 point . . . c6-c5 - q u ite a logical pla n .
1 8 e3 c5 It h a s been suggested that 1 1 . . . tt::le4 should
19 tt::l d 3 cxd4 be considered . Then there can follow 1 2
D raw ag reed . After the game Akopian and I tt::lxe4 ( 1 2 cxd5 cxd5 1 3 tt::l xe4 dxe4 should
found a long, forcing variation, which con­ also be checked ) 1 2 . . . fxe4 1 3 f3 exf3 14 exf3
cluded 1 5 moves later i n a d rawn endgame, (noth ing is g iven by 1 4 .:txf3 l:txf3 1 5 ..txf3
but now I am unable to remember it. tt::l d 7 1 6 cxd5 tt::l x e5 ! ? 1 7 dxe6 ..tcS+). The
critical reaction for Black is 1 4 . . . ..txe5 1 5
dxe5 c5 (th reatening 1 6 . . . d4) 1 6 cxd5 .i.xd5.
Let us turn to the set-up with the bishop on
If he should succeed i n completi ng his
b2.
development, he wil l , of cou rse, be q u ite
8 tt::l b d2 b6 alright, but on the other hand after 1 7 f4 he
9 tt::l e 5 ii.b7 seriously has to reckon with f4-f5 . Here
1 0 ..tb2 0-0 add itional analysis is req uired .
1 1 .l:tc1 The position after 1 1 . . . aS has occurred many
1 74 � Modern I nterpretation of the Dutch Defence
times in practice.
Chiburdanidze-Agdestein (Haninge 1 988):
12 ltJd3 ltJa6 ( 1 2 . . . ttJd7 is also possible) 1 3
ltJf3 ltJb4 1 4 c5 (there is noth ing else - the
a2-pawn is attacked ) 14 . . . bxc5 1 5 dxc5 il.. c 7
16 a3
Of course, White was counting on 1 6 . . . ttJxd3
1 7 exd3, but Simen Agdestein unexpectedly
retreated his knight - via a6 and b8 to d7. A
strong and original manoeuvre!
16 ... ltJa6! 1 7 ltJfe5 ltJb8
Wh ite appears to stand wel l , but in fact she
has nothing. Black will now exchange the
knight on e5, then place his bishop on a6 and
rook on b8, and endeavour to play . . . e6-e5.
18 f3 ltJbd7 19 ltJxd7 ttJxd7 20 e4 fxe4 21
fxe4 .l:!.xf1 + 22 'il'xf1 il.. a 6 23 'i!fd 1 �b8
Black has seized the in itiative .
Tu kmakov-Agdestein (Dortmund 1 987):
12 ltJb1
White plays his knight to c3, in order then to
exchange pawns on d5 and in the event of
. . . c6xd5 to occupy the b5-sq uare with the
knight. But this plan is slow, and now it is not
obligatory for the black knight to go to a6.
12 ... ltJbd7 1 3 cxd5 cxd5 14 ltJc4
(see diagram)
1 4 . . . b5!
Black is agreeable to the exchange of a
couple of minor pieces. After a l l , his pawn will
now advance to b4 , sh utting the opponent's
bishop and knight out of the game.
15 ttJxd6 'i¥xd6 16 ltJc3 il.. a 6 17 'ifd2 �feB
1 8 f3 b4 1 9 ltJd 1 a4 20 ltJe3
In the event of 20 bxa4 both wh ite pawns on
the a-file will soon be lost.
20 . . . a3
Black has an obvious advantage. The result­
ing position illustrates well the thoughts
expressed earlier about the unenviable fate
of the wh ite bishops in this variation. Com­
pare the light-square bishops on g2 and a6 -
which of them is bad?
The following is a more solid continuation for
Wh ite:
1 2 e3 ltJa6
1 3 'ii'e2
Black usually repl ies 1 3 . . . ltJe4 . I n the g a me
Petursson-Dolmatov (Aku reyri 1 988) Black
carried out a different, although strategically
rather risky pla n . I have to admit that it does
not g reatly appeal to me.
13 . . . il.. xe5?! 14 dxe5 ltJd7 1 5 .Ufd 1 ttJac5 1 6
ltJf3 l:tac8 1 7 il.. a 3 .l::t fe8 1 8 'ifb2
(see diagram)
 Modern Interpretation of the Dutch Defence 1 75

1 8 fxe4 axb3
1 9 axb3 �xeS
20 dxeS l:iad8
21 exdS exdS

1 8 ... gS!
At first sight a senseless decision. But i n fact,
with the centre closed it is hardly possible to
exploit the weakening of the kingside. And
Black, by advancing his pawn to g4, will
threaten the e5-pawn . Black's position is preferable. He will now
1 9 ttJe1 g4 play . . . c6-c5 and later either follow u p with
The next move plan ned is 20 . . . 'i!Vg7 . True, if . . . d 5--d4, or force the advantageous ex­
the wh ite q ueen retreats to a 1 , it will be very change on d 5 . He will place one of his pieces
dangerous to captu re the e5-paw n . On the on e6, blocki ng the e5-pawn and thereby
other hand, the mobil ity of the g2-bishop is restricti ng the mobil ity of the bishop on b2 .
now still fu rther restricted , and it is u n l ikely The position of the wh ite king is somewhat
that the wh ite knight will manage to reach f4 . weakened and in the future it may come
At any event, Margeir Petu rsson hastened to u nder attack.
simpl ify the play. Here is the conti n u ation of the game, without
20 �xeS ttJxcS 21 lLld3 ttJxd3 22 l:Ixd3 any notes.
The position would appear to be roughly 22 'ii'e 3 cS 23 'ife2 �a6 24 .l::t a 1 d4 2S �c1
equal . bS 26 �f4 'i¥e6 27 'iVhS?! bxc4 28 �h3
'i¥b6 29 bxc4 �xc4 30 l::!. d b1 'i!Vc6 31 .l::ta 7
Petu rsson-Short 'i!Ve4 32 .Uba 1 �dS 33 l:xc7 'ii' h 1 + 34 �2
Reykjavi k 1 987 l:.xf4+ 3S gxf4 'ii'x h2+ 36 'lt>e1 'iig 3+ 37
Black won this game is classic style. 'it>e2 i.. c4+ 38 'it>d1 'it'd3+ Wh ite resigned .
13 . . . ttJe4 As you see , i n all the examples we have
1 4 l:.fd 1 ttJc7 examined n oth i n g special was demanded of
Black - he simply completed his develop­
Later the provocative 1 4 . . . ttJb4 ! ? 1 5 a3 ttJa6
ment and fully prepared h imself for the
was also tried , when Wh ite has problems
opponent's active possibil ities. The e3-e4
with his a3-pawn.
advance did not promise any particu lar
15 f3 ttJxd2 benefits and was double-edged , but Wh ite
1 6 'i!Vxd2 a4 does n ot appear to have any other effective
1 7 e4 fxe4 pla n .
1 76 � Modern I nterpretation of the Dutch Defence

After it became clear that the set-up with

i.. b2 and lt:Jbd2 does not promise anyth i n g , a
new idea appeared .
8 lt:Je5
White wants to prevent the development of
the bishop on b7. If 8 . . . b6? he has the strong
reply 9 cxd5! cxd5 10 lt:Jc4 . I should mention
that I consider the position arising after
9 . . . exd5 to be u nfavourable for Black. It
leads to a structu re typical of the Queen's
I ndian Defence, but with the pawn worse
placed on f5 , where it is doing noth i n g . Of
course, this evaluation only appl ies with the
pawn on b6 - if it is still o n b7, there are no - position after 1 2 tt:Jd3 -

objections to the captu re . . . e6xd5 .

8 . . . 0-0 Black acted more logically i n the following
9 i.. b2 game.
9 lt:Jd3 has also occurred , but then Black
again has the opportu n ity to fianchetto his Kir.Georgiev-Knaak
bishop. East Germany-Bulgaria Match 1 987
9 . . . i.. d 7 12 . . . i.. h 5
1 0 'ili'c1 i.. e 8 If Black does place his bishop o n f7 , it will be
1 1 i.. a 3 lt:Jbd7 only after l u ring the knight to f4 .
In contrast to the similar situation which we 1 3 l:!.e1
analysed earl ier, when the bishop has been Or 1 3 lt:Jf4 i..f7 1 4 i.xd6 'ii'x d6 1 5 'ii'a 3 'iic 7,
moved to e8 the transition into the endgame and Black has the idea of . . . d5xc4 followed
by 11 . . . i.. x a3 1 2 'ii'x a3 'ii'x a3 1 3 lt:Jxa3 by . . . e6--e5.
condemns Black to a lengthy and passive
13 . . . .l:!.ae8
defence. It will hardly be possible for him to
1 4 i.. x d6 'ii'x d6
play . . . c6-c5, and Wh ite will beg i n a pawn
offensive on the q ueenside. 1 5 'ii'a 3 'iic 7
1 2 lt:Jd3 1 6 lt:Jd2 lt:Je4
1 7 lt:Jf3 dxc4!
1 8 bxc4 c5
(see diagram) A transformation of the centre wh ich is
typical of this variation .
1 9 l:tac1 a6
I n the game Ti mman-5hort (Brussels 1 987) 1 9 . . . cxd4 ! ? .
the Engl ish grandmaster made the rather
20 e3 i.. xf3 !
pointless move 12 . . . i..f7 . After 1 3 i.. x d6
'Wxd6 14 'ii'a 3 'Wxa3 1 5 lt:Jxa3 .l:!.fe8 (intend­ Only now, when Wh ite can n ot recaptu re with
ing 16 . . . e5) 16 f4 Jan Timman reached a the e-pawn .
better endgame and later he converted it i nto 2 1 i.. xf3 e5
a win . 22 i.. x e4 fxe4
 Modern I nterpretation of the Dutch Defence ttJ 1 77

23 ttJxc5 exd4 correct 12 ... a4! 1 3 ttJxd7 (in the event of 1 3

24 exd4 ttJxc5 bxa4 Black now begins exchanging: 1 3 . . .
25 'ifxc5 'ifxc5 i.xe5, and the pawns o n the a-file are very
weak) 1 3 . . . axb3 1 4 axb3 i.xd7 Wh ite's
26 dxc5 .l:tf5
position is in no way better.
27 .l:tb1 .U.e7
The double rook ending is d rawish .
To conclude this section I should l ike to show
you one of my own games.
Later an even more convincing plan of
defence was fou n d .
9. . . tiJbd7 ! ?
Van Wely - Kra m n i k
1 0 tiJd2 a5
E u ropean J u n ior Championsh i p ,
Arn hem 1 990
7 b3 'ife7
8 i.b2 b6
9 tiJbd2 i.b7
1 0 ttJe5 0-0
11 tiJdf3
Apparently my opponent was not very skil led
in the subtleties of the opening - he al lows
Black to develop his knight at d7 without any
hindrance.
11 . . . tiJbd7
1 2 'ifc2 .l:.ac8
The . . . c6-c5 advance is in the air.
An amusing situation : it is not easy for Black
to continue his development (he can only 1 3 cxd5 cxd5
aim for exchanges with . . . ttJe4), but for 14 'ii'd 3 ttJe4
White too no concrete plan is apparent. It is a Now Wh ite has to reckon with 1 5 . . . ttJxe5 1 6
kind of mutual zugzwang position , immedi­ dxe5 i.a3 1 7 i.xa3 'ii'x a3 with a strong
ately after emerging from the open ing! in itiative on the q ueenside.
Adorjan-Moskalenko ( H u ngary 1 990): 1 1 15 ttJxd7 'ti'xd7
a 3 ttJe4 1 2 tiJdf3 ttJxe5 1 3 ttJxe5 i.xe5 1 4 1 6 ttJe5 'fie7
dxe5 b6 (with the idea of 1 5 . . . i.a6). 1 7 f3 tiJf6
Draw agreed . White has noth i n g : his bishops It remains for Black to play 1 8 . . . tiJd7 ,
are so useless , that one can not talk about distu rbing t h e knight on e5, and he w i l l n o
him having the advantage of the two bishops longer have a n y problems.
- rather the opposite!
1 8 .Uac1 tiJd7
Ruban-Meister ( H u ngary 1 990): 11 tiJdf3
19 tLlxd7 'ifxd7
tt'le4 1 2 'ifc2
20 .l:txc8 l:.xc8
The game conti nued 1 2 . . . ttJxe5 1 3 ttJxe5
.ixe5 14 dxe5 b5 1 5 f3 tLlg5 1 6 cxb5 cxb5 1 7
.l:!.fc1 with advantage to Wh ite . But after the (see diagram)
1 78 � Modern I nterpretation of the Dutch Defence
White has a choice between the cautious 2 1
l:k1 and the more active 2 1 e4 . Which is
correct?
21 e4?!
After playing this, my opponent offered a
draw. If he had done this after 2 1 l:.c1 , I
would probably have had to agree.
21 . . . dxe4
22 fxe4 ..txe4
23 ..txe4 fxe4
24 'ifxe4 ..te7 !
A very strong manoeuvre, which was under­
estimated by my opponent. Here too he
should have offered the exchange of rooks:
25 .Uc1 .
25 .l:l.e1 ?! ..tf6
I n the event of the e6-pawn being captu red ,
the black rook invades at c2 . But otherwise
White is condemned to passive defence .
After 2 6 11e2 I defended t h e pawn with my
rook from c6 , advanced . . . b6-b5-b4 , ex­
changed queens by . . . 'ii'd 5, and placed my
rook on a6, my king on f7 and my pawn on
h5. This led to an i nteresting bishop end i n g ,
which I managed t o win .
Plan with i.c1 -f4
7 ..tf4 ..txf4
Of course, White's kingside pawns must be
spoiled . Otherwise he will play 8 e3 and
subsequently recaptu re on f4 with the e­
pawn.
8 gxf4 0-0
A u n ified theory of this variation does not
exist. The continuations which occur here - 9
lDbd2, 9 tDe5 and 9 e3 - usually transpose
into the same positions.
Since Wh ite's kingside has been weakened
by the exchange on f4 , the bishop manoeu­
vre to h5, which is not very effective after 7
b3, is much stronger here. There is now the
idea of opening the g-file by . . . g7-g5.
Black's standard scheme of action is: his
bishop goes to h 5 , knight to e4 , other knight
to d 7 , then . . . 'itt h 8 and . . . g7-g 5 . Of cou rse,
the opponent must try and oppose this plan.
Let us make the moves 9 lDbd2 'ife 7 1 0 l:.c1 .
Why is 1 O . . ..td7? a mistake here? White
.
continues 1 1 'ii' b 3! .tea 1 2 tDe5 ..t h 5 1 3 e3,
and the development of the black pieces is
h i ndered . The knight cannot be played to d7,
and a6 is the wrong place for it.
1 3 . . . 'itt h 8 1 4 .l:.c3 ( 1 4 'itt h 1 ! ? ) 1 4 . . . tDa6 1 5
'ii'a 3 lDb4 1 6 c5. Black's position i s worse,
since his knight is roaming about on inappro­
priate squares.
Thus, if we want to develop our bishop at d7,
we must seriously reckon with 'ii' b 3. And if
we choose this pla n , we should do so
immediately, without 9 . . 'ii'e 7. .
 Modern I nterpretation of the Dutch Defence
Nikol ic-Salov (Leningrad 1 987): 9 tt::l b d2
..id7 1 0 'ii' b 3 'ifb6 1 1 e3 .ie8
Now Black's knight wants to go to d7, and his
queen , incidentally, can retreat to c7 . There­
fore Wh ite h u rried to exchange q ueens.
12 'ifxb6 axb6 1 3 tt::l e s ...i h5 1 4 .if3 .ixf3 1 5
4Jdxf3 tt::l a 6
In the endgame even this route is not bad .
Although White's position is slig htly more
pleasant, he has no real win n ing chances.
The game ended i n a d raw.
Kal i n ichev-Giek ( U S S R 1 987): 9 tt::l b d2
4Jbd7
Along with . . . .ic8--d 7 , also not a bad pla n .
Black aims for a n y exchanges o f knights i n
the centre, after which h e usually does not
have any problems.
10 .Uc1 ( 1 0 e3 and 1 1 'ii'c2 is more accu rate)
1 0 . . . tt::le4 1 1 e3 'i!Ve7
After 1 2 a3 in one game there followed
1 2 . . . 4Jdf6 1 3 tt::l e 5 ...i d 7 1 4 f3 tt::l d 6 1 5 'it>h 1
...i e 8 1 6 �g 1 .ih5. Then Black played . . . 'it>h8
and . . . .Ug8, and after c4-c5 he retreated his
knight to f7 and prepared . . . g7-g 5 . For
Wh ite it is simply not apparent what he can
do. As usual , the g2-bishop is no better than
its opponent on h5.
12 tt::lxe4
ttJ 1 79
1 2 . . . dxe4? !
An original decision ! The standard 1 2 . . . fxe4
is sounder. Then events can develop roughly
as fol lows: 1 3 tt::l d 2 tt::lf6 14 f3 exf3 1 5 tt::l xf3
...i d 7 1 6 tt::l e 5 .ie8. Black is pla n n i ng to play
1 7 . . . tt::l d 7 and after the exchange of knig hts
to bri ng his bishop via g6 to f5 . If he should
succeed i n doing th is, it is Wh ite who will be
having to try to equal ise.
1 3 tt::l d 2?!
The only way to cast doubts on the oppo­
nent's strategy was by 1 3 tt::l e 5. Now Black is
entirely successfu l .
1 3 . . . c5 1 4 tt::l b 3 (altogether the wrong place ! )
1 4 . . . b6 1 5 dxcS ttJxcS 1 6 tt::lx cS bxcS
Wh ite's position is worse . The . . . e6-e5
advance is immi nent, the g2-bishop is shut
in, and if Wh ite should try to activate it by f2-
f3, then after the exchange of bishops the
weakening of his king's position may tell .
The game which I now wish to show you
ended in a crushing defeat for Wh ite in j ust
23 moves. And this is not surprising - on an
examination of it one gains the impression
that the player with Wh ite simply did not
know where to place his pieces, or which
changes i n the structu re were advantageous
to h i m , and which were not.
1 80 \t> Modern I nterpretation of the Dutch Defence

Shabalov - Vyzhmanavin The rook on c7 1ooks well placed , but that is

Tashkent 1 987 a l l , whereas it would seem that Black's
attack on the kingside is already i rresistible.
9 tbc3 lbbd7
1 a �h1 ..i.hS
9 . . . ..i.d7? is incorrect: 1 0 'ifb3 'ii' b 6 1 1 tba4
'i!fxb3 1 2 axb3, and Black has a d ifficult 19 'i!Va3 ..i.e2
endgame. 20 l:tg1 .l:.xf2
10 tOes tbe4 21 ..i.xe4 lif1
22 .l:txf1 .i.xf1
23 'ii'd 6 'iVg4!
White resig ned i n view of 24 .i.xd5 .i.g2+! .

I n the next example i t was Black who played

inaccurately in the opening. Let us see what
this led to .

Beliavsky - Van der Wiel

Amsterdam 1 990
9 e3 lb bd7
10 tOes tbxeS?!
1 0 . . . tbe4 .
1 1 tbxe4? fxe4 1 1 dxeS!
The exchange of knights on e4 is advanta­
geous to Black. Now comes a further position­
al mistake - White allows the exchange on
e5. He should have played 1 2 tbxd7 .i.xd7
1 3 e3 .tea 14 f3 , although after 14 . . . exf3 and
1 5 . . . ..tg6 Black's position is preferable.
12 e3? tbxeS
1 3 fxeS
1 3 dxe5 g5 was even worse for White. [A
dubious assertion - in view of the reply 1 4
"ikg4. 13 . . . .i. d 7 is sounder - Dvoretsky.)
13 . . . ..td7
If 14 f3 , then 14 . . . exf3 followed by the
manoeuvre of the bishop to g6. However, this
Earlier i n such situations everyone, for some
was probably the lesser evi l . But for some
reason , captured with the f-pawn , including
reason Wh ite took his queen to the queen­
Bel iavsky h imself. The second game of his
side - away from the the main events.
match with Salov (Vilnius 1 987) went 1 1
1 4 'ii' b 3? b6 fxeS?! tbg4 ( 1 1 . . . tbd7 1 2 f4 b 6 i s also good)
1 S cxdS cxdS 1 2 lLld2 .i.d7 1 3 h3 lbh6 1 4 f4 .tea 1 s 'iii> h 2
1 6 .l:tac1 1\VgS �ha 1 6 'ilke2 g5 with excellent prospects for
1 7 .l:tc7 .tea Black.
 Modern I nterpretation of the Dutch Defence ltJ 181

11 . . . lDd7 .i.f4 and 9 e3 against me - previously he had

As Bel iavsky remarked , after 1 1 . . . tDe4 1 2 preferred other set-ups.
b4! (the knight is i n danger! ) 1 2 . . .1i'b6 1 3 a3 9 . . . 'ike7
a5 ( 1 3 . . . dxc4 14 .i.xe4 fxe4 1 5 lDd2 with an I wanted a fig ht, and therefore I did not play
obvious advantage) 14 c5 'ii'a 7 1 5 f3 axb4 1 6 9 . . . .i.d7 1 0 "ii b 3 'ifb6.
fxe4 'ii'x c5 1 7 'ifd2 Wh ite's extra piece is
1 0 lDbd2 lDbd7
more valuable than the pawns. But after the
11 'ii'c 2 tDe4
move in the game Black's position also
remains sign ificantly worse. 1 2 tDxe4
The conclusion : Black can not exchange on It is not clear what else to suggest for White.
e5 with the knig hts on b1 and f6. This Whereas I was intend ing to continue 12 ... 'itt h 8
someti mes happens - you com m it just one or 1 2 . . . lDdf6 1 3 lDe5 ii.d7 14 f3 lDd6 , and
inaccuracy, and it then becomes difficult to then to fol low the usual plan: . . . .i.e8-h5,
save the game. . . . <it;lh8 and . . . g7--g5.
12 lDd2 'fie7 12 . . . fxe4
1 3 l:tc1 l:.d8 Here the captu re 1 2 . . . dxe4?, as in the
14 'ii'c 2 lDf8 Kal i nichev-Giek game, would have led to a
1 5 tDb3 lDg6 d ifficult position after 1 3 lDe5 tDxe5 1 4 dxe5
or 1 3 . . . c5 14 'iic 3.
By placing his knight on h4, Van der Wiel
tries to create th reats to the wh ite king . 1 3 lDd2
Noth ing comes of th is, because the c8- Of cou rse, not as Shabalov played : 1 3 tDe5?
bishop is too far away from the kingside. tDxe5 with the better chances for Black.
16 cxd5 exd5 13 . . . tl)f6
17 lDd4 lDh4 1 4 f3 exf3
18 ii.h3 g5 1 5 tDxf3 ii.d7
19 �h1 g4 1 6 lDe5 .i.e8
I n the event of 1 9 . . . gxf4 20 exf4 the wh ite
rook goes to g5.
20 l:.g1 h5
2 1 .i.f1 llf8
22 f3 l:.f7
23 'ii'f2 lDg6
24 .i.d3 �g7
25 b4!
Black stands badly on both wings. Withi n a
few moves the game concluded .

Novi kov - Kra m n i k

Moscow 1 99 1
9 e3
Igor Novikov later told m e that i t was under
the impression of the Beliavsky-Van der
Wiel game that he chose the variation with 7
You a l ready know Black's plan: the ex­
change of knights by 1 7 . . . lDd7 followed by
playing the bishop to g6. White must fight
against it.
1 82 � Modern I nterpretation of the Dutch Defence
1 7 l:tf3 lLld7
1 8 .Uh3! h6
19 tLlg4
Draw agreed . There is no other move apart
from 1 9 . . . lLlf6 , but then the knight retu rns to
e5.
It may seem that, by playing 1 9 tLlxd7 , Wh ite
would have retained the better position, but
in fact this is not so. I reply 1 9 . . . -i.xd?
(intend ing 20 . . . e5).
After 20 .l:th5 lif5! 2 1 .l:txf5 exf5 22 cxd5
'il'xe3+ 23 'iff2 'il'd3 White most probably
has to repeat moves: 24 .i.f1 (24 l:e 1 .l:r.e8)
24 . . . We4 25 .i.g2.
And if 20 l:tg3 Black has 20 . . . c5! .
[20 11g3 c5 2 1 dxc5! 'il'xc5 22 'il'g6 'J:. f7 23
cxd5 exd5 24 nd 1 gives White a clear
advantage. It is probable that somewhere
earlier Black's play can be improved. Per­
haps he should manage without 9 . :jke 7 - .
Dvoretsky.]
In this game both players apparently made
the best moves, and neither managed to
gain an advantage. Such a cou rse of events
is in accordance with the present-day state
of theory in the 'stonewall' variation it is -
perfectly playable for Black!
 tLJ 1 83

PART IV

Complicated Strateg y in P ractice

Artur Yusu pov

C rux of the Pos ition

O topics with the former world champion

n one occasion , w h e n d iscussing chess nua nce or a positional idea , by hold ing o n to
this guiding thread he is often able to change
Boris Spassky, the author heard a very the cou rse of the play i n his favour.
interesti ng descriptio n of the play of another
great chess player, Robert Fischer. I espe­ Wa ng Zili - Yusu pov
cially remember that Spassky singled out Olympiad , Novi Sad 1 990
only one relative weakness i n the play of the Ruy Lopez
American champio n : even Fischer did not 1 e4 e5
always manage to sense the critical , turn i ng­ 2 lt:lf3 lt:lc6
points of a game. What, the n , ca n be said
3 .i.b5 a6
about mere mortals!
4 .i.a4 lt:lf6
How often do we ponder over a position and
5 0-0 lt:lxe4
seek a solution, where we simply need to
make a normal move, or, on the contrary, 6 d4 b5
quickly skip over a situatio n in which there is 7 .i.b3 d5
a direct way to our goa l . The ability to 8 dxe5 ..te6
concentrate at the req uired moment, and to 9 lt:lbd2 lt:lc5
understand or sense when it is possible to
1 0 c3
decide the game or at least direct it along
This is a well-known position from the Open
desirable li nes - this is a distinguishing trait
Variation of the Ruy Lopez. Now B l ack has a
of a g reat player.
choice between 1 0 . . . d4 and the move i n the
How many times have I heard or myself said , game.
in explanation of a failure: ' I d i d n 't under­
10 .i.e7
stand what was going o n ! ' I n nearly every
. . .

game of chess one can pick out decisions, 11 .i.c2

moves, which rad ically i nfl uenced the result The pawn sacrifice 1 1 lt:ld4 is a standard
(here, of cou rse, we are not talking about idea , but here it is clearly prematu re because
blu nders or serious positional mistakes). B ut of 1 1 . . . lt:lxe5 1 2 f4 lt:lc4 1 3 lt:lc6 'i!Vd7 1 4
when a player is able to g rasp the essence of lt:lxe7 lt:le3 ! .
a position , whether it be a veiled tactical 11 . . . .i.g4
1 84 � Crux of the Position

A forced move, since now 1 1 . . . 0-0 is very Wh ite's idea : in this way he regains the
strongly met by 1 2 liJd4 ttJxe5 1 3 'ii' h 5 tiJg6 pawn.
14 f4 f5 1 5 h4. 15 . . . <t>xh7
12 .Ue1 0-0 1 6 'ifc2+ �g8
Nowadays 1 2 . . . 'ifd7 1 3 tiJf1 l:d8 , su pporti ng It is extremely dangerous to play 1 6 . . . tiJed3
the d5-pawn , is more usual. 1 7 tiJxg4 f5 ( 1 7 . . . <t>g8 ! ) , since, along with 1 8
13 tiJf1 tiJge5 ttJxe 1 1 9 'ii'xf5+ with a guaranteed
The other developing move is 1 3 tiJb3 ! ? . d raw, there is also 1 8 l:.d 1 fxg4 1 9 tiJe5, and
Wh ite stands better.
13 . . . J:te8 ! ?
1 7 ttJxe5 .i.e6
A n interesting continuatio n , the point of
which is to begin play against the e5-pawn . 1 8 tiJc6

1 4 tiJe3 It is importa nt to deprive Black of the

advantage of two bishops.
White has a variety of possibilities: 1 4 h3, 1 4
i..f4 , 1 4 b4? ! and 1 4 tiJg3. The Chinese 18 . . . 'ii'd 6
player begins a tactical operation , but Black 1 9 tiJxe7+ J:txe7
is well enough prepared for the immediate The outcome of Wh ite's tactical operation
clash of the forces. can be assessed : he has not gained any
advantage. Black is not badly developed and
g iven the opportun ity he is ready to seize the
i n itiative.
20 b3!
A subtle move , creating the unpleasant
threat of a pin by 2 1 i.. a 3.

14 . . . ttJxe5
After serious thought Black decided to go
along with his opponent's idea, rightly as­
suming that the resulting position would
promise him at least equal chances. The
alternative 14 . . . .i.xf3 1 5 'ii'xf3 ttJxe5 1 6
'iVxd5 is also acceptable for Black, only now 20 . . . tiJd7!
he should not continue 1 6 . . . tiJcd3 in view of
One of the most d ifficult moves i n the game.
1 7 l:td 1 , but 1 6 . . . i..f8 , obtaining cou nterplay
It is always u npleasant to have to retreat, but
thanks to the weakness of the d3-square.
the natural 20 . . . tiJe4 is weaker, since after 2 1
15 i.. x h7+! .i. b 2 .l:r.ae8 ( i f 2 1 . . . c 5 , then 22 c4) 22 liad 1
This intermediate check is the point of the black centre is under fire: c3-c4 is
 Crux of the Position
threatened , and 22 . . . c6 is n ot possible i n
view o f 23 'ifxe4 . It transpires that the knight
on e4 prevents Black from consolidati ng his
positio n i n the centre.
21 �b2?!
Played too abstractly. Wh ite will be unable to
advance c3-c4 i n view of the reply . . . d5-d4.
It follows that at b2 the bishop stands badly.
21 a4 came into consideration, bringing the
rook i nto play. After 21 . . . �b8 22 axb5 axb5
the position is roughly equ a l . Perhaps Black
should change his plan by choosing 21 . . .
.l:r.ee8 ! ? , and if 2 2 ii.a3, then 2 2 . . . c5.
21 . . . l:tae8
22 l:tad 1 c5
23 f3? !
Wh ite made t h i s move after a serious
consideration of the position . However, if the
Chinese player had taken Black's counter­
actions i nto accou nt, he would probably
have restricted h imself to the safe move 23
'YWd2 , mainta i n i n g approximate equal ity.
At this, perhaps the decisive moment of the
game, I thought for more tha n half a n hour.
The tra i n of my th i n king was roughly as
follows:
What does the opponent want? It would
appear that Wh ite has devised a n i nteresting
pla n : he wants to play his q ueen to g3, force
the exchange of queens, and obtain a good
l2J 1 85
endgame i n view of the weakness of the d5-
and c5-pawns.
How can the opponent's plan be prevented?
The direct 23 .. .f5 (24 tL!xf5?? ..txf5 ) did not
appeal to me because of the simple 24 'ii'd 2,
and Black has somewhat weakened his
position.
How can I strengthen my position? The black
pieces are not badly placed , but the position
of the knight is rather passive .
What are the d rawbacks of Wh ite's set-up?
Thinking in this way, I grad ually managed to
d iscover a weak point in the opponent's
position - the d3-square! I n deed , the inva­
sion there of the knight after . . . c5-c4 ,
combi ned with action on the e-file and on the
a7-g 1 diagona l , may decide the outcome of
the battle. This means that 23 . . . ttJe5 fol­
lowed by . . . c5-c4 and . . . tLld3 is temptin g .
Even s o , i t w a s n ot easy for me t o evaluate
the consequences of this pla n , u ntil I
d iscovered a strong possibility on the 26th
move . General considerations may prove
fau lty, therefore they should be supported by
specific calculation .
23 . . . ttJe5!
24 'ii'f2
If 24 c4 ! ? Black had the reply 24 . . . d4 (but not
24 . . . bxc4 because of 25 �xe5! ) , and after 25
cxb5 (25 b4 'i!i'c7! ) n ot 25 . . . axb5 i n view of 26
b4! 'fie? (or 26 .. .'i!fb8) 27 'ii'x c5 , but the
subtle 25 . . . 'ii' c7 ! 26 tLlf1 (26 ltJc4?? ..txc4 27
bxc4 tL!xf3+) 26 . . . ..td5 with the advantage.
[ This is not altogether so: White maintains
the balance by 27 11xe5! 'iixe5 28 'ii'xc5 or
27 . . J1xe5 28 �xd4. Therefore Black does
better to play 26. . . iLd7 (with the threat of
27 . . . ttJxf3+) 27 <i;h 1 iLxb5. Even stronger,
apparently, is 25. . . iLc8! (instead of 25. . . "illc 7)
26 iLa3 (the knight cannot move because of
26. . . ttJxf3+) 26. . . "ikh6! 27 ii.xc5 ttJxf3+ 28
gxf3 l:txe3 29 if.. b 4 (the only defence)
29 . . . axb5 with a great advantage - Dvoret­
sky.]
1 86 \t> Crux of the Position
24 . . . c4!
25 'ii'g 3?!
I n such a situation it is difficult to bring
oneself to sacrifice the exchange, but this
was probably the best practical solution :
after 2 5 !Dc2! !Dd3 2 6 �a3 (26 .l:txd3 cxd3 2 7
�a3 can b e met b y either 27 . . . "ikc7 2 8 �xe7
'ii'xc3 29 !Dd4 .l:txe 7, or 27 . . . 'il'f4 28 �xe 7
dxc2 29 �a3 .l:tc8 30 'ii'xc2 b4! ) 26 . . . !Dxf2
(26 . . . 'ikc7 27 ild2 !Dxe 1 28 .l:txe 1 .l:r.d7 29
!Dd4 with compensation) 27 �xd6 !Dxd 1 28
:Xd 1 .l:td7 29 �b4 Wh ite would have
blocked the position and retained good
drawing chances. Thus if 29 . . . .l:ta8 there
follows 30 �a5, and 30 . . . cxb3 3 1 axb3 d4? is
not dangerous in view of 32 .l:txd4 .
25 . . . f6
26 .l:td4? !
The most principled continuatio n , with which
I had to reckon with first of all when
beginning the operation to play my knight to
d3. But 26 !Dc2 'il'b6+ 27 !Dd4 !Dd3 28 l:te2
was nevertheless stronger, although after
28 . . . a5!? Black's advantage is undisputed .
[Instead of 28 l:te2 possible is 28 i.a3! with
good equalising chances, and therefore it
makes sense to advance the a-pawn a move
earlier: 27 . . . a5!? - Dvoretsky.]
26 . . . 'ii'c 5!
After th is key move the strategic battle is
won , although Wh ite can still put u p a tough
resistance. 26 .. .'ii' b 6 was less accu rate on
account of 27 �a3 with cou nterplay. But now
Black renews the th reat of . . . !Dd3 and at the
same time prevents the activation of the
white pieces: if 27 !Dc2 there follows
27 . . . cxb3! 28 axb3 i.f5 29 i.a3 (or 29 l:.e2
i.xc2 30 :Xc2 !Dc6) 29 . . . 'iWxc3 30 J.. xe7
�xc2 3 1 i.xf6 !Dxf3+! 32 gxf3 'ii'xe 1 + 33
'iWxe 1 l:txe 1 + 34 'iitf2 .Ue6 35 i.g5 i.xb3 and
wins.
27 bxc4 dxc4
28 !Dc2 a5!
U n n ecessary complications would have re­
sulted from 28 . . . �f5 29 �a3 !Dxf3+ 30 'i'xf3
(30 �2? ! !Dxe 1 3 1 �xc5 .l:[e2+ 32 'iitf1 .l:!.xc2
with a decisive advantage for Black) 30 . . .
.lbe 1 + 3 1 !Dxe 1 'iWxa3 3 2 'ii'xf5 I!xe 1 + 33
�2 . The q u iet 28 . . . .l:.d7 29 �a3 'ilia? was
not bad , but here it is still not easy to breach
the opponent's defences. By undermining
White's centre , Black achieves his goal more
simply and quickly.
29 i.a3 b4
30 cxb4 'il'a7
The point of Black's idea . The rook is still
pin ned and deprived of support; the threat is
3 1 . . . !Dc6 .
31 b5 .l:td7
The simplest. 31 . . . i.f5? 32 �xe7 �xc2 did
not work because of 33 i.xf6 ! .
3 2 .l:tee4
Wang Zili was in serious time-trouble, but in
any case Wh ite's position can no longer be
defended . After 32 l:r.ed 1 Black would have
won by 32 . . . !Dd3, while if 32 l::txe5 the
simplest was 32 . . . .l:lxd4.
32 . . . i.f5
33 'ilf2 i.. xe4
34 .l:txd7 'ii'x d7
35 fxe4 'ii'd 1 +
36 !De1 !Dd3
At last the knight has reached the square for
 Crux of the Position lLJ 1 87

which it was aiming 1 3 moves earlier! White
resigned .
I n this game I was able t o concentrate at
critical moments and come to successfu l
decisions on the 1 4th , 20th and 23rd moves.
After the 23rd move the game was controlled
by Black, who found and carried out the good
plan of playing his knight to d3. This
positional idea became the guiding thread
which led him to his goa l .
Yusu pov - Gavri kov
Zu rich 1 994
Griinfeld Defence
1 d4 ltlf6
lDa4 Wd8 1 4 Wb4 or 1 2 . . . lDxd5 1 3 lDxd5
cxd5 14 .tc7 'iWe6 1 5 'iWb4 lDb6 1 6 0-0 Ilfc8
1 7 .l:lc5 with the better game. What operates
here is a standard consideration in such
structures: White should delay the exchange
of c-pawns as long as the knight is still on b8 ,
in order not to allow its development on the
active c6-sq uare.
12 lDxe4 .txe4
The captu re with the pawn is less natu ra l :
then 1 3 c 5 'ilkd8 1 4 lD e 5 is good , as is the
i m mediate 1 3 lDe5 , since after 1 3 . . . c5 1 4
dxc5 'iWxc5 1 5 b4 Black cannot play 1 5 . . . 'iWc7?
1 6 lDxf7 e5 1 7 lDxh6+.
13 0-0

2 c4 g6 1 3 c5 'il'd8 1 4 'i!i'b4 is weaker in view of

1 4 . . . b6 (as my opponent pointed out, the
3 lDc3 d5
simple 1 4 . . . 'ii'c8 is also good enough to
4 .tf4 .tg7 equal ise) 1 5 cxb6? ! axb6 1 6 Jl.xb8 c5! .
5 .l:lc1 0-0 However, 1 3 lDe5! ? came i nto consideration,
6 ltlf3 c6 with favourable complications.
The qu iet system of development chosen by 13 . . . ltld7
Black is not altogether in the spirit of the
Grunfeld Defence. In contrast to the main
variations, here he does not aim to u nder­
mine the opponent's centre immediately, but
restricts h imself to a solid but rather passive
position.
7 e3 .te6
8 lDg5 .tf5
9 .te2 'ii' b 6
The alternative was the modest 9 . . . h6 1 0
tt:lf3 Jl.e6. Wh ite can hope for a slight
advantage after both 1 1 cxd5 lDxd5 1 2 .i.e5
f6 1 3 Jl.g3, and 11 b3 lDbd7 1 2 h 3 .
1 0 'i!i'd2 ! ?
An important moment i n the game. Wh ite is
A new move, the idea o f which i s seen in the
at the cross-roads: he can either block the
variation 1 0 . . . lDbd7 1 1 cxd5 cxd5 1 2 lDa4 !
q ueenside, or exchange on d5. However, for
'i'd8 1 3 'ikb4 with advantage to White .
the moment neither of these continuations
10 . . . h6 promises a clear advantage. After 14 c5 'ii'd 8
1 1 lDf3 lDe4 1 5 'ii' b4 the simple reply 1 5 . . . 'il'c8 is good ,
Now too after 1 1 . . . lDbd7 Wh ite can advanta­ but Black can also consider 1 5 . . . .txf3 1 6
geously exchange on d 5 : 1 2 cxd5 cxd5 1 3 .i.xf3 e5 1 7 dxe5 lDxe5 1 8 .txe5 .txe5 1 9
1 88 � Crux of the Position

1Wxb7 1i'f6 b3 a5. And in the event of 1 4 cxd5 I n stead of the move in the game, 16 ... .l::!. a d8
i.xd5 15 e4 i.xe4 16 i.xh6 i.xh6 17 'ii'x h6 was more accu rate, fig hting against 1 7 e4,
i.xf3 1S i.xf3 l"Df6 1 9 .l:.fd 1 .l:.adS the on which there would have followed 17 . . . g5
opponent begins attacking my weak pawns. 1 S .te3 t"De5 1 9 i.e2 l"Dg6, although in this
What does Black want? H is most natural case too after 20 d5 White would have
plan is to exchange on f3 and c4 and then retained the better chances. 1 7 'ii'c2 'it>h 7 1 8
attack the centre with . . . e7-e5, simpl ifying 'itt h 1 e5 1 9 dxe5 t"Dxe5 20 i.xe5 i.xe5 2 1 f4
the game. It turns out that with a simple was also not bad , with a minimal advantage
developing move I can parry this th reat and for Wh ite.
simultaneously prepare 1 5 cxd5.
1 4 .l:.fd 1 !
Such prophylactic moves are always un­
pleasant for the opponent. Especially since
he cannot reply in the same style: the
careless 14 . . . l:.adS?? loses immed iately to
1 5 c5.
14 . . . dxc4
After 1 4 . . . a5 Wh ite can now play 1 5 cxd5
i.xd5 1 6 e4 i.xe4 1 7 .txh6 .txh6 1 S 'ii'x h6
i.xf3 1 9 i.xf3 , when after 1 9 . . . l"Df6 he has
the unpleasant 20 h4, while if 1 9 . . . 'ii'x b2 ,
then 20 h4 'ii'x a2 2 1 h5 'ife6 22 d5 1i'f6 23
hxg6 is good , with the possible variation 1 7 'ii'c 2
23 . . . fxg6?! 24 dxc6 t"De5 25 cxb7 .:tabS 26
I was wrong to avoid the consistent 1 7 e4,
l:tc7 l"Dxf3+ 27 gxf3 with adva ntage to Wh ite.
si nce after 1 7 . . . 'it>h7 1 S .te3 'ii' b4 1 9 'i'e2
1 4 . . . .txf3 1 5 i.xf3 dxc4 1 6 .l:.xc4 e5 is no
the thematic 1 9 . . . e5 can be strongly an­
better in view of 1 7 .tg3 exd4 1 S exd4 with
swered by both 20 dxe5 t"Dxe5 2 1 i.b3 'ike?
the th reat of 1 9 l:tb4 , and if 1 S . . . a5, then 1 9
22 f4 t"Dd7 23 e5, and 20 d 5 . But now the play
d5. With the move in the game Black makes
takes on a closed , manoeuvring character,
concessions in the centre, but retains a solid
where it is harder for White to probe the weak
position.
poi nts in his opponent's defences.
15 i.xc4 i.xf3
17 . . . e6
1 6 gxf3 a5 1 8 'itt h 1 l:ifd8
The point of White's idea is revealed in the 1 9 .l:!g1
variation 1 6 . . . e5 1 7 dxe5 t"Dxe5 1 S i.xe5 With the primitive threat of 20 .txe6 , which ,
.txe5 1 9 'ii'c2 ! with advantage. For example, of course, Black easily parries.
19 . . . 'ittg 7 20 .l:.d7 'ii'x b2 (20 . . . .l:.adS 2 1 .l:.cd 1
19 . . . l"Df8
:Xd7 22 lhd7 .txb2 23 i.b3 with the better
game) 2 1 'ii'x b2 .txb2 22 lic2 ! (but not 22 20 i.b3 ! ?
l:tb 1 ? in view of 22 . . . b5!) 22 . . . .ta 1 (22 . . . b5 Prophylactic play b y Wh ite. He prevents the
23 i.b3) 23 .l:.xb7 a5 24 l:td2 .l:!.abS 25 Ita? opponent from gaining space on the q ueen­
.:.as 26 .l:.dd7 l:txa7 27 .l:txa7 .tc3 , and now side with 20 . . . a4 followed by . . . .l:ta5 .
the best is 2S h4 ! , to be able to answer 20 . . . 'ii b 5
2S . . . g5 with 29 h5. 21 a4 'ii' b4
 Crux of the Position t2J 1 89

l:!xc4 j_xb2 30 c6 bxc6 31 .:.Xc6 are in his

favou r.
26 .l:!.c5 j_e7?
My opponent falls i nto the trap. Carried away
by the struggle for . . . c6-c5, for an instant he
drops his guard and forgets about the
weakening of his king's position. 26 . . . .l::td 7
was more cautious, with a minimal advan­
tage for Wh ite .

With the threat of 22 . . . c5. This advance is

the key to the position : after it my adva ntage
my completely vanish. Wh ite's su bsequent
actions become understandable: he actively
fights against the opponent's i ntentions.
What will look like an attack on the kingside
is in fact a defence of the q ueen side.
22 l:tg4! llac8
If 22 . . . c5 Black did not like the reply 23 j_c7 .
However, after the exchange sacrifice 23 . . . h 5
24 l:.e4 ( o r 2 4 j_xd8 l:.xd8) 24 . . . cxd4 25 27 l:.xh5! !
i.xd8 .l:.xd8 he would have retai ned q u ite An u nexpected tactical blow on the kingside.
good defensive chances. Thus if 26 l:.d 1 , I n this game, which would also be fully
then 26 . . .f5 is q u ite acceptable. He could appropriate for the theme of manoeuvring, I
also have considered the more modest was successfu l with play by my advanced
22 . . . l:.d7!? 23 j_g3 We7 with a minimal rook.
advantage for Wh ite .
27 . . . gxh5?
23 j_g3 'ii' b 6
The sharp change in the situation affects my
If 23 . . . h5 I would not have played 24 l:!g5 in
opponent and he chooses a poor move ,
view of 24 . . . c5 25 .U.xc5 .U.xc5 26 dxc5 l:td2 ,
which leads by force to his defeat. 27 . . . c5! ?
but 2 4 .l:te4 . Now after both 2 4 . . . c 5 25 d 5 (25
was much more tenacious, after which Wh ite
dxc5 'i!fd2 26 l:tc4 j_xb2 is unclear) 25 . . . c4
would have had to play accu rately: 28 j_e5!
26 l:txc4 l:.xc4 27 1Vxc4 'ii'xc4 28 j_xc4 and
(u nconvincing is 28 dxc5 'i!fc6 or 28 l:.xc5
24 .. ."ii' b 6 25 j_e5 Wh ite retains a slight
j_xc5 29 dxc5 'i!fb4) 28 . . .f6 29 l:!.g 1 cxd4
advantage.
(29 . . . fxe5 30 l:.xg6+ cJilf7 3 1 'ii'f5+ �e8 32
24 h4 h5 j_xe6 ltJxe6 33 l:.xe6 is hopeless for Black)
25 l:!.g5 j_f6 30 l:txg6+! cJi;f7 31 .U.xf6+ j_xf6 (31 . . . �e8 32
Here too Wh ite is well prepared for 25 . . . c5. l1xf8+) 32 lih 7 + cJile8 (32 . . . cJilg8 33 j_xf6!
After this there follows simply 26 .l:txc5 .l:ixc5 l:1xc2 34 .llg 7+ �h8 35 .U.Xb7+ �g8 36 .l::t x b6
27 dxc5 'ii' b4 28 'ii'c4 , when both 28 . . . j_xb2 .l:tc1 + 37 'it>g2) 33 j_xffi! (th reatening 34 .U.e7
29 'ii'x b4 axb4 30 :c4 and 28 . . . 'ii'xc4 29 mate) 33 . . . 'il¥d6 34 'i!fe4 ltJxh7, and now not
1 90 � Crux of the Position

35 'ii'x h7? 'ii'f8! , but 35 .i.xe6! �f8 (35 . . . tt:Jxf6 Here the knight is more passively placed
36 'ii'g 6+) 36 'ii'f5! - the 'ambush' threats than at f6 , and Wh ite gains the advantage of
created by the queen are deadly. the two bishops without any particular
28 l:tg1 compensation for Black. 4 . . . c5 is another
orig inal continuation , which after 5 cxd5 exd5
Threatening to move the bishop to c7 with a
6 dxc5 (in the game Psakhis-Korzubov,
discovered check.
U S S R Championship First League 1 983,
28 . . . lt:Jg6 Black equalised after 6 .i.b5+ tt:Jc6 7 tt:le2
28 . . . �h8 29 .i.e5+ f6 would also have lost to tt:Je 7 8 0-0 0-0 9 dxc5 .i.xc5 1 0 b3 .i.g4! 1 1 h3
30 �xe6 tt:Jxe6 31 'fig6 . .i.h5 1 2 �b2 l:tc8) 6 . . . tt:Jf6 7 .i.b5+ i.d7 8
2 9 .i.xe6! �xd7+ tt:Jbxd7 9 tt:Je2 1ed to a more pleasant
All the wh ite pieces swiftly join the attack. position for Wh ite in the game Bandza­
Aieksandrov (Fru nze , 989).
29 . . . �h7
5 .i.d2 ! ? 0-0
After 29 . . .fxe5 30 'ii'x g6+ �f8 any move by
the bishop from g3 is decisive. After 29 . . . 'iti>g7 5 . . . c5 6 a3 .i.xc3 7 �xc3 cxd4 8 'ifxd4 f6
Wh ite wins by 30 .i.e5+, while if 29 . . . �f8 - came into consideration, as in the game
30 'ii'f5. Psakh is-Kupreichik ( U S S R 1 984 ). However,
after 9 l:ld 1 Wh ite's position is preferable.
30 'it'f5 l:tg8
6 a3 .i.xc3
Or 30 . . . fxe6 31 'fif7+ �h6 32 .i.f4+.
7 .i.xc3 b6
31 .i.e5 �h6
8 tt:Jf3 .i.a6
32 �f4+ �h7
9 b3 c5
It was no better to play 32 . . . �g7 33 'ii'xf7+
�h8 34 .i.e5+ tt:Jxe5 35 'ii'x h5 mate, or
32 . . . tt:Jxf4 33 'ifxf4+ .i.g5 34 hxg5+ �h7 35
'ii'xf7+ �h8 36 'ifxh5+.
33 'ifxh5+ �g7
34 �h6+
Black resigned .

Yusupov-Lautier
Amsterdam 1 994
Queen 's Gambit
1 d4 d5
2 c4 e6
3 tt:Jc3 �b4 The fi rst serious problem for Wh ite: he has to
Joel Lautier plays the opening in an original decide what pawn structu re it would be
way, employing an u nusual hybrid of the desirable to obta i n . Despite his advantage of
Queen's Gambit Decli ned and the N imzo­ the two bishops, he has to take serious
l ndian Defence. But if Wh ite does not object account of the opponent's cou nterplay, as
to a Nimzo-lndian with 4 e3, this does not shown by the followi ng variations:
bring Black any benefits. A) 1 0 .i.d3 cxd4 11 exd4 lt:Jd7 1 2 0-0 l:tc8 1 3
4 e3 tt:Je7 'ii'e 2 lt:Jg6! 1 4 g 3 "i/c7 1 5 lt:Jd2 e5 (or
 Crux of the Position ltJ 1 91

1 5 . . . dxc4 1 6 bxc4 e5); probably not have repl ied 1 1 . . . cxb3 in view of
B) 1 0 l:tc1 dxc4 11 bxc4 cxd4 1 2 exd4 tiJd7 1 2 ..tb4, but 1 1 . . . tiJd5! ? . However, after 1 2
13 ..td3 l::tc 8 14 ..tb2 tiJg6! 1 5 'ii'd 2 ..tb7 1 6 bxa7 ttJxc3 ( 1 2 . . . .l:r.xa7 1 3 ..td4 ; 1 2 . . . tiJbc6
'ife3 'ili'f6 . 1 3 'ii'd 2) 1 3 'ii'x d8 Jbd8 1 4 axb8'ili' l:taxb8 1 5
..txc4 (or 1 5 b4 ttJa4 1 6 llc1 ) 1 5 . . . ..txc4 1 6
Usually with the two bishops you should not
bxc4 it i s not clear whether he would have
avoid hanging pawns, but in the g iven
been able to demonstrate that Black's
instance the black knight obtains good
in itiative is worth the sacrificed pawns. 1 1
prospects on g6. In view of this featu re of the
bxc4 ! ? also deserved serious consideration .
position, I preferred a sound conti nuation ,
After 1 1 . . . 'ii' x d 1 + ( 1 1 . . . bxc5 is weaker be­
which secu res a slight but enduring advan­
cause of 1 2 ..td3) 1 2 l::. x d 1 bxc5 1 3 ..te2
tage.
tiJbc6! 1 4 tiJd2 .l:.ab8 1 5 l:tc1 ( 1 5 0-0 is of
10 dxc5! dxc4 equal merit) 1 5 . . . tiJd8 ! ( 1 5 . . .f5 is weaker i n
This intermed iate capture was the main view o f 1 6 f4! followed b y g2-g4) 1 6 ttJe4
problem . After 1 0 . . . bxc5 Wh ite would have tiJb7 White has a slight advantage .
secu red the better pawn structu re by 1 1 11 . . . ..txc4
llc1 ! ? (the simple 1 1 ..td3 is also not bad ). If
1 2 bxc4 bxc5
now 1 1 . . . ttJbc6?, then 1 2 cxd5 ..txf1 1 3 dxc6 .
The reaction to 1 1 . . . tiJbd7 is simila r - after 1 2 1 3 'ili'xd8 llxd8
cxd5 ..txf1 1 3 dxe6 Wh ite has a g reat 1 4 'it>e2
advantage. If 1 1 . . . 'ii' b 6 there is the adequate Of cou rse, in the endgame the king should
reply 1 2 ..td3. It remains to consider the most remain i n the centre .
critical reply 1 1 . . . dxc4 . After 1 2 'ii'x d8 .l:r.xd8 14 . . . tiJd7
1 3 ..ta5 Wh ite's idea becomes clear: he
1 5 tiJd2
takes play into an endgame, where the
weakness of the opponent's pawns begins to
tell . There can follow:
A) 1 3 . . . .l:!.d5 14 bxc4 ( 1 4 ..txc4 is also not
bad ) 14 . . . .l:td6?! ( 1 4 . . . �d7 is better) 1 5 tiJe5
i.b7 1 6 ..tc7 .l:!.a6 1 7 .l:!.b1 with a decisive
advantage;
B) 1 3 . . . llc8 14 ..txc4 ..txc4 1 5 l:txc4 tiJd7 1 6
'it>e2 tiJb6 1 7 llc2 c4 1 8 ..txb6 axb6 1 9 .l:!.xc4!
(1 9 bxc4 is less accu rate because of the
reply 1 9 . . . b 5 ! ) 1 9 . . .llxc4 20 bxc4 l:txa3 2 1
.l:!.b 1 with advantage;
C) 1 3 . . . .l:!.d6 14 ..txc4 ..ixc4 1 5 .l:lxc4 tiJd7 1 6
'iti>e2 l:!.a6 1 7 l:ld 1 .:Xa5!? ( 1 7 . . .tiJd5 1 8 b4)
18 .l:!.xd7 tiJd5 1 9 .l:!.b 7 .:xa3 20 ttJe5 with
Here it was Lautier's tu rn to take an
somewhat the better game.
important decision: he had to decide on a
11 ..txc4 plan of action . Generally speaking, here
I made this natural move without any Black has little possibil ity of infl uencing the
particular hesitation , althoug h perhaps it character of the subsequent play. His only
would have been worth giving it some plan of exerting some pressu re on my
thought. After 1 1 cxb6 !? the opponent would position is by active play with his knights.
1 92 � Crux of the Position
Therefore here the placing of the knights is of
great importance. My opponent did not
sense the importance of the moment, and
without much thought he made an outwardly
natural and 'pretty' move . . .
15 . . . ttJc6?
The c6-square is only apparently a good one
for the knight - in fact here it is restricted by
the opponent's bishop, it is not attacking
anything, it is not preventing anyth ing and it
lacks any clear prospects. When I discussed
and analysed the game with Mark Dvoretsky,
we came to the conclusion that this su per­
ficial and aimless move was the main cause
of Black's defeat.
The knights should have been deployed
more actively, aiming for pressure on the c4-
pawn. 1 5 . . . ttJc8!? was correct, in order to
play one knight to d6, and the other to b6. As
the variations given below demonstrate, at
best Wh ite would have retained only a slight
advantage:
A) 1 6 a4 ltJd6 1 7 a5 f6 1 8 g4 .Uab8 1 9 h4 e5
(with the idea of . . . e�4. but 1 9 .. .<lt>f7 20 f4
h5!? is also not bad) 20 f3 �f7 and by playing
his knight via f8 to e6, Black equal ises ;
B) 1 6 g4 ltJd6 1 7 .:tac1 f6 (if 1 7 . . . ltJb6 1 8
i.e5 f6, then 1 9 i.xd6 .l:.xd6 20 ltJe4 with the
better chances for Wh ite) 1 8 h4 ltJb6 1 9 i.a 1
e5 20 g5 � 2 1 lihg 1 h5 22 gxh6 gxh6 23
h5 l::t g 8 24 f4 exf4 25 exf4 l:tae8+ 26 �d3 (if
26 �f3 there follows 26 . . . ttJbxc4! 27 ltJxc4
ltJxc4 28 Itxg8 liJd2+) 26 . . . l:Id8 27 .l:.xg8
.l:.xg8 28 ltJe4 ttJxe4 29 �xe4 .l:.e8+
(29 . . . <ii?e 6!? 30 f5+ �f7) 30 �f5 ltJc8 3 1 l:id 1
ltJe7+ 32 �g4 lig8+ 33 �f3 �e6 with an
eq ual position;
C) 16 l:.hb1 ltJd6 17 i.a5 .l:.dc8 1 8 l:.b3 <ii?f8
1 9 f4 (less is promised by 1 9 .l:.d3 �e7 20
lld 1 in view of 20 . . . ltJb6 21 .l:.c1 ltJa4)
1 9 . . . �e7 20 e4 f6! 2 1 .U.ab 1 l:.ab8 with a
minimal advantage for Wh ite .
1 6 .l:thb1
White would have secured a slight advan-
tage with 1 6 g4 ltJb6 1 7 .l:.ab 1 f6 18 h4 ( 1 8
.l:tb5 ltJa4) 1 8 . . . �f7 1 9 h 5 lid7 20 lihc1 liad8
2 1 llc2 .
16 . . . l:!.ab8
Now if 1 6 . . . ltJb6 there would natu rally have
followed 1 7 a4 .
1 7 lib5!
I n principle it i s advantageous for White
to exchange one pair of rooks, si nce he
holds the in itiative and it will be easier for him
to i nvade the opponent's rear, while at the
same time it will more d ifficult for Black to
gain cou nterplay. However, every trifle is
important, and before exchanging rooks it is
not bad to provoke a weakening of the b6-
square.
17 . . . a6
1 8 lixb8 .l:.xb8
What to do now? The d i rect 1 9 .l:tb1 does not
promise anyth ing, since the exchange of all
the rooks merely eases Black's defence: he
will not have to worry about the invasion of
the enemy rook. I need another target to
attack. I n this position the superiority of
bishop over knight is somewhat camou­
flaged , but it only req uires Wh ite to begin
exploiting his qual itative advantage on the
kingside (the fact that he can advance his
pawns there, whereas Black is forced to
remain passive), when the long-range power
of the bishop will tell .
 Crux of the Position ctJ 1 93

1 9 g4! f6 The final touch : there is no defence against

20 h4 <i;f1
21 h5 lt:Jb6
22 l:d1 tt:Ja4
Disconcerted by the advance of the enemy
pawns on the kingside, Lautier looks for
counterplay on the q ueenside, but he choos­
es an unfortu nate moment for th is. He should
have opposed rooks by playing 22 .. J::td 8 ! ? ,
although here too Wh ite would b e clearly
better.
23 i.. a 1 tt:Ja5??
With his previous move Black prepared this
mistake. Of cou rse, 23 . . . tt:Jb2 was u nattrac­
tive in view of 24 l:t b 1 (or immediately 24
l:g 1 ! ? ) 24 . . . tt:Ja4 25 l:g 1 followed by f2-f4
and g4-g5, as was 23 . . . l:d8 24 f4 e5 25 fxe5
tt:Jxe5 (if 25 . . . fxe5 there is the strong reply 26
l:tf1 + <i;e6 27 tt:Je4 h6 28 g5 - less
convincing is 26 .l:tb1 tt:Ja 5 ! , when 27 .txe5?
is not possible in view of 27 . . . l:xd2+) 26
.ixe5 fxe5 27 :f1 + <i;e7 28 lt:Je4 - i n every
case Wh ite has a g reat advantage. But the
move in the game leads to a rapid defeat.
24 tt:Je4 lt:J b6
It was hopeless to play 24 . . . tt:Jxc4 25 .l:r.d7+
�8 26 h6 or 24 . . . 'it>e7 25 h6, when Black's
kingside pawn chain is destroyed .
25 g5
Wh ite has successfu lly carried out his
strategic a i m : he has created targets to
attack on the kingside.
25 . . . f5
25 . . . tt:Jaxc4 would also have lost in view of 26
gxf6 gxf6 27 tt:Jxf6 h6 28 lt:Jg4 .
26 tt:Jxc5 tt:Jaxc4
27 tt:Jd7 .:tc8
The point of Wh ite's manoeuvre is that
27 . . . l:tb7 28 tt:Jxb6 tt:Jxb6 loses because of
the pin: 29 .l:!.b1 l:b8 30 .te5.
28 tt:Jxb6 tt:Jxb6
29 .l:td6
the i nvasion of the 7th rank by the rook, since
if 29 . . . llb8 Wh ite decides matters with 30
i.. d 4.
29 . . . l:tc2+
30 <i;f3 .l:!.a2
31 i.. x g7 tt:Jc4
32 .l:!.d7+ <i;e8
33 .l:!.a7 tt:Jd6!?
Black defends imaginatively, but his cou nter­
attacking resou rces are clearly insufficient.
34 g6 hxg6
35 h6 tt:Je4
36 Wg2 .l::t xf2+
Or 36 . . . tt:Jg 5 37 i.. f6 .
37 �g1 l:td2
38 h7 lld 1 +
39 �g2 .:td2+
40 ..t>f1
Black resigned .
Only a superficial glance will suggest that
Black lost because of his blu nder on the 23rd
move . I th i n k that the reader will agree with
the author's opin ion : the real reason for his
defeat was that my young opponent lost the
thread of the game, by failing to fi nd the
correct defensive pla n , i nvolving the more
active placing of his knig hts.
I recommend the readers to compare this
episode with another, thematically very simi­
lar one, which occu rred in the 3rd game of
the A. Sokolov-Yusu pov Candidates Match
(Riga 1 986). It is analysed in my lectu re
'Surprises in the Opening' in the book
Secrets of Opening Preparation by Dvoret­
sky and Yusupov. There Black faced a
similar problem . By fi nding the correct
arrangement of his knig hts , he not only
successfully solved h is opening problems,
but also, after exploiti ng his opponent's
mistake , seized the in itiative.
1 94 �
Evgeny Bareev
Strategy i n G ra n dmaste r Games
I were played at the recent tou rnament in
am going to show you a few games that
Linares. Or more , correctly, we will look at
them together. After I returned home I
analysed my game with Anatoly Karpov, in
which at first sight all was simple and clear,
but even on this one I had to spend a mass of
time. This appl ies even more to the other
games, which I have not yet looked at. I am
afraid that many of my prel iminary impres­
sions of them will prove false. With your help
we will try to investigate them , so you will
have to do some real work, and to solve the
problems which the players faced during the
game.
We will focus mainly on positional problems.
However, in modern chess everything is
mixed up: even qu iet situations are usually
full of tactical nua nces, and in sharp posi­
tions one cannot forget about strategy. Now
you have to be a un iversal player, since it is
impossible to achieve good resu lts thanks
only to a good knowledge of the open i n g , or,
say, mastery of the endgame - comprehen­
sive and versatile preparation is req uired .
The lead ing players in the world have a
mastery of all the different chess weapons,
and what often come to the fore are
psychological factors . For example, flexibil ity
of th inking - the rapid switching during the
cou rse of a game from the solving of
positional problems to the fi nding of tactical
subtleties, and vice versa . Of enormous
importance are competitive form , reserve of
energy, and the abil ity to perform at one and
the same high level throughout an event. In
Linares, towards the fi nish there for a decl ine
in the standard of play even with you ng
players such as Boris Gelfand and Vishwa­
nathan Anand (and to a lesser extent with
Vasily lvanchuk, who was in ind ifferent form
from the very start). And the experienced
grand masters Artur Yusupov and Alexander
Bel iavsky, whose class is widely recognised ,
were simply nowhere to be seen at the end .
They weakened physically and were i n capa­
ble of solving problems with which they
easily coped at the start.
Only Garry Kasparov was able to play evenly
throughout the tournament. It is not just that
he calculates variations excellently a n d
possesses enormous experience; the world
champion's opening knowledge is universal­
ly recog nised , and on top of everything his
nerves at the board are usually stronger than
even those of his younger opponents.
Therefore his success is fully merited . But,
i ncidental ly, in cases when Kasparov does
not devote sufficient attention to special
preparation before an event, it is possible to
fight with h i m on eq ual terms, and for this
reason i n some recent tou rnaments he did
not ta ke fi rst place.
The whole range of questions on wh ich I
have touched here makes up a separate and
very important topic, to some aspects of
which I will retu rn . But now it is time to turn to
the games.
I will beg in in chronological order with my
game against Valery Salov, which took place
in the th ird rou n d . I n the fi rst two games I had
scored just half a point.
What should you do i n such a position? How
should you base your play? To hope for
overall competitive success, you have to win
 Strategy in Grandmaster Games '2J 1 95

quite a number of games. But if your

opponent does not make any very serious
errors, it will be very difficult to outplay h i m .
Should you play i n your usual manner,
circumspectly, aiming to exploit isolated
inaccu racies by the opponent, or try to
overcome him quickly, by playing extremely
sharply? It is my impression that players
have now stopped launching forwa rd des­
perately - they all prefer to play their own
game , and this is probably correct.

Bareev - Salov
- position after 1 0 b4 -
Linares 1 992
Bogo-lndian Defence
1 d4 ttJf6 Three suggestions have been received : 1 )
. . .'i!i'e7 and . . . e6-e5; 2) . . . c7-c5 ; 3): . . a7-a6
2 c4 e6
and . . . b6-b5 . It is evident that you lack
3 ttJf3 .ib4+ experience in such situations, otherwise you
Today the Queen's I ndian Defence and the would have known that fi rst you should place
Bogo-lndian Defence 3 . . . .i b4+ are to be your bishop on e4 . Not your knight, but your
found in the repertoire of nearly every player. bishop! Then you can play . . . c7-c5 . This is
Without a deep study of Queen's I ndian set­ how Black acted in most of the games I
ups it is impossible to play the Closed know, and Salov made his next move
Games, j ust as a knowledge of the Ruy practically without th inking.
Lopez is necessary for the Open Games. 10 . . . .ie4
4 ttJbd2 b6 1 1 l:!.c1
Positions of qu ite different types arise after For the moment the play is of a qu iet,
4 . d5 or 4 . . 0-0 .
. . .
manoeuvring character, and the players are
5 a3 .ixd2+ solving purely positional problems. Ti me for
6 .ixd2 .ib7 thought is restricted - only two hours. It is not
7 g3 known when the maximum output of time
7 �g5 is more often played , but the move and effort will be demanded , but in the future
made by me is no weaker. such a moment is bound to come. And for the
present you must comparatively qu ickly -
7 . . . d6
with in one or two m i n utes - make sensible
8 .tg2 ttJbd7 moves. If on each move you spend , say,
9 0-0 0-0 three, five or ten minutes , then later, when
1 0 b4 the crisis is reached and it becomes impor­
ta nt to think perhaps for half an hour, you will
(see diagram) not have a reserve of time. And even if you
outplay your opponent and obtain the better
Do you know the typical ways of playing such position, because of approach ing time-trou­
positions? What plans does Black usually ble, accu mulated fatigue and lack of fresh­
carry out? ness, mistakes will become simply inevitable
1 96 � Strategy in Grandmaster Games
and you will be unable to conclude the game
successfully. This is why Wh ite's last move
was made very qu ickly. At a1 the rook was
doing nothing, whereas at c1 it may come i n
usefu l . Quite sufficient grounds for t h e taking
of a decision.
11 . . . 'i!i'e7
Why didn't my opponent play 1 1 . . . "1!t'c8 with
the idea of placing his queen on b7? I th ink
that then the reply 12 c5 is u npleasant.
12 'i!Vb3
I had to think about this move, after which I
had already used 35 minutes, whereas Salov
had spent just five. An appreciable advan­
tage on the clock, which might well tel l in the
future! But here my opponent made a
serious mistake - he did not sense that one
of the key moments in the game had now
arisen . If you miss a key moment, later it will
be very difficult to make up for this.
I have set a positional trap. I make on that I
want to play 1 3 .i:!.fd 1 , but in fact I am
planning someth ing qu ite different. If Salov
had thought here for ten min utes or so, he
would have real ised that it was now time to
advance his c-pawn . But he made another,
rather unexpected move, although one that
sometimes occu rs in similar positions. Can
anyone suggest what this might have been?
No? What does the expert - Igor Khenkin -
say? He suggests 1 2 . . . c6 . Yes, this is a
typical idea , and not a bad one. U lf Anders­
son simply adored playing like that. Howev­
er, even so I would have preferred 1 2 . . . c5 .
But Salov, who was expecting the rook move
to d 1 , overlooked my main threat c4-c5!
12 . . . l:.fc8?
1 3 c5!
Here Salov quickly began catch ing me up on
the clock. By advancing my pawn to c5 I
have gained space. It is obvious that the
captu re on c5 does not work and Black now
has serious problems. Try to assess the
resulting position and decide how Black
should defend .
13 . . . b5
Correct! Black has to fig ht for the l ight
squares, and try to exchange a few pieces, to
leave the opponent with a passive bishop on
d2. If 14 c6 he now has the reply 14 ... tt'lb6.
And what plan do you recommend for White?
It is appropriate to have a proper think - if we
find the correct plan, it will be possible to
make the next few moves practically without
th inking.
I n cidentally, for the present we are solving
strategic problems. But at any moment
tactics may beg i n . It is important to be
constantly ready (in particular - psychologi­
cally) to switch from positional manoeuvring
to tactical l i nes.
I don't like the suggestion of 14 ..ih3. This is
sometimes played by Karpov in similar
positions, but at h3 the bishop is badly
placed . This move can be good when it
creates the strong threat of attacking the
bishop on e4 . This is clearly not the case
here.
Play the bishop to g5? Well , you and I h ave
different views on life . I never give up a
bishop - only if you put a kn ife to my throat.
Yes, one of the possible plans is .l:!.a1
followed by .l:!.fc1 and a3-a4 . Such play is
especially appropriate when the centre is
closed . I didn't want to beg in it now, not
because it is bad , but because it is neverthe-
 Strategy in Grandmaster Games
less a pity that the centre is not closed . The
opponent has a strong sq uare at d5. It is
desirable to force h i m somehow to declare
his intentions in the centre .
T h e correct decision was found b y our expert
Igor Khenkin : 1 4 'ir'b2 . For the moment
Wh ite does not determine anything. He
intends I1fd 1 , he retains the possibil ity of
playing :a 1 , and the reply . . . .i.d5 will be
made without gain of tempo. The modest
queen move contains another veiled idea,
wh ich will become clear from what follows.
14 'iVb2! h6
1 5 l:tfd 1 a6
16 .i.e1 d5
Wh ite has calmly improved the placing of his
pieces; his opponent has finally been unable
to withstand the tension and has declared his
intentions. Why did he play this? He wants to
move his passive d7-knight via b8 to c6 , but
with the pawn on d6 it is bad to play . . . lt:Jb8
because of the reply c5xd6 . Do you now
understand the reason why I was not in a
hurry to move my rook from c1 ?
1 7 .i.f1 !
Only here did Valery see to his horror the
strategic danger th reatening h i m . The knight
is going not to e5, but to a5 via the b3-
sq uare , which the q ueen has vacated for it.
The bishop on f1 is req u i red for an attack on
the b5-pawn after the thematic a3-a4 .
There is a suggestion for Black to remove his
bishop in advance from e4 , i n order to meet
the knight manoeuvre with his own knig ht: 1 8
li:Jd2 li:Jb8 1 9 li:Jb3 lt:Jc6 . However, for the
moment I will not move my knight, but will
play .l:!.a 1 and a3-a4 . And only after the
opponent defends his b5-pawn with . . . c7-c6
will I transfer my knight to a5. The threat of
the sacrifice lt:Jxc6 also arises.
17 . . . c6
If Black takes his knight to c6 , it is not clear
how he can hold his pawn on b5.
1 8 a4
CtJ 1 97
Now, as our second expert Volodya Kramnik
rightly remarked , Black should have serious­
ly thought about the exchange on f3 , in order
to forestall the un pleasant knight manoeu­
vre . The position is a closed one, and in such
cases bishops are no stronger than knig hts.
So, 1 8 . . . .i.xf3 1 9 exf3 g6 20 .:ta 1 . What
should Black do? After 20 . . . h5, apart from
the manoeuvre .i.d2-g 5 , he has to reckon
with the pawn offensive f3-f4 , f2-f3, h2-h3
and g 3-g4 . Then the bishop will go to h4 or
g3, with the constant threat of f4-f5. Moreo­
ver, it is not essential to do all this immediate­
ly - Wh ite can first play on the queenside, by
doubling rooks on the a-file.
After 18 ... .i.xf3 19 exf3 Khenkin suggests
playing 1 9 . . . e5. U nfortunately for Black, it is
not clear what he wants. The bishop comes
out to h3, and the rook endeavours to go to
e1 as soon as possible. In the event of . . . e5-
e4 the wh ite pawns are undou bled .
Thus the exchange on f3 does not promise
Black an easy life. What other plan of
defence can he choose?
1 8 . . . I1a7 1 9 .l:!.a 1 li:Jb8 (to captu re on b5 with
the c-pawn and place the knight on c6) is an
unpromising idea. It is too passive - Wh ite
will not hu rry with the exchange on b5, and
the knight on b8 will be so pitifu l !
Mark Dvoretsky's suggestion o f . . . lt:Jf6--e8-
c7 looks much better. But how is it, Mark
1 98 � Strategy in Grandmaster Games

lzrailevich , that you seem to know every­ 20 . . . axb5

thing, but your pupils can't discover a simple 21 lDd2 �g6
knight manoeuvre? 22 lb b3 'ii'd 8
Here Salov qu ickly took a difficult decision, What to do next - how can Wh ite strengthen
and, it would appear, a very strong one - he his position?
deferred the suggested manoeuvre and
made a pawn move on the kingside. It is
probable that previously he had already
encou ntered this idea.
18 . . . g5!?
I n itially I did not understand the opponent's
idea and I continued in the same direction .
1 9 .l:ta1 lbe8
It is hard to imagine, but . . . g7-g5 and
. . . lbf6-e8 are links in the same pla n . In the
event of 20 lbd2 �g6 21 f3 it will be possible
to prevent the e3-e4 break by 21 . . . f5 . And if,
after playing g3-g4 , Wh ite tries to prepare
�g3, the bishop will encou nter the reply
. . .f5-f4 . Generally speaki ng, for active play
It would appear that Wh ite should seek a
in the centre my bishop on e 1 is badly
way of exploiting the a-file and the weakness
placed . As you see, Salov not only hopes to
on c6 . The seizu re of the file would be a
hold the queenside, by placing his knight on
considerable ach ievement, but who is going
c7 , but he also takes measu res beforehand
to concede it? In the event of the exchange
to eliminate my future play in the centre and
of all the rooks I did not see how to ach ieve
on the kingside. It would appear that a
anyth ing real . For the moment the bishops
serious war in the futu re ca nnot be avoided .
are not taking part in the play, and diagonals
The situation had seemed to be completely for them must be fou n d .
safe. I was manoeuvring my pieces with in
T h e solution t o t h e position would seem t o be
my own territory and , comfortably seated in
to block the file as soon as possible (in the
my chair, I was th inking of how to win the
hope of again opening it in the future) and to
game most simply. I will strengthen my
in itiate play on the kingside.
position here, strengthen it there, and g radu­
ally break through the defence. But now it 23 lba5 'ikc7
becomes clear that there will be no easy 24 .:!.a3 .Ua6
success, and I will defi n itely have to devise 25 l:!.da 1 l:tca8
someth ing. But what to do now? Since I myself am not
However, my next few moves do not require sure that what I did next was correct, let's
any particular thought and are easily found. discuss it together.
20 axb5 Wh ite would like to play f2-f3 and g3-g4, in
Just in case, I hu rry to exchange pawns order to bring his bishop out to g3. The
before the appearance of the knight on c7 . opponent will obviously meet this with . . . f7-
But on the other wing too the knight has a f5.
comfortable square at g7. It would be desirable to shake the enemy
 Strategy in Grandmaster Games lLJ 1 99

defences as strongly as possible. From this Salov has deployed his pieces very skilfully.
point of view it is worth considering the 30 . . . tt:Jf7
attempt to latch on to the g5- and h6-pawns I was expecting only 30 . . . tt:Jd7, so that the
by 26 h4 . If 26 . . . gxh4, then 27 gxh4, and the knight should guard the f6-pawn . It turns out
rook ca n be switched to the kingside along that this is by no means obligatory. But which
the third rank. is more correct, 30 . . . tt:Jd7 or 30 . . . tt:Jf7 ? I
It would not be bad first to include the q ueen don't know - I fear that this is an irresolvable
in the attack, by placing it, say, at c1 . But then problem.
it will cease to oppose the . . . e6-e5 break. 3 1 i.. c 3 �e8
I did not in fact manage to fi nd a move which This was Salov's idea. I am forced to lau nch
completely satisfied me, but i n a practical into tactical compl ications.
game this is often a very d ifficult, al most 32 i.. xf6 gxf4
unrealistic task. Time is limited , and a
33 i.. x g7
reserve of it is needed for the solving of
purely tactical problems after the inevitable If 33 gxf4 there is the u npleasant reply
opening of the position . Thus sometimes you 33 . . .'�Jh5.
must quickly make some logical move, 33 . . . 'it>xg7
realising that it may not be the strongest. 34 gxf4
26 "iVc1 ! ? f6 I am playi ng on the knight at f7 and the rook
Black decides against playing . . . f7-f5 be­ at a6, which are not too well placed , and I
cause of h2-h4 , but now the plan of f2-f3 , hope to be the first to switch my pieces to the
g3--g4 and i.. g 3 gains in strength . kingside.

27 f3 tt:Jg7 34 . . . �xe2
35 �g3
28 �g2
A natural move, but 35 "iVf1 !, intending f4-f5
The bishop is no longer needed at f1 , and
and �g3+, was probably stronger. After
from g2 it will support the e2-e4 advance. If
some five m i nutes' thought I did not find this
Wh ite should succeed in playing th is, the
possibility, but I saw another idea.
opponent will be deprived of . . . e6-e5, his
only real cou nter-chance, and therefore he 35 . . . 'it>h7
must exploit it immediately.
28 . . . es
29 dxe5
Strategy g ives way to tactics. I exploit the
fact that 29 .. .fxe5? is bad in view of 30 h4 ! ,
when the g5-pawn can not b e held. If
necessary, the pressure on the c1 -h6 d iago­
nal can be fu rther intensified by i.. d 2 (now it
will be understandable why the q ueen went
to c1 ).
29 . . . tt:JxeS
30 f4
Strange: White's actions look logica l , but the
opponent's position is still holding out - 36 i.. f1
200 � Strategy in Grandmaster Games
A question: why did I play this? Was my flag
about to fall? Wel l , we each did indeed have
only about a minute left. Perhaps you don't
like my move, but you realise that I did have
some idea? That's right, the switching of my
rook to g2 - I very much wanted to g ive mate
on the g-file.
The black rook has to retreat on the e-file
(36 . . . Iic2 37 'ii'e 3 is obviously not worth
considering). But to which square? To e4 , e7
or e8 , or perhaps even e6?
The correct reply is e 7, in order to defend the
7th rank. And from a8 the second rook will
defend the 8th rank. But defi n itely not the
other way round!
Let's look concretely at 36 ... l:.e7 . Sacrifice
on g6? This gives noth ing - the knight on f7
secu rely defends the king . That means 37
l:!.a2 ltJh8 38 l:!.ag2 l:!.a8.
How should White conti nue the attack?
That's right, 39 f5! �xf5 40 �d3, and if
40 . . . �xd3 White has the decisive 41 'ifxh6+ ! .
But Black fi nds the defence 4 0 . . . .:tf8 . After 4 1
l:th3 there i s 4 1 . . Jie6, s o 4 1 lif3! must be
tried , and if 41 . . . l:lef7?, only then 42 l:ih3,
and mate is inevitable. But how to win the
game after 4 1 . . . .:te5 , do you see? There is a
pretty idea suggested by Kramnik: 42 l:lg4
(renewing the threat of 43 \!Vxh6+), and if
42 . . . ltJg6, then 43 .l:!.h3 .l:!.e 1 + 44 'ifxe 1 �xg4
45 l:!.g3 with an attack. But Black has the
reply 42 . . . �xd3, and if 43 .l:!.xf8 the double
attack 43 . . . 1i'e7. There is no mate : 44 .l:!.xh8+
�xh8 45 'i1Vxh6+ �h7. [It is not essential to
take the h6-pawn: 45 "ik c3! 'ii'e 6 46 I!g3 - the
bishop is attacked, and the deadly 4 7 ltJxc6
is threatened. But after 42. . . ltJg6! 43 i1h3
Black parries the attack by 43. . . 'iig 7! -
Dvoretsky.]
Instead of 42 l:!.g4 I would probably have
preferred 42 'iff1 ! (or 42 "ikb 1 ! ) .
If m y opponent h a d played 3 6 . . . Iie7 ! , in time­
trouble I would have been unlikely to see the
mating idea f4-f5 and 'iix h6+ - this is with in
the powers of perhaps only our expert
Volodya Kramnik. Besides, Black's defence
can be improved . I n stead of 38 . . . l:Ia8? he
can choose 38 . . . �e4 ! .
It is good i f you have a t least half an hour in
reserve - you can sit calmly and calculate
variations. But for some reason time is
invariably lacking when you most need it. If
only one could learn to play chess, so that at
the required moment there is always suffi­
cient time!
36 . . . .:te8?
37 .l:!.a2
It is probable that the position is now lost - it
is hard to defend against the mate on g7.
37 . . . �e4
38 'ii'c 3 ltJhB
In the event of 38 . . . .l:!.g8 the simple 39 'ii'f6 is
strong . It transpires that Salov had his own
idea - the switching of his knight to g6. The
ideas of g rand masters sometimes tu rn out to
be bad , but it is very rarely that they play
without ideas.
39 �d3 ltJg6?
39 . . . �xd3 40 'ii'x d3+ .l:te4 41 .l:tag2 .l:!.a7 was
essential.
40 .Uag2! l:.xa5??
A typical fortieth move in time-trouble. In the
tou rnament bulletin it is awarded two ques­
tion marks, but u ndeservedly - all the same
there was no way of saving the game.
[No, quite deservedly - since Black could
have seized the initiative with 40 . . . �xd3! 41
�xd3 .l:!.e 1 +! 42 � f2 l:te4. Therefore White 's
last move should be awarded not an excla­
mation mark, but a question mark. He could
have won by 40 �xe4! l:!.xe4 (40. . . dxe4 4 1
::1ag2) 4 1 :txg6 or 4 1 "ikf6. In a sharp time­
scramble such mutual tactical errors are,
alas, almost inevitable - Dvoretsky.]
41 Iixg6
Black resigned .
 Strategy in Grandmaster Games l2J 201

It anyone should tel l you that g randmasters Bareev - Kasparov

play without making mistakes - don't bel ieve Linares 1 992
it! In Linares, an el ite tou rnament ( 1 7th F I D E
King 's Indian Defence
Category), there was a huge n u m ber of
blunders, and not only in time-trouble. Artu r 1 d4 l2Jf6
Yusupov, for example, blundered away an 2 c4 g6
importa nt pawn against Miguel l l lescas as 3 l2Jc3 �g7
early as the 8th move. And here is a nother 4 e4 d6
example of a surprising 40th move. 5 �e2 0-0
6 i.g5
Ti mman - Karpov
This is the Averbakh Variation . Which of you
Linares 1 992 King's I ndian players can tel l me, what is the
most popular move for Black? This was the
one that was made.
6. . . l2Ja6
7 h4
A novelty? No, it's not. Here, you can read in
the bulleti n : Budnikov-Kruppa, USSR Cham­
pionship 1 99 1 .
What is the idea of the move h2-h4 ? Do you
think that I want to g ive mate? No, here there
is no mate . On the contrary, Wh ite is hoping
to block the kingside, in order to deprive the
opponent of active possibilities. Khenkin
adds: . . . and develop the knight at h3'. Yes,
·

For a start, in a good position Jan Tim man for
some reason gave up his central pawn : 39
J:tc1 ?? l:txd4 40 l2Jxf6. And now, instead of
the easily winning 40 . . . .l:txh4+ 41 'itt g 2 �h3+,
Karpov chose 40 . . .'it>xf6?? 41 l:.xe6+ �xe6
42 �xe6+ <;tJg7 43 'ii'f7 + 'itt h 6 44 'it'f6+ l2Jg6
(44 . . . .U.g6 allows mate on h8) 45 'ii'x d4.
Although he held on for a fu rther ten moves,
it was exclusively thanks to his personal
qual ities.
here the knight is more actively placed - I
th ink it was Tarrasch who fi rst said this?
Here the world champion fell asleep. I n
principle, o f cou rse, he knows everything
and he usually plays the open ing qu ickly. But
if you set him some problem, he becomes
the same as anyone else - he sta rts to think,
and sometimes for a long time.
7 . . . h6
After the game the champion timidly sug­
gested that he should have played 7 . c5. . .

In the eighth round my game with the world

8 �e3 e5
champion took place. At that point I had just
9 d5 l2Jc5
two poi nts, but I was still wanting to fi nish in
the prize-win ners. This meant that I had to 1 0 'ii'c 2 c6
beat Kasparov. But how to achieve this? 1 1 h5

(see diagram)
202 cJ;t Strategy in Grandmaster Games
When in a joint analysis after the game I
mentioned the move made by Yu ri Kruppa ,
Kasparov immediately pooh-poohed it. Try to
guess what decision was condem ned (and ,
probably, rightly) by the world champio n . The
move . . . g6-g5 ? No, that's in fact what he
played . The reply that he didn't like was
1 1 . . .gxh 5? ! . It not only weakens the f5-
square, but the h6-pawn comes under attack
and the h 1 -rook is immed iately in play. The
King's I ndian Defence is a resil ient opening,
and even here not all is clear, but neverthe­
less this is not what Black should play.
11 . . . g5
[A year later in the Interzonal Tournament
(Biel 1 993) Gelfand tried against Bareev a
clever idea of Mikhail Shereshevsky: 1 1 . . . cxd5
1 2 cxd5 i.d7!? 1 3 hxg6 fxg6. After 1 4 b4 (the
capture of the h6-pawn would allow Black to
gain excellent counterplay along the newly­
opened h-file: 1 4 .i.xh6 �xh6 1 5 1J.xh6 <J;; g 7
1 6 .l:!.h 1 '1:1 h 8 1 7 :xhB 'iixhB) 1 4 . . . lLJ a 6 1 5 a 3
h5 1 6 f3 a complicated position was reached
with chances for both sides. And later
1 1 . . . cxd5 12 cxd5 'fia5!? was also employed
- Dvoretsky.]
12 f3 a5
1 3 g4
As was plan ned , the kingside has been
successfully blocked . But what would you
say about attempts such as 1 3 0-0-0 or 1 3
.l:!.d 1 ? The idea of them is to hinder the
development of the bishop at d7, and to force
the prel iminary exchange of pawns on d5.
More specifically: 1 3 .l:i.d 1 (or 1 3 0-0-0)
1 3 . . . cxd5 1 4 cxd5 .ltd? 1 5 �xc5 dxc5. What
is the assessment of the resulting position?
Can't I succeed with the plan of seizing
control of the light squares?
Khenkin thinks that strateg ically White stands
better. Perhaps strateg ically this is so, if, say,
we make two successive moves for Wh ite -
a2-a4 and .i.e2-c4 . But dynamical ly, after
1 6 a4 c4! (with the idea of . . . 'it'b6, . . . .l:!.fc8
and . . . .i.f8-c5 ) the situation is far from clear.
The dark squares are very weak, and
nowhere will my king feel comfortable. Of
cou rse, it is possible to play this as White,
but I would not recommend it. And not with
every opponent can you decide to go in for a
position of this sort. Against Kasparov this is
completely hopeless - he has too keen a
feeling for the in itiative. I chose a completely
different strategy, ai med at restriction - I tried
not to al low h i m to calculate variations, and
to remind h i m what it was l i ke to fight i n his
matches against Ka rpov.
13 . . . .i.d7
1 4 lt:Jh3
Let's try and devise a plan for Black.
There is a suggestion that he should play
 Strategy in Grandmaster Games
14 . . . 1!fb8. I remember that a similar manoeu­
vre was carried out i n 1 982 in a game
Bareev-Khenki n , and i n the meantime it has
not become any better. At that time Igor was
not yet a King's I ndian player and he can be
excused .
Another proposal is 1 4 . . . cxd5 1 5 cxd5 �c8
(threatening 1 6 . . . tt:Jxh5) 1 6 lt:Jf2 tt:Ja4 . I will
tell you in secret that I myself was d reaming
of fi nding some way to exchange the knight
on c5.
Here, finally, is the correct idea : 14 . . . a4,
preparing . . . "�a5 followed by the exchange
on d5, . . . Itfc8 and . . . b7-b5. Black only
needs to delay slightly, when he would fi nd
himself in a positional bind, and therefore he
needs urgently to take the offensive on the
queenside. If the b7-pawn should reach b4
and drive away the knig ht, White will suffer
from a lack of space and will be unable to
arrange his pieces normally.
14 . . . a4!
What should I do now? In a game there can
be key moments, on which its entire subse­
quent course depends. Either you fi nd the
correct plan and you hold the in itiative, or
you fail to solve the problem and you begi n to
drift.
Perhaps the move made by me is not too
good , but someth ing had to be done, and I do
not see any other way of combating the
opponent's pla n . At any event, against the
world champion my idea justified itself. If you
are able to find it, you can confidently play
the Averbakh Variation with White.
Take on c6? This is what many have done
and for none of them did it tu rn out wel l .
Black recaptu res 1 5 . . . bxc6 a n d t h e weak­
ness of the d6-pawn is of no sign ificance.
Yes, you' re right, you must play the q ueen to
d2. A difficult move! Only, don't th ink that I
wanted to sacrifice a piece on g5. The idea is
to exchange the q ueens.
1 5 1ifd2! cxd5
ttJ 203
1 6 cxd5 ifa5
The champion sat for 25 minutes, calculating
the variation 1 6 . . . ..ltxg4 1 7 fxg4 tt:Jfxe4 1 8
tt:Jxe4 tt:Jxe4 . After the game he regretted that
he had not gone in for this conti nuatio n ,
althoug h he also agreed that after 1 9 �b4
the piece is worth more than the pawns.
17 tt:Jb1
And what would you now play with Black?
First, let's assess the position . You th ink that
Wh ite stands slightly better? Apparently
that's what Kasparov also assumed , but I am
not so sure. I think that the chances are
roughly eq ual.
Who suggested 17 . . . �c7 ? But why immed i­
ately capitulate in this way? I reply 1 8 tt:Ja3,
then Itc 1 , tt:Jf2 and at some point tt'lc4. Black
is left with no cou nterplay.
Khenkin recom mends 1 7 . . . �xd2+ 1 8 tt:Jxd2
( 1 8 'it>xd2 tt:Jfxe4+ is dangerous for Wh ite)
1 8 . . . b5 1 9 tt:Jf2 Itfc8. Correct! Black should
probably have gone i n for this position - to
me it seemed roughly eq ual. To fight for an
adva ntage the wh ite knight should sta nd not
at d2, but at a3.
I n view of th is, 18 i.. x d2 can also be
considered . But I would not have ventu red
th is, since then the sacrifice of a piece is
un pleasant. If 1 8 . . . tt'lfxe4 1 9 fxe4 tt'lxe4 , then
20 tt:Jc3; therefore I suggest 1 8 . . . �xg4 1 9
204 <;!? Strategy in Grandmaster Games
fxg4 4Jcxe4 (with the idea of . . . 4Jxd5), and if
20 i..f3 - 20 . . . 4Jxd2 followed by 21 . . . e4 .
[Apparently, after 20 t0c3!? t0xc3 2 1 bxc3
t0xd5 22 c4 White would nevertheless have
stood better - Dvoretsky.]
Each of us has his own style, his own way of
playi ng. In Black's place I would have agreed
to the exchange of queens. But Kasparov
does not like positions in which he has no
counterplay. He did not want to go i nto a
qu iet and slightly inferior (as it seemed to
him) endgame, and so he decided on a
rather dubious piece sacrifice. I n add ition he
took into account my reputation of being a
player who is incapable of calculating a
single variation , and he wanted to exploit
this. In the end I did indeed go wrong , but
here he was simply lucky - the resulti ng
position was not one of those where I often
go wrong, since in fact there was hardly
anything to calculate.
17 . . . 4Jfxe4
1 8 fxe4 4Jxe4
1 9 'il'xa5 llxa5
What move should White make?
After the game Kasparov showed me many
interesting variations. I hadn't seen them, but
there was no need to.
20 4Jc3!
A simple and very strong move. If 20 . . . 4Jxc3
2 1 bxc3 lixd 5 , then 22 lLlf2 with a subse­
q uent blockade on the light squares. Black
has th ree pawns for the knight, but without
cou nterplay this is insufficient - he needs
some dynamic factor, such as two connected
passed pawns.
White would have lost (or almost lost) after
20 tLlf2? tLlg3 2 1 .l:tg 1 4Jxe2 22 'iti>xe2 e4, and
the b2-pawn can not be defended .
Khenkin suggests 20 .i.b6 .l:l.xd5 2 1 i.f3. I
don't l i ke this - I wanted to make normal ,
solid moves, whereas here, with his develop­
ment incomplete , White scatters his pieces
around too much. Apart from 2 1 . . ..l:lb5, he
even has to reckon with 2 1 . . . 4Jf6 22 i.xd5
4Jxd5. He is a rook up, but the g4-pawn is
lost, and in the future also, possibly, the h5-
pawn . Too complicated !
20 . . . tLlg3
21 .l:tg1
I thought for a long time about 21 tLlf2 - an
attempt, by also giving u p the exchange, to
play for a blockade. But a rook is neverthe­
less stronger than a knight.
21 . . . 4Jxe2
2 1 . . . 4Jxh5 has been suggested . Then , if
there is noth ing better, 22 4Jxg5 lLlf4 23
4Jge4 4Jxd5 24 4Jxd5 .Uxd5 25 g5 and White
has an obvious advantage. 21 . . .f5 22 l:!.xg3
f4 also did not work: 23 i.. b 6 ! .
22 '1t>xe2
What move should Black make? Of course,
he has to open the diagonal of his dark­
square bishop.
22 . . . e4
And what should Wh ite play?
(see diagram)
 Strategy in Grandmaster Games
He has to reckon not only with 23 . . . a3, but
also 23 . . . i.xc3 24 bxc3 l:txd5, then . . . i.b5+
and . . . i.d3 . His king is not too secure, and
the e4-pawn can never be captu red - i n
general, he h a s problems in converting h i s
advantage. A n d yet Wh ite's position i s
already close t o being won .
23 l:lac1 !
Now to 23 . . . a3 there is the reply 24 b4 , while
if 23 . . . i.xc3, then either 24 l:l.xc3 , or 24 bxc3
l:txd5 25 c4. It is by such u ncomplicated but
accu rate moves that a game is held together.
Many of them have to be made - miss one
somewhere, and the result can change.
The other natural continuation 23 l:!.ad 1
would apparently have been weaker be­
cause of 23 .. .f5! 24 gxf5 a3 25 i.d4 (25 b4
i.xc3 26 bxa5 i.b5+ 27 'it>f2 l:txf5+ 28 'it>g2
i.e2 ! ) 25 . . . i.xd4 26 l:!.xd4 axb2 with an
unclear game.
23 . . . f5
24 gxf5
One of the critical moments of the game.
Here the champion thought for a long time.
I ncidental ly, good players also d iffer from the
not so good in that, for some reason , it is at
the critica l , most important moments that
they th ink.
Black stands worse . If he hopes to save the
game, he must all the time fi nd the strongest
moves, which do not worsen his position any
lb 205
fu rther and maintain cou nterplay - otherwise
the opponent's task will become too easy. He
has th ree possibil ities: the captu re on f5 with
bishop or rook, and also the attack on the h5-
pawn.
The last possibil ity is the strongest. If Black
should manage to obta i n two con nected
passed pawns on the kingside, and also post
his bishop at f3 , noth ing will remain of
Wh ite's advantage. Kasparov saw 24 . . . i.e8 ! ,
o f course, b u t he rejected it. Why?
Dvoretsky suggested that it was because of
25 lt:lf4 . It is u n l i kely that Kasparov consid­
ered this, si nce he did not mention it i n our
joint analysis after the game. 25 . . . gxf4 26
i.d4 is certainly dangerous for Black, but he
has the simple 25 .. J::txf5.
The world champion was concerned about
the position arising after 25 lt:lxg5! hxg5 26
l:!.xg 5 'it>h8. But he had no need to - here at
least the n u m ber of pieces is equal. For
example, if 27 lt:lxe4 B lack can consider
either 27 . . . i.xb2 , or 27 . . . i.e5, intending
. . . l:txd5 or . . . .l:!.b5.
24 . . . lbf5?
25 lt:lf2 .tea
Which move did I make, without any thought
at all? That's correct, the pawn has to be
defended - it is j ust too important.
26 l:l.h1
At this moment Kasparov's wife became
anxious. Apparently she can't play chess,
but she can clearly determine the situation
on the board either from the expression on
her husband's face , or by the movement with
which he straightens his tie. Evidently at this
moment he pul led his tie the wrong way.
26 . . . i.b5+
Here I bel ieved that I was going to win . If the
opponent himself gives u p such an important
bishop, it means that all is not well in his
position.
27 lt:lxb5 l:txb5
28 l:!.c8+ 'it>h7
206 � Strategy in Grandmaster Games
If you can now fi nd one more correct move,
you will win the game, irrespective of whom
you are playing against.
After 26 .l::t h 1 ! it should not have been so
difficult to play again in the same style: 29
l:tb 1 ! . At the board after 29 . . . l:tfxd5 I consid­
ered only the reply 30 tt:Jxe4 . But it is not the
strongest: after all, Black is dreaming of
placing either of his rooks on h4 and picking
up the h5-pawn , and the captu re of the e4-
pawn makes it easier to carry out this pla n .
3 0 b4! axb3 (otherwise 3 1 a3) 3 1 axb3 i s
much stronger. It is true that there are few
pawns left on the board , but this is unimpor­
tant. It only remains for me to beg in an attack
- place my rook on c7 and then direct the
second rook into the opponent's position .
Unfortunately, I did not see as far as 30 b4.
As a result I rejected 29 .l:l.b1 ! and th rew
away the win .
2 9 l:td1 ? l:txb2+
30 .l:l.d2
That day I would have been better playing
draug hts, because I only calculated varia­
tions with captu res: 30 . . . l:txd2+ 31 .txd2
l:txd5 32 ..ic3. After the exchange of bishops
there is noth ing to defend the king and my
attack would be bound to succeed .
30 . . . a3!
With a sigh of relief Kasparov's wife sat back
in her chair - it was sensed that she was now
relaxed about the outcome of the game.
31 .l:.c7 l::!. x d5
32 tt:Jxe4
What should Black play?
The correct course was 32 . . . l:tdxd2!+ 33
..ixd2 'i!?g8 followed by 34 .. J:txa2. But
Kasparov preferred to move his king immedi­
ately.
32 . . . 'itlg8?!
How did Wh ite reply?
33 .l:tcc2 !
This move does not change the evaluation of
the position. But whereas after the exchange
of rooks on d2 the position would have been
simply equal, now, in time-trouble, to hold the
game Black had to demonstrate everything
he was capable of. To the credit of the world
champio n , who was again under the th reat of
defeat, he began playing with redoubled
strength .
Wh ite wants to exchange a pair of rooks: 34
l:txd5 l:txc2+ 35 l:td2 , and if 35 . . J�b2 - 36
'itld3, intending 37 l:txb2 axb2 38 �c2 . The
h5-pawn is al ive, and the black pawns a re
weak - in general I retain excellent chances
of success.
33 . . . b5!
34 l:txd5 l:txc2+
35 l:td2 .l:l.b2
 Strategy in G randmaster Games
36 c;;t> d 3
Of cou rse , not 36 tt:lxd6? �c3 .
36 . . . d5
Now I am the one who has to exercise
caution. To where should I move my knight
(in my own time-trouble)?
Why to c5 , rather than g3 or d6? You want to
hold the b3-sq uare? I did not particularly
want to play this (it would have been
tempti ng to transfer the knight to f5), but I
had to - the threat of . . . b4-b3 was too
serious. Bad was 37 tt:ld6? d4 38 �f2 b4 39
l2lc4 b3! or 39 l:!.xb2 axb2 40 '>t>c2 d3+.
37 tt:lc5!
What should Black do?
37 . . . l:!.b4!
The No. 1 enemy is the h5-pawn , and it must
be attacked as soon as possible.
38 tt:le6 I!.h4
39 tt:lxg7 '.t>xg7
40 �d4+ 'it>g8
A rather strange move (not surprising - it
was the fortieth ! ) . Si nce childhood we have
been taught that in the endgame the king
should be moved towards the centre , and
therefore 40 . . . '.t>f7 suggested itself. What
would you now recommend for Wh ite?
Yes, 41 I!.c2 can be played , but what for? To
activate the rook? Breaking through with the
lZJ 207
rook is not a problem, but the rook a n d
bishop on their own can not achieve any­
th ing. If only the king could be included in the
attack! I magine the wh ite king at e6 and the
bishop at e5 - Black would have to resign
even if I had no pawns left at all. Therefore I
should have had the sense to play 4 1 'it>c3 ! ,
trying t o reach t h e d5-pawn with a by-pass
via the wing . The position would probably
have remai ned drawn, but Wh ite would not
have risked anyth ing and he could have set
his opponent some problems.
41 �e5? !
The idea is the same: 4 1 . . . I!.xh5? 42 '.t>d4.
U nfortu nately, we are not playing draug hts ­
it is not obligatory to captu re.
41 . . . '.t>f7 !
A few more moves were made, but they were
no longer able to change anything.
42 I!.e2
To stop the king from going to e6.
42 . . . I!.xh5
43 '.t>d4 'it>e6!
Not fearing ghosts!
44 � b8+ '.ttf5
45 l:!.e3 l:!.h1
46 I!.f3+ �g4
47 I!.g3+ '>t>f5
48 I!.f3+
Draw ag reed .
I n cidentally, what do you th ink, what kind of
game did Kasparov and I play, positional or
combinative? There was seemingly a sharp
struggle, and a piece was sacrificed . But at
the same time, most of the decisions were
taken by the two players on purely strategic
grounds. So that even I do not know!
In conclusion , I will show you my game with
Anatoly Karpov.
208 � Strategy in Grandmaster Games

Karpov - Bareev A very natural move , but in my opinion , not a

Linares 1 992 very good one. I n the game Nikolic-Schwartz­
man (Wij k a an Zee 1 995) 1 2 .ll e 1 !? 0-0 1 3 e3
Slav Defence
was tried .
1 d4 d5
12 . . . 0-0
2 c4 c6
3 lt:Jc3
Delaying the development of the g 1 -kn ight
has some point. For example, in the variation
3 . . . lt:Jf6 4 cxd5 cxd5 5 �f4 lt:Jc6 6 e3 e6 7
i.d3 .id6 Wh ite can play 8 �xd6 'iVxd6 9 f4 ! .
But it i s well known that noth ing comes for
free. The move order chosen by Karpov also
contains a drawback: the opponent has a
sharp gambit conti nuation, which has now
come into fashion .
3. . . e5!?
4 dxe5 d4
5 lt:Je4 'iVa5+
6 lt:Jd2 Let's beg i n from here! I wonder if anyone can
6 .id2 !? would seem to be stronger. But I suggest the move chosen by the former
don't want to dwell on the su btleties of the world champion?
opening - it would take too much time to 1 3 lt:Jg5 easily comes to mind. It is not very
investigate the nu merous complicated and good , since after 1 3 . . ..�:Jc5 Black is threaten­
non-standard positions which can arise . So ing 1 4 . . . h6. All the pieces are exchanged on
let's restrict ourselves to just one - that e4 , and the result is a d raw.
which arises on emerging from the opening You want to play 1 3 c5, placing the pawn en
in th is game. And if you want to work on the prise? To say the least, this is not at all a
theory of the open ing, you can do that Karpov move.
you rself. No other suggestions? A pity. He made his
6... lt:Jd7 move very q u ickly.
7 e6! ? 1 3 b3
Here 7 lt:Jf3 tt:Jxe5 h a s also been played , but It can happen that you are th inking after a
usually Black has managed to maintai n the move made by Karpov, and you fi nd an idea
balance. which you want to carry out. But first you
7. . . fxe6 have to understand his plan. He does not
8 g3 e5 want to exc�ange pieces prematu rely; he will
9 �g2 lt:Jgf6 develop his bishop at b2, place his queen's
rook at e 1 or d 1 , and attack the centre with
1 0 lt:Jgf3
e2-e3. All his pieces will be harmoniously
1 0 lt:Jh3. placed . Black must somehow oppose the
10 . . . �e7 opponent's pla n , and fi nd a cou nter-plan.
1 1 0-0 'iVc7 I see two ideas for Black. Neither of them is
1 2 'iVc2 immed iately apparent.
 Strategy in Grandmaster Games
The first begins with the obvious 1 3 . . . lbc5,
but after 14 �b2 he needs a move which
does not i mmed iately suggest itself: 1 4 . . . g 6 ! .
After e2-e3 t h e bishop goes with gain of
tempo to f5, and then the knight to d 3 .
You suggest preparing t h e development of
the bishop by 1 4 . . . lbg4 (instead of 1 4 . . . g6)?
To be honest, I didn't see this move. Let's
have a look: 1 5 h3 �f5 1 6 'ili'd 1 lbf6 1 7 g4
i.g6 - everything would seem to be alrig ht.
Perhaps this was even stronger.
I chose a different plan - an interesti ng one,
but on another occasion I would not have
gone in for it ( 1 3 . . . lbc5 is simpler and safer).
13 . . . .l:leB
1 4 �b2 �f8 !
Now if 1 5 e3 there follows 1 5 . . . d3! and
1 6 . . . e4 .
1 5 lbg5 lbc5
If the e4-square is conceded to Wh ite's
kn ight, he will gain an obvious advantage,
and for free. My idea is first to exchange the
knights, then the light-sq uare bishops, to
leave the opponent with only a bishop on b2
and not to allow it to become active, by
meeting e2-e3 with . . . c6-c5.
It is impossible to guess Karpov's next move.
Dvoretsky's suggestion is 1 6 a3. Close! This
was the second move suggested by the ex­
champion after the game.
16 h3
Typical Karpov! After th i n king what he was
defending against, I fi nally rea l ised what I
myself want - 1 6 . . . g6 and 1 7 . . . ..if5 . I am
ashamed to say that even after 1 5 . . .lbc5 I did
not yet see the idea of bringing out the
bishop to f5. After mentally thanking my
opponent for the h i nt, I continued playi ng.
16 . . . g6
1 7 g4
Karpov has taken control of the f5-sq uare
and is continuing the fight for the light
squares.
ctJ 209
When you are playing such a complicated
game, you constantly have to chose from
continuations that are al most equ ivalent.
One move is slig htly stronger, another is
slightly weaker - how do you sense this?
After analysing the position later, I came to
the conclusion that instead of 1 6 . . . g6 it would
have been more accu rate to play 1 6 . . . a5!.
The move is a sensible one, of course, but
why it should be made at this precise
moment - d u ring the game it was completely
i mpossible to understand th is. I will explai n :
after lbg5-e4 i t is important t o have t h e reply
. . . a5-a4 . There is the th reat of driving away
the bishop by . . . a4-a3, and if the wh ite rook
is no longer at a 1 , it will be possible to
exchange pawns on b3 and adva nce the
rook to a2.
All this is rather complicated , and I cannot
condemn myself for the move made in the
game, 1 6 . . . g6. But my next inaccu racy was
altogether more serious.
An important and rather complicated mo­
ment in the game. Can you work out what is
happen ing here and what should be played?
Often , for some reason , it is bad moves
which instantly come to m i n d . However, it is
easier to fi nd a bad move than to make it,
because you can not bring you rself to play it
and you beg i n to have doubts .
And here we have the fi rst not altogether
210 � Strategy in Grandmaster Games
successful suggestion: 1 7 . . . b6. Why weaken
the c6-square? In order to combat Wh ite's
plan? But what is his plan? You th ink it is to
place the knight on e4? Noth ing of the sort!
His main idea is the undermining move e2-e3!
You would have played your bishop to g7?
You have found someth ing to boast about.
Many would have played this, and that's
what I played . But I shouldn't have done.
We will fi nd the correct solution if we
understand why the seemingly rather natu ral
move 17 . . . �g7 is bad . No, it is not a matter of
1 8 b4 tt::le 6 1 9 tt::l g e4 - after all, here I have
the excellent f4-point.
I ncidentally, the problem also proved to be
too difficult for Karpov. He is a mig hty
prophylactic player, and is especially strong
in moves such as 16 h3, but here there is
noth ing for Wh ite to defend against - he
simply needs to fi nd an active pla n .
After 1 7 . . . �g7? the reply 1 8 e 3 ! is very
strong . If 1 8 . . . d3, then 1 9 'ii'd 1 . The idea was
to support the d3-pawn by . . . e5-e4 , but now
1 9 . . . e4 is dangerous because of the ex­
change sacrifice 20 �xffi �xf6 21 tt::l g xe4 , or
first 20 b4!?. If 1 9 . . . h6 20 tt::l g e4 , while after
1 9 . . . h5 there is the strong reply 20 b4 . I n
every case the d3-pawn , lacking support by
the other pieces and pawns, is lost.
Black would probably have had to reject
1 8 . . . d3 in favour of 1 8 . . . dxe3 1 9 fxe3 , when
White has achieved exactly what he wanted .
He has deployed his forces wel l and has won
the strategic battle .
One of the experts suggested that, although
the strategic battle has been lost, the
position can be held by tactics: 1 9 . . . �h6. Let
us check: 20 .l:txffi �xg5 21 .l:l.xg6+! hxg6 22
'ii'xg6+, and in whose favou r are the tactics?
Of course , it does sometimes happen that a
chance tactical opportun ity changes the
evaluation of a position. But normally, if you
are strateg ically lost, here everything col­
lapses and all the variations turn out not in
your favou r.
1 8 . . . �h6 (instead of 1 8 . . . dxe3) hardly helps
Black. Wh ite has a pleasant choice between
1 9 exd4 and 1 9 tt::lg e4 .
How should Black defend? 1 7 . . . �h6 is better
than 1 7 . . . �g7, but it involves a loss of time
(the bishop is doing noth ing on h6). After 1 8
tt::lg e4 tt::l xe4 1 9 tt::lxe4 tt::l xe4 2 0 'ii'xe4
Wh ite's position is preferable i n view of the
same idea of the underm i n i ng move e2-e3.
The correct solution is 1 7 . . . tt::l e 6 ! . Now if e2-
e3 Black always has the reply . . . c6-c5. I n
the event of 1 8 tt::l g e4 tt::lxe4 1 9 tt::lxe4 i.g7
we reach a position which occu rred in the
game.
I will once again emphasise: sometimes i n a
game a critical moment is reached . If you
miss it, if you fail to fi nd an accu rate move,
the game changes cou rse and may begin to
go down h i l l .
T h e problems which Karpov and I were
trying to solve were seemingly purely strate­
gic, but at the same time the play was full of
tactics, little concrete variations, which need­
ed to be discovered and calculated . Some­
times you have to delve i nto a position and
th ink for at least ten minutes. You can 't
spend ten minutes on every move - other­
wise, before you know it, you end up in time­
trouble. But at important, critica l moments,
time should be spent. The q uestion is merely
to decide which moment is critica l . Anatoly
Evgenevich and I did not manage to do this,
and we both missed it.
17 . . . Ji.g7?
1 8 tt::l d e4? tt::lfxe4
1 9 tt::l x e4 tt::l e 6
It is now time for Wh ite to forget about
striving for an advantage and to th ink about
how to equalise. The underm i n i ng move 20
e3 is now pointless because of 20 . . . c5 , and
the exchange on d4 will g ive me the very
important f4-point.
20 b4
Apart from 20 e3, I also considered 20 c5,
 Strategy in Grandmaster Games
but the move in the game did not occur to
me. Meanwhile it is q u ite thematic. Why was
Karpov not in a hu rry to advance his pawn to
c5? To avoid conced ing the d 5-point to the
bishop (20 c5 lDf4 21 e3 lbxg2 and .. .i.e6-
d5).
20 . . . lDf4
21 e3 lbxg2
22 �xg2
How would you assess this position? More
pleasant for Black? But it is not enough to
give such an assessment - you must also
demonstrate it, and this is not easy. For
example, noth ing is g iven by 22 . . . .i.e6 in
view of 23 lDg5. Here I engaged in some­
th ing rather pointless - I tried to breach the
opponent's defences immediately, by analys­
ing 22 . . . c5 23 lbxc5 b6 24 lDe4 .i.b7. Finally
I sensed : this does not work, White's position
is too solid. In add ition I have an extremely
stupid bishop on g7. It was very usefu l to
realise that at the very first opportun ity I
should try to get rid of this useless piece.
22 . . . aS!
23 a3
Of course , I advanced my a-pawn . If you
don't play this, you simply lose you r self­
respect! But what next?
I still can't understand why I didn't make the
natu ral move 23 . . . 'i¥e7 ! . At c7 the q ueen is
l2J 21 1
doing noth ing, whereas from e7 it covers two
important diagonals, defends the f6-square
and attacks the b4-pawn . If 24 c5, then
24 . . . .i.e6 25 lDd6 .l:lf8 . It is probable that I
was instinctively afraid of placing my queen
at e7 because of some tactics involving
.l:lae 1 - but I was wrong! 24 l::ta e1 axb4 25
axb4 'i!i'xb4 26 exd4 does not work in view of
26 . . . �e6 , or, even better, 26 . . Jla2 . Perhaps
Wh ite should play 24 exd4 immed iately, but
then he has to reckon with the interposition
of 24 . . . axb4 ! ? .
23 . . . l:tf8
The next problem is what to do as Wh ite. It is
like th is, when you a re playing a game, and
there is no time to relax - one problem after
another!
24 c5 is obviously illog ical - it was against
this that Black's last move was aimed : now
the knight goes to d6 without gain of tempo.
The correct reply was 24 f3! . White must
consolidate. If Ka rpov had found it, he wou ld
not have lost. But he made a move from the
ra nks of 'how not to play chess'.
24 .l:lad 1 ?
He wants to captu re on d4, but he plays this
in the wrong order. For a start I open the file
which has been conceded to me.
24 . . . axb4
25 axb4 dxe3 !
Now the two possi bil ities available to White
are of roughly equal merit, and neither gives
clear eq ual ity. If he captu res 26 fxe3 , then
after 26 . . . J:.xf1 27 .:f.xf1 'ii'e 7 the b4-pawn is
hanging. I develop my bishop at e6 and at
some point play . . . h7-h5.
26 f3
If Wh ite were to make another couple of
moves in succession and pick up the e3-
pawn , he would stand better. But who is
going to allow that? Can you guess what I
played?
26 . . . �e7
212 � Strategy in Grandmaster Games
At last! 26 . . . .l:Ia2 was premature - the
opponent would have replied 27 .l:Ide 1 and
then driven away the rook with 28 'i!Vb3. But
now the b4-pawn is hanging , and in the event
of 27 c5 ..ll. e6 I am now securely controlling
the a2-sq uare.
27 �c1
'What candidate moves are there?' Artu r
Yusupov would have asked you here.
Excellent - 27 . . . �h6 and 27 . . . 'i!Vxb4 . And
which one shall we make?
27 . . . ..ll. h 6!
After 27 . . . 'ii'x b4 28 ..ll. x e3 Wh ite's position is
not worse, despite being a pawn down .
28 .l:Ide1
Karpov finally real ised that his rook on d 1
was doing nothing. What should Black play
now?
28 . . . ..ll. e 6
And again in the event of 28 . . . 'ii' x b4? I would
have been tricked : 29 g5 �g7 30 ..ll. x e3 with
equal ity. Now, however, things become bad
for White - he has noth ing better than 29
�xe3, but then he will not have sufficient
compensation for the b4-pawn .
You wish to try 29 lt:Jc5 ? Black repl ies
29 . . . i..f7 30 i..x e3 �xe3 31 .l:Ixe3 b5, and
how is Wh ite to defend? Let us continue the
variation: 32 .l:Ife 1 �xc4 33 .l:Ixe5 'ii'f7(f6).
There is a double attack on f3 and a2, and
later I will add the bishop on d5.
32 'it'c3! (instead of 32 .l:Ife 1 ) is a cleverer
idea : 32 . . . �xc4 (32 . . . .l:Ia2+ 33 .l:If2 ) 33 .l::i.a 1 ,
but after 33 . . . .l:Ixa 1 34 'i!Vxa 1 �d5 Wh ite stil l
stands worse, a n d worse t h a n in t h e game.
After all, a bishop is stronger than a knight!
29 �xe3 �xe3
30 .l:Ixe3 'ii'x b4
31 .l:!.b1
Again there are two candidate moves :
3 1 . . . "it'xc4 and 3 1 . . . 'i!Ve7. Which would you
prefer?
I thought that in the endgame after 31 . . . 't!Vxc4
32 'it'xc4 ..ll. xc4 33 .l:Ixb7 I would have only a
slight advantage thanks to my extra pawn ,
but Karpov said that it was altogether drawn.
You may not bel ieve me, but you should
bel ieve Karpov!
31 . . . 'it'e7 !
Wh ite's position would appear t o b e tenable,
but he needs to make accu rate moves. And
here Karpov had ru n i nto time-trouble.
32 lZ'lf2
This move did not even occu r to me - why
remove the knight from the centre? I thought
that it was i mportant for Wh ite to pick up the
b7-pawn and therefore I was expecting 32
'i:Vb2 ! �xc4 33 'it'xb7. What can I do?
Exchange q ueens and after 33 .. ."iVxb7 34
.l:Ixb7 attempt by 34 . . . .l:Ia2+ 35 '.t>g3 ..ll.f1 to
g ive mate? This is clearly a utopian dream . It
is probably better to play 33 . . . 'it'h4 (threaten­
ing a rook check) and then 34 . . . �d5. But
how to win here, I don't know - it is extremely
difficult to shake Wh ite's defences. In the
game I had trouble in breach ing a similar
position even though I was the exchange up.
Why, thoug h , did the opponent play his
knight to f2? He wanted to captu re the e5-
pawn , but here he was u nfortunate - I found
a strong reply. What was it?
No, not 32 . . . 'it'c5. Here he had something
prepared - 33 'i:Vc3 or 33 .l:Ibe 1 , I don't
 Strategy in Grandmaster Games
remember. But the idea i s correct: Black
needs to attack the c4-pawn , to obtain the
d5-square for his bishop.
32 . . . 'it'f7 !
Grandmaster Razuvaev very much likes
moves which fulfi l two or more fu nctions.
From f7 the q ueen attacks f3 and c4 , and
defends the b7-pawn , so that Yu ri Sergeevi­
ch would have praised me here. If now 33
tt'ld3, then 33 . . . e4 ! 34 fxe4 (things are also
not easy for White after the more tenacious
34 tt'le5 exf3+ 35 llxf3 'i¥g7 ) 34 . . . ..txc4 35
tt'le5 �a2 36 "i!Vxa2 .1i.f1 +.
33 't!Vb2
'Better late than never! '
33 . . . ..txc4
34 'i¥xb7
34 'it'xe5 is hopeless - the b7-pawn is stil l
al ive, and t h e weakness a t f 3 is perceptible.
Now Wh ite seemingly achieves his aim, but
in fact he ru ns into a blow prepared
beforehand.
34 . . . 'ikf4!
An extremely u n pleasant surprise . It is said
that Karpov plays very well i n time-trouble.
He confi rmed this opinion i n the present
game, mainly by the fact that he managed to
make all his moves, on the way losing only
the exchange.
35 'ikxc6 !
35 .l:tbe 1 ..td5 (or 35 . . . .l:i.a2) is hopeless for
White, and therefore the exchange sacrifice
is forced .
35 . . 'ikxe3
.
36 "i!Vxc4+
A very simple q uestion : what should Black
play, 36 . . . 'it>h8 or 36 . . . .t:tf7 ?
You suggest 36 . . .'�h8. Why? So as not to
come under a pin? And what do the experts
say? Their opinions are divided . So, 2-1 in
favour of the king move - this means it must
be deemed the strongest. Why? I ndeed , to
avoid coming under a pin (37 'it'd5).
213
36 . . . �h8
37 llb3
You may not bel ieve me, but here White has
g reat d rawing chances. Enormous! U ntil I
reached the time contro l , I didn't real ise this
- I thought I was winning easily. Fortunately,
over the remaining moves to the control
Karpov managed to thoroughly spoil his
position.
Yet the idea of the defence is very simple.
White must play his q ueen to e2 and his rook
to e3 - to hold the th ird rank, to hold the
second, defend the f3-pawn and th reaten a
further attack on the e5-pawn by tt'ld3.
37 . . . "i!Vd2 !
I try t o confuse my opponent - I do not allow
the q ueen to go to e2 .
38 'i!Vc5 �ae8
39 .i::!. d 3
It is possible that the position is drawn even
after the exchange of q ueens: 39 'it'e3 'it'xe3
40 l::t x e3 . The weakness of the e5-pawn is
too serious. But if Black moves his queen,
Wh ite carries out the necessary regrouping ­
�e2 and I:te3. Here it would have been very
useful for my opponent to have a reserve of
time, to work out what he should be aiming
for, which positions were draw, and which
were not (say, after the exchange of q ueens
on f4 Black takes with the pawn , and the
214 � Strategy in Grandmaster Games
endgame is now lost). But Karpov's flag was
hanging, he had no time to th ink, and for the
moment he did not offer the exchange of
queens.
39 . . . 'ii'f4
40 'ii'c 3?!
40 'ii'c2 ! should have been played , followed
by ife2 and l:Ie3. Perhaps my opponent was
instinctively afraid of 40 . . . e4 , but he had no
need to be.
40 . . . l:tc8
41 'ii' b2
Benefiting from the bitter experience of the
previous rou nds, after the time control I
immediately thought for 25 minutes. Of
course, it will not be possible for you to
determine qu ickly that very strong move,
which in the game I sought for so long . But
nevertheless try suggesting some ideas for a
playing for a win. How can the opponent's
position be shaken?
Yes, it is tempting to play . . . h7-h5. If Black is
able to place his pawn on h4, his winning
chances will be sharply improved . Therefore
White is obl iged to reply h3-h4 ! .
41 . . . lic4 ! !
A very strong move . What do you th i n k the
point of it is? In fact it simu ltaneously
contains three ideas (again Razuvaev would
be happy).
Here we have the first explanation : it blocks
the a2-g8 d iagonal, the king obtains the g8-
square, and after the inclusion of . . . h7-h5
and h3-h4 the advance . . . e5-e4 ! becomes
possible. This idea is correct, but what is
more important is that this advance follows if
Wh ite places his q ueen on e2 (as in fact
occu rred in the game). Therefore my move is
prophylaxis against the reg rouping of the
wh ite pieces , about which I have already
spoken. I sat and came to real ise that it was
necessary to defend against it - this was the
main reason why Black's move was made.
I n stead of 41 . . . 1lc4 Kramnik suggested
playing 4 1 . . . l:r.fd8, but then 42 'i!fe2 ! .l::t d4 43
l:te3 . However, now I am ready to carry out
his pla n : 42 . . . .l::td 4 followed by 43 . . . l:!.fd8.
The third idea is very primitive , but it also has
to be reckoned with - 42 . . . l:.fc8 with the
threat of . . . l:tc2 , and if 43 .l:ld2 , then 43 . . . .l:l.c3.
Karpov did not guess my idea, and he
q u ickly placed his q ueen on e2 . After my
reply he sat and thought for 45 min utes , but
he was unable to fi nd any defence.
42 'il'e2 h5!
Of course, my opponent was planning 43 h4,
but now he saw the refutation : 43 . . . hxg4 44
tt'lxg4 (44 fxg4 'ii'f6 ! followed by 45 . . . 'ii'c 6+)
44 ... e4 (the breakthrough has become pos­
sible thanks to the rook on c4) 45 fxe4 l:txe4
46 'ifb2+ �g8 47 'ii'a 2+ l:c4. There is no
time to exploit the pin on the rook, si nce the
wh ite king is threatened with mate . [If 48
tt'le5!? there follows 48 . . . 1We4+ 49 �g 1
'l:.f1 +! 50 <J:;xf1 "ikh 1, winning the queen -
Dvoretsky.]
43 tt'le4 h4
A new and very serious weakness has
appeared in Wh ite's position - the g3-
square. Dvoretsky l i kes to emphasise the
i m portance of a second weakness when
converting an advantage. It was easy to
defend the f3-pawn while simultaneously
attacking e5, but now, to cover the second
 Strategy in Grandmaster Games ttJ 21 5

weakness, the knight will be forced to keep order to pin the opponent's pieces and place
guard on e4 , where it is doing noth ing. If, by h i m in zugzwang.
playing his q ueen to f2 , Wh ite were able to 48 . . . Vc1 !
force the reply . . . g6--g5, from e4 the knight
49 'ii'x h4 llc2+
would at least be attacking a pawn, but this
50 liJf2 g5
hope cannot be real ised .
51 'it'g3 'ike1
44 .l:.e3 l:Ic8
52 h4
45 'ikf2 l:!.c2
To where is it more accu rate to play the king
46 lle2 lixe2
- g6 or h6?
47 ..Wxe2 �g7!
52 . . . �g6?!
In completely winning positions one should
be particularly watchfu l . After the move made
by me Wh ite has acquired a chance tactical
opportun ity: 53 f4! ? with the threat of 54
'it'd3+. Fortu nately, here Black's advantage
is too g reat and he wins by 53 . . . .:1.c3 54 'iih 2
'ii'e 2, but in other circu mstances such an
oversig ht might have cost h i m very dearly.
However, Karpov was again in severe time­
trouble and was forced to make his moves
i nstantly.
53 hxg5 �xg5
54 f4+
Karpov set me many traps in this game, but I There are no other moves.
too set h i m qu ite a few. Black's last move not 54 . . . exf4
only comes into his plans (it rei nforces the 55 'itf3 l:f.d2!
g6- and f6-squares, and prevents the cap­
56 �h3 'ii'f1 +!
ture of the h4-pawn with check), but it also
contains a trap, i nto which my opponent falls. An important check - I saw it earlier when I
But all the same things were bad for h i m - if played 47 . . . �g7 ! . The captu re with the
he had marked time, at some point I would q ueen on f2 would have led to stalemate ,
have jumped in with my queen at c1 , for while the captu re with the rook would have
example: 48 ..Wd3 llc7 49 'ite2 Vc1 . allowed nu merous checks. Also not possible
was 56 . . . 'ike3? 57 ltJe4+ and 58 liJxd2 .
48 ..Wf2
57 'ii'g 2 .l:lxf2
A natural move - intending to captu re on h4
or give a check on a7. What had I prepared Wh ite resigned .
here? The game concl uded , just in time for us to go
That's right, the pawn can be sacrificed , in off for d i n ner.
216 �
Mark Dvoretsky
Whose Strategy wi l l tri umph?
AYusupov
t the first session o f our school Artur
showed two of his games
(against Anatoly Karpov and Jan Timman), in
which over a nu mber of moves a tense
struggle revolved around one pivotal featu re
- some central strateg ic problem . I n this
struggle the outcome of the entire battle was
effectively decided . Here it is very important
that you skilfully maintain the tensio n , not
conced ing an inch to the opponent, and that
in the protracted strategic dispute you
continually find new tactical arguments.
1 should like to invite you to look at two of my
games of the same sort. The fi rst of them
was exceptionally sign ificant for me in the
competitive sense . It was played fou r rounds
before the end of the Moscow Champion­
ship. Grandmaster Anatoly Lein and I had
broken away from all our other rivals and
were lead ing with eight and a half points out
of eleven. Our individual meeting could
decide (and did indeed decide) who won the
title of champion.
Lei n - Dvoretsky
Moscow Championship 1 973
French Defence
1 e4 e6
2 lL'lf3 d5
3 lL'lc3 lDf6
4 e5 lDfd7
5 d4
This position also sometimes arises via the
'normal' move order: 2 d4 d5 3 lbc3 tbf6 4 e5
tt::lfd7 5 tt::lf3 , althoug h 5 f4 is considered to be
more dangerous for Black.
5... c5
6 dxc5 lbc6
7 .if4 i.xc5
8 .id3 f6
8 . . . 0-0? 9 i.xh7+ ! .
9 exf6 "ii'xf6
Opening books recommend 9 . . . tt::Jxf6. But I
did not like playing in accordance with theory
(partly because I did not know it very well)
and 1 used to take the first convenient
opportun ity to diverge from the main theoret­
ical paths.
Objectively the capture on f6 with the knight
is sounder, and the move made by me is
rather risky. But I had already tested it in two
games played not long before the Moscow
Championsh i p , so that I had managed to
gain some experience . Whereas my oppo­
nent was probably encou ntering it for the first
time.
Success in the opening struggle is often
determi ned not by objective factors, but by
you r (and you r opponent's) prepared ness for
the events which develop on the board .
Real ising th is, in this decisive game I
un hesitatingly took a risk in the open ing.
10 i.g3
Another possibil ity is 1 0 i.g5 ! ? . The game
Bareev-Dvoretsky (Minsk 1 972) developed
as fol lows : 1 O . . . "ii'f7 1 1 'ii'd 2?! (the start of an
incorrect plan - 1 1 0-0 0-0 1 2 .ih4 is
stronger) 1 1 . . . 0-0 1 2 0-0-0 tt::l d e5 1 3 tt::l xe5
tt::Jx e5 1 4 f3 .id7 1 5 llhe 1 tt::l x d3+ 1 6 'ifxd3
.l:tac8 1 7 .ie3 .ib4 1 8 .id4 :c4 1 9 <it>b1 .l:tfc8
20 'i¥e3 b5, and Black had the in itiative .
10 . . . 0-0
1 1 0-0
 Whose Strategy will triumph?
Al ready here one can trace the strategic
motifs around wh ich the struggle will revolve
over the cou rse of many moves. By advanc­
ing his e-pawn , Black would like to set up a
powerful pawn centre . But this is still a long
way off - fi rst he must complete his
development, exchange the most dangerous
enemy pieces (the f3-kn ight and the d3-
bishop) and rei nforce his dS-point. For his
part, Wh ite d reams of establishing control
over the d4- and aS-sq uares, blocki ng the
black pawns and making the c8-bishop 'bad ' .
11 . . . lt::l d 4
11 . . . lt::l d eS? did not work because of 1 2
ttJxeS lt::lx eS 1 3 .txh 7 + .
1 2 lt::l x d4 .txd4
1 3 'ii' d 2
Rudolf Kimelfeld (Moscow 1 972) played
more pu rposefully against me: 1 3 'i!fe2 lt::l c 5
14 l:tae1 (nothing is g iven by 1 4 lt::l b S .txb2
1 S l:tab 1 .td7) 14 ... lt::l x d3 1 5 cxd3. Here the
win of a pawn by 1 S . . . .txc3? 1 6 bxc3 'it'xc3
would be su icidal - after 1 7 .teS, relying on
his powerful bishop, White develops an
attack on the g7-point, whereas Black's
'opposite-colour' bishop is completely use­
less.
There followed 1 5 ... .td7 16 .te5 .txe5 1 7
'it'xe5 'it'xe5 1 8 l:txe5.
(see diagram)
ctJ 217
At fi rst sight it seems that White has been
successful : he has seized control of the aS­
square and left h imself with a knight against
a 'bad' French bishop. I ndeed , imagine if he
were to play f2-f4 and switch his knight to d4
- my position would i m mediately become
strategically hopeless. But the opponent
does not have time for th is, and for the
moment the bishop is not i n fact as bad as it
appears. Black has dynamic resources at his
disposa l , i nvolving . . . d5-d4 and play on the
open c-file. To be honest, I do not even see a
safe way for Wh ite to gain equality.
1 8 . . . .l:tac8 1 9 f4
The followi ng variation is typical (although by
no means forced ): 1 9 d4 l:tc4 (the immed iate
1 9 . . . bS! is more accu rate) 20 l:.d 1 bS 2 1 l:ld2
b4 22 lt::l e 2 l:tfc8 23 'it>f1 l:tc2 24 'it>e 1 cJ;;f7
(intend ing 2S . . . .tbS) 2S 'it>d 1 ? (2S lt::lf4 is
better) 2S . . . .ta4 ! 26 b3?! (26 lbc2 .txc2+
27 'it>e 1 is necessary, although after 27 . . . .tfS
the initiative is with Black) 26 . . . .Uxa2! . There's
the bad bishop for you !
Nearly th irty years later the diagram position
was reached i n a bl itz game which I played
against the Cuban g rand master Rivero Be­
serra (Miami 2000). My opponent chose the
cautious 1 9 l:te2 , but he too was u nable to
extingu ish Black's i n itiative: 1 9 . . . bS 20 l:tfe 1
b4 2 1 lt::l d 1 l:tc1 22 f3 (White is hoping to
tra nsfer his knight via f2 or e3 to g4 and then
218 � Whose Strategy will triumph?
to e5) 22 . . . l:tfc8 23 lL'le3? ! �b5 24 lL'lg4
�xd3 25 l:txe6 i.f5 26 l:te8+ l:txe8 27 l:txc1
�xg4 28 fxg4 l:te2 29 l:tc7 a5 30 l:ta7 d4 3 1
h 4 (31 l:txa5 l:txb2 3 2 l:td5 l:txa2 3 3 l:txd4 b3)
31 . . . l:txb2 32 ltxa5 d3 33 l:td5 d2 34 'Wti>f1
l:.xa2 Wh ite resigned .
1 9 ... d4! 20 lL'le2 l:tc2 21 f5?!
The simple 2 1 lL'lxd4 l:.xb2 22 lL'lxe6 was
preferable, when I would have had to go into
a sharp double-rook end ing, since 22 . . . �c6
23 llf2 or 22 . . . .l:r.c8 23 lL'lc5 followed by l:if2 is
hopeless for Black.
21 ... exf5 22 lL'lxd4 .l:.xb2
If now 23 .Ue7, then 23 .. J:tf7 24 lL'lxf5 i.xf5
(24 . . .<it>f8!? 25 l:!.xf7+ <Ji;xf7 also comes into
consideration) 25 l:te8+ l:tf8 26 .U.xf8+ <Ji;xf8
27 l:txf5+ <i;e7 with the better rook endgame
for Black.
23 l:tc1 g6!
And I remained with an appreciable advan­
tage.
Let us return to the game with Lei n .
13 . . . ttJcs
14 .l::ta e1 lL'lxd3?!
1 4 ... �d7 was more accu rate , with an excel­
lent position for Black. The prematu re ex­
change affords the opponent add itional
possibilities.
1 5 cxd3 �d7
From this moment there develops an excep­
tionally tense battle for control of the central
squares.
1 6 lL'le2 �b6
1 7 lL'lxd4 'ii'x d4 1 8 �e5 was th reatened .
1 6 . . . �xb2?! is u nfavourable, since after 1 7
l:tb 1 the rook penetrates onto the 7th rank.
Now Wh ite could play 1 7 d4, but after
1 7 . . . �b5 the chances are roughly equal.
This is not enough for Lein - he wants finally
to win the battle for the centre and he finds a
subtle idea .
1 7 <i;h1 !
A multi-purpose move! It creates a direct
positional threat: 1 8 i.d6 l:tf7 1 9 f4 , then
�e5 and at some point lL'ld4 . 1 7 . . . e5? is not
possible because of 1 8 lL'lc3 , and both
central pawns are under attack.
17 . . . .U.ae8
Black parries the opponent's th reat ( 1 8 i.d6
ltf7 1 9 f4 e5!) and prepares . . . e6-e5.
18 lL'lg1 !
A logical development of the idea begun with
the previous move. Again I am forced to
reject 1 8 . . . e5? in view of 1 9 tL'lf3 i.c7
( 1 9 . . . e4 20 dxe4 dxe4 2 1 'ii'x d7) 20 'ii'c 3!
(weaker is 20 d4 e4 ). At the same time
measures have to be taken against the
seizu re of the centre by 1 9 lL'lf3 or 1 9 �e5
followed by 20 f4 (20 lL'lf3).
18 . . . �b5!
A cou nterattack against the d3-point: 19 i.e5
'i¥f5 or 1 9 lL'lf3 'i¥f5 20 lL'le5 i.d4.
Take note: the two players are engaged in a
strateg ic war, but the means employed are
purely tactical - concrete strong moves,
short variations, th reats, double attacks . . . In
chess, tactics and strategy are inseparably
linked - deficiencies in either of these fields
will i nevitably tell on you r qual ity of play and
resu lts.
19 i.d6 l:tf7
20 f4 i.d4!
 Whose Strategy will triumph?
B lack has defended against 2 1 �e5 (after
which 21 . . . �xe5 22 l:txe5 'ii'x e5 was i ntend­
ed ). Less accu rate was 20 . . . 'ii'f5 21 .l:tf3
followed by �e5 or r!e5.
Now the culmi nati ng point of the fierce battle
for the central squares has been reached . By
attacking the b2-pawn, I try to divert my
opponent from the natu ral knight move to f3 .
Even so, it deserved serious consideration .
True, after 2 1 lDf3 !? �xb2 noth ing is g iven
by 22 .l:tb 1 �c3, and 22 d4? �xf1 22 l:txf1
does not work, as this can be strongly
answered by either 22 . . . �xd4 !? 23 ii'xd4 (23
ltJxd4 e5! 24 tiJb5 a6 25 ttJc7 .l:td8) 23 .. .'iVxd4
24 ltJxd4 e5! (P.Wolff), or 22 . . . �d8!? 23 �b4
a5! 24 �xa5 'i!ka8 . However, the simple 22
�e5! �xe5 23 ttJxe5 llff8 (23 . . . .l:tc7? 24
'ii'a 5) 24 'iib4 or 24 'ii'a 5 would have secu red
Wh ite excellent positional compensation for
the sacrificed pawn .
21 ttJe2
Not the best square for the knig ht, but Wh ite
has a specific idea . Here 2 1 . . . �xb2? 22 lib1
is now bad for B lack, while if 21 . . . ..tb6 there
follows 22 �e5 ii'f5 23 .l:.f3 , and the battle for
the centre concludes not in my favour.
21 . . . l:td8!
An important intermediate move. It would
have been a mistake to attack the bishop
with the other rook: 21 . . . l:Id7? 22 �e5! �xe5
23 fxe5 'ili'xe5? 24 ttJc3. But now 22 �e5 is
tD 21 9
not possible (the rook on f1 is hanging ) , and
22 ltJxd4 'ii'x d4 23 .l:!.xe6 iixd3 24 'ii' x d3
�xd3 leads to a somewhat better ending for
B lack.
It was necessary to reckon with 22 llt'b4 . I n
the event o f 22 . . . �xd3?! 23 ltJxd4 i.xf1
(23 . . . .l:txd6? 24 'il'xd6 �xf1 25 ttJxe6 or 25
.l:.xe6) 24 �e5! �xg2+ 25 �xg2 the advan­
tage is with Wh ite - his pieces are very
strongly placed . However, the cool-headed
22 . . . ..txb2 ! would have changed the pictu re ,
for example: 23 .l:t b 1 �xd3 24 l:Ixb2 :Xd6 , or
23 d4 ..txe2 24 ..te5 �e7 ! 25 'it'xe7 l:txe7 26
l:Ixe2 �a3. There is also 23 .l:tf3 , which was
later analysed by grandmaster Patrick Wolff.
In the event of 23 . . . �c6? Wh ite gains an
advantage by 24 d4! a5 25 'ii'c 5 a4 26 a 3 ! ,
b u t t h e two other possible replies, 23 . . . �a6! ?
and 23 . . . a5!?, are sufficient t o parry the
opponent's aggression .
22 �a3 �b6
I breathe a sigh of relief. The wh ite pieces
have been d riven away from the key e5-
sq uare, which means that I have won the
strategic battle. But, of cou rse, not yet the
game.
Grandmaster Milan Matulovic used to pay
enormous attention to opening theory. It is
said that he kept a special score of the
outcome of the open ing duels i n his own
games, and, to his delight, more often than
not he was successfu l . B ut things were much
worse as regards overall successes in
tournaments. For me it was not enough to
win an intermediate stage - I also wanted to
win the race as a whole. And for this it was
necessary, without relaxing , to carry on
worki ng.
23 ttJc3 �c6
24 'ili'e2
A new problem. The e-pawn is attacked , and
if Black defends it by 24 . . . .l:te8 , then 25 �d6
- the bishop returns to the e5-square and
memories alone are all that remain of my
positional achievements.
220 � Whose Strategy will triumph?
24 . . . �c7!
An important intermed iate move. After 25
'i!t'xe6 'i!t'xe6 26 .l:!.xe6 .ltxf4 or 26 . . . �f4 the
ending favours Black i n view of his two
bishops.
25 i.. c 5 :tea
26 .ltg1
Apparently White is now ready to admit
defeat in the piece battle for the centre and
he intends to play d3--d4, which will prevent
for ever the . . . e6-e5 break. My opponent
had little time left, and I decided that this was
the moment to switch from strategy to
tactics. Try to find and calculate Black's
combination.
26 . . . .ltxf4 ! ?
I fairly quickly saw t h e variation 27 g3 'i!t'xc3 ! !
2 8 bxc3 d4+ 2 9 'i!t'g2 ! (weaker i s 2 9 'i!t'e4
i..d 2 30 l:l.xf7 i.. xe 1 ) 29 . . . i.. x g2+ 30 '>itxg2
i..d 2 31 .Uxf7 i..xe 1 32 .l:!.xb7 dxc3 and I was
relying on the strength of my passed c3-
pawn . I ndeed , White loses after 33 l':!.xa7?
.Uc8 34 i..e 3 �d2. Also bad for him is 33
Wf3?! .Uc8! 34 .lte3 c2 35 .i.c1 �c3 36 .l:.xa7
(36 'it>e2 i.. d4 37 'it>d2 .l:!.f8) 36 . . . .Uf8+ 37 'it>e2
.ltd4 38 .l:.c7 .l:.f2+ 39 'it>e 1 .l:.xh2 or 37 'it>g2
.ltd4 38 Uc7 .l:.f2+ 39 '>ith3 h5! (analysis by
Wolff). However, the cautious 33 .l:.c7!
enables him to gain a draw.
Even so , from the practical point of view I
thi n k that Black's decision was justified . With
time-trouble imminent, my opponent had
insufficient time for an accu rate calculation
of the variations and he simply took me at my
word .
27 .i.xa7?! 'i!t'h4
28 .i.g1 .i.d6
29 .Uxf7 '>itxf7
Black has a clear advantage. He has two
strong bishops, and the . . . e6-e5 advance,
about which I have been d reaming since the
very opening, can no longer be prevented .
30 d4 '>itg8
31 .l:!.f1 h6
The th reat is often stronger than its immedi­
ate execution ! Black does not h u rry with the
breakth rough in the centre, preferring first to
make all the usefu l moves to improve his
position . Such tactics are especially effective
in the opponent's time-trouble.
32 a3 e5!
33 'i!t'f2?
An oversight, but Wh ite's position is already
difficult. My light-square bishop is threaten­
ing to come very strongly i nto play.
33 . . . 'i!t'xf2
34 �xf2 .Uf8 !
35 'lt>g1
When he exchanged the q ueens Lei n was
cou nting on this move . 35 dxe5 .ltxe5 36
'>itg 1 d4 was also hopeless.
35 . . . exd4
Only here did my opponent see that he could
not take on d4 because of 36 . . . i.. x h2+ .
36 tt:Je2
(see diagram)
U nfortunately, my play was typified by a
serious deficiency, which I was q uite unable
to overcome - a tendency towards hasty
decisions, especially when the main prob­
lems in the game had already been solved . I
don't even wish to try and remember how
 Whose Strategy w i l l triumph? lLJ 221

40 .l:!.xd3 i.. d 6
41 h3 .l:!.a8
Here the game was adjourned , and White
sealed his next move. I am a sound pawn to
the good , and in add ition I have the
advantage of the two bishops. It appears that
the win is a matter of straightforward
tech nique. At any event, that is what I
assumed at the adjourn ment. Alas, analysis
did not confirm this evaluation - for a long
time I was unable to fi nd a convincing plan
lead ing to a wi n .
4 2 g4 .l:!.a2
- position after 36 lLle2 -

many important points were lost as a result!

That was also the case here . I realised that
my position was completely won and I
considered two tempting moves: 36 . . . �b5
and 36 . . . d3. I instantly weighed u p the
variation 36 . . . �b5 37 l:.e 1 �xh2+ 38 Wxh2
:Xf2 39 lt::l x d4 and decided that the oppo­
nent would gain good positional compensa­
tion for the pawn (strong knight against
passive bishop) - and I promptly played
. . . d4-d 3 . But this 'calcu lation' was complete­
ly wrong: fi rstly, after 39 . . . l:txb2 I would be
not one, but two pawns u p , and secondly, it is
possible to transfer the bishop via d3 to the 43 lLlf3? .l:.a 1 + 44 'it>g2 �b5 was completely
excellent square e4 . bad for Wh ite , but I had to reckon seriously
with 43 lLlf5 ! ? . Analysing this continuation
36 . . . d3?
d u ring the adjourn ment, I made a serious
37 lt::l d 4 mistake , which could have cost me dea rly.
Here I discovered to my su rprise that it was I was intending to go i nto a rook end ing,
not at all easy to breach my opponent's relying on a variation which , as it seemed to
defences. He wants to play h2-h3, then l:td 1 , me, led to a forced wi n . Because of this I did
and what can I do? If 37 . . . �c5 there follows not bother to analyse seriously the bishop
38 lLlf3. ending arising after 43 . . . i.h2+ ! ? 44 Wg2 !
37 . . . �a4 ! ? �xf5 45 gxf5 i.f4 4 6 'it>f3 (46 l:txd5?? �e3)
38 b3 �d7 46 . . . l:ld2 47 l:txd2 (bad is 47 l:lc3 �e5)
The bishop wants to go to g4 a n d , with his 4 7 . . . i.xd2 48 i.d4 'it>f7 49 'it>e2 �g5 50 'it>d3
flag about to fal l , this th reat seemed so .tf6 51 �f2 �e5 52 .t h4 g6 53 fxg6+ 'lt>xg6
dangerous to Lei n that he decided to part 54 �f2 h 5 .
with his a-pawn .
39 l:td1 �xa3 (see diagram)
222 \t> Whose Strategy will triumph?
And indeed , the evaluation of this position is
not obvious. The opponent's dark-sq uare
bishop and his pawns arranged on light
squares create a barrier in front of my king
which is difficult to cross.
Here is the variation which I was i ntending to
go in for:
43 . . . �xf5 44 gxf5 (44 l1xd5? l:ta 1 +) 44 . . . l:ta5
(with the positional th reat of 45 . . . ..te5) 45
..td4 ..tc5! (now, before the wh ite king has
reached e3) 46 �2 �xd4+ 47 .l:lxd4 .l;tb5 48
b4 'iti>f7 49 'iti>e3 .
49 . . . 'iti>f6 50 'iti>f4 h5 5 1 h4 b6. Wh ite is in zug­
zwang and he has to g ive u p a second pawn .
Studying a book of mine in which this
variation was reproduced , grandmaster l lya
Smirin noticed that after 52 ..ti>f3 <bxf5 53
'iti>e3 it is not possible for Black to convert his
advantage because of the tragic-comic posi­
tion of his rook, shut in at b5. Black is also
unsuccessful with 5 1 . . . l:.b6 (instead of
5 1 . . . b6) 52 .Uxd5 l:.xb4+ 53 'iti>g3 b6 54 'iti>h3
l:tb3+ 55 'iti>g2.
The third session of the Dvoretsky-Yusu pov
School ( 1 99 1 ) was devoted to the improve­
ment of endgame mastery. As one of the
homework tasks for those taking part, it was
suggested that they should check my analy­
sis of 43 lt:'!f5.
I assu med that after the mistake had been
fou n d , in search of a win the students would
have to analyse the bishop ending. However,
Vad i m Zviagintsev and Maxim Boguslavsky
found a simpler solution to the problem -
they improved Black's play in the rook
end ing.
I n stead of 49 ... ..ti>f6? they suggested 49 . . .
'iti>e7 ! . Then 5 0 ..ti>f4 'iti>f6 5 1 h 4 h5 1eads to the
familiar zugzwang position , but with the
pawn on b 7. After 52 'iti>f3 <bxf5 53 <be3 'iti>e5
the rook comes i nto play via b6. And if 50
�3 . then 50 . . . 'iti>d6 5 1 'iti>f4 ( 5 1 .l::t g 4 ..ti>e5 52
l:.xg7 .l:txb4 is also hopeless) 5 1 . . . .l::t b 6,
intending . . . l:!.b6-c6-c4 .
As you see, although objectively it should
have led to defeat, 43 lt:'!f5 ! ? would have set
Black serious problems. But it turns out to be
no easier to fi nd a win after the qu iet
conti nuation chosen by my opponent.
43 'iti>g2 ! ?
Over t h e next few moves Lein h a s a clear
plan of actio n . He wants to bring his king to f3
and then , after playing ..tg3 or ..te3-f4, offer
the advantageous exchange of da rk-sq uare
bishops, and if Black avoids the exchange -
place his bishop on e5. Then all the wh ite
pieces will be ideally placed , the d5-pawn will
remain secu rely blockaded , and i n addition
all the time Black will have to reckon with the
th reat of lt:'!f5 .
And how can I strengthen my position? Of
course, if I were able to transfer my bishop to
 Whose Strategy will triumph? 223

e4, the game would be decided . But can this

be achieved , when tt::l f5 is a constant
problem for Black?
I n itially I was pinning my hopes on the
variation 43 . . . h5 44 gxh5 �e8 (intending
45 ... �xh5, then . . . �g6 and . . . �e4) 45 ltJe6
(45 tt::l f5 �c5) 45 . . . �f7 ! 46 ltxd5 �e7 ! . But I
did not find anyth ing convincing after 45
'it>g 1 ! �xh5 46 tt::lf5 .
I also examined 43 . . . �c5 44 'it>f3 �e8
(44 . . . �b5? 45 tt::lx b5 .l:i.xf2+ 46 �g3 ) 45 �e3
�g6, but the rook ending arising after 46 tt::l f5
�xf5 4 7 gxf5 �xe3 48 'it>xe3 is most
probably drawn . - position after 45 . . . h4! -

Only after studying these and many other

variations did I finally discover the correct variation 46 tt::lf5 1Ih2! 47 �xc5 Itxh3+ 48
plan . '.te2 �b5 49 �e3 (or 49 tt::l e 3) 49 . . . l::. h 2+.
43 . . . �c5 46 .l:tc3 �b6
44 'itf3 I n the event of 46 . . . b6?! Wh ite would have
In the event of 44 'it>g3 there can fol low gai ned cou nterplay by 47 b4! �xb4 48 l::.c7
�a4 49 tt::lf5 (but not 49 tt::l e 6? �d 1 + 50 �f4
44 . . . g5! (44 . . . �e8 is weaker: 45 �e3 �g6
�d6+ ).
46 tt::lf5 �xf5 47 gxf5 �xe3 48 l:txe3 �f7 49
�f4) 45 �e3 �d6+ 46 �f3 .l:i.h2, or 45 '.tf3 47 tt::l f5? !
h5!. Lein nevertheless overlooks his opponent's
44 . . . h5! ! tactical idea . He could have put up a tougher
Here my opponent thought for a long time. It defence with 47 tt::l e 2 (after which I was
became clear that he was not prepared for intending 47 . . . �d8) or 47 l:tc2 .
th is turn of events. 47 . . . 1Ih2!
45 �e3 48 tt::l e 7+
45 gxh5 �xh3 (or 45 . . . �e8) was clearly bad 48 �xb6? l::.x h3+ .
for White. If 45 �g3 I was planning the 48 . . .
waiting move 45 . . . 'it>h7, and if 46 �f4 or 46 49 tt::l x d5 Itxh3+
�e5, then , as in the game, 46 . . . h4! . The game is decided ! The position has
45 . . . h4! opened u p and the two bishops can at last
demonstrate their true strength .
(see diagram)
50 'iitf4 �dB!
Black has fixed the h3-pawn and it has
51 .l:tc1 �c6
become a real weakness: 46 . . . l::. h 2 is threat­
52 �b6
ened (the king has been deprived of the g3-
square). If the bishop moves from e3, Wh ite 52 .l:!.d 1 �e6 ! .
has to reckon with . . . �b5!, since the capture 52 . . . ltf3+
of the bishop is not possible in view of mate Of cou rse, it was also possible to play
with the rook on f2 . 52 . . . �xb6 53 tt::l x b6 l:txb3 54 tt::l c4 'it>f6! 55
The tactical basis of Black's plan is the g5+ 'it>g6 56 tt::l e 5+ 'it>h5.
224 � Whose Strategy will triumph?
53 'it>e5 ii.g5
54 I:lc3?
And Wh ite resig ned .
What was the strategic basis of Black's
winning plan in the endgame, and which
positional considerations could have hel ped
it to be discovered at the board?
When defending, your opponent endeavours
to protect secu rely all his weaknesses. I n the
broad sense of the word a weakness i n his
position may be not only a vulnerable point
or a badly-placed piece, but also, for
example, an invasion sq uare which must be
covered , or an enemy passed pawn , which
needs to be blockaded .
With skilfu l defence it is usually not too
difficult to hold one weakness. I n such cases
the correct strategy for the stronger side is
always to seek or create a second weakness
in the opponent's position . By attacking this
second weakness, and then if necessary
again switching our attack to the fi rst, we
stretch and finally break the opponent's
defences.
Look at how the great masters of the
endgame convert an advantage. You will see
that nearly always at some point they open a
'second front' .
I n the above ending White initially only had to
solve the problem of the passed d5-pawn .
His pieces were excellently placed for
combati ng it and , of course, Black could not
hope for success with the passed pawn
alone. By the advance of his h-pawn to h4 he
was able to fix a second weakness in the
opponent's position - the h3-pawn . Wh ite's
position immediately became critical .
The followi ng game, played a t the start o f a
tournament in the small Estonian town of
Vilyandi, is also a memorable one for me.
First of all, it was my first win against a
grandmaster. After beati ng another g rand­
master two rounds later (also with Black!), I
was so inspired by my success that I began
winning game after game and with a score of
1 1 points out of 1 3 I took first place, ahead of
the g reat Mikhail Tal . It was probably the best
tou rnament i n my l ife, i n both the competitive
and the creative sense.
It is usefu l to th ink about the reasons for both
your own failures, and your successes, in
order to pick out the factors which influence
your results. I had only just fin ished U n iversi­
ty, I had received my diploma, put it away in a
drawer and switched to chess. Before the
event I held a training session with the
Moscow youth team , I played football with
the lads and perhaps we even stud ied a bit of
chess. Mentally and physically I was in
excel lent shape. And although from the
opening I often obtai ned d u bious positions
(the present game is a good example), this
did not h i nder me. If you have sufficient
energy for the su bseq uent play, you can
often manage to repair the conseq uences of
unsuccessful opening strategy.
Shamkovich - Dvoretsky
Vilya ndi 1 972
Nimzo-lndian Defence
1 d4 tLlf6
2 c4 e6
3 tLlc3 ii.b4
4 e3 b6
5 tLlge2 tLle4?!
· This move does not enjoy a good reputation .
Correct play leads to a position which is
somewhat inferior a n d , more important,
passive for Black.
6 f3
A dynamic reply. Leonid Shamkovich allows
the doubling of his pawns i n order to q u ickly
create a strong pawn centre .
6 'ii'c2 is more often played . I n the game
Gul ko-Dvoretsky (Moscow Championship
1 972) after 6 . . . ii.b7 7 a3 ..ixc3+ 8 tLlxc3
tLlxc3 9 'ifxc3 0-0 1 0 b3 d6 1 1 i.b2 tLld7
Wh ite gained some positional advantage (he
 Whose Strategy will triumph?
has the two bishops and more space).
However, Black has no real weaknesses and
in the end I gained a d raw, although it will be
realised that there is little pleasure in
defending such a position .
Events developed in more lively fashion in
the game Vaisser-Dvoretsky (Kiev 1 970):
6 .. .f5 7 a3 i.xc3+ 8 tt::l x c3 tt::lx c3 9 'iVxc3 i.b7
1 0 d5! 0-0 ( 1 0 . . .'ife7 is better, intending after
1 1 dxe6 dxe6 to play . . . tt::ld 7 and castle on
the queenside) 1 1 b4! (an excel lent position­
al pawn sacrifice) 1 1 . . . exd5 1 2 i.b2 .l:tf7 .
Let's try to decide how Wh ite should develop
his in itiative.
There followed 1 3 .ll d 1 ? ! dxc4 1 4 ..txc4 d5
15 0-0 c6 16 b5 cxb5 1 7 i.xb5 tt::l d 7 18 f3
l:i.c8 1 9 'it'd4 'iff6 with roughly equal chanc­
es. 1 3 0-0-0! was much stronger: 1 3 . . . dxc4
(otherwise 1 4 c�d5 with an obvious advan­
tage) 14 i.xc4 d5 1 5 b5! , and the threat of 1 6
l:.xd5! i.xd5 1 7 .l:td 1 is extremely u npleas­
ant.
6. . . tt::l x c3
The game Meulders-Winants (Belgian Cham­
pionship 1 983) went 6 . . . ..txc3+ 7 bxc3 tt::ld 6
8 tt::lg 3 i.a6 9 'i!i'a4 'ii' h 4 1 0 i.d3 tt::l f5 11 i.xf5
exf5 1 2 0-0 'ii'f6 1 3 e4 1kc6 1 4 'ika3 ! 'ifxc4 1 5
.l:te 1 , and the black king was in g reat danger.
7 bxc3 ..te7
7 . . . i.d6! ? 8 e4 i.a6 was interestin g , aiming
ttJ 225
to l u re the pawn to e5 in order later to attack
it by . . . d7-d6 or . . . f7-f6 .
8 tt::l g 3 tt::l c 6
9 i.d3 ..ta6
10 e4 tt::l a 5
11 'ife2
The same pawn structu re has been reached
as in the Samisch Variation of the N imzo­
l ndian Defence (4 a3). There Black is left
with his knight on f6 , whereas here he has
his dark-sq uare bishop, which , generally
speaki ng, is a slig htly stronger piece. But on
the other hand I have lost several tempi in
the opening ( . . . tt::lf6-e4xc3, . . . i.b4-e7), and
in addition the opponent has managed
without the move a2 a3, and so Black has not
succeeded in creating cou nterplay against
the c4-pawn .
11 . . . 0-0?!
The king will not feel too comfortable on the
kingside. In the game I . Sokolov-Johansen
(Olympiad , Manila 1 992) Black retained an
acceptable position after 1 1 . . . d6 12 0-0 'ii'd 7
1 3 I:t b 1 ( 1 3 a4! ? ) 1 3 . . . h5. A recommendation
by Alexey S h i rov also deserved considera­
tion : 1 1 . . . c6 ! ? (Black prepa res 1 2 . . . d5) 1 2 e5
d5 1 3 exd6 'i¥xd6 14 tt::l e4 'ii'd 7.
12 0-0 d5?!
I n the event of 12 . . . c5 1 3 d 5 e5 14 f4 (or 1 4
tt::lf5) Black's position i s clearly worse. 1
should probably have tried 1 2 . . . g6!? 1 3 i.h6
l:i.e8 , subseq uently choosing between . . . d7-
d5 and . . . f7-f5 .
1 3 cxd5 i.xd3
14 'ii'x d3 exd5
(see diagram)
Wh ite's plan is obvious: e4-e5 and f3-f4-f5,
creating a powerfu l attack on the kingside.
1 5 tt::l f5 also looks q u ite good . What do you
th ink, which move is stronger? To answer
this q uestion you must also take Black's
cou nter-actions into account.
226 � Whose Strategy will triumph?
15 eS?I 'ii'd 7
1 6 lLlf5
Practically forced - since otherwise I would
have played 16 .. .f5 ! , halting the opponent's
attack. Now it is clear that he should have
begun with 1 5 lLlf5 ! , since this would have
created the add itional th reat of 1 6 lLlxe7+
'ii'x e7 1 7 exd5. I would have had to make
some not very useful move such as 1 5 . . . c6 ,
and then the e4-e5 advance would have
gained in strength .
16 . . . g6!?
Wh ite wants to play f3-f4 , then lLlxe7+ and
f4-f5, and to crush me with these pawns.
Therefore I decided to weaken my kingside,
in order to drive the knight from f5 and meet
the opponent's attack with . . . f7-f5 ! .
We see that, a s in the previous game, a
battle of plans has developed : Wh ite d reams
of obtaining a powerful pawn pair at e5 and
f5, wh ile Black tries to disrupt this plan by
advancing his f-pawn , after which he will be
able to breathe more easily. But whereas i n
the game with Lein I was an equal partner,
here my opponent has far more chances of
success. However, Wh ite also has his
problems - on every move he has several
tempting possibilities, and it is not at all easy
to make the correct choice.
1 7 lLlh6+
After 1 7 lLlxe 7 + 'ii'xe 7 1 8 .i.h6 .:tea noth ing is
g iven by either 1 9 f4 f5! , or 1 9 g4 (taking
control of the f5-square ) 1 9 . . . 11Vh4 ! . And in
genera l , when the opponent's position is
cramped , it is natu ral to keep as many pieces
as possible on the board . Bad now is
1 7 . . . �g7? 1 8 lLlg4 �h8 1 9 .i.h6.
17 . . . �hB
I n the event of 1 8 f4 f5 1 9 exf6 .i.xf6 20 f5 I
still stand worse , but rather freer than before.
It is desirable to prevent . . . f7-f5, but if 1 8 g4
there follows 1 8 . . .f6 .
After t h e g a m e Mikhail Tal suggested 1 8
.i.f4 ! ? . A q uestion for you : how should the
reply 1 8 . . .f5 be evaluated?
After any other move by me ( 1 8 . . . lLlc4, for
example) there follows 1 9 .l:.ae 1 , and then
the bishop retreats, renewing the th reat of
f3-f4-f5 . The knight on c4 is prettily placed ,
but in fact it is not much use - here it does not
have the slightest influence on the defence
of the kingside. I n the positional sense 1 8 . . .f5
is, of cou rse, the correct reactio n , but in reply
Wh ite has the possibil ity of a clever combina­
tion .
1 9 e 6 'ii'x e6 20 :ae 1 'ii'd 7 2 1 l:txe 7! 'iixe 7 22
.i.e5+ l:tf6 23 lLlg4! (23 g4? lLlc4 24 g5
lLlxe5) 23 . . . fxg4 24 fxg4 - White regains his
rook and would appear to gain the advan­
tage.
If you r opponent has devised something,
 Whose Strategy will triumph?
from the practical point of view it sometimes
makes sense to deviate , and not go along
with his ideas. But here the . . . f7-f5 advance
is too important for Black for h i m to reject it at
the fi rst sign of d ifficulties. Let's continue
checking the combinatio n . There are two
possible repl ies: 24 . . . ltf8 and 24 . . .<�g8.
After 24 . . . .l:tf8 25 .l:txf6 (25 g5? 'iix e5! )
2 5 . . . .:txf6 2 6 g5 �g8 2 7 gxf6 Wh ite stands
better, but not so much better that because of
this Black should reject the strategically
necessary move 1 8 . . . f5.
Besides, there is also a second possibil ity:
24 .. .<it>g8!?. After 25 llxf6 the reply 25 . . . lt'Jc4?
is incorrect i n view of 26 l:txg6+! hxg6 27
'it'xg6+ �8 28 'it'h6+ �f7 (28 . . . �e8 29
'iic6+ ) 29 'it'h5+! �8 30 'iVh8+ �f7 3 1 1i'xa8
lt'Jxe5?! 32 'iixd5+. Well , before placing the
knight on c4 , let's move the rook to the safe
square d8 (or e8), and the position becomes
unclear.
Many years later the young Polish player
Kamil Miton found an important strengthen­
ing of the attack. He suggested including the
move 1 9 g4 ! ! . In the event of 1 9 . . . 'ii'e 6 20
gxf5 gxf5 21 �h 1 followed by 22 lig 1 Black's
position is difficult, since his pieces are tied
to the defence of the weak f5-pawn . And after
1 9 . . . fxg4 the idea of 20 e6! 'iix e6 21 .l:f.ae 1
gains in strength . 2 1 . . . 'iid 7 no longer works
in view of 22 l:txe7! 'iix e7 23 .i.e5+ :f6 24
fxg4, and Wh ite also gains a g reat advan­
tage after 21 . . . 'iic6 22 .l:txe7 .l:txf4 23 'i¥e3.
I magine to you rself that, deliberating over
the move 1 8 . . .f5, you i n itially ascertained
that the combination beg i n n ing with 1 9 e6
was harmless, but you then discovered the
idea of 1 9 g4! ! . Should you still decide on the
advance of the f-pawn? It is hardly possible
to g ive a defi nite reply, but I would probably
have taken the risk. Black's position is
anxious, in most variations the move . . . f7-f5
solves his defensive problems, and the
probably of the opponent finding a latent
refutation is not too g reat.
l2J 227
1 8 l:te1
Now 1 8 .. .f5? is not possible because of 1 9
e6 and 20 lt'Jf7+. I n add ition White has
created the th reat of 1 9 lt'Jxf7+ llxf7 20 e6.
And yet this last move causes some doubts.
The rook moves off the f-file, where it could
have come i n usefu l .
18 . . . l:tae8 !
Tactics at the service of strategy! I prepare
. . . .lidS followed by . . . f7-f5 , and in the event
of 1 9 lt'Jxf7+ l:txf7 20 e6 I have the saving
resou rce 20 . . . .i.h4! 21 .i.g5! l:txe6 .
1 9 �f4 �h4! ?
20 g3 �d8
21 lt'Jg4 h5!
I am obl iged to fu rther weaken my kingside ­
otherwise the opponent will place his bishop
on h6, depriving me of any hope of cou nter­
play.
22 lt'Jf6?!
Shamkovich overesti mates his attacking
chances. After the exchange of mi nor pieces
he hopes to establish his bishop on e5 and
then , after opening li nes on the kingside , get
at my king . However, he is not able to put this
plan into effect - Black has adequate
defensive resou rces. At any event, psycho­
logically I felt far easier. After a l l , now Black
has only one concern - to su rvive , avoid
being mated . In any q u iet endgame his
228 � Whose Strategy will triumph?

knight will be stronger than the opponent's 33 'Ot>xh3 'it>h7, transposing i nto the equal
bishop. position which occurred i n the variation with
Wh ite should have simply played 22 lt:Jf2 ! 28 f4 .
(but not 22 lt:Je3? g5), and then prepared g3- 30 . . . 'ifxg4+!
g4. 3 1 'it'xg4 lt:Je3+
22 . . . �xf6 32 '>t>f3
23 exf6 lt:Jc4 32 ;t>g3 lt:Jxg4 33 f5 �xe5 (33 . . . .l::i.xf6!?) 34
At last the time has come to place the knight dxe5 �xe5 35 fxg6 fxg6 was hopeless for
on its lawful square. White.
24 �e5 c6 32 . . . lt:Jxg4
The queen must be freed from the need to 33 .l::i.a g1
guard the c7-pawn . What would you now play as Black?
25 '>t>g2 'iif5!
26 �e2
26 'iixf5 gxf5 27 'it>h3 'lt>h7 28 'it>h4 Wg6 29
.l::i.e2 .l::i. e6 30 .l::i. a e1 would have led to an
unclear ending.
26 . . . .l::i. e 6
I have already begun to take an interest in
the f6-pawn : I am i ntending 27 . . . l::i.fe8 28 f4
lt:Jxe5 and 29 . . . 'ir'xf6 (or 28 . . . Wg8 with the
th reat of 29 . . . 'iixf6). Therefore Wh ite hurries
to open lines.
27 g4 l!fg5
28 h4
The start of a forci ng variation, at the end of The exchange of minor pieces by 33 . . .
which my opponent overlooked a small lt:Jxe5+?! 3 4 dxe5 looks prematu re; also
tactical subtlety. The balance would have dubious is 33 . . . lt:Jxf6?! 34 f5! .l:!.xe5 35 dxe5
been maintai ned after 28 f4 'iVxg4+ 29 �xg4 .l:!.xe5 36 fxg6 . But after 33 . . . lt:Jh6! 34 .Ug5
hxg4 30 h3 (30 'lt>g3!?) 30 . . . gxh3+ 3 1 'it>xh3 ;t>h? Wh ite's attack would have come to an
'lt>h7! 32 l::i. h 1 M.h8 33 'lt>g3+ 'lt>g8 34 .l::i. x h8+ end and Black should have calmly begun the
'it>xh8 35 f5! gxf5 36 'it>f4 . conversion of his two extra pawns ( . . . c6-c5
28 . . . 'i!Vxh4 etc . ) . U nfortunately, here the same story
29 f4! occurred as in my game against Lei n . Having
29 .l::i. h 1 was poi ntless in view of 29 . . . 'iig 5. gained a winning position after g reat emo­
.l::i.fe8 tional stress, I relaxed and quickly made a
29 . . .
superficial move with my king , underestimat­
30 .Uh 1 ?
ing the exchange sacrifice which my oppo­
Shamkovich is still i n the grip of a n illusion . nent had prepared .
Of course, 30 gxh5? gxh5 or 30 f5? .l::i.x e5 3 1
33 . . . ;t>g8?
dxe5 .l::i.x e5 was unfavourable for h i m , b u t it
was better to play 30 'iff3 hxg4 31 'ifg3 34 .l:!.xg4! hxg4+
'iih 3+! 32 'ifxh3+ (32 Wf2 lt:Jd2 33 'ifxh3+ 35 'it>xg4
gxh3 34 .l::i. h 1 lt:Je4+ and 35 . . . lt:Jxf6) 32 ... gxh3+ It is staggering but, deep in the endgame and
 Whose Strategy will tri u m p h ?
the exchange and a pawn down , Wh ite still
retains an attack. He wants to prepare f4-f5
and pursue my king with his rook. Of course,
here 36 f5 is not yet a th reat in view of
36 . . . .l:i.xe5 (it is Black's dream to return the
exchange at an appropriate moment), but
soon such a threat will become real . It is not
possible for Black to strengthen his kingside
defences - all that remains is a cou nter­
attack on the queenside.
35 . . . c5
36 �g5
If now 36 . . . cxd4 37 cxd4 .l:i.c6 , then 38 f5 !
(since Black no longer has a double captu re
on e5) 38 . . . .l:i.c2 39 .l:i.g 1 ! with an u nclear
position. I found another idea - I tried to
undermine Wh ite's pawn chai n , to weaken
the defence of the bishop on e5.
36 . . . b5!
37 a3 a5
38 .l:i.b1 ?
The decisive mistake , after which Black's
idea triumphs. Meanwhile, after 38 �h6 or
38 .l:i.e 1 followed by f4-f5 the win for me
would still have been very much i n doubt.
38 . . . b4
39 axb4 axb4
40 cxb4 cxd4
41 i.. x d4
Variations such as 41 b5 d3 42 b6 d2 43 b7
Qj 229
.l:i.b6! are easy and pleasant to calculate.
41 . . . .l:i.e1 !
42 .l:i.b3 .l:i.e4
Here the game was adjourned . Wh ite sealed
the move 43 i.. c 5 and then resig ned without
resuming . I was intending 43 . . . .l:i.f1 ! (43 . . . d4? !
is weaker in view of 44 .l:i.a3 ! ) 44 i.. d 6 .l:i.g 1 +
45 'it>h4 (45 'it>h6 .l:i.e2 or 45 . . . .l:i.ee 1 ) 45 . . . g5+.
We have analysed two tense games. I n both
of them the outcome depended on the
correct perception by the two players of the
strategic problems facing them, but at the
same time on their tactical resou rcefu l ness,
without which it is not possible to successful­
ly put your plans into effect, and later also on
their endgame tech nique. Good g rounds for
once more thinking a bout the need for every
player to rid himself of his playing weakness­
es and to ach ieve harmony in his play, by
developing and perfecting all aspects of his
mastery.
What other conclusions should be drawn
from these games? Remember: in each of
them there came a moment when it should
have become clear to Wh ite that he would be
unable to ach ieve complete strategic suc­
cess - the opponent had sufficient cou nter­
chances. I n such situations it is important to
display a sober and flexible approach , to be
able to cal l a halt a n d , giving up your
u n realisable aims, seek a way to achieve an
acceptable, relatively safe position. Neither
of my opponents coped with this problem .
Finally, as we have see n , the loss of a
strategic battle by no means sig n ifies the
loss of the game. Nearly always there remain
sufficient opportu n ities to make l ife difficult
for the opponent, by continually setting him
new problems.
And, on the contrary, after outplayi ng your
opponent you must not relax, but continue
playing at full i ntensity. Otherwise you risk
missing a deserved win , as almost occu rred
with me in both of these games.
230 �

PART V

Artur Yusu pov

F rom Games by P u p i ls of the S c h ool

Tprevious books will know that the au­

hose readers who are familiar with our the pawns on the kingside, where she has a
'qualitative majority' (compare this example
thors consider one of the main ways of with the Yusupov-Lautier game on p. 1 90).
improving at chess to be the analysis of you r After 1 7 . . . f6 1 8 lLlf1 �f7 1 9 lt::l g 3 it.d7 20
own games. Before each session o f our 'ili'c2 g6 Black continues with . . . l:lh8, . . J:tag8,
school, the pupils did some 'homework' by . . . h7-h5 and . . . g6-g 5 . Of cou rse , such a
annotating several of their games. The most pla n , which somewhat weakens the castled
interesti ng of these were then discussed position , should be carried out carefully, but
during the lessons. The present chapter in this case Wh ite is unable to create
contains some episodes where instructive counterplay on the open e-file: note how well
positional problems occu rred . The ages of Black's pieces cover all the invasion squares.
the pupils are given in brackets . The piece attack carried out by Kadymova
led merely to simplificatio n .
C hoice of plan 17 . . . lt::l g 6?
Kovalevskaya - Kadymova ( 1 4) 1 8 lt::l f1 lt::l f4
USSR 1 990 1 9 lt::l e 5! lt::l x e2+
20 l:.xe2 l:!.e8
21 'i¥f4 lt::l x e5
22 .:txe5 .l:.xe5
23 dxe5 'i¥e6

l lakha Kadymova has successfully solved

her opening problems and obtained the freer
position. But what to do next? Black can use
her advantage in space to g radually advance
 From Games by Pupils of the School ct:J 231

The situation has changed sharply. Black I n this example too , both players made
has been left with a 'bad' bishop. Wh ite mistakes. Here Black should have continued
should activate her knight and bring it to d4. his development with 1 4 . . ."it'a5 followed by
I n the game she chose the i ncorrect route. . . Jk8 and . . . ..tc5 or . . . ..tb4. I n stead of this
24 tL:lg3? Ji.d3 Baklan offered the exchange of q ueens,
which spoiled his pawn structu re.
The knight did not in fact reach its goal.
24 tL:ld2 ! suggested itself. If 24 . . . �e8 25 �e 1 14 . . . 'ir"b6?
f6 26 tL:lf3 ..te4 , then 27 exf6 'it'xf6 28 'i!kxf6 1 5 'ir"xb6 axb6
gxf6 29 tL:ld4 'i.t'f7 (the more tenacious 1 6 ..te2
29 . . . .l:!.e5 is also u npromising : 30 f3 ..tg6 3 1 The simple 1 6 tL:lb5 would have led to an
'.t>f2) 3 0 f3 Ji. d 3 3 1 .l:i.xe8 'i.t'xe8 3 2 tL:lxb5! and advantage for Wh ite.
wins. As Dvoretsky pointed out, instead of 16 . . . 0-0
26 . . . ..te4 Black has the stronger 26 . . . ..tg4 !
1 7 ..tf3?
27 tL:ld4 fxe5! 28 'it'g5 exd4! (after 28 . . . 'iVg6
29 'i!kxg6 hxg6 30 tL:lc6 e4 31 f4 followed by Wh ite conti nues to play inaccu rately and
'i.t'f2-e3 Wh ite retains the better chances, without a pla n . He clearly does not under­
despite the pawn deficit, in view of the stand the essence of the position and simply
control over the weakened dark squares) 29 makes 'solid ' moves. The bishop is doing
.l:!.xe6 ..txe6 30 cxd4 c3 , and the passed noth ing on f3 , whereas the thematic . . . b6-b5
pawn has to be blocked by the quee n , which advance should have been. prevented .
means that the position favou rs Black. 17 . . . �a5?!
Therefore it makes sense for Wh ite to play Of course, the i m med iate 1 7 . . . b5! was
not 25 lle 1 , but 25 tL:lf3 ! ? , and after 25 . . . ..tg4 stronger.
26 tL:ld4 'i!Vxe5 27 'i!Vxe5 �xe5 28 f3 Ji.d7 29 1 8 a3 b5
�f2 the excellent position of the knight on d4
19 tL:la2 .Ua4
fu lly compensates for the sacrificed pawn,
although not more. 20 g3 f6?!
As was shown by Dvoretsky, Black should
have played 20 . . . b4 ! , since if 21 axb4??
Arrangement of the pieces
there is 21 . . . �fa8.
Morozov - Baklan ( 1 2) 21 f5? !
Kiev 1 990 T h e preparatory 2 1 tL:lc3! .l::!. a 5 was more
accu rate , and only now 22 f5 .
21 . . . d4! ?
22 exf6 ..txf6
23 fxe6 ..txe6
24 ..txb7 ? !
A typical mistake : Wh ite is tempted by
material g a i n , underestimating the oppo­
nent's cou nterplay. 24 lbb4! was sounder,
with the advantage.
24 . . . �b8
25 ..tf3 b4! ?
26 tL:lxb4 .Ubxb4
232 w From Games by Pupils of the School

27 axb4 .ta2+ Arba kov - Boguslavsky (1 6)

28 �c1 i.g5+ Moscow 1 99 1
29 .:!.d2

A rook move to d B suggests itself. But

29 . . . .te6? sometimes it is very d ifficult to decide which
The bishop should have moved to another rook to place on a n open file. In this case too
square: after 29 . . . .tc4 30 b3 l:ta 1 + 31 �b2 Black did not guess right.
lixh 1 Black would have retained excellent 16 . . . .:!.bd8?!
drawing chances. But, as sometimes hap­
16 ... llfd8! was stronger, and if Wh ite plays
pens, a mistake which should have led to a
1 7 'ii'e 3 as in the game, then after 1 7 . . . 'ii'xe3
loss, in fact leads to a win for Black!
1 8 fxe3 li:Jf6 1 9 I:tad 1 e5 (or 1 9 . . . 'it>f8) 20
30 c3?? dxc5 bxc5 21 h3 Black has the important
Wh ite would have won by 30 �d 1 ! .l:r.a 1 + move 2 1 . . . �f8 , consolidating his position.
(30 . . . .tc4 3 1 <be 1 .:!.a 1 + 32 .:!.d 1 ) 3 1 c;t>e2 17 'ife3 ! 'i!fxe3?!
i.c4+ 32 l:.d3.
1 7 ... 'ii'ffi or even 1 7 . . . '�g4 ! ? was preferable.
30 . . . d3
1 8 fxe3 lt:Jf6
Or 30 ... .tb3 3 1 �b 1 .txd2.
1 9 l:r.ad 1 e5
31 b3 l:ta2
1 9 . . . cxd4 20 exd4 l:.d7 came i nto considera­
32 l:td1 i.xb3 tio n .
White resigned . 20 dxc5! bxc5
21 h3 llfe8
22 i.c6! l:txd 1
23 l:.xd 1 l:.c8
24 .l::. d 6!
Wh ite stands sign ifi ca ntly better in view of
the threat of retreating his bishop followed by
.:!.a6.
 From Games by Pupils of the School ttJ 233

Exchanging 3 1 l:txd6+ ltJxd6

32 l::t d 5 .:l.b8
Exchanging is undoubted ly one of the most
complicated elements of positional play. It 33 il.g3
was no accident that an experienced Soviet
trainer offered roughly the following advice:
'If you are playing a weaker opponent,
exchange a few pieces. It is probable that he
will not understand which pieces he should
exchange, and which he should keep on the
board . '

Baklan (1 3) - S h iyanovsky
Kiev 1 99 1

33 . . . l:tb5?!
A tech nical inaccu racy. Before changing the
structu re , it is usefu l to improve your position
as much as possible, or worsen the oppo­
nent's position. As shown by Dvoretsky,
Black should first have g iven a series of
checks: 33 . . . l:tb2+ 34 'it>f1 .l:tb1 + 35 �2
l:tb2+ 36 'it>g 1 , and only now played 36 . . . l:tb5
with equal ity.
34 l:txb5 axb5
I nstead of calmly strengthening his position 35 il.f2 ltJb7
by playing 20 b4! , then a2-a3, lld3, l:tad 1 36 'it>d2 h5
and at a conven ient moment c3-c4 , White 37 il.e3 g6
hu rries to exchange his active knight and 38 'it>c2 'it>d6
loses nearly all his advantage.
39 h4 ltJc5?
20 ltJb7? l:tab8 When offering to go i nto a pawn ending , you
21 ltJc5+ il.xc5 must calculate the variations very accu rately.
22 il.xc5 ltJcB I n this case Black got it wrong! 39 . . . ltJd8 was
23 f3 ltJd6 correct.
24 ltd3 ltJb7 40 il.xc5+ 'it>xc5
25 il.f2 .:thd8 41 g3 'it>b6
26 l:tad 1 l:.d6 42 f4 c5
27 'it>f1 l:!.bd8 43 a41 exf4
28 'it>e2 c5 44 gxf4 bxa4
29 b3 c4 45 'it>b2
30 bxc4 bxc4 And Wh ite soon won .
234 � From Games by Pupils of the School
Zviagi ntsev (1 5) - Galkin
Kramatorsk 1 99 1
The problem of exchanging was central to
this game. Black should have retai ned and
made use of his active knig hts. 1 9 . . . a4
suggests itself, after which White would have
to reckon with 20 . . . l2Jb3+ . In the event of the
exchange on c5 (of which , possibly, Black
was afraid, but qu ite unnecessarily) White
creates a weakness for himself on f2 . If 20
'ili'e 1 , then 20 . . . �h7 with good play.
19 . . . l2Jfe4?
20 l2Jgxe4 l2Jxe4
21 l2Jxe4
Or 21 'ifc2 l2Jxc3 22 'it'xc3 .
21 . . . 'ii'xe4
22 'ifc3!
It is probable that this simple move was
overlooked by Black. 23 �d3 is threatened .
22 . . . 'ii' b 1 +
23 'it>d2 'iff5
With the idea of 24 . . . e4 .
24 �g4 'ii'f6
25 l:tg3 h5!
A good plan: Black wants to retreat his
bishop to h7 and then exchange the dark­
square bishops, after which he will gain
counterplay on the f-file. After 25 . . . e4 26 �e6
'fie? Vadim Zviagintsev had prepared a
q ueen sacrifice: 27 l1xg6! �xc3+ 28 bxc3
with a winning attack.
26 �e2
As shown by Zviagintsev's analysis, the
active 26 �e6 also came i nto consideration:
A) 26 . . . �f7 27 �xf7 (White consolidates his
advantage more simply with 27 l:ldg 1 !? �h6
28 �xh6 'ii'x h6+ 29 'ike3 ) 27 . . . l:lxf7 (if
27 . . . �h6?, then 28 �xh5 'it'xf2+ 29 �e2 ) 28
l:r.g5! 'iff3 (28 . . . �h6? is bad in view of 29
.UXh5 Vxf2+ 30 'it>c1 ) 29 l:lg6! with the
advantage;
B ) 26 . . . �h7! 27 l:.dg 1 �h6, retaining coun­
terplay.
26 . . . h4
If 26 . . . �h7?! Wh ite was not intending to
reply 27 J.xh5 i n view of 27 . . . �h6 with
chances for both sides, but 27 l::tg 5 ! .
27 l:tg4!
If 27 l:tg5?, then 27 . . . �h6.
A critical moment in the game. 27 . . . J.h5?
loses to 28 :Xg7! �xe2 29 l:tdg 1 �h5 30
l:!.xc7 . I n order to hold the position , without
fail Black must ach ieve the exchange of his
passive dark-square bishop for its active
opponent. Therefore correct was 27 . . . J.h7!
28 l1dg 1 �h6 29 �xh6 (29 f4? ! l:tae8)
29 . . . 'ili'xh6+ 30 'iVe3 'it'f6! with a double­
edged game.
However, as Dvoretsky pointed out, Wh ite
 From Games by Pupils of the School tLJ 235

would nevertheless have retained the better Meanwh ile, Diana Darchiya could have
chances by offering the exchange of the immediately decided the outcome with an­
light-square bishops: 28 ..td3 ! , and if 28 . . . ..th6 other exchange: 25 'iVb3 ! . It is important to
29 �c1 . eliminate the q ueen , the defender of the e6-
27 . . . ..tf5? pawn , and thereby completely disorganise
28 l:1g5 l:.f7? ! the opponent's defences. Black has no
defence, for example: 25 .. ."ilkxb3 26 axb3
Black recog nises his mistake and tries to
i.d7 27 l:tf7 .
prepare 29 . . . i.h6 (if 30 l:th5 he has the reply
30 .. Jih7). However, he should have recon­
ciled himself to the loss of a tempo and Rasted - Kadymova (1 5)
played 28 . . . i.h7 29 .l:.h5! 'ilg6 . Now Wh ite Germany 1 992
easily prevents the exchange of the dark­
sq uare bishops and q uickly decides the
game with an attack on the g- and h-fi les.
29 l::. h 5+! �g8
30 l:tg1 �f8?!
3 1 ..tg5
Black resig ned .

Darch iya ( 1 4) - Gedyev

Moscow 1 99 1

Even in such a completely level situation one

must take care over exchanges, so as not to
be left in the end with bad pieces.
11 ..txf5?!
Wh ite should have exchange her 'bad' g5-
bishop (wh ich otherwise is doomed to ru n u p
against its own pawn cha i n ) for t h e oppo­
nent's more active knig ht. In this case the
game would have remained completely
level, whereas now Black gains the in itiative.

Often the assessment of a position depends 11 . . . lLlxf5

on slight nuances. Thus if the a5-pawn were 1 2 l:l.fe1 h6
moved to a4, Black would have a defensible 1 3 ..te3 l:tae8
position . 14 lLlf1 l2Jxe3?
25 ..tg6? Retu rning the favou r. Of cou rse, Black
The exchange of bishops is anti-positional . should have played 1 4 . . . l:l.e7 followed by the
Although Wh ite later won the game, this doubling of rooks, and if 1 5 tLlg3 captured
occurred only thanks to a blu nder by the the knight with the bishop, aiming to exploit
opponent. her knight's greater manoeuvrabil ity com-
236 � From Games by Pupils of the School
pared with the passive bishop. Now, howev­
er, not a trace of Black's advantage remains
and the game soon ended in a draw.
Pawn structu re
Chekaev (1 3) - Goldaev
USSR 1 989
Wh ite could have parried the threat of
1 3 . . . bxc4 with the simple 1 3 'i!i'c2 ! and then
carried out a standard plan: l:.ae 1 , lt:Jd 1 and
f2-f4 with the advantage. I nstead he ex­
changes pawns, which merely increases the
opponent's chances on the queenside.
13 cxb5? axb5
1 4 f4
Now Black would have obtained the more
pleasant game by continuing 1 4 . . . b4! 1 5
lt:Ja4 (or 1 5 lt:Je2 lt:Jg4 1 6 .l:tf3 c4! with the
th reat of 1 7 .. .'S'b6+) 1 5 . . . lt:Jd7 followed by
. . . ..ta6 , . . . ..tf6 and at a convenient moment
. . . c5--c4.
Smirnov - Emelin (1 3)
Len ingrad 1 989
What distingu ishes a grand master from a
master? Such a q uestion is often asked by
chess enth usiasts. Many th ink that a grand­
master calculates variations fu rther or has a
better knowledge of opening theory. I n fact
this is not the real difference. One ca n pick
out two important traits , in which a higher­
ran ked player is normally superior to one of
lower rank: this is a feeling for the critical
moment in a game and a more su btle
understanding of various positional prob­
lems.
The diagram position is an excel lent illustra­
tion of this last trait. When Sergey Dolmatov
looked at the game, he gave the following
assessment, which is very importa nt for such
structures:
'When in Spanish positions Wh ite closes the
centre, the placing of his a-pawn becomes
important. If Wh ite has already played a2-
a4, then Black gains cou nterplay by advanc­
ing his pawn to c4 and occupying the c5-
square with his knight. However, if the wh ite
pawn is still at a2, the . . . c5-c4 plan is now
weaker in view of b2-b3, weakening the c4-
pawn . Wh ite, in tu rn , should prepare the
opening of the a-file with the preparatory b2-
b3, and only then play a2-a4 . '
 From Games by Pupils of the School
Subtle eval uations of this type a re g radually
accu mulated by a player, and make u p his
store of positional knowledge.
Now it will be easy for the reader to
understand why the fol lowing natural move
by Wh ite has a q uestion mark attached to it.
20 a4?
Correct was 20 b3 followed by a2-a4 , when,
in Dolmatov's opinion , Wh ite's chances are
slightly better.
20 . . . c4!
After this standard reply Black seized the
in itiative .
Bogus lavs ky ( 1 6) - Cherniak
Moscow 1 99 1
Wh ite's pawn structu re has been spoiled . H e
should have taken the opportun ity not only to
rectify it, but also to open up the game, which
is undoubtedly advantageous to the side with
the two bishops. After the correct 1 1 f5! White
would have gai ned the advantage. The
natu ral castl ing move proved to be a serious
mistake, and the situation changed sharply.
1 1 0-0? CiJe7
1 2 �e1 0-0
1 3 CiJe4 CiJf6
1 4 CiJxf6+ �xf6
1 5 g4 �e8
ctJ 237
1 6 f5?
Excessively sharp ! 1 6 �d2 followed by 1 7
�c3 was better.
16 . . . gxf5
1 7 gxf5
Now the simple 1 7 . . .'it'd7 followed by . . . 'it>h8
would have placed Wh ite in a critical
position.
Material-g rabbing
This is a common mistake: many young
players, on seeing an opportu nity to win
material, often do not pay due attention to the
opponent's cou nterplay and disregard sound­
er alternatives ( remem ber the Morozov­
Baklan game). Here is another example on
the same theme .
Darch iya (1 1 ) - Welcheva
World G i rls Championship,
Fond d u Lac 1 990
20 �xeS?!
Wh ite should simply have strengthened her
position with 20 f3, when against the th reat of
2 1 b4 followed by the invasion of the knight
at c6 Black would have no adequate de­
fence. The move in the game al lows her
opponent more saving chances.
20 . . . dxc5
238 � From Games by Pupils of the School

21 tLlb7 'ili'b6 28 'ii'e 8 'ii'a 7

22 ttJxc5 l:tc8 29 .i.xb5?
23 b4 a5 It was possible to play for a win only by
24 lLld3 %bc1 + making the prophylactic move 29 .i.f1 ! .
25 'ii'x c1 axb4 Wh ite was fortunate still to have a d raw after
her opponent's strong reply.
26 ttJxe5? !
White repeats the same mistake, g radually 29 . . . b3!
making things more difficult for herself. The 30 ii.d7! ii.xe5
modest 26 g3 was stronger. 3 " 'i'xe5+ 'it'g8
26 . . . ii.f6 32 'ii'e 8+
27 'iVc8+ 'it'g7 Draw.

I n dex of P layers a n d Analysts

Ad ia nto 35, 47 B l u menthal 1 55 1 44 , 1 46, 1 65, 1 82 , 1 85, 1 86,

Adorjan 64, 1 77 Bogolj u bow 77 1 97, 200, 202, 204 , 205, 209,
Agdestein 1 7 4 Bog uslavsky 87, 1 6 1 , 222, 2 1 4 , 2 1 6, 224, 231 , 233, 234
Akopian 1 72 232, 237 E l i s kases 1 28
Alburt 1 7 1 Boleslavsky 1 49 E m e l i n 236
Alekh ine 76, 77, 78, 79, 1 29 Bolog a n 1 23 , 1 56 Ervin 30
Aleksa n d rov 1 90 Botvi n n i k 7, 27, 42, 84, 1 38 E uwe 27
Alexa nd ria 1 30 B u d n i kov 201
Anand 7 1 , 75 F a rago 1 03
Andersson 1 9 Capa b l a n ca 7 Fedore n ko 42
An ikaev 1 8 C h a rush i n 1 1 4 F i l ipenko 1 20 , 1 2 1 , 1 22
Arba kov 232 Cheba n e n ko 40 Fischer 1 9, 50, 67, 1 1 3, 1 42
Amason 1 20 Chekaev 236 Franke 114
Ashley 66 Chern iak 237
Averbakh 1 47 Chernosvitov 64 Gabdra k h m a nov 11
C h i b u rd a n idze 1 7 4 Gaj i c 57
Bagirov 44 C h i styakov 1 3 1 G a l k i n 234
Baklan 64 , 67, 70, 92, 1 57, Conq uest 1 63 Gavrikov 70, 1 87
1 60, 231 , 233 Gedyev 235
Balashov 95 Darchiya 67, 235, 237 Gelfa n d 1 5, 202
Ba ndza 1 90 Dj u ric 10 Geller 54, 63, 67, 69, 1 27, 1 56
Bareev 1 67, 1 94 , 1 95, 201 , Dl ugy 57, 58, 59 Georg iev, Kir. 1 76
202 , 203, 208, 2 1 6 Dolmatov 1 60 , 1 7 1 , 1 73 , Gertler 58
Becerra Rivero 2 1 7 1 74 , 236 Ghinda 1 3
Bel iavsky 66, 72, 1 80, 1 8 1 Donner 50 G l eizerov 1 22
Bel l i n 55 Dorfm a n 95 Glek 1 63, 1 79
Blumenfeld 9 Dvoretsky 29, 1 1 7 , 1 1 9, G l u s h n iov 1 4 1
 I ndex of Players and Analysts l2J 239

Goldaev 236 Kond ratiev 1 07 Moldagal iev 1 55

Goldin 171 Korchnoi 30, 69, 1 64 Morozov 231
Gonzalez 93 Korneev 98 Moska l e n ko 1 77
Gorelov 1 62 Korzu bov 1 90 Movsesian 67, 1 60 , 1 6 1
Gottschall 47 Kosi kov 98, 1 00 , 1 05 Muller 51
G rau 1 28 Kotov 23
G riffiths 1 57 Kovalevskaya 230 N i kolic 1 79 , 208
G rue nfeld 59 Kozul 56 , 1 67 N i mzowitsch 7, 27, 47, 7 1 ,
G ufeld 37 Kram n i k 1 6 1 , 1 62 , 1 68 , 1 69 , 101
G u l ko 30, 1 1 9 , 1 2 1 , 1 24 , 224 1 72 , 1 77 , 1 8 1 , 1 97, 200, 2 1 4 Nisman 1 46
G u revich , M. 1 64 Kruppa 201 NN 1 55
K u ij pe rs 1 55 Novi kov 1 81
Haba 171 Kupch i k 1 1 6 Nunn 1 5, 5 1 , 79, 9 1 , 1 09 ,
Ha rtston 1 55 K u p re i c h i k 1 24 , 1 90 1 34 , 1 35, 1 57
Herzog 66
H o rt 46, 54 , 67, 80
Larsen 54 Pach m a n 73, 74
H u bner 1 0, 49, 50, 1 55
Lasker, E d . 1 27 Panno 93
l l i nchenko 1 05
Lasker, E m . 7, 8 P e n rose 1 55
l lyi n-Genevsky 1 07
Lautier 1 90 Perl is 1 42
lvanch u k 30, 42, 1 94
Laza rev 42 Peters 30
Lei n 59, 2 1 6 Petrosia n , A . 1 55
Johansen 225 Leko 93 Petrosi a n , T. 37, 99, 1 1 3, 1 43
Lerner 68 Petu rsson 1 74 , 1 75
Kadymova 1 57 , 1 60 , 230, Levenfish 1 50 P i nter 64
235 Lisitsyn 65 Polugayevsky 1 1 3, 1 43
Ka l i kshte i n 89 Liti n s kaya 1 30 Psakhis 55, 93, 1 28 , 1 90
Kal i n ichev 1 79 Littlewood , P. 1 27
Kamsky 7 1 , 75 Log i nov 80 Rabar 68
Kan 1 50 Ragozi n 1 07
Karpeshov 80 McG u n n ig l e 1 55 Rash kovsky 1 20
Karpov 30, 33, 44, 46, 60, Ma ka rychev 55 Rasted 235
67, 68, 72, 1 36 , 1 55, 201 , M a l a n y u k 1 64 Razuvaev 60, 1 00 , 2 1 3
208 M a rja novic 56 Reshevsky 27
Kasparov 1 5, 46, 47, 68, M a rs h a l l 1 1 6 , 1 27 , 1 42 Reti 84
1 36, 1 72 , 201 M a rtin Gonzalez 93 R i b l i 55
Keene 1 27 Mats u l a 1 20 Roma n i s h i n 1 03 , 1 28, 1 55
Kengis 9 Matulovic 2 1 9 Romanovsky 1 52
Keres 27, 42, 56 Mayer 5 1 Rossol imo 1 55
Khenkin 1 62 , 1 63, 1 64 , 1 66 , Med n i s 1 27 R u b a n 1 77
1 67, 1 96 , 1 97 , 201 , 202, Meister 1 77 R u b i n etti 23
203 Merku lov 1 8 R u b i n stei n 79, 86
Kholmov 54, 1 56 Mestel 54
Kimelfeld 2 1 7 Meulders 225 Sadler 54
Kingston 1 56 M ichel 1 55 Saigin 1 35
Kinsman 89 M iles 1 28 S a k h a rov 1 4 1
Kiryakov 70, 1 57 , 1 60 M i r u m i a n 92 Salov 1 79 , 1 80 , 1 95
Knaak 63, 1 76 M iton 227 Savch e n ko 1 64
240 � Index of Players and Analysts

Schm idt 1 71 Sterk 78 70, 1 1 0

Vito l i n s
Schwartzman n 208 Sterner 1 49 89
Vysoch i n
Shaba lov 69, 1 80 Suetin 99, 1 1 9 Vyzhmanavin 1 80
Shamkovich 224
Shcherbakov 1 2 1 Ta i m a nov 1 7, 64, 1 47 Wachtel 1 55
Shereshevsky 1 6, 202 Ta kacs 86 Wade 1 55
S h i pov 1 56 Ta l 55, 1 38, 226 Wa ng Z i l i 1 83
S h i rov 89, 225 Ta rrasch 201 Welcheva 237
Shiyanovsky 233 Tel m a n 94 Wilson 1 64
Short 1 67, 1 7 1 , 1 72 , 1 75 , Ti m m a n 1 2 , 1 5, 33, 54, 1 76 , W i n a nts 225
1 76 201 W i rthen s o h n 80
Simagin 27, 1 30 , 3 1 3, 1 32 , Ti moshch e n ko 2 0 Wolff 2 1 9, 220
1 33, 1 35, 1 39 , 1 4 1 Tol u s h 6 5
Smej kal 72 Tsesh kovsky 1 7 , 29, 30, 1 1 7 Ya ku bovich 1 32
S m i rnov 236 Tu kmakov 1 1 0, 1 67 , 1 73, 1 74 Yus u pov 9, 1 0, 1 1 , 1 2 , 1 3 ,
Smyslov 29 1 4 , 1 5, 1 7 , 20, 23, 30, 80,
Sokolov, A. 60, 1 93 U lybin 1 72 1 36 , 1 6 1 , 1 83 , 1 87, 1 90,
Sokolov, I . 225 Uusi 1 33 1 93
Soloviov 1 00
Spasov 1 4 Va isser 1 67 , 225 Zagorya n s ky 84
Spassky 1 39, 1 83 Van der Wiel 72, 1 57 , 1 80 Zaitsev, I . 1 04
Speelman 55 Van Wely 1 77 Z i l berman 64
Spraggett 1 2 Vasyu kov 1 66 Zviagi ntsev 67, 70, 1 47, 1 6 1 ,
Stei n 56 Vi ku lov 1 44 222, 234

I ndex of O pe n i ngs

Bogo-l ndian Defence 1 1 0 , 1 95 Pi rc-Ufi mtsev Defence 33

Caro-Kann Defence 20, 57-60 Queen's Gambit 42, 86, 99, 1 90
Dutch Defence 1 62- 1 82 Queen's I ndian Defence 1 44 , 1 46
Engl ish Opening 1 7 , 1 05 Reti Opening 84 , 1 03
French Defence 29, 92, 1 00 , 1 07, 2 1 6 Ruy Lopez 7 1 , 1 83
GrOnfeld Defence 1 87 Semi-Slav Defence 27
King's I ndian Defence 37, 95, 1 24, 201 Simagi n-Larsen Opening 1 9
Nimzo-lndian Defence 224 Slav Defence 89, 1 1 9 , 1 2 1 , 208
Old I ndian Defence 23 Torre Attack 80