1 d4 d5 2 c4 c6

To my parents.
The Slav
CHESS PRESS OPENING GUIDES

Other titles in this series include:

1 90 1 25 9 05 6 Caro-Kann Advance Byron Jacobs

1 90 1 25 9 06 4 Closed Sicilian Daniel King
1 90 1 25 9 0 3 X Dutch Leningrad Neil McDonald
1 90 1 25 9 1 0 2 French Advance Tony Kosten
1 90 1 25 9 02 1 Scandinavian John Emms
1 90 1 25 9 OS 0 Semi-Slav Matthew Sadler
1 90 1 25 9 0 1 3 Sicilian T aimanov James Plaskett
1 90 1 25 9 09 9 Trompowsky Joe Gallagher

For further details for Chess Press titles, please write to The Chess Press
c/o Everyman Chess, Gloucester Mansions, 1 40a Shaftesbury Avenue,
London WC2H SHD.
Chess Press Opening Guides

The Slav

Matthew Sadler

ir
[1Illj

The Chess Press, Brighton

First published 1 997 by The Chess Press, an imprint of First Rank Publishing,
23 Ditchling Rise, Brighton, East Sussex, BN1 4QL, in association with
Everyman Books plc
Reprinted with corrections 1 999

Copyright © 1 997 Matthew Sadler

Distributed by Everyman Chess, Gloucester Mansions,

1 40a Shaftesbury Avenue, London WC2H SHD.

All rights reserved. No part of this publication may be reproduced, stored in a

retrieval system or transmitted in any form or by any means, electronic,
electrostatic, magnetic tape, mechanical, photocopying, recording or
otherwise, without prior permission in writing from the publishers .

A CIP c atalogue record f o r t h i s b o o k i s available from the British Library

ISBN 1 90 1 259 00 5

Cover design by Ray Shell Dcsign

Printed and bound in Great l3ritain by
l:3iddles Ltd, Guildford and King's Lynll
 CONTENTS

1 d4 d5 2 c4 c6

Bibliography 8
Introduction 9

The Old Main Line (3 ttJf3 ttJf6 4 ttJc3 dxc4 5 a4 �fs 6 e3 e6

7�xc4 �b4 8 0-0) : Black plays to prevent e4 13
2 The Old Main Line (3 ttJf3 ttJf6 4 ttJc3 dxc4 5 a4 �fS 6 e3 e6
7�xc4 �b4 8 0-0) : Black allows e3-e4 27
3 The New Main Line (3 ttJf3 ttJf6 4 ttJc3 dxc4 5 a4 �fS 6 ttJeS) :
Black fights for control of e4 39
4 The New Main Line (3 ttJf3 ttJf6 4 ttJc3 dxc4 5 a4 �fs 6 ttJeS) :
Black counterattacks 53
5 The Smyslov Variation (3 ttJf3 ttJf6 4 ttJc3 dxc4 5 a4 ttJa6) 62
6 The Bronstein Variation (3 ttJf3 ttJf6 4 ttJc3 dxc4 5 a4 �g4) 72
7 The 4 . . . a6 Slav (3 ttJf3 ttJf6 4 ttJc3 a6) : White plays 5 e3 87
8 The 4 . . . a6 Slav (3 ttJf3 ttJf6 4 ttJc3 a6) :
Aggressive options for White 98
9 The Exchange Variation (3 cxds cxdS) 1 08
10 Move-Orders and Transpositions 1 18
11 Odds and Ends 130

Index of Complete Games 142

 BIBllOGRAPsH Y

Books

Encyclopaedia o/ Chess Openings vol. D (ECO), Sahovski Informator 1987

Bats/ord Chess Openings 2 (BCO), Kasparov & Keene (Batsford 1989)
Winning with the Slav, Schipkov & Markov (Batsford 1994)
The Slav for the Tournament Player, Flear (Batsford 1988)

Periodicals

In/ormator
ChessBase Magazine
New In Chess Yearbook
British Chess Magazine
Chess Monthly
 INfRODUCrlON
1 d4 d 5 2 c4 c6
The skill of preparing an opening is
frequently misunderstood: many play­
ers (including some strong grand­
masters) believe that to play an open­
ing well, it is necessary to analyse a
great many variations; that no prepa­
ration is complete without at least one
queen sacrifice and that to stop before
move 20 is akin to criminal negli­
gence. I know from experience that
the sheer volume of opening theory
can be overwhelming, and this is es­
pecially true for the non-professional
player who has little time to keep up
with the latest fashions.
And yet, having been a professional
player now for six years, I know that
I have won more games from 'nor-
mal' openings than from any 30-move
piece of analysis (and it's not because I
haven't done any!) . The brilliant 'I
had this position after move 80 on my
board at home' games that we see in
magazines are the exceptions: beauti­
ful, treasured by every chessplayer,
but very, very rare. Chess is a sport
and most games are a struggle, and we
win games because we fight harder
than our opponents, or because we
understand the position better.
In my opinion, opening prepara­
tion can be successfully reduced to
three simple steps:
1. Knowing the main aim of our
opemng.
2. Knowing the value of move­
orders.
3 . Understanding typical positions.
Therefore, let's apply these ideas to
the Slav.
Opening Aims
With 2 c4, White challenges the black
centre. The natural 2 . . . e6, allowing
Black to develop his kingside pieces,
has the drawback of blocking the
light-squared bishop inside the pawn
chain. 2 . . . c6 aims to hold the centre,
to develop the light-squared bishop
9
Th e S l a v
outside the pawn chain, and then to
play . . . e7-e6 and conclude the black
development. However, the course of
chess ideas, like love, never runs
smoothly! Black must be careful
when he develops his light-squared
bishop: after 1 d4 ds 2 c4 c6 3 CDf3
CDf6 4 CDc3
Black would like to play 4 . . . �fS , but
he will have great difficulty defending
b7 after S cxds cxds 6 'iYb3 !
Now 6 . . . 'iYb6 loses a pawn to 7
CDxds 'iYxb3 8 CDxf6+ exf6 9 axb3 and
6 . . . b6 weakens the queenside light­
squares too much: 7 e4! dxe4 8 CDeS e6
(to stop 'iYxf7+ mate) 9 �bs+ CDfd7 10
g4 �g6 1 1 h4! , intending h4-hS, trap­
ping the bishop. The general rule is
10
that Black can only play a quick
. . . �fs if he can successfully defend b7
with his queen. Thus, 1 d4 ds 2 c4 c6
3 CDf3 CDf6 4 e3 �fs
is fine for Black since S cxdS cxdS 6
'iYb3 can easily be met by 6 . . . 'iYc7;
however 1 d4 ds 2 c4 c6 3 CDc3 CDf6 4
e3 �fS?!
S cxds cxds 6 'iYb3 is not good,
since 6 . . . 'iYc7 loses a pawn to 7 CDxdS .
So how can Black carry out his main
idea? Black either has to stop White
from playing 'iYb3 , or he has to find a
good way to defend b7. This is a typi­
cal opening dilemma: whether to pre­
vent an opponent's threat directly, or
whether to arrange the pieces in such
a way that the threat is nullified.

The main line of the Slav runs 1 d4
dS 2 c4 c6 3 tt'lf3 tt'lf6 4 tt'lc3 dxc4.
First, Black wins a pawn and
threatens . . . b7-b5, making this gain
permanent. Second, the b3-square is
cunningly taken away from the
queen, which means that White can­
not attack b7, and hence that . . . �f5
becomes possible. While White recap­
tures the c4-pawn, Black will develop
the light-squared bishop to f5 or g4
and will be looking to complete his
kingside development: 5 a4 (surroun­
ding the c-pawn by preventing . . . b7-
bS) 5 . . . �f5 6 tt'le5 (intending tt'lxc4) or
6 e3 (intending �xc4) 6 . . . e6 are the
main lines. So far I have been very
enthusiastic about Black's strategy,
but now I have to reveal the downside
of his play. This sort of schizophrenia
is necessary when you play both sides
of the Slav, as I do!
4 . . . dxc4 relinquishes control of e4,
which makes it easier for White to
cramp Black with two central pawns
on d4 and e4. But White must be care­
ful that his pawns do not become
weaknesses as Black first immobilises,
then attacks them. The bottom line is
that the player who has the better
In t r o du c tio n
understanding of the line will get the
best results with either colour.
The second idea is to play 1 d4 d5 2
c4 c6 3 tt'lf3 tt'lf6 4 tt'lc3 a6.
The move 4 . . . a6 was first played in
Britain by Grandmaster Jon Levitt,
but it is Grandmaster Julian Hodgson
who has upheld this variation at the
highest level, and introduced the most
significant ideas.
The original idea of 1 d4 d5 2 c4 c6
3 tt'lf3 tt'lf6 4 tt'lc3 a6 was to meet iVb3
with . . . b7-b5, moving the pawn to a
protected square. However, ways
were found to exploit the drawbacks
of . . . b7-b5: the dark-square weak­
nesses on c5, b6, as and the slightly
exposed black queenside. Then, in a
brilliant piece of unstereotyped think­
ing, Hodgson realised that . . . a7-a6
could allow the rook to defend b7
from a7. No one had dared to play
such a strange move before, but Julian
did, and this has made some previ­
ously dodgy lines completely viable.
However, although avoiding
. . . d5xc4 helps to prevent e2-e4, Black's
position is less dynamic than in the
. . . d5xc4 lines, as it is much harder to
break against White's centre with
11
Th e Sla v
. . . c6-cS and . . . e7-eS. My own prefer­
ence as Black is for the 4 . . . dxc4 lines,
as they are richer in content and offer
a wider range of possibilities to suit
many different styles.
Move-Orders
Move-orders are a much underrated
part of opening preparation. Oppo­
nents don't always play fair! Imagine
the scene: you sit down to play, con­
fident that you know your opening at
least as well as your opponent, and
what happens? He plays the opening
in some unusual move-order, and you
emerge a bit dazed to find yourself
playing a different line to the one you
wanted! And unless you work out
your move-orders thoroughly, this
will continue to happen, time and
time again.
12
S o how can this happen in the Slav?
If you want to play the 4 . . . a6 Slav,
then there is nothing that White can
do to muddy the water, which is one
of the attractions of this line. 4 . . . dxc4
lines, however, require some care.
First, White can try to sidestep them
by playing an early e2-e3, protecting
c4, e.g. 3 !bc3 !bf6 4 e3 . Black has
many reasonable moves here, but
none of them fit in with the idea we
want to play. The other way for
White to play is c4xdS, leading to the
Exchange variation; 4 . . . a6 Slav fans
should study these positions particu­
larly carefully, since this sort of posi­
tion is very typical of this line, and
there are many transpositions.
Understanding Typical Positions
Well, for this part, read on . . .
 CHAPTER ONE
The Old Main Line:
Black plays to prevent e3-e4
1 d4 d5 2 c4 c6 3 LiJf3 LiJf6 4 LiJc3
dxc4 5 a4 .§i.f5 6 e3 e6 7 .§i.xc4 .§i.b4
8 0-0
This traditional system of devel­
opment for White is especially popu­
lar at club level. White quietly recap­
tures the pawn and puts his king to
safety before he starts his plan of e3-
e4, to obtain a 'perfect' pawn centre.
This chapter examines Black's at­
tempts to prevent White from achiev­
ing this goal.
First, we need to ask ourselves a
few questions in order to establish our
approach:
Question 1: How will White try
and achieve e3-e4?
Answer: White has two major ap­
proaches:
a) 'i'e2. This is the most dangerous
idea, which we shall examine first.
b) ctJh4, to remove the bishop on
f5, which is helping Black to prevent
e3-e4.
Question 2: How can Black fight
against 'i'e2 and e3-e4?
A nswer: Black has three pieces at­
tacking the e4-square: the bishops on
b4 and f5 and the knight on f6. When
White plays 'i'e2 he is supporting the
e3-e4 push with only two pieces: the
queen and the knight on c3. How­
ever, he will achieve the e3-e4 advance
with tempo because the e-pawn at­
tacks the bishop on f5. If Black wants
to, he can simply pre-empt this by re­
treating the bishop to g6, so that e3-e4
no longer attacks the bishop. Now if
White plays e3-e4 regardless, Black
can win a pawn by playing . . . �xc3
and . . . ctJxe4.
Question 3: What move-order
should I play this in?
Answer: My own favourite has
been to play 8 . . 0-0 9 'i'e2 �g6;
.
8 . . ctJbd7 9 'i'e2 �g6 is sharper since
.
White can offer a dangerous pawn
sacrifice.
Gamel
Richardson-Sadler
Islington Open 1995
This was a crucial game for me: I was
leading the Islington Open by only
half a point and only a win would
guarantee first place. However, even
more importantly, only a win would
be good enough to pip Keith Arkell
for the Leigh Grand Prix!
1 d4 d5 2 c4 c6 3 LiJf3 LiJf6 4 LiJc3
dxc4 5 a4 .§i.f5 6 e3 e6 7 .§i.xc4 .§i.b4
13
Th e Sla v
S 0-0 0-0 9 'iVe2 jL g 6 1 0 LDe5
10 e4? ! �xc3 ! 1 1 bxc3 tbxe4 wins a
safe pawn. Consequently, White re­
verts to 'Plan B ' : he exchanges his
knight on f3 for my bishop on g6 and
removes an attacker of e4.
1 0 . . . LDbd7 1 1 LDxg6 hxg6 1 2 .lci.d 1
12 e4 tbb6! wins a pawn as 13 e5 (or
13 Mdl �xc3 14 bxc3 tbxc4 15 'i'xc4
tbxe4) fails to 13 . . . 'i'xd4 14 �xe6 fxe6
15 exf6 'i'xf6 with a clear extra pawn.
1 2 ...'iVe7!?
An interesting move, though I
imagine that it is not the most accu­
rate - for Kramnik's 12 . . . 'i'a5, see the
next game.
1 3 e4
Wait a minute! Wasn't Black play-
14
ing t o prevent e3-e4? Well, smce
White's aim was to push a pawn
within his own territory, it was al­
ways unlikely that we could prevent
it for ever. However, by resisting for
as long as possible, we have forced
White to make a concession, namely
that he has had to play tbe5xg6 before
being able to play e3-e4. Although
White gains the two bishops with this
manoeuvre, he exchanges off the piece
that would be most affected by the e3-
e4 advance; on g6, the bishop has little
scope if White can maintain his pawn
on e4. Moreover, the departure of the
knight from f3 means that White loses
some control over the central dark
squares, d4 and e5 . This last point is
seen to great effect in this game.
1 3 . . . e5 1 4 d 5 nacS !
I spent a great deal of time at this
stage and realised that I had to force
White to release the tension in the
centre and play d5xc6. The explana­
tion for this has a lot to do with the
central dark squares: without the d­
pawn, I can transfer a knight to e6
(via c5) and exploit the outpost that
my pawn on e5 creates on d4. By
placing my rook on c8, I was hoping
to get my opponent worried about
possible threats on the c-file, in order
to tempt him into playing d5xc6.
1 5 jL g 5 :!:;lfdS 1 6 dxc6?
Here it is! After this mistake, White
has to be very careful to avoid a dis­
advantage. The correct plan is ex­
tremely ingenious. Black has two ave­
nues of pressure: he has possible
threats along the c-file and he can de­
velop some pressure against e4 by
means of . . . tbc5. How can White deal
 Th e O ld M a in L in e : B l a c k pla ys to pre v e n t e 3- e 4
with both these threats? With the
manoeuvre 16 l':rd3 ! 4.Jc5 17 l':re3 ! ! On
e3, the rook covers e4 and defends the
knight on c3 along the rank, thus pro­
tecting White's queenside against c-file
play. White is slightly better after 16
l':rd3, but the game is still very compli­
cated.
1 6 .� . bxc6 !
My opponent had underestimated
this recapture. Although it weakens
Black's queens ide pawn structure,
Black protects the central light
squares, over which he previously had
little control due to the exchange of
his light-squared bishop. This move is
so strong, because White's queenside
is so weak: the pawn on a4 gives
Black a comfortable slot on b4 for the
queen, from which it can attack the
a4- and b2-pawns.
1 7 .!:Id 3 ? CLlc5 1 8 .!:Ih 3 ? CLle6 !
White's 17th and 1 8th moves were
excessively optimistic as he had no
chance of an attack along the h-file.
Meanwhile, Black threatens . . . 4.Jd4.
With his control of the d-file, and
White's weakened queens ide as a clear
target, I believe that Black can already
be thinking about victory. During the
game, I was very impressed with
White's attitude: realising that his po­
sition had worsened considerably,
White regrouped and concentrated
totally on defence.
1 9 iLxe6 �xe6 20 .!:Id 3 gxd 3 2 1
�xd 3 �b3 2 2 .t:!.b 1 l:!:b8 2 3 iL d 2
iLa 5 !
Protecting the d8-square, so that the
rook can use either of the open files
on the board.
24 f3 .!:Id8 25 �e2 CLle8 !
Black's knight is the least active of
his piec€s. The text prepares to bring
it to d4 via c7 and e6. When this hap­
pens, all of Black's troops will be on
their optimal squares.
26 iLe 1 CLlc7 27 l:!:c 1 CLle6 28 �c2
�b6+ 29 �f2 CLld4
Intending 3o . . . ihc3 31 �xc3 4.Je2+
winning the exchange, as 32 '>jVxe2 is
impossible since the queen is pinned
to the king.
30 �h 1 �b3 31 �b 1 l:!:b8 3 2 h3
�c4 3 3 f4 exf4 34 �xf4 ge8 3 5
�d 1 iLxc3 3 6 iLxc3 CLle2 3 7 �f3
gxe4
The first weak pawn falls.
38 a5 a6 39 �d3 CLlxc3 40 �xc4
gxc4 41 bxc3 .!:!.c 5 !
15
Th e Sla v
I think that the rook ending is
winning now. 4 1 . . .Mxc3 42 Md8+ �h7
43 Ma8 would have regained the a­
pawn, but now after 42 Md8+ �h7 43
Ma8 , Black can simply play 43 . . . Mxa5.
42 .l:i.d8+ 'it> h 7 43 gc8 g 5 44 'it>g 1
'it>g6 45 'it>f2 �f5 46 �c7 f6 47
.l:i.xg 7 .l:i.xa 5 48 �e3 .l:!.e5+ 49 'it>d4
�f4 50 ga7 ge6 !
5 1 �b7
5 1 Mxa6 c5+! wins a rook.
5 1 . . .'it> g 3 5 2 .l:i.b2 .l:i.e 5 5 3 llb6 c5+
54 'it>c4 f 5 5 5 gxa6 'it>xg 2 5 6 .llg 6 f4
57 h4 f3 58 .llx g 5+ .l:i.xg 5 59 hxg5
f2 60 �xc5 f 1 � 61 c4 �f5+ 6 2
'it>d6 � g 6 + 0 - 1
This was a very important game for
me, and an instructive example of
16
what both sides should be aiming for
in this variation.
... .
. Game 2
Karpov-Kramnik
Monte Carlo (blindfold) 1995
1 d4 d 5 2 c4 c6 3 CZJf3 CZJf6 4 CZJc3
dxc4 5 a4 jLf5 6 e3 e6 7 jLxc4 3l.b4
8 0-0 0-0 9 �e2 jLg6 1 0 CZJe 5 CZJbd7
1 1 CZJxg 6 hxg6 1 2 .lld 1 �a 5 !
This move i s more active, and
probably more logical, than 12 ... 'iYe7.
While Black is not threatening to win
a pawn immediately with 13 . . . �xc3
due to 14 bxc3 'iYxc3 15 �d2 'iYc2 16
�d3 ! 'iYb2 17 Mdb 1 , winning the
queen, it does prevent 13 e4, as with
White's centre slightly weakened,
Black can get away with taking the
pawn: 13 . . . �xc3 14 bxc3 'iYxc3 1 5
�d2 'iYxd4 (15 . . . 'iYc2!?) 16 �b4 'iYe5
17 �xf8 Mxf8 when with two pawns
for the exchange, Black stands very
well. Note that 13 ctJa2 allows
13 . . . 'iYxa4 14 ctJxb4 'iYxa 1 15 ctJa2
(hoping to trap the queen) 15 . . . 'iYb 1 ! ,
escaping t o f5!
1 3 jLd2
Protecting c3, and intending e3-e4,
 Th e O ld M a in L in e : Bla c k pla ys to pre v e n t e 3- e 4

but . .
. 3 3 �a2 �h2+ 34 \t>f 1 �xh3+ 3 5
1 3 . . . e5 1 4 d 51Iad8 W e 2 tLJ e 5 36 iL e 2 �xg4+ 3 7 W d 2
14 . . . cxds? 15 ctJxds ctJxds 16 �xds !!ed8+ 38 We 1 �xd 1 + 3 9 iLxd 1 �g 1
'i'xds 17 �xb4 wins for White. 40 iLd2 tLJd3+ 4 1 <;t>e2 ldb2+ 0 - 1
1 5 dxe6 bxe6 1 6 iLe1 e4 !
Securing an outpost on d3 for the S o White's plan of 9 'iVe2 and 10
knight. ctJes seems harmless. Let us take a
look at the more direct 9 ctJh4.

Game 3
Yusu pov-Kram n ik
Riga 1995

1 d4 d 5 2 e4 e6 3 tLJf3 tLJf6 4 tLJe3

dxe4 5 a4 iLf5 6 e3 e6 7 iLxe4 iL b4
8 0-0 0-0 9 tLJh4!
The most testing idea: White elimi­
nates the bishop on fs without wast­
ing time on 'iVe2.
1 7 iLb3 �e5 1 8 1Le2 iLd6 1 9 g 3
tLlc5 2 0 \t> g 2 �f 5 2 1 lId2 1L e 5 ! 22
gad 1 J::i: b 8 !
The white queenside is looking
very weak.
23 iL b 1 iLxe3 24 bxe3 tLJxa4 25
ga2 tLJb2 26 U d 2 tLJe4 2 7 1::i: d 1 tLJe 5
28 h3 tLJf3 29 lIa4 tLJ g 5 30 g4 �e5
3 1 �e2 lIfe8 3 2 !!xa7 tLJf3

9 . . . tLJbd7 ! ?
A typical stratagem: Black's dou­
bled f-pawns will take over the
bishop's task of preventing e3-e4.
Question 4: What is wrong with
9 . �g6?
. .

Answer: White can play 10 ctJxg6

hxg6 1 1 'iVc2!
Question 5: Why is it important to
Threatening . . .'iVh2+. Now Black is put the queen on c2 and not e2?
just winning. Answer: First, the queen neutralises

17
Th e Sla v
Black's most active plan of . . . 'iYa5,
threatening . . . jl,xc3 . Second, with the
queen on c2, Black must be careful
that when he plays . . . e6-e5 he does
not allow 'iYxg6! ( . . . e6-e5 has opened
up the a2-g8 diagonal and the f7-pawn
is now pinned to the king, so Black
cannot recapture on g6) . Of course,
we are dealing with subtle nuances
rather than big differences, but it is
important to understand them none­
theless.
1 0 tLlxf5 exf5 1 1 �c2 g 6 1 2 f3 �b6
Preventing e3-e4 by attacking the
d4-pawn, which has been weakened
by the absence of the white knight
from £3 .
1 3 �h 1 �aeS 1 4 �f2 c5 1 5 �h4
.ld:cS ! ? 1 6 jLa2 �fdS 1 7 jLd2 tLlfS 1 S
a 5 �a6 1 9 �fd 1 c4
Shutting out White's light-squared
bishop.
20 jLe 1 .l:!.eS 2 1 e4 jLxc3 22 jLxc3
fxe4 23 d5 tLlSd7 24 �d4 Y2 - Y2
The draw was agreed in a very
murky position.
It is now time to consider the other
move-order: 8 . . . tLlbd7, intending to
meet 9 'iYe2 with 9 . . . jl,g6 as above.
18
Game 4
Ivanchuk-Bareev
Dortmund 1 995
1 tLlf3 d 5 2 d4 tLlf6 3 c4 c6 4 tLlc3
dxc4 5 a4 jLf5 6 e3 e6 7 jLxc4 jL b4
S 0-0 tLlbd7
Question G: What does Black gain
from delaying castling?
Answer: 8 . . . tLlbd7 is directed against
the plan of an early tLlh4, which we
saw in Yusupov-Kramnik. After 9
tLlh4, Black will reply 9 . . . jl,g6, as 10
tLlxg6 hxg6 is extremely dubious for
White. Since Black has not castled, his
rook is well placed on the semi-open
h-file, pointing towards White's king!
Black will play . . . 'iYc7 (attacking h2) ,
castle queenside and then double
rooks on the h-file, which is not what
White was hoping for when he sensi­
bly (he thought!) took the bishop
pair!
So what does White do after
8 . . . tLlbd7 9 tLlh4 jl,g6? Give up? Cry?
Well, if he's a genius like Ivanchuk,
he chooses a third option: he gets
sneaky.
9 tLlh4 jLg6 1 0 jLe 2 ! ?
 Th e O ld M a in L in e : Bla c k pla ys to pre v e n t e 3-e 4

White wants to take on g6 only

once Black has castled; so he plays a
useful consolidating move while he
waits for Black to commit his king.
The text prevents the bishop on g6
from escaping the knight's attentions
by 10 . . . ]LhS!? The alternative waiting
move, 10 h3 , is considered in the next
game.

1 2 . J:lc8 ! ?
.

Since the white queen i s o n the c2,

Black tries to inconvenience it by
opening the c-file. The immediate
12 . . . cS would be met by 13 l2la2 ! , net­
ting Black's other bishop since
13 . . . ]LaS loses a pawn to 14 dxcS, in­
tending b2-b4. Bareev therefore plays
the rook to the c-file in order to facili­
10...0-0 tate . . . c6-cS . The . . . c6-cS break is
Fans of tactics can investigate played less often than . . . e6-eS in the
10 . . . ]Lxc3 1 1 l2lxg6 (not 1 1 bxc3 l2lds Slav, but it is a typical idea that is well
12 l2lxg6 l2lxc3! 13 iVc2 l2lxe2+ 14 worth remembering.
'iYxe2 hxg6, winning a pawn) 1 3 e4 ! ?
1L.]Lxb2 12 l2lxh8 ]Lxa1 13 ]La3 or
13 iVc2 (unclear - Ivanchuk) and
when they have, 1'd be grateful if they
could tell me what is going on! How­
ever, more positional players can be
happy with Bareev's move. Although
White's queen will go straight to c2,
the bishop is more passive on e2 than
on c4: after . . . e6-eS, Black no longer
has to fear iVxg6 (in fact he'd be quite
pleased to see it!) as the white bishop
is not on the a2-g8 diagonal; and this
also means that White cannot reply so Since White does not want the c-file
easily with d4-ds after . . . e6-eS or . . . c6- to be opened, he prepares to meet
c5. 13 ... cS with 14 ds . If only he still had
11 ct:Jxg6 hxg6 1 2 'iVc2 his bishop on c4! This move

19
Th e Sla v

introduces a sharp pawn sacrifice that ous play due to his threat of . . . Me2.
is probably not quite good enough, so 1 3 . . . c5 1 4 d 5 exd 5 1 5 exd 5 ! ? l:l:e8 ? !
13 Md1 was suggested by Ivanchuk as The start of a series of slight inaccu­
an alternative, when he claims a slight racies that Ivanchuk exploits brutally.
advantage for White. 1 3 . . . cS 14 ds Is . . . iLxc3 16 bxc3 ttJxds 17 Md1
exdS 15 ttJxds ttJxds 16 Mxds VlJie7 1 7 ttJ7f6! 1 8 iLf3 (18 c4 ttJb4! unpins)
b3 ! i s indeed rather better for White. 1 8 . . . VlJie7! is Ivanchuk's recommenda­
Black's main problems are the weak­ tion, as 19 iLxdS ttJxdS 20 MxdS al­
ness of his light squares and his bishop lows mate after 20 . . . VlJiel .
on b4, which is shut off from the rest 1 6 IId 1 c4
of Black's pieces by the pawn on cS .
The immediate 1 3 . . . VlJie7 is stronger:
14 e4 (also interesting is 14 b3!?,
intending 14 ... cS [ 14 ... eS is more
sensible] 15 dS iLxc3 16 d6! and VlJixc3
with advantage) 14 . . . cS 1 5 dS exdS 1 6
ttJxds ttJxds 1 7 exdS (17 Mxds loses
the e-pawn to 17 . . . ttJf6)

1 7 d6!
A very strong move: the d6-pawn
exposes the light-squared weaknesses
in the black position by opening the
hl-a8 diagonal and freeing dS for the
knight on c3 .
1 7 . . . l:l:e6 1 8 i.f4 "i'b6 1 9 l2lb5 i.c5
20 i.g3!
1 7 . . . c4! , freeing the bishop and pre­
venting White from establishing a
light-squared blockade of the queen­
side with b2-b3 and iLc4. After 1 8
iLxc4, Black can regain the pawn with
18 . . .tZJb6 19 b3 ttJxc4 (19 . . . ttJxdS 20
iLb2! [20 iLxds Mxc2; 20 Mxds VlJiel+!J
leaves White more active due to his
two raking bishops) 20 bxc4 VlJih4!,
when 2 1 f4 (the only move to save the
c-pawn) 2 1 . . .Mfe8 (22 . . . Mel+ is now a
threat) 22 g3 VlJig4 gives Black danger-

20
 Th e O ld M a in L in e : Bla c k pla ys to pre v e n t e 3- e 4

Calmly protecting f2.

20 'ctJe4 2 1 .liLg4 f5 22 .liLf3 ! <;t>h7
. .

23 CiJc7 ge5 24 tLJd 5 gxd 5 2 5 !!xd 5

lLlxg 3 2 6 a 5 1 -0
26 hxg3 would also have won.
White is just the exchange up with a
winning pOSltlOn.

Game 5
T opalov-Gelfand
Belgrade 1995
1 d4 d5 2 c4 c6 3 tLJf3 tLJf6 4 tLJc3
dxc4 5 a4 .liLf 5 6 e3 e6 7 .liLxc4 .liLb4
8 0-0 tLJbd7 9 tLJh4 .liLg6 1 0 h 3 ! ?
Another waiting move. However,
Black's bishop has a square!
Black's passive play has gIven
White a huge space advantage.
1 9 . . . tLJd6 20 tLJxg 6 hxg6 2 1 .liLf4
tLJe8 22 \\Wf3 .liLd 6 23 .liLe3 tLJdf6 24
g5 tLJh 5 25 !:tfd 1 tLJc7

10 . . . .liLh 5 1 1 g4 tLJd 5 ! 26.liLc2? !

Black makes a discovered attack by
the queen on the knight on h4. Once
the knight retreats, Black will again
have g6 for his bishop.
12 tLJg 2 .liLg6 1 3 tLJa2 .liLe7 ? !
Topalov rightly suggests 1 3 . . . �d6
14 f3 hS ! as an improvement, striking
immediately against the exposed king­
side pawns.
1 4 \\We2 tLJ 5 b 6 1 5 .liLb3 c 5 1 6 tLJc3
0-0 1 7 a 5 cxd4 1 8 exd4 tLJcs 1 9
lLlf4
26 dS! eS 27 liJe4 (Topalov) would
have led to a clear edge for White.
26 . . .f5! 27 h4 b 5 2S axb6 axb6 29
l:haS \\WxaS 30 \\WxaS J::!. x aS 3 1 d 5
exd 5 3 2 tLJxd 5 tLJxd 5 3 3 .liLb3 <;t>fS
34 .liLxd 5 l:!.a4 35 .liLxb6 l:!.xh4 36
.liLc6 .liLe7 3 7 J::!. a 1 tLJf4 3S .l:!.aS+ <;t>f7
39 .liLeS+ �e6 40 J::!. a 6 �d 5 41 .liLf7+
<;t>e4 42 il..e 3 Yz - Yz
Instead of 9 liJh4, White has a more
testing plan: 9 ik'e2 and 10 e4!

21
Th e Sla v
Game 6
Ivanchuk-Lautier
Linares 1994
1 d4 d 5 2 e4 e6 3 ctJf3 ctJf6 4 ctJe3
dxe4 5 a4 iif5 6 e3 e6 7 iixe4 iib4
8 0-0 ctJ b d 7 9 'iWe2 iig 6 1 0 e4 ! ?
This pawn sacrifice is the problem
with this move-order.
1 0 . . . iixe3
The more restrained 10 . . . 0-0 is con­
sidered in the next chapter.
1 1 bxe3 ctJxe4 1 2 iia3 !
Question 6: What i s going on?
Answer: In return for the sacrificed
centre pawn, White has gained the
two bishops and prevented Black
from castling kingside.
Question 7: Can't Black just grab
another pawn with 12 .. .ct:Jxc3?
Answer: NO!! 13 'i'b2 (hitting the
knight and the pawn on b7) 13 . . . QJxa4
14 'i'b3 ! , threatening both 'i'xa4 and
.ixe6, is horrible for Black.
Question 8: Can't Black just block
the a3-f8 diagonal with 12 . . . c5 and
then castle kingside?
Answer: This is logical, but 13 dxc5
is awkward, meeting 13 . . . 0-0 with 14
22 ,
c6! and 13 . . . QJexc5 with 14 .ib5 ! 0-0
15 .ixd7 QJxd7 16 .ixf8 .
Question 9: How then can Black get
his king to safety?
Answer: Black can castle queenside
instead of kingside.
Question 1 0: You don't sound very
impressed!
Answer: Black's position is horrible!
White has a simple and extremely ef­
fective plan: a4-a5-a6, softening up the
black queenside, and then, after mov­
ing the bishop on c4, c3-c4-c5 finish­
ing off the job. In reply, Black must
seek to play . . . e6-e5 and activate his
kings ide pawns.
1 2 . . .. �e7 1 3 afe 1
From el, the rook protects the c3-
pawn and supports the c3-c4-c5 push.
1 3 . . 0-0-0
.
13 . . . c5 14 QJe5 ! QJxe5 (14 . . . 0-0 15
QJxg6 wins a piece) 15 dxe5 'i'xe5 (or
else White plays f2-£3 and h2-h4, trap­
ping the knight) 16 £3 wins a piece,
while 13 . . . QJd6 (blocking the a3-f8
diagonal) 14 .ixe6! 0-0 (14 . . . fxe6 15
'i'xe6+ is crushing) 15 .ib3 favours
White due to his bishop pair.
1 4 a5
1 4 . . . ctJd6 ! ?
 Th e O ld M a in L in e : B l a c k pla ys to pre v e n t e 3- e 4
The young Rumanian player Gab­
riel Schwartzman tried 14 .. J;fhe8
against Razuvaev in Dortmund 1993,
but after 1 5 a6! b6 16 ttJh4 ttJd6 17
jLb3 es 1 8 ttJxg6 hxg6 19 iVg4!
(preventing Black from activating his
kingside pawns with .. .f7-fS by attack­
ing g6) 19 . . -,,�?b 8 20 jLxd6! iVxd6 2 1
jLxf7 �e7 2 2 iVxg6, White stood
clearly better. 19 . . . cS cutting out the
bishop on a3 , was suggested as an im­
provement, but after 20 dxcS bxcs 2 1
�ab 1 ! (preventing 2 1 . . .�b8 due t o 22
jLxf7+) 2 1 . . .�e7 (defending f7) 22
�cd1! (eyeing the knight on d6) the
only positive course of action open to
Black is to wring the neck of the man
who made this suggestion!
1 5 1l.b3 1l.h5
An idea of Bareev's I believe, trying
to inconvenience White by the pin on
the knight. Note that 1 s . . . �he8 is met
by 16 ttJh4!
1 6 h3
16 'iVe3 unpinning, and eyeing the
a7-pawn, is also interesting.
16 . . J!he8
17 a6!
Softening up the protection around
the black king.
1 7 . . . b6 1 8 c4 !
Threatening c4-cs.
1 8 . . . c 5 1 9 � a 4 e 5 20 dxc5 b x c 5 2 1
�e3 ge6 22 tLlg 5 .l::rf 6 23 � b 2 h 6 24
tLle4 tLlxe4 25 �xe4 ge6
I would not recommend this posi­
tion to anyone. The game finished:
26 ge 1 �g6 27 �a8+ �b8 28
1l.xd7+ gxd 7 29 �xb8+ 'it>xb8 30
gxe5 gdd6 3 1 .l:!.xc5 .l:!.xa6 3 2 gxa6
gxa6 3 3 �xg7 h5 34 f4 f 5 35 �e5+
'it>b7 36 gc7+ 'it>b6 3 7 'it>h2 'it>a5 38
'it>g3 �e8 39 'it>h4 'it>b4 40 g 3 gg6
4 1 .l::rx a7 'it>xc4 42 1::!. a 8 �c6 43 .l:!.a3
�e8 44 ge3 �d 5 45 �f6 � d 7 46
�c3 �e6 47 � b4 �d7 48 �xh 5
gg8 49 �c3 �d6 50 'it>h6 �e6 5 1
�h7 ga8 5 2 � b4+ �d7 5 3 �g7
l::!. a4 54 'it>f6 � d 5 5 5 �c3 �e4 56
g4 fxg4 57 hxg4 �c2 58 .l:!.e7+ �d8
59 �e5 ga6+ 60 l::!. e 6 gxe6+ 6 1
�xe6 �e8 6 2 �d6 �d 1 6 3 g 5 1l.c2
64 f5 1 -0
Game l
Ehlvest-Schwartzma n
New York Open 1996
1 tLlf3 d5 2 d4 tLlf6 3 c4 c6 4 tLlc3
dxc4 5 a4 �f5 6 e3 e6 7 �xc4 �b4
23
Th e S l a v
8 0-0 ttJ b d 7 9 '¥We2 � g 6 1 0 e4 �xe3
1 1 bxe3 ttJxe4 1 2 �a3 '¥We7 1 3 J:l:fe 1
0-0-0 1 4 a 5 �b8 1 5 �e7 ! ?
An interesting manoeuvre, transfer­
ring the bishop to the annoying h2-b8
diagonal. The bishop has already ful­
filled its task on a3 by preventing the
black king from castling kingside.
1 5 . . . .i:!.de8 1 6 �h4 �a8 1 7 '¥Wb2 f5
1 8 ttJd2 f4 1 9 f3 ttJxd2 20 '¥Wxd 2
J:l:hf8 2 1 �f 1 e 5 22 d x e 5 J:l:xe 5 2 3
a6
Ehlvest criticises this move, prefer­
ring White after 23 Ma4 Md5 24 'i'a2
ctJc5 25 Md4. This may well be more
accurate, but the essential point is that
White will always have very good
chances because Black's king is weak
and White's bishops are strong.
23 . . . b 6 24 :!la4 .i:!.d 5 2 5 :!ld4 :!lxd4
26 '¥Wxd4 ttJ e 5 2 7 .i:!.d 1 :!le8 28 �e4
iLe2 29 J::!. e 1 llxe 1 + 30 1i.xe 1 1i.f5
3 1 1i.d 2 '¥Wd 7 32 'li'xf4 b5 3 3 �f 1
ttJxa6 34 �e3 � b 7 3 5 �f2 �g6 3 6
h4 e 5 3 7 '¥W e 3 �b6 3 8 � h 2 '¥We8 3 9
'¥Wg 5 h 6 40 '¥Wd 2 '¥W e 6 4 1 �g3 �f6
42 �d7 e4 43 �f2+ ttJe5 44 g 3 !
After a time-scramble and a little
confusion, White re-establishes con­
trol with this evil little move, which
24
prepares to activate the light-squared
bishop on the long diagonal.
44 . . . 'i'xf3 loses simply to 45 'i'd6+.
44 . . . a 5 45 iLg2 �e6 46 '¥Wd8+ �a6
47 '¥We7 ttJd7 48 f4 '¥We7 49 �e6+
ttJb6 50 '¥Wxg 6 a4 51 '¥We8 a3 5 2 'i1\Vf8
a 2 53 'i1\Va3+ ttJa4 54 'i1\Vxa 2 'i1\Vd 7 55
�d4 1 -0
If this isn't enough to convince you
of the danger in accepting the pawn
sacrifice, then try this!
Game 8
Hubner-Beliavsky
Munich 1994
1 d4 d 5 2 e4 e6 3 ttJe3 ttJf6 4 ttJf3
dxe4 5 a4 �f5 6 e3 e6 7 �xe4 � b4
8 0-0 ttJbd7 9 'i1\Ve2 �g6 1 0 e4 �xe3
11 bxe3 ttJxe4 1 2 �a3 'i1\Ve7 13
J::!. f e 1 ! ?
A very aggressive alternative to the
old 13 Mfc l . White sacrifices yet an­
other pawn, reasoning that this will
merely open more lines for his pieces.
1 3 . . . ttJxe3
Best and the most critical. 13 . . . 0-0-0
was extremely unpleasant for Black in
the game Beliavsky-Akopian, No­
vosibirsk 1993: 14 'i'b2 Mhe8 15 as e5
16 Mab 1 c5 17 �f1 f6 and now 1 8
ctJh4! ctJxc3 19 'i'xc3 exd4 2 0 'i'b3
JiLxb 1 21 Mxb 1 gives White an over­
whelming initiative, as Beliavsky
pointed out. Clearly in such lines, the
king's rook is much more actively
placed on e 1 than on c 1 (as in lines we
have seen previously). This is also true
of 13 . . . c5, as in Hubner-Hertneck,
Munich 1994, when 14 d5! e5 15 �d3!
ctJef6 1 6 ctJxe5 0-0-0 17 ctJxd7 Mxd7 1 8
 Th e O ld M a in L in e : B l a c k pla ys to pre v e n t e 3- e 4

c4 (Hubner) is the (unpleasant) best now becomes a little random, due to

that Black can hope for. mutual time pressure, but White pulls
14 'i'b2 ctJe4 1 5 a 5 ! ? through in the end.
T o break up the black queenside
with a5-a6. 15 I[ac 1 is also interesting.
15 . . . ctJdf6
15 .. .liJd6 (intending . . . 0-0) is met by
16 .!xe6! (16 Vib4 c5 ! [not 16 . . . lZ'lxc4??
1 7 'ife7 mate] 17 dxc5 lZ'lxc4 1 8 c6
4Jxa3 ! 19 cxd7+ Vixd7 20 Vixa3 Vie7
21 'ifa4+ Vi d7 22 Vi a3 [preventing
kingside castling], which leads to a
draw by repetition after 22 . . . Vie7)
22 . . 0-0 17 ltxd7 Vixd7 1 8 lZ'le5 Vic7
.

19 a6! , breaking up the queenside with

an advantage.
Instead 1 5 ... a6 (preventing a5-a6) is
best, when Hubner suggests 16 I[e3
4Jd6 17 .!xe6 0-0-0 18 ltxd6 Vixd6 19
.!c4 and I[b3 with a dangerous attack.
16 tLJe5 a6 1 7 .l:i.ac 1 .l:i.d 8 1 8 jLxa6 ! !
22 . . . ctJd6 23 �c7 .l:i.d 7 24 �c3 0-0
25 ctJe5 Il:.dd8 26 �c6? �xc6 27
.l:i.xc6 ctJb5 28 jLxf8 �xf8 29 gxa6
ctJxd4 30 f3 f6 ? ! 31 ctJxg6+ hxg6 3 2
f 4 Il:.d 5 3 3 .l:i.e4 g 5 3 4 �a4 e 5 3 5
fxe 5 fxe 5 3 6 � a 7 � g 8 3 7 JJ.e7

1 8 . bxa6 1 9 ctJxc6 ctJg4 20 ctJ e 5 !

. .
'i'xa5 2 1 ctJxg4 �b5 2 2 �c2 ? !
This i s White's first inaccuracy in
this fascinating game! Hubner notes
that the simple 22 Vixb5 axb5 23 I[c7
I\,a8 24 ltb4 I[a4 25 I[c8+ 'it'd7 26
I\,xh8 I[xb4 27 lZ'le5 'it'c7 28 I[c1 + 'it'b7
29 f3 lZ'lf6 30 lZ'lxg6 would have been
clearly better for White. The game
N ow White is winning again.
37 . . . ctJc6 38 JJ.e6 gc5 39 �g6 �f7
40 �xg 5 g 6 41 h4 �f6 42 �h2 ctJe7
43 �a4 �c6 44 .l:i.g 3 ctJf5 45 gf3
�g7 46 l:te4 �e6 47 h5 ctJe7 48
hxg6 ctJxg6 49 �h3 ctJe7 50 �g3
ctJg6 51 �f5 �e7 5 2 �g4 l:ta7 5 3
� g 5 ctJh8 5 4 � g 4 ctJ f 7 + 5 5 <;t> h 5 +
� f 8 56 �g6 � e 8 5 7 � h 4 1 -0

25
Th e S l a v

S u m m a ry

The alert reader will have noticed my profound mistrust of the line 8 . . . 'bbd7
9 'i'e2 �g6 10 e4 �xc3 1 1 bxc3 'bxe4. I honestly cannot understand the at­
traction of these lines for Black. Therefore, if Black wishes to try to prevent
e3-e4, then Kramnik's 8 . . . 0-0 9 'i'e2 �g6 is the line for you; Yusupov's plan of
a quick 'bh4 is the most testing response.

1 d4 d 5 2 c4 c6 3 ttJt3 CUt6 4 ttJc3 dxc4 5 a4 .ltf5 6 e3 e6 7 .ltxc4 .ltb4

8 0-0

8 . . . 0-0
8 . . . 'bbd7 (D)
9 'bh4 �g6
10 �e2 - game 4
1 0 h3 - game 5
9 'i'e2 �g6 (9 . . . �g4 - see next chapter) 10 e4 �xc3 (10 . . . �g6 -
see next chapter) 1 1 bxc3 'bxe4 12 �a3 'i'c7
13 .sfe 1 - game 8
1 3 .sfc1 0-0-0 14 as (D)
14 . . . 'bd6 game 6
-

14 . . . c,t>b8 - game 7
9 �e2
9 'bh4 game 3
-

9 . .lt g 6
. .

9 . . . 'bbd7 - see next chapter

1 0 ttJe5 ttJ b d 7 1 1 ttJxg6 hxg6 1 2 .l:td 1 (D)
12 . . . 'i'e7 - game 1
12 . . . 'i'aS game 2
-

B... ttJ b d7 14 a5 12 'iJ.d1

26
 CHAPTER TWO

The Old Main line:

Black allows e3-e4

1 d4 d 5 2 c4 c6 3 l2lf3 l2lf6 4 l2lc3 Answer: Good question! Generally,

dxc4 5 a4 i.f5 6 e3 e6 7 i.xc4 i.b4 black players play 9 . . iLg6 to pretend
.

8 0-0 that they are willing to take on the

This chapter deals with lines arising pawn sacrifice after 10 e4. Even if
from 8 . . .cubd7 9 'Wie2 0-0 10 e4 iLg6 they don't intend to take the pawn,
(or 9 . . . iLg6 10 e4 0-0, turning down the idea is to make White waste a lit­
the dangerous pawn sacrifice on the tle time on the clock thinking about
way) and 9 . . . iLg4. his variations! You never know - a
Question 1 : When White plays e3- few minutes might be handy later!
e4, he gains a commanding central
presence with pawns on e4 and d4. Game 9
Why is Black playing this position? G ofshtein-Sadler
Isn't he just worse? Ischia 1996
Answer: Central pawns are strong if
they are dynamic and able to advance 1 d4 d 5 2 c4 c6 3 l2lf3 l2lf6 4 tLlc3
and chase away the opposing pieces. dxc4 5 a4 i.f5 6 e3 e6 7 i.xc4 i.b4
Otherwise, they can present easy tar­ 8 0-0 tLlbd7 9 �e2 0-0 1 0 e4 i.g6
gets for the enemy pieces. In this case, 1 1 i.d3
White cannot advance d4-dS, and e4-
eS leaves a hole on dS for the black
pieces. Moreover, Black is threatening
to win a pawn with . . . iLxc3 and
tiJxe4, now that his king is safely cas­
tled. Therefore, while the d4-e4 centre
gives White a definite space advantage,
Black has plenty of threats against the
white centre, which is the basis of his
counterplay.
Question 2: What is the difference
between playing 9 . . . iLg6 first or
9 . . . 0-0 10 e4 iLg6? White had to deal with the threat

27
Th e S l a v

of . . . iLxc3 and . . . tbxe4. It is a general the knight on f3 to the queen and

rule that the longer you can delay thus threatens . . . e6-eS. The alternative
committing your centre, the better, 1 1 . . .h6 is considered in Game 12.
since the later you reveal your hand, 1 2 itf4
the less time your opponent has to Trying to avoid e4-eS for a while
adjust to it. longer, White brings another piece to
11 . . . it h 5 bear on eS. Strangely enough, this is
probably not the best move. The di­
rect 12 eS is considered in the next
game.
12 . . . 'Vj'e7 ! ?
Black threatens 1 3 . . . iLxf3 1 4 'iVxf3
eS! , equalising comfortably. I think
that this is a novelty: 12 . . . Me8 had
been played before.
1 3 e5
Absolutely necessary.
13 . . . tZ:l d 5 1 4 tZ:lxd5 cxd 5

Question 3 : What is the pomt of

this move?
Answer: The first place to look for
counterplay, is with your pawn
breaks. Pawn breaks have two func­
tlons:
a) They attack the opponent's
pawn structure and force him to react
to you.
b) They are a 'breakout': they gain
space and therefore give more room
for your pieces to become active. White has a space advantage due to
Black has two pawn breaks in this his pawns on d4 and eS. Normally,
position: . . . e6-eS and . . . c6-cS . Usually with his bishops, knight and queen
he prefers to aim for . . . e6-eS, since this pointing towards the black king,
stops White from playing e4-eS him­ White could consider launching a
self, inconveniencing the black kingside attack, but here Black's
knight. For example, 1 1 . . .cS 12 eS! bishop on hS interferes with this plan:
tbdS 1 3 tbxds exdS (13 ... iLxd3 14 it can exchange itself for the knight
'iVxd3 exdS 1 5 tbgS ! is unpleasant) 14 on f3 or return to g6 to block any
iLxg6 hxg6 15 tbgS, with threats of white attack on the b 1-h7 diagonal.
eS-and 'iVg4-h4 and 'iVh7 mate, is nasty Meanwhile Black will challenge for
for Black. The move in the game pins the c-file, exchanging the bishop on

28
 Th e O ld M a in L in e : B la c k a lia ws e 3 - e 4

d3 for the one o n g6 i n order t o free

c2 as an entry square for the black
major pieces. Black will also transfer
his knight to c6 via b8 from where it
not only attacks d4, but can invade
the white queenside by as-b3 or via
b4. So what on earth can White do?
Stay calm! White does not want to
exchange pieces on the c-file since this
would help Black to free his cramped
position, so he has two plans. First
(my own favourite) , he can concen­
trate on the kingside where White 1 7 . . . Jixd 3 1 8 �xd 3 .l:!.e4 !
holds most of the trumps: a space ad­ Black is now planning . . . b7-bs and
vantage and a large concentration of 19 b3 is met by 19 . . . Mc7 ! , when 20
minor pieces. I would try to push my 'ik'bS does not win a piece as Black has
kingside pawns: 1 5 h3 !!fc8 16 g4 ltg6 20 . . . ltc3 , when he stands well. Realis­
17 h4, intending h4-hs. This plan ing that his activity on the queenside
demonstrates the drawback to ltf4, has come to naught, White goes back
however: White would like to throw to 'Plan A' and expands on the king­
the f-pawn forward as well, but the side, but he is several tempi down on
bishop gets in the way. The chances what he could have had earlier.
after 15 h3 are, I believe, about equal. 1 9 h4 h 6 20 h5 ctJ e 5 !
White's choice is interesting, but there
is always a danger in choosing plans
based mainly on tactical points: if
there is just a little hole in your calcu­
lations, then you often find that you
have just wasted time and must retreat
in disarray. On the other hand, such
plans are often the most unexpected
and the most difficult for the oppo­
nent to deal with!
1 5 �e3 1:He8 1 6 a 5 ! ? Jig6 1 7 .l:!.a4 ! ?
At first I thought about playing
17 . . Mab8, intending 18 . . . bS to drive
. Oops! My opponent had missed
away the white rook. Then to my that one. However, after a big think,
horror I noticed 18 ltbS ! White is he came up with an active defence.
threatening to take the knight on d7 2 1 dxc5 nxf4 22 a6 ! bxa6 23 c6 a 5
and then take my bishop on b4, so 2 4 .l:!.e 1 g e 8 2 5 b 3 ge7 26 .l:!. e 2
18.. .ctJf8 is natural, but then 19 'ik'b3 ! ne4? !
and my bishop i s trapped! A rather casual move. 26 . . . Mfs 27

29
Th e S l a v

Mal Mxh5 28 g4 Mh3 29 'It>g2 �h4!! 35 'It>gl Mc1+ leads to immediate

(this lovely move was pointed out to mate.
me by Julian Hodgson after the game)
was the way to play. So 12 �f4 doesn't seem all that
27 '!::!'a 1 jL c 5 ? ! 28 .!::!.x a5 �xc6 ? ? promising for White. What about 12
e5 instead?

Game 10
Gelfand-Lautie r
.t-urich 1994
1 d4 d 5 2 c4 c6 3 lLlf3 lLlf6 4 lLlc3
dxc4 5 a4 jLf5 6 e3 e6 7 iLxc4 jL b4
8 0-0 lLlbd7 9 �e2 jLg6 1 0 e4 0-0
1 1 jLd3 iLh5 1 2 e 5
Probably the best move.

29 .!::!. b 5 ? ?
My God! I had missed that White
could simply win a piece with 29
�b5 , forking rook and bishop. I had
thought that I could play 29 . . . �xf2+
with a discovered attack on the rook
on c2, but White just plays 30 Mxf2!
Luckily White shared the same blind
spot! After 29 Mb5, White is just lost.
29 . . . �c7 30 �a2 iLb6
Trapping the rook.
3 1 <;t>h2 .l::!. c3 3 2 �d2 �c6 33 gxb6 1 2 . . . lLld 5 1 3 lLlxd5
axb6 34 lLld4 .!::!. h 4+ 0- 1 The alternative, 13 liJe4, 1S dealt
with in the next game.
1 3 . . . cxd 5
13 . . . exd5 has been suggested, but
since White already has a space advan­
tage on the kingside, I am sceptical
about conceding a pawn majority as
well in that area.
1 4 �e3 h 6 ? !
A debatable decision. I would pre­
fer 14 . . . �e7, followed by a rook to
the c-file and . . . liJb8-c6.
1 5 lLle 1 !

30
 Th e O ld M a in L in e : Bla c k a I/a ws e 3- e 4

pletely dominates the black knight.

29 . . . 'i'xg6 30 'i'e3 � h 7 3 1 3i.f4 'i'f7
3 2 3i.e 5 'i'd7 3 3 h4 \tlg8 34 h 5 !

A very instructive plan: White in­

tends to play f2-f4 and f4-f5, which is
particularly dangerous once Black has
weakened his kingside with . . . h7-h6. Fixing the g 7-pawn.
Moreover, the knight will eventually 34 . . . 'i'd8 3 5 'i'g3 'i'd7 36 3i.d6 'i'f7
come to d3 with tempo, hitting the 37 'i'e5 'i'f5 38 'i'e2 b4 39 g4 'i'f6
bishop on b4. 40 3i.e 5 'i'g 5 41 �g2 b3 42 3i.d 6 !
15 .. .f5 1 6 exf6 'i'xf6 1 7 3i.b5 ttJb8 Preventing the knight from becom­
1 8 tLld 3 ! a6 mg active.
18 ... 1td6 19 tLJe5 is not pleasant for
Black.
1 9 tLlxb4 axb5 20 a5 ttJa6 2 1 ttJ d 3 !
l:tfe8 22 3i.d 2 g e 2 23 gae 1 g a e 8 24
l:txe2 gxe2 25 ge 1 gxe 1 + 26 3i.xe 1
iLg6 2 7 ttJe5 'i'f5 2 8 h 3 'i'e2 29
lLlxg 6 !

42 . . . �f7 43 'i'd 1 \tlg8 44 3i.g 3 'i'f6

45 3i.e 5 'i'h4 46 3i.g 3 'i'f6 47 'i'd2
'i'e7 48 'i'e3 'i'f6 49 3i.d 6 � h 7 50
'i'x b3 'i'xd4 51 'i'e2+ \tlh8 5 2 'i'e8+
�h7 53 'i'e2+ �h8 54 'i'e8+ �h7
5 5 'i'xe 6 !
Protecting g4.
Excellent judgement. In the result­ 55 . . . 'i'xb2 56 3i.e 5 'i'c2 57 'i'e7
ing position, White's bishop com- 'i'e4+ 58 �g3 'i'd3+ 1 -0

31
Th e S l a v
After 59 f3 , Black cannot stop
mate. His knight has not moved since
move 20!
Black should be fine after the ex­
change of knights on dS , providing he
avoids weakening his kingside. Let us
take a look at 1 3 tbe4.
Game 1 1
Xu Jun-Akopian
Moscow Olyrt;lpiad 1994
1 d4 d 5 2 c4 c6 3 CLlf3 CLlf6 4 CLlc3
dxc4 5 a4 �f5 6 e 3 e 6 7 �xc4 � b4
8 0-0 CLlbd7 9 �e2 �g6 1 0 e4 0-0
1 1 � d 3 �h5 1 2 e 5 CLld 5 1 3 CLle4 ! ?
Many white players d o not enjoy
the positions that we have seen in the
first two games of this chapter. With
the centre closed, and the prospect of
exchanges on the c-file, they feel un­
easy about their winning prospects; so
recently the plan with 13 tbe4 has
come to prominence. Of course,
Black keeps his knight outpost on ds
and his pawn-break against the centre
with . . . c6-cS. However, White's space
advantage remains and he retains e4 to
transfer first his knight, then his
32
queen to the kingside. Yes, this is the
hacker's option!
1 3 . . . �e7
13 . . . cS is unpleasantly met by 14
ltgS! 'i'aS 15 ltbS! Once Black moves
the knight on d7, he will lose the cs­
pawn, and he cannot protect it with a
rook due to the bishop on gS. If he
protects the knight with l S . . . 'i'c7,
then 16 Mac1 is unpleasant. The text
prevents ltgS and prepares . . . c6-cS .
1 4 CLlg3
The direct approach. 14 as has also
been tried.
Question 4: Why does White play
14 as?
Answer: a4-aS gains queenside
space, preventing Black from using
the as or b6 squares for his pieces.
Question 5: So what's the verdict? Is
it a good move?
Answer: Absolutely not! These
aims are completely irrelevant. 14 . . . cS
is logical, striking at the d4-pawn.
Neither 1 5 dxcs tbxeS nor 15 tbxcs
tbxcs (lS . . . ltxcS!?) 16 dxcS MCS (of
course not 16 . . . ltxcs 17 ltxh7+ �xh7
1 S iVc2+ �gS 19 tbgS ltg6 20 iVxcs,
winning a pawn) , followed by . . . ltxcs
or . . . MxcS, promise White anything.
 Th e O ld M a in L in e : B l a c k a //o ws e 3 - e 4
14 . . JLg 6
. 1 7 tZJc3
1 5 JLxg6 hxg6
15 ... fxg6 used to be played almost
automatically, primarily for defensive
reasons: black players were worried
that if they recaptured with the h­
pawn, White would play his knight
on g3 to gS via e4 and his queen to h4,
when Black would have no defence to
'Ih7 mate. By taking with the f-pawn
Black retains the option of ... h7-h6 to
keep a knight out of gS, and of course,
he gains the semi-open f-file for coun­
terplay. And then people realised that
White's attack was hardly automatic
after 1 5 . . . hxg6, so this move gradually
became the main line!
1 6 tLle4 c 5
This is a perfectly reasonable plan,
challenging Black's knight on ds once
Black has weakened its support by
playing . . . c6-cS, but why did the Chi­
nese player avoid 17 CLlegS? 17 . . . cxd4
18 'iVe4 's'e8 19 'iVh4 CLlf8 seems to be
a good reason. The knight on f8 de­
fends against 'iVh7 and if by some
miracle White manages to threaten to
get a rook on h3, then Black can hit
the 'panic button' and chase the
knight away with .. .£7-f6. White
could, however, try and open up the
black kings ide with h2-h4-hS , possibly
after 16 'iVe4 cS 17 h4 cxd4 18 hS !?
And now it's up to you, the reader!
We'll have to wait for practical tests
before a conclusion can be reached.
1 7 . . . �b6
17 . . . CLl7b6 18 as CLlxc3 19 bxc3 CLlds
20 c4 CLlb4 also seems reasonable.
1 8 tZJxd 5 exd 5 1 9 dxc5 tZJxc5 20
JLe3 �e6 21 a5 a6 22 .l:!.ac 1 .l:!.ac8
23 .l:!.fd 1 tZJb3 24 .l:!.xc8 .l:!.xc8
Black has an isolated d-pawn, but
White's queenside is weak. The posi­
tion is about equal.
25 JLb6 JLd8 ? ! 26 JLxd 8 .l:!.xd8 27
�e 1 ? !
33
Th e S l a v
Black's slightly incautious 25th
move allowed White the chance to
activate his queen by the lovely 27
'iYe4!, intending 12Jg5 and 'iYh4, as
27 . . . clxe4 allows 2 8 lIxd8+ �h7 29
ttJg5+ winning the queen (analysis by
Xu Jun) . The rest is hard-fought, but
it was always going to be a draw.
27 . . :Viii e 7 28 'Viii c 3 tLJ c 5 29 'iVb4 �f8
30 �f 1 tLJe6 3 1 'Viiix e7+ c!;xe7 3 2
J::i d 3 d4 3 3 J::i b 3 J::i d 7 34 g 3 tLJd8 3 5
J::i b 6 J::i d 5 3 6 b4 d 3 3 7 c!;e 1 c!;d7 38
�d2 c!;c7 3 9 J::i d 6 J::i x d6 40 exd6+
�xd6 41 �xd 3 �d5 42 tLJ d 2 tLJc6
43 c!;c3 tLJe5 44 tLJb3 tLJc4 45 f4 f6
lh - lh
Game 12
Sadler-M iles
British Championship 1 998
1 d4 d 5 2 c4 c6 3 tLJf3 tLJf6 4 ct:Jc3
dxc4 5 a4 �f5 6 e 3 e6 7 �xc4 Si.b4
8 0-0 tLJ b d 7 9 'Viii e 2 Si.g6 1 0 e4 0-0
1 1 � d 3 h6
A slightly risky idea. Black devel­
ops quietly and waits for an opportu­
nity to break with either . . . c6-c5 or
... e6-e5. The problem is that, as in this
game, Black can really get sat on!
34
l1.. .'iYaS has a similar idea. After 12
�f4! lIfe8 (12 . . . �xc3 13 bxc3 'iYxc3
loses to 14 �d2 'iYc2 15 �d3 'iYb2 16
lIfb l) 13 h3 lIac8 14 12Ja2!, White had
a slight advantage in Beliavsky-Short,
Linares 1995, as 14 . . . �f8 (14 . . . 'iYxa4
15 ttJc3! 'iYb3 16 �c4 wins the queen)
15 b4! gains queenside space with
tempo: 15 . . . 'iVxa4 16 lIfb l and 12Ja2-c3
traps the queen, while 15 . . . �xb4 16
12Jxb4 'iYxb4 17 lIfb l 'iYaS 18 lIxb7 is
unpleasant for Black.
1 2 �f4 J::i c 8 1 3 l:Ifd 1 J::i e 8 1 4 h3 a6
1 5 l:Iac 1 !
After 15 ttJa2 �f8 16 b4 Back has
16 . . . aS! (an excellent idea, breaking
White's grip on the c5-square) 17 bxaS
'iYxaS 18 �d2 'iYc7 19 e5 �xd3 20
'iYxd3 12JdS 21 ttJc3 ttJxc3 22 �xc3
ttJb6 23 as ttJdS 24 �el c5 and Black
had no problems in LSokolov-Oll,
Moscow Olympiad 1994.
1 5 . . . �b6
No better is 15 . . .'iYaS 16 ttJd2! b5
(16 . . . �xc3 17 bxc3 'iYxa4 18 lIal traps
the queen) 17 axb5 axb5 18 12Jb3 'iYb6
19 e5 12Jd5 20 12Jxd5 exdS 21 �xg6
fxg6 and Black was probably happy
he couldn't see his position in Lautier­
Gelfand, Monaco (blindfold) 1999.
1 6 �b 1 !
I really like this development
scheme. White consolidates his queen­
side and mobilises all his pieces, ready
for any of Black's breaks.
1 6 . . . �h 7 ? !
Black should really have taken the
plunge with 16 . . . c5, although 17 12Ja2!
wins the bishop pair with a nice ad­
vantage for White.
1 7 tLJe5?
This is rather careless though!
 Th e O l d M a in L in e : B l a c k a ll o w s e 3 - e 4

17 . . :i'd8 ? tremely solid and has no pawn weak-

17 ... ctJxeS 1 8 dxeS "iVc7! 19 "iVc4 nesses.
�f8 is okay for Black.
1 8 ClJc4!
N ow Black is suffering.
18 . . . ClJb6 1 9 ttJ a 2 ! .ll. f 8 20 b 3 !
Maintaining the knight on c4, as
Black will now have to improve
White's structure to get rid of it.
20 . J::i. a 8 2 1 �h 1 ttJc8 22 a5 ttJd7
. .

23 ClJc3 i.b4 24 ttJa4 Wiie 7 25 Wiig 4

itJf6 26 �f3 ttJa7 27 .ll. g 3 ttJd7 28
d5! exd 5 2 9 exd 5 i.xb 1 30 J::i. x b 1
cxd 5 3 1 ttJcb6 ttJxb6 3 2 ttJxb6 gad8
33 ClJxd 5 'i'f8 34 ttJc7 ge7 3 5 J::i. x d8 1 5 . . . a6 1 6 ..\te2 J::i. f d8 1 7 ga4 ..\te7
�xd8 36 ttJ d 5 J:i.e 1 + 3 7 J::i. x e 1 i.xe 1 1 8 f4 CLle8 1 9 .ll. e 3 gac8 20 .ll. f3
38 b4 Wiie 8 ? ? 3 9 ttJc7 ! �c6 40 'iVe3 ttJd6 21 i.e2 g 6 22 J::i. a a 1 1h - Y2
ixb4 41 Wiix a7 .ixa5 42 �b8+ �h7
43 ClJe8 b 5 44 ttJd6 f6 45 Wii a 7 i.b4 Here Kasparov shows a more criti­
46 ClJf5 i.f8 47 �f7 'i'c8 48 CLld4 cal approach for White.
ia3 49 ttJe6 'i'g8 50 'iVd7 �h8 5 1
ic7 .ib4 5 2 f4 .ic3 5 3 f5 b4 54 Game 14
id6 1 -0 Kasparov-Bareev
Novgorod 1994
Occasionally 9 . . . .,tg4 is seen instead
of 9 .. .,tg6, as in the next two games.
. 1 d4 d5 2 c4 c6 3 ttJf3 ttJf6 4 <1Jc3
dxc4 5 a4 i.f5 6 e 3 e6 7 i.xc4 i. b4
Game 13 8 0-0 ttJbd7 9 'iWe2 i.g4 1 0 h 3 !
Khalifman-Kir Georgiev
.

Elenite 1994

1 d4 d5 2 ttJf3 ttJf6 3 c4 dxc4 4

itJc3 c6 5 a4 Jl. f5 6 e3 e6 7 i.xc4
ib4 8 0-0 ttJ b d 7 9 �e2 i.g4 ! ?
A rather unusual move that aims to
bore White to tears by exchanging off
into a dull ending.
10 J::i. d 1 'i'a 5 ! 11 e4 'i'h5 1 2 h3
ixf3 1 3 'i'xf3 'i'xf3 1 4 gxf3 0-0 1 5
a5
White has the two bishops and a The most aggressive idea: now if
space advantage, but Black is ex- 1 0 . . . .,thS, White can play for e3-e4

35
Th e S l a v
without fear of the . . . iVaS-hS ma-
noeuvre.
1 0 . . . SLXf3 1 1 jVxf3 0-0 1 2 .i:::l. d 1 jVa5
1 3 e4 e5 1 4 d 5
This bears a distinct similarity to
Richardson-Sadler from Chapter 1 .
White's queen i s better placed o n f3
than e2, though of course Black's
queen is more actively placed as well.
1 4 . . . ttJb6 1 5 �b3 �xc3
cxd 5 1 7 exd5 .i:::l. a c8 1 8 c4! !
Brilliant. 1 8 . . . lLJxc4 19 �gS iVa6 20
d6! lLJxd6 21 �xf6 gxf6 22 iVxf6 lLJc4
23 iVxa6 bxa6 24 Md7 is clearly better
for White (Kasparov) .
1 8 . . . .i:::l. f e8 1 9 � d 2 jVa6 20 d 6 ! ttJbd7
21 �e3 .i:::l. c 6 2 2 a 5 ! .i:::l.x d6 23 �a4
.i:::l. x d H 24 .i:::l. x d 1 .i:::l. d 8 25 c 5 !
Black has been tied up in quite bril­
liant fashion.
25 . . . h 6 26 jVf5 g 6 27 �c2 ! �xa5
28 .i:::l. d 6 <;t>g7 29 jVd 1 �c7 30 h4
�h8 31 h 5 g 5 32 'i¥f3 <;t>g7 33 'i¥f5
b6 34 SLxd 7 ! ttJxd7 3 5 �xg 5 ! 1 -0
Kasparov gives 3 S . . . hxgS 36 iVxgS
'.tf8 37 h6 bxcS 38 h7 winning. A
really magical game.
Question 6: In your own games you
have played both 8 . . . 0-0 and 8 . . . lLJbd7.
36
Which is the better move?
A nswer: I can offer no definitive
conclusion: in theory, 8 . . . 0-0 and
8 ... lLJbd7 are equally good. However,
practical chess is not only about find­
ing the very best move: when making
a final decision whether to play 8 ... 0-0
or 8 ... lLJbd7, it is also necessary to
take the strength of your opponent
1 6 bxc3 mto account. The following game
shows why.
Game 15
D . Strauss-Lakdawala
USA 1992
1 d4 d5 2 c4 c6 3 ttJf3 ttJf6 4 ttJc3
dxc4 5 a4 � f5 6 e3 e6 7 �xc4 �b4
8 0-0 ttJbd7 9 'li'b3
This move forces a draw by repeti­
tion if White desires after 9 . . . a5 (best)
10 lLJa2 (chasing the bishop away)
10 . . . �e7 1 1 iVxb7 Mb8 12 iVa6 (12
iVxc6 Mb6!) 12 ... Ma8 1 3 iVxc6 Mc8 14
iVbS Mb8, as the queen cannot escape
from the rook's attack. So if your op­
ponent is much weaker than you, or
you desperately need a win, you must
play 8 . . . 0-0, as 9 iVb3 iVe7! (protecting
b7 and facilitating ... c6-cS) is nothing
 Th e O l d M a in L in e : Bla c k a ll o w s e 3 - e 4

for White. Alert readers will observe 1 2 ....txa5 1 3 "lixe6+ '>t>dS 1 4 e5

that I chose 8 . . .0-0 against Richardson tiJe4 1 5 tiJxe4! .txe4 1 6 "lif7 ! !
in just such a must-win situation. The
attempt to avoid the repetition in this
game is brutally dealt with.
9 :tlVb6 1 0 e4
..

A brilliant discovery of David

Gliksman. The queen moves to allow
the e-pawn to advance.
1 6 . . . MfS
10 . . . .tg6 The situation is already desperate.
10 .. .ctJxe4 1 1 CLlxe4 �xe4 1 2 �xe6! 1 6 ... �xf3 1 7 e6! (D.Strauss) 1 7 ... CLlf6
is clearly better for White. 1 8 e7+ �c8 19 �e6+ �c7 ( 1 9 . . . CLld7
11 �xe 6 ! fxe6 1 2 a 5 ! 20 e8� 20 �f4 + wins, while instead
1 6 . . . �g6 17 �xg7 Me8 1 8 dS ! ! (again
D.Strauss) , opening more central
lines, is crushing as 1 8 . . . cxdS 19 �g5 +
�c8 20 Mfc 1 + is appalling for Black.
17 'llIVx g7 Si.d5 1S e6! .txe6 19
"lig5+!
Now the point of 1 2 as is revealed:
this check picks up the loose bishop
on as !
1 9 .. . '>t>cS 20 Mxa5 �gS 2 1 "lih5 .tg4
22 Mg 5 ! .txh5 23 MXgS+ tiJfS 24
MxfS+ '>t>d7 25 tiJe5+ '>t>c7 26 MxaS
Diverting the bishop to an inferior 'i'xd4 27 Me 1 .te2 2S MeS c5 29
square. �e7+ '>t>cS 30 tiJc6 1 -0

37
Th e S l a v

S u m mary

After 8 . . . LtJbd7 9 "iVe2 0-0 10 e4 �g6 11 �d3 �hS, 12 eS is probably White's

best try and after 12 . . . LtJdS, hackers should choose 13 LtJe4 and positional
players should favour 13 LtJxds and 14 "iVe3 . These positions are complicated
and interesting in all cases. After Kasparov's brutal treatment, 9 . . . �g4 should
probably be avoided unless you spot a flaw in 'Gazza's' analysis. If you do,
then you're probably Mr Karpov! But remember, if you desperately need a
win as Black, 8 . . 0-0 is the only way to play.
.

1 d4 d5 2 c4 c6 3 CL'lf3 CL'lf6 4 CL'lc3 dxc4 5 a4 3l.f5 6 e3 e6 7 3l.xc4 3l.b4

8 0-0

8 . . 0-0
.

8 . . . LtJbd7 (D)
9 "iVe2 �g4
10 �dl - game 13
10 h3 - game 14
9 "iVb3 - game 1 5
9 �e2 CL'l b d 7 1 0 e 4 iL g 6 1 1 iL d 3 iLh 5 (D)
11. . .h6 - game 12
1 2 e5
12 �f4 - game 9
1 2 . . . CL'l d 5 (D)
13 LtJxds - game 1 0
1 3 LtJe4 - game 1 1

B . . . CL'lbd 7 11 . . . iLh 5 12 . . . CL'ld5

38
 CHA PTER THREE

The New M ain Line:

Black fig hts for contro l of e4

1 d 4 d5 2 c4 c6 3 iZJf3 iZJf6 4 iZJc3

dxc4 5 a4 iLf5 6 iZJe5
The sequence 1 d4 d5 2 c4 c6 3 ttJf3
ttJf6 4 ttJc3 dxc4 5 a4 �f5 6 ttJe5 is the
latest fashion at the highest level. 6
ttJe5 truly is a 'modern' move: White
does not worry about putting his king
into safety; he tries immediately to
achieve e2-e4 by moving the knight to
e5 and playing f2-£3 . In principle
Black should not allow White to
achieve e2-e4, since here it is ideally
defended by a pawn on f3 , leaving the Question 1 : What is the material
white pieces free for active operations. balance?
In this chapter, Black continues the A nswer: Black has four pawns for a
fight for e4 with an interesting piece piece, though it is touch and go
sacrifice: whether he can hang on to the c4-
6 e6 7 f3 iL b4
. . . pawn. Even three pawns, however, is
Pinning the knight on c3 and thus good material compensation for the
fighting for e4. temporarily rather inactive bishop on
8 e4 f1 .
... and after this there is no going
back! Game 16
8 .lil.xe4 9 fxe4 iZJxe4 1 0 iLd2
. . . Lal i c-Sadler
10 'i'f3 leads to a well-known draw Hastings 1995/96
by repetition after 10 . . . 'i'xd4 1 1
'fxf7+ �d8 1 2 �g5+! ttJxg5 1 3 'i'xg7 1 c4 c6 2 d4 d 5 3 iZJf3 iZJf6 4 iZJc3
.ixc3+ 14 bxc3 'i'xc3+ 15 �e2 'i'c2+ dxc4 5 a4 iLf5 6 iZJe5 e6 7 f3 iLb4
16 �el 'i'c3+, etc. 8 e4 iLxe4 9 fxe4 iZJxe4 1 0 iLd2
10 . . . �xd4 1 1 iZJxe4 �xe4+ 1 2 �e2 �xd4 11 iZJxe4 �xe4+ 12 �e2
iLxd 2 + 1 3 �xd 2 �d 5 + iLxd2+ 1 3 �xd2 �d5+ 1 4 �c3

39
Th e Sla v

After 14 �c2 lLJa6, which is consid­

ered in the next game, White always
has to worry about . . . lLJa6-b4+, which
can be irritating. Therefore white
players began to experiment with 14
�c3 , avoiding this sort of counter­
play. The one drawback to the king
on c3, however, is that it is just within
reach of the black queenside pawns,
so Black can play an aggressive con­
tinuation that would not succeed
against 14 �c2.
24 �xc2 'i'xb2+ 25 �d3 Mb3+ 26
�e4 f5+ 27 �f4 'i'd4+ 28 �g5 'i'd8+
29 �h5 'i'e8+ 30 �h4 'i'e7+ 3 1 �h5
g6+ 32 �h6 'i'g7+ 33 �g5 h6+ 34 �f4
g5 mate!

1 4 . . . 0-0 ! 1 5 �e3
This looks incomprehensible - what
is wrong with 15 lLJxc4? The problem
is 15 . . . b5! 16 lLJe5 and now 16 . . . b4+!
The king's exposed placement gives
Black a vital extra tempo for the at­
tack. 17 �xb4 is met by 17 . . . lLJa6+! 1 8
�a3 ( 1 8 '»Yxa6 l'hb8+! 19 �c3 '»Yxe5+
winning) 18 . . . Mab8 19 'i'e3 'i'd6+ 20
�a2 lLJb4+ 21 �b 1 'i'd1+ 22 'i'c1
'i'd4! (threatening . . . '»Yxe5 and
. . . 'i'e4+) 23 'i'e 1 lLJc2!
see follo wing diagram
Thanks to this powerful knight
thrust, Black is now winning by
force. Fasten your seat belts, a rather
long variation lies ahead!
The attempt to turn down the gift
with 17 �c2 is met by 17 . . . b3+ 1 8
�c3 lLJa6! 19 'i'e3 ( 1 9 '»Yxa6 '»Yxe5+ 20
�xb3 Mab8+ wins for Black)
19 . . . Mab8 (intending . . . lLJb4) 19 �xa6
'i'a5+ 20 �d3 '»Yxa6+ 2 1 �e4 with a
crazy position, where anything could
happen (especially to the white king!) .
15 'i'e3 first of all aims to take con­
trol of some dark squares; secondly,
White frees the bishop to develop and
threatens �xc4.
1 5 . . . b 5 ! 1 6 .i1L.e2

40
 Th e N e w M a in L in e : Bla c k fig h ts fo r c o n t r o l o f e 4
Instead 1 6 axb5 cxb5 1 7 �e2 tiJd7!
18 tLlxd7 'li'xd7 19 �f3 l:I.ac8 20 l:I.xa7
b4+! 2 1 �c2 (Hubner points out that
2 1 �xb4 loses to 2 1 . . .l:I.b8+ 22 �a3
'i'd6+ 23 �a2 l:I.b3 24 'li'f2 l:I.fb8)
2 1 . . .'li'b5 gave White nothing in the
stem game Gelfand-Hubner, Munich
1992. Even if White wins both of the
black queenside pawns for his b­
pawn, and manages to swap off the
queens and both sets of rooks, the
resulting ending is likely to be drawn
since White has the wrong-coloured
rook's pawn for his bishop . Hence, all
Black needs to do is to aim for a posi­
tion like
and he will draw Slllce the best
White can achieve is
which is stalemate! The game con­
tinued 22 l:I.d1 b3+ 23 �b1 g6 24 g4
(taking f5 from the black queen)
24 . . . 'li'b8 25 l:I.a4, when 25 . . . 'li'b5 26
l:I.a7 'li'b8 would have led to a draw by
repetition according to Hubner, while
25 . . . 'li'xh2!? 26 l:I.c 1 c3 27 bxc3 l:I.fd8
led to great complications.
1 6 . . . CLld7 1 7 CLlxd 7 'iVxd 7 1 8 'iVc5
A new idea. Piket had earlier
played 18 �f3 against Kramnik, but
Black's strategy is similar in both
cases.
1 8 . a6! 1 9 Uhd 1 Yz - Yz
. .
Exciting stuff, this grandmaster
chess! 18 . . . a6! was in fact a strong new
idea and showed (I think) good under­
standing of the position. Black has
41
Th e Sla v

four pawns for the piece, which is The more active alternative,
ample. However, his pieces are pas­ 16 . . .'!iJcS, is considered in Game 22.
sive and he has to find a way to acti­
vate them. White is strong on the
light squares (he has a light-squared
bishop) , but weak on the dark
squares, so I have to put my queen in
contact with some dark squares. The
c7-square is the obvious spot since
from there the queen eyes as, eS , f4
and the h2-pawn. However, I obvi­
ously couldn't play 1 8 . . :iWc7, as 19
axbs would win for White. Therefore
1 8 . . . a6! seemed logical, and after a lit­
tle calculation I saw that it was indeed 1 7 1le2
the best move. For example, after 17 . . . 'i'xg2 wins a pawn, but after 1 8
19 . . . 'i'c7, 20 Md6? Mad8 ! 21 Mxc6 'i'f4! Mhg1 'i'xh2 19 Mxg7 h e has problems
is extremely worrying for White. defending his second rank.
1 7 . . . �a8 1 8 g4!
The next game is intended as a cau­
tionary tale for black players, and I
hope that my opponent will forgive
me for using it in this way. Cynics
may point out that I am demonstrat­
ing one of my rare wins from a cata­
strophic British Championship!

Game 1 7
Sadler-Ferguson
British Championship 1996

1 d4 d 5 2 c4 c6 3 lZlf3 lZlf6 4 lZlc3 Black has played rather slowly, put­

dxc4 5 a4 llf5 6 lZle5 e6 7 f3 ll b4
8 e4 llxe4 9 fxe4 lZlxe4 1 0 ll d 2
�xd4 11 lZlxe4 �xe4+ 12 �e2
llxd 2+ 1 3 �xd 2 �d5+ 1 4 �c2 lZla6
Black cannot hold on to the c4-
pawn with 14 . . . bS as 15 Md1 'i'cs 1 6
axbS axbs 1 7 'i'f3 ! , attacking f7 and
the rook on a8 , is just one way of ex­
ploiting Black's mistake.
1 5 lZlxc4 0-0-0 1 6 �e3 �b8
ting his king to safety in the corner,
so White begins to take control. This
nice move takes away the annoying
check on fS from the black queen.
1 8 . . . �d 7
A novelty. Kramnik-Kir.Georgiev,
Moscow Olympiad 1994, had contin­
ued 18 . . .f6 19 Mhd1 'i'g2 20 Mxd8
Mxd8 2 1 'i'xe6 with a clear advantage
for White.

42
 Th e N e w M a in L in e : Bla c k fig h ts fo r c o n t r o l o f e 4
1 9 Bad 1 tLJ b4+ 2 0 � b 1 tLJ d 5
Black has got his knight to a central
outpost on dS , but it is hard to suggest
another active thing for him to do.
2 1 �a3 �c7 22 a5 �b8 23 a 6 !
Softening up the black queenside.
23 . . . b6 24 �f3 .l::!. h e8 2 5 .l::!. h e 1 f6 26
�b3 �a8 2 7 h4! b5 28 tLJe3 �a5
A desperate attempt to break out,
but one that is easily refuted.
29 tZlxd 5 exd 5 30 nxe8 nxe8 3 1
!:!'xd 5 !
31 . . . �b8
3 1 . . .exdS is met by 32 'iYxds + and
mates.
32 nd6 �xa6 33 �xc6 Be 1 + 34
It>c2 �c7 3 5 n d 7 + �b6 36 �xb5
1 -0
Question 2: What went wrong?
Why did Black lose without seem­
ingly being able to put up any sort of
fight?
Answer: Black knew very little
about this line and was extremely un­
fortunate that this is one of those
variations where knowledge is essen­
tial: the big decisions for Black are
strategical - there is almost no chance
to calculate your way out of trouble.
Black must understand where and
when to exchange queens; where to
aim to put his knight; and how to
arrange his pawns. Without this
knowledge, Black has little chance of
success.
Question 3: How does one acquire
this knowledge?
Answer: One must study games in
this line and draw conclusions from
them.
Game 18
K ra m n i k -Lautier
Linares 1994
1 tLJf3 d 5 2 d4 tLJf6 3 c4 c6 4 tLJc3
dxc4 5 a4 �f5 6 tLJe5 e6 7 f3 � b4
8 e4 �xe4 9 fxe4 tLJxe4 1 0 .i1Ld2
�xd4 11 tLJxe4 �xe4+ 12 �e2
�xd2+ 1 3 �xd2 �d5+ 1 4 �c2 tLJa6
1 5 tLJxc4 0-0-0
Black has castled queenside, and not
kingside. Why? First, castling queen­
side brings a rook immediately to the
open d-file; second, if White ex­
changes queens, the black king is well­
placed to protect the queenside pawns
from a potential attack by the white
knight.
1 6 �e5 f6 1 7 �xd 5
43
Th e Sla v

The more critical 17 'i¥e3 lS dis­

cussed in Games 20 and 2 1 .
1 7 . . . exd 5

Here Black has exchanged queens
on his own terms: he has forced
White to take on ds . We can conclude
that the exchange of queens is only
acceptable to Black if it improves the
black pawn structure. Also, Black
should recapture on ds with the c­
pawn: after . . . e6xdS, Black has just a
4-2 majority on the queenside; after
. . . c6xdS, Black has a pawn chain of
five against just two white kingside
pawns on g2 and h2. Black is more
likely to be able to create passed
pawns and a pawn chain that will re­
strict the white pieces with the latter
rather than the former.
1 8 tZ:la3 tZ:l b4+ 1 9 W d 2 Wd7 20 !;le 1
!;le8 2 1 !;le3 b6 22 .li b 5+ Wd6 23
.lie2 a6
Preventing lLlbs+.
24 h4 !;lhf8 2 5 h 5 f5 26 h6
A typical attacking idea for White.
Although White runs the risk of los­
ing this pawn, as it is now cut off
from the rest of its troops, if White
can get a knight to gS or a bishop to
g8 . . .
26 . . . g 6 27 �h e 1 �e5 28 a5 tZ:la2 ! 29
�xe5 bxe5 30 �a 1 tZ:lb4 3 1 tZ:le2
tZ:le6 3 2 .lixa6 !;la8 3 3 .li b 5 !;lxa5 34
!;lxa5 tZ:lxa5 3 5 tZ:l e 1 tZ:lb7 3 6 tZ:lf3
We7 3 7 tZ:lg5 tZ:ld6 3 8 tZ:lxh7 tZ:lf7 !
Just in time!
39 g4 tZ:lxh6 40 gxf5 gxf5 4 1 We3
e5
Black's pawns are now very dan­
gerous and White must play accu­
rately.
42 .lie6 Wd6 43 .lib7 tZ:lg4+ 44 Wf3
e4+ 45 Wf4 tZ:le5 46 .lia6 e4 47 tZ:lg5
tiJd3+ 48 We3 f4+ 49 Wd4 e3 50
tiJf3 e2 51 .lib7 tiJe5 52 tZ:le 1 tZ:ld3
53 tiJf3 tiJe5 54 tiJe 1 f3 5 5 We3 We5
56 Wf2 Wd4 5 7 tiJxf3+ tiJxf3 58
Wxe2 tZ:le5 59 Wd2 tiJd3 60 .lixd5

44
 Th e N e w M a in L in e : B la c k fig h ts fo r c o n tro l o f e 4

Iiix d5 6 1 b 3 Y:l - Y:l squares, which makes it easy for

So what if White does not swap White to blockade them with his
queens? bishop.

Game 19
Karpov-Hjartarson
Tilburg 1988

1 d4 d 5 2 c4 c6 3 t2Jf3 t2Jf6 4 t2Jc3

dxc4 5 a4 .ltf5 6 t2Je5 e6 7 f3 itb4
8 e4 iLxe4 9 fxe4 t2Jxe4 1 0 itd2
�xd4 11 t2Jxe4 �xe4+ 12 �e2
ltxd2+ 1 3 'kt>xd2 �d 5+ 1 4 'kt>c2 t2Ja6
1 5 tLlxc4 0-0-0 1 6 �e 5 f6 1 7 �e3 ! ?
At the time, this was a new idea.
White is claiming that he has forced 23 . . . �xb3+ 24 �x b3 t2Jxb3 25
his opponent to weaken the pawn 'kt> x b 3 .l:! d 4 26 h4! .!:!.hd8 2 7 itc4
structure around his king. 'kt>c7 28 h5 .!:!.g4 29 h 6 !
1 7 . c5 ! ?
. .

17 . . . �b8 is considered in the next

two games.
1 8 'kt>b3 t2Jb4

It's that plan again! Now the black

Black's plan seems very logical: he
is aiming to put a knight on d4.
19 gc 1 ! t2Jc6 20 'kt>a3 ! t2Jd4 2 1 t2J a 5 !
e5 2 2 �c3 ! b6 23 t2J b 3
This game is still the model for
dealing with . . . c6-c5 and . . . lZJb4.
White weakens Black's light squares
by forcing all his pawns to dark-
kingside pawns are softened up and
White gradually assumes complete
control.
29 . . . l:!.xg2 3 0 hxg7 .l:!.xg7 31 .l:!.cf 1
l:!.d6 32 .l:!h6 e4 33 .l:!hxf6 h 5 34
l:i:6f4 .!:!.d4 3 5 gf7+ .!:!. d 7 36 .l:!.xg7
.l:!.xg7 3 7 .l:!.f4 .!:!.g3+ 38 'kt>c2 .!:!.g2+ 3 9
'kt> c 3 .!:!. g 3 + 40 'kt>d2 It g 4 4 1 l:!.f7+
'kt>d6 42 'kt>e3 a6 1 -0
A really impressive game from
Karpov.

45
Th e Sla v

21 �b3 ctJd5 22 �f3 �f4 23 1;;!; f 1

Game 20 �d4 24 �f2 �xf2 25 llxf2 �c8 !
S h i rov- Ba reev
Biel 1991

1 d4 d 5 2 c4 ctJf6 3 ctJc3 c6 4 ctJf3

dxc4 5 a4 lLf5 6 ctJe5 e 6 7 f3 lL b4
8 e4 lLxe4 9 fxe4 ctJxe4 1 0 lLd2
�xd4 11 ctJxe4 �xe4+ 12 �e2
lLxd2+ 1 3 �xd2 �d5+ 1 4 �c2 ctJa6
1 5 ctJxc4 0-0-0 1 6 �e5 f6 1 7 �e3
�b8 1 8 1Le2 �xg 2 ! ?

The key move, preparing . . . J:rd7 to

contest the second rank.
26 lLg4 f5 27 lL h 5 lld7 28 1;;!; f g2
1;;!; h d8 29 ctJe5 llxg 7 30 llxg 7 ctJf4
31 llxh7 lld 5 ! 32 ctJg6 ctJxg6 3 3
lL x g 6 lld 7 !

Seeking t o reduce White's winning

chances by exchanging as many
pawns as possible. The less greedy
1 8 . . e5 is considered in the next game.
.

1 9 1;;!; h g 1 �xh 2 20 1;;!; x g 7 ctJ b4+

Centralising the knight with tempi.

The ending is equal according to

Bareev.
34 �c4 1;;!; x h 7 35 lLxh 7 � d 7 36 �c5
b6+ 3 7 �c4 a 5 38 lLg6 �d6 39
lLe8 e 5 40 lL h 5 e4 4 1 lLe8 c 5 42
�c3 �e5 43 lLd7 f4 44 lLb5 f3 45
lLa6 Y:z - Y:z

This is probably fine in general for

Black, but not very inspiring. I prefer

46
 Th e N e w M a in L in e : B la c k fig h ts fo r c o n t r o l o f e 4
the more active plan that Black
adopted in the next game.
Came 21
K ra m n i k -Ivanchuk
Linares 1994
1 CLlf3 d 5 2 d4 l2lf6 3 c4 c6 4 l2lc3
dxc4 5 a4 iLf5 6 l2le5 e6 7 f3 iLb4
8 e4 iLxe4 9 fxe4 l2lxe4 1 0 iLd2
�xd4 11 l2lxe4 �xe4+ 12 �e2
�xd 2+ 1 3 �xd2 �d5+ 1 4 �c2 l2la6
1 5 4:Jxc4 0-0-0 1 6 �e5 f6 1 7 �e3
Wb8! 1 8 iLe2 e 5 !
This i s the best set-up for Black:
pawns on f6 and eS, keeping the
queenside pawns where they are,
while transferring pieces to the two
outposts d4 and dS . Sometimes, Black
will transfer the knight to d4 via cS
and e6.
1 9 93 �e6 20 b3 .nd 5 21 �b2 Rhd8
22 :gad 1 l2l b4 23 �c3 �e7 24 l2le3
Sxd 1 25 Rxd 1 .Q.xd 1 26 iLxd 1 9 6 2 7
�d2 �c7 28 l2lc2 a 5 !
An important move, securing the
knight on b4. If White could success­
fully play a4-aS, then the knight could
become vulnerable and Black would
have to be careful.
29 l2le3 �c5 30 l2lc4 b6 3 1 � b 1
Y:, - Y:,
Neither side has an obvious way to
contmue.
It is time to draw a few conclusions:
1. Black should castle queenside.
2. If the queens are to be ex­
changed, Black wants them to be ex­
changed on dS, when he will improve
his pawn structure with . . . c6xds .
3 . Black's ideal outpost for his
knight is on d4 and not ds .
4. Black's best pawn structure in
the middle game is to place his pawns
on a7, b7, c6, eS, f6, g7 and h7, since
this creates two central outposts, d4
and ds .
But this is not the end of the story.
Came 22
K ramni k-Shirov
Dortmund 1996
1 l2lf3 d5 2 d4 c6 3 c4 l2lf6 4 l2lc3
dxc4 5 a4 iLf5 6 l2le5 e6 7 f3 iL b4
8 e4 iLxe4 9 fxe4 l2lxe4 1 0 iLd2
�xd4 11 tLlxe4 �xe4+ 12 �e2
iLxd2+ 1 3 �xd2 �d5+ 1 4 �c2 l2la6
1 5 l2lxc4 0-0-0 1 6 �e3 !
47
Th e S l a v
With hindsight, very obvious! This
is the very latest idea in this line:
White avoids giving Black the extra
tempo .. .f7-f6, forcing Black to look
for another defensive formation.
1 6 . . . eiJc 5 1 7 �e2 �xg2 1 8 ghg 1
�xh2 1 9 I:Ixg7 !
This move was originally thought
to be impossible due to 19 . . . .:cI3, but
20 ':h 1 ! is very strong for White after
20 . . . .:xe3 2 1 ':xh2 or 20 . . :iUxh 1 2 1
ilxcl3 .
1 9 . . . I:Id4? ! ?
An amazing attempt that j ust falls
short. When I started to analyse this
position, I w anted to play 1 9 . . . .:hg8,
aiming to meet 20 ':xf7 with 20 . . . .:g2,
but 2 1 'iYxcs ':xe2+ 22 �b3 ':cI3+
48
�a2 is j ust winning for White: his
king is fact perfectly safe. 20 . . . .:gS ! ?
and 2 0 . . :i'g2 (threatening . . :i'g6+) are
both interesting, but the onus is
clearly on Black to find a reasonable
continuatIOn.
20 �xd4 �xe2+ 21 ct:Jd2 gd8 22
�c5 I:Ixd 2+ 23 �b3 I:Ixb2+ 24 �a3
Amazingly White's king is quite
safe, and now it is j ust a matter of the
material telling in the end.
24 . . . I:Id2 25 I:Ig3 '>iVe4 26 .i::i. b 3 b6 27
�f8+ �b7 28 �xf7 + �a6 29 �f3
�xf3 30 gxf3 �a5 3 1 .i::i. f4 gd3+ 32
�b2 .i::i. h 3 33 ge 1 I:Ih2+ 34 �b3
.i::i. h 3+ 3 5 �c2 I:Ih2+ 36 �d3 I:Ih6 37
I:Ife4 c 5 3 8 I:Ixe6
The black pawns are insufficiently
advanced to cause White any real
problems.
3 8 . . . I:Ih3+ 39 g 1 e 3 I:Ih 1 40 I:I3e4
I:Ih3+ 4 1 �c4 g g 3 42 gh6 a6 43
I:Ixh7 gg5 44 gb7 gh5 45 ge6
I:Ih4+ 46 �d5 g b4 47 I:Ic6 gd4+ 48
�e6 gb4 49 .i::i. b 8 1 -0
White will win easily by attacking
the black pawns from the rear with
his king.
23 Postscript: Since the first edition of
 Th e N e w M a in L in e : Bla c k fig h t s fo r c o n t r o l o f e 4
this book went t o press, there have
been several important games in this
variation, which have focused mainly
on the previously relatively neglected
idea of Black castling kingside
(15 . . . 0-0) instead of queenside. This is
based on the fine idea of
1 5 . . 0-0 1 6 "Vje5 l:!.ab8 !
. 19 . . . 'iWgS 20 'iWg4! 'iWcs 2 1 Ma3 Md4 22
This wonderful idea has two
points: first, Black supports the . . . b7-
bS advance to open up the queenside
against the exposed white king; and
second, Black protects the pawn on
b7 against the typical White manoeu­
vre 'iWxdS followed by ctJaS or ctJd6.
So far, White has even been struggling
in this position!
17 .§Le2
17 Md1 ctJb4 + 18 �el ctJa2 + 19
�c2 ctJb4 + was a draw by repetition
in P.Cramling-Hector, Malmo 1998.
1 7 . ctJb4+ ! 1 8 Wc3 b 5 1 9 l:!.hd 1 f6 !
. .
20 'iVg 3 �e4 21 ctJe3 ctJ d 5+
tLlxd 5 cxd 5
Yes, it really is as bad as it looks! I
don't know how, in the game Kram­
nik-Van Wely, Tilburg 1998, White
managed to hang on!
23 .§Ld3 l:!.fc8+ 24 Wd2 �b4+ 25
'iit> e 2 bxa4 26 'it>f 1 l:!.b6 2 7 �e3 �b3
28 l:!.d2 g 6 2 9 "Vjd4 �b4 3 0 Si.e2
'i'xd4 % - %
A few months after that game,
these two players, Kramnik and Van
Wely, continued their debate of this
variation at Wijk aan Zee 1999 - but
this time with colours reversed! Van
Wely varied from the standard 16
i"eS with
1 6 �f3
The latest idea. The white queen
exercises less influence from f3 than
eS, but on the other hand, White
threatens the immediate Md1 , chasing
the queen from dS.
1 6 . . . �c5 1 7 Si.e2 ctJ b4+ 1 8 Wb1
l:!.ad8 1 9 l:!.c 1 l:!.d4 !
Kramnik's improvement over
'iWhS, when Black had wasted rather a
lot of time in Beliavsky-Shirov, Bel­
grade 1997.
20 l.1a3 l.1fd8 21 g4 �g 5
Black is beautifully mobilised.
22 l:!.b3 a5 23 h4 "Vjg6+ 24 Wa 1
ctJc2+ 25 Wa2 ctJ b4+ 26 Wa 1 ctJc2+
% -%
So the idea of castling kingside
may, after all, be Black's best course
of action.
To finish this section, here are two
games featuring slightly offbeat at­
22 tempts by White.
Game 23
Adianto - K ramnik
London (Intel Grand Prix) 1994
1 d4 d5 2 c4 c6 3 ctJf3 ctJf6 4 ctJc3
dxc4 5 a4 .§Lf5 6 ctJe5 e6 7 f3 .§Lb4
8 Si.g5
Fighting for e4 by pinning the
black knight, so that 9 e4 is now a
threat. Instead 8 ctJxc4 0-0 9 �gS h6
10 �h4 was met by 10 . . . cS! 1 1 dxcS
'iWxd1+ 12 �xd1 (12 Mxd1 �c2! 1 3
Mel �xa4! 14 �xf6 gxf6 1 5 Mal �b3
16 ctJb6 ctJc6 17 ctJxa8 Mxa8 gave
Black excellent compensation for the
exchange in Beliavsky- Bareev, USSR
1986) 12 . . . Md8+ 13 �el ctJc6! 14 e4
�h7 1 5 �f2 ctJd7! in Akopian-Oll,
49
Th e Sla v

New York Open 1994. 1 5 .. .'�Jd7! in- 22 lLle3 0-0 23 !:!'a 1 !:!.b3 24 !:!.xa4
tends . . . tLJxc5, highlighting the weak- SLxe3+ 25 \tg3 z:!.xb2 26 SLf 1 f 5 !
ness on b3, and . . .f7-f5 activating the
light-squared bishop on h7. After 16
�d3 �xc5! 17 �xc5 tLJxc5 18 �c2 f5!
1 9 exf5 tLJd4! Black stood clearly bet­
ter.

White is in big danger, but some­

how he just hangs on.
27 !:!.xe6 f4+ 28 \th3 �e8 29 !:!.xe8
!:!.xe8 30 �c4+ \th7 3 1 � d 5 �e5 32
!:!.c4 !:!.h5 3 3 g3 �e5 34 !:!.c7 !:!.b6 35
8 . . . h6 9 SLh4 c5 1 0 dxc5 �a 5 ! 1 1 !:!.f7 \tg6 36 !:!.xa7 !:!.b8 3 7 !:!.d7 h5
l'i'd4 lLlc6 1 2 lLlxc6 bxc6 1 3 e4 38 �e7 l:ib2 39 �f8 fxg 3 40 !:!.xg7+
itxc 5 ! \tf6 41 !:!.f7+ \tg6 42 !:!.g7+ \tf6
1 3 . . . �g6 14 �f2 i s less good for lh - lh
Black.
1 4 �xc4 �g6 1 5 �a6 �xa6 16 Game 24
SLxa6 !:!. b 8 ! I .Sokolov-Bareev
Leon 1 995

1 d4 d 5 2 c4 c6 3 lLlc3 lLlf6 4 lLlf3

dxc4 5 a4 SLf5 6 tLle5 e6 7 g 3

White already has trouble with his

queenside.
1 7 lLld 1 lLl d 7 ! 1 8 !:!.C 1 � b4+ 1 9 \tf2
lLlc5 20 �e2 lLlxa4 2 1 !:!.xc6 �c5+

50
 Th e N e w M a in L in e : Bla c k fig h ts fo r c o n t ro l o f e 4
White fights for the e4-square in a
different way: by putting the bishop
on g2. This also helps to dissuade
Black from playing the pawn break
... c6-c5, as b7 will be hanging.
7 . �b4 8 -li g 2 �e4!
. .
Forcing White to block the long
diagonal, which will allow Black to
play ... c6-c5 without fear of �xb7.
9 f3 -lig6 1 0 0-0
10 e4 c5 11 �e3 cxd4 12 �xd4
'iVxd4 13 �xd4 ctJc6 14 ctJxc6 bxc6 1 5
0-0-0 0-0-0 i s the theoretical recom­
mendation, but 16 �f1 is more pleas­
ant for White due to his superior
structure and Black's inactive bishop
on g6. 13 . . . ctJfd7!?, instead of
13 ... ctJc6, was my first idea in order to
reactivate the bishop on g6 with .. .f7-
f6 and . . . �f7. However, 14 ctJxc4 f6
15 0-0-0 ctJc6 16 �f2 (intending
... ttJc6) 16 . . . �e7 17 ctJa2! wins the
bishop pair, giving White a small ad­
vantage, as 17 . . . �c5 loses to 1 8
lhd7+! �xd7 1 9 �xc5, winning two
pieces for a rook. In fact, 1 1 .. .�c7! is
stronger: 12 ctJxc4 cxd4 (attacking the
knight on c4) 1 3 �xd4 ctJc6 is fine for
Black and 1 2 0-0 cxd4 13 �xd4 (13
ctJb5 �xe5 14 �f4 �c5 1 5 ctJc7+ �e7
16 ctJxa8 ctJa6! wins for Black)
1 3 . . . ctJc6 is also good.
1 0 . . . c5 1 1 4:Ja2 �a5 1 2 d x c 5
Or 12 ctJxc4 cxd4 13 ctJxa5 �xa5 14
�xd4 ctJc6 1 5 �c4 �b6+ 16 �h 1
ctJa5!, intending . . . ctJb3 .
1 2 . . . �d 5 1 3 �xd 5 exd 5 1 4 QJxg6
hxg6 1 5 l::!. b 1 4:Jbd7 1 6 �e3 l::!. c 8 1 7
b4 cxb3 1 8 4:J c 1 b2 1 9 4:J b 3 �c3 20
.!::!. f d 1 4:J e 5 21 �d4 -lixd4+ 2 2 '!::!' x d4
b6 23 l::!. x b2
Better was 23 cxb6 axb6 24 's'xb2
's'c4 with an equal position according
to Ivan Sokolov.
2 3 . . . 4:Jc4 24 l::!. b 1 bxc5 25 QJxc5 0-0
26 4:Jd3 4:Ja3 27 l::!. b 7 4:Jc2 28 l::!. f4
l::!. f e8 29 � h 3 l::!. c 3
29 . . . ,S,b8! (I.Sokolov) gave chances
for an edge for Black.
30 �f2 g5 3 1 l::!. f 5 g4 3 2 �xg4 4:Je3
3 3 l::!. x f6 4:Jxg4+ 34 fxg4 gxf6 3 5
l::!. x a7 l::!. a 3 36 4:Jf4 l::!. a 2 3 7 a 5 l::!. e 4
38 l::!. a 8+ � h 7 39 l::!. d 8 l::!. x a5 40
4:Jxd 5 l::!. x d 5 41 l::!. x d 5 l::!. x g4 42 �f3
.!::!. a 4 43 g4 �g6 44 h4 '!::!' a 1 45 h 5+
�g7 46 l::!. f 5 l::!. g 1 47 e3 l::!. f 1 + 48
�e4 l::!. g 1 49 l::!. f4 .!::!. g 3 'h - 'h
A tough endgame.
51
Th e Sla v

S u m m a ry

The sidelines do not seem to cause Black any problems, but undoubtedly the
most crucial line at the moment is Kramnik's 14 Wc2 ctJa6 15 ctJxc4 0-0-0 16
'iVe3 . In general, such positions are easier to play for White than for Black.

1 d4 d5 2 c4 c 6 3 lLlf3 lLlf6 4 lLlc3 dxc4 5 a4 �f5 6 lLle5

6 . . . e6 7 f3 (0)
7 g3 game 24
-

7 . . . �b4 8 e4
8 �g5 - game 23
8 . . . �xe4 9 fxe4 lLlxe4 1 0 �d2 �xd4 1 1 lLlxe4 �xe4 1 2 �e2 �xd2+
1 3 'It>xd 2 �d5+ 1 4 'It>c2
14 Wc3 game 1 6 -

1 4 . . . lLl a 6 1 5 lLlxc4 0-0-0 1 6 �e5 (0)

16 'iVe3
16 . . . Wb8 game 1 7
-

1 6 . . . ctJc5 game 22
-

1 6 . . . f6 1 7 �e3
17 'iVxd5 - game 18
1 7 . . . 'It>b8
17 . . . c5 game 1 9
-

1 8 �e2 (0)
1 8 . . . 'iVxg2 - game 20
18 . . . e5 - game 21

7 f3 1 6 �e5 18 �e2

52
 CHA PTER FOUR

The New Main Line:

Black co unterattacks

1 d4 d 5 2 c4 c6 3 ltJf3 ltJf6 4 ltJc3 man-Salov, 199 1 , but had not been

dxc4 5 a4 i.f5 6 ltJe5 tried since, as, in his notes, Khalifman
In this chapter we consider lines in had pointed out a continuation that
which Black does not doggedly fight seemed to equalise for Black.
on for e4, but switches his attention
to the d-pawn, arguing that 6 ctJe5 has
weakened White's control of d4. Re­
cent attention has concentrated on
6 . e6 7 f3 c5! ? 8 e4 cxd4, which is a
. .

specialty of both Boris Gelfand and

Alexei Shirov. In the following games
you will see the wildly different ways
in which they handle this line!

Game 25
Piket-Gelfa n d
Wijk aa n Zee 1996 9 . . . .§Lg6 1 0 i.b5+ ltJfd 7
10 . . . ctJbd7?? simply loses to 1 1 �g5 .
1 d4 d 5 2 c4 c6 3 ltJf3 ltJf6 4 ltJc3 1 1 'i¥xd4 a 6 !
dxc4 5 a4 i.f5 6 ltJe5 e6
In the 1920s and 1930s, 6 . . . ctJbd7 7
ctJxc4 -&rc7 8 g3 e5 (attacking d4) was
popular, but after 9 dxe5 ctJxe5 10
1£4 gd8 11 -&rc 1 �d6 12 ctJxd6+
'ixd6 13 �g2, White stands better.
He has the two bishops and Black
cannot activate his queenside pawn
maJonty.
7 f3 c5 8 e4 cxd4 9 .§Lxc4 ! ?
A n unusual move that was success­
ful in its first appearance in Khalif-

53
Th e Sla v
And this was it. If now 12 iLe2,
then 1 2 . . . CDxeS 13 'ilxeS CDc6 is very
good for Black, so White is forced
into exchanges.
1 2 .liLxd 7 + tLlxd 7 1 3 �xd 7 + �xd 7
1 4 tLlxd 7 �xd 7 1 5 � e 2 .liL b 4 1 6
gd 1 + ctJe7 1 7 .liLf4 g h c 8 1 8 gac 1
%-%
White obviously has more crucial
possibilities. Who better to test the
black position than Garry Kasparov?
Game 26
Kasparov-Shi rov
Dos Hermanas 1 996
1 d4 d 5 2 c4 c6 3 tLlf3 tLlf6 4 tLlc3
dxc4 5 a4 .liLf5 6 tLle5 e6 7 f3 c 5 8
e4 cxd4 9 exf5
The main line and the only real test
of Black's play.
9 . . . .liL b4 ! ?
Black could not take the knight on
c3 , regaining his piece, because after
9 . . . dxc3 10 'ilxdS+ �xdS 1 1 CDxf7+
White wins a rook. However, if we
imagine that it is Black's move after
9 . . . iLb4, then 10 . . . dxc3 is possible be­
cause after 1 1 'ilxdS+ �xdS 12 CDxf7+
54
�e7 13 CDxhS, 13 . . . cxb2 is discovered
check, so Black wins. White can try
10 �f2!?, since after 10 . . . dxc3 1 1
'ilxds+ �xdS 1 2 CDxf7+ �e7 13
CDxhS, 13 . . . cxb2 is no longer discov­
ered check, but Black can exploit the
other exposed piece in White's posi­
tion: the knight on eS. He can play
10 . . . 'ilc7!, threatening both 1 1 . . .'ilxe5
and 1 1 . . .dxc3 , as White can no longer
exchange queens with 'ilxdS+. It
seems that White can stop both these
threats with 1 1 'ilxd4, but Black has
the last laugh after 1 1 . . .iLcs, picking
up the queen. Sadly Black is not com­
pletely winning after 1O . . . 'ilc7, as
White can play 1 1 CDa2, attacking the
bishop on b4, but after 1 1 . . .'ilxeS 12
CDxb4 'ilcS ! Black has powerful com­
pensation for the piece: two pawns
and the exposed white king. I think
that Black is better here. Garry played
the morc natural. . .
1 0 .liLxc4 �d 6 ! ?
An amazing move, adding t o the
confusion by attacking another piece.
1 1 .liLb5+ tLlc6 1 2 tLlc4 �c5?
A serious and, in such a sharp posi­
tion, fatal mistake. Black could simply
have retreated with 12 . . . 'ild7 when,
 Th e Ne w M a in L in e : Bla c k c o u n t e ra t t a c k s
due to the threat o f . . . d4xc3, White
probably has nothing better than to
repeat moves with 13 ctJe5 �d6.
1 3 .1Ld2!
Black cannot regain the piece now,
as 13 . . . dxc3 14 bxc3 �a5 loses a piece
to 15 ctJxa5 . Obviously Shirov did not
miss this move; but I believe that he
overlooked something extremely
cunning later on.
13 0-0 1 4 ctJa2 ..txd2+ 1 5 �xd 2
. . .
CiJ e 7
Black's is threatening ... a7-a6, win­
nin g the bishop, while he can also try
... LZJxf5, intending . . . ctJe3 . White
seems to have problems but . . .
1 6 �b4 !
Forcing the exchange of queens as
16.. .'i'xf5 allows 17 �xe7.
16 . . . �xb4+ 1 7 ctJxb4 a6 1 8 ctJ b 6 !
White does not lose the bishop af­
ter all! The rest is easy for Kasparov.
1 8 axb5 1 9 ctJxa8 �xa8 20 fxe6
. . .
bxa4 21 exf7+ �xf7 22 �d2 �e6
23 J:i.hc 1 �d6 24 b3 b 5 25 bxa4
bxa4 26 .8.c4 ctJf5 27 ctJc2 ctJd7 28
gcxa4 .8.xa4 29 .l:1.xa4 ctJb6 30 ctJxd4
1 -0
A fine game by Kasparov, but not
one that refutes Black's idea. Let us
take another look.
Game 27
Gelfand-Shirov
Dortmund 1996
1 d4 d 5 2 c4 c6 3 ctJc3 ctJf6 4 ctJf3
dxc4 5 a4 ..tf5 6 ctJe5 e6 7 f3 c5 8
e4 cxd4 9 exf5 ..tb4 1 0 ..txc4 �d6
Shirov could not resist trying this
idea a second time, but this time he is
convincingly mauled. I do wonder
why Black has been avoiding
10 . . . dxc3 1 1 �xds+ �xds 12 0-0 (12
ctJxf7+ �e7 1 3 ctJxhs cxb2+) 12 . . . cxb2
13 �xb2 �e7 14 fxe6 fxe6.
White's two bishops, the weak (but
extra) pawn on e6 and the slightly
exposed king on e7 obviously offer
compensation for the pawn, but I
don't see an advantage for White.
Black's bishop on b4 prevents his op­
ponent from playing Me 1 and ganging
up on e6, so 15 ctJd3 is tempting:
15 . . . �d6 is met by 16 Mfe 1 while
1 5 . . . �a5 16 �a3+ is also sub-optimal,
as Jon Speelman would say! However,
15 . . . McS ! , attacking the bishop on c4,
is the best defence: 16 ctJxb4 Mxc4 is
55
Th e Sla v

good for Black and 1 6 iLxe6 �xe6 17 Suddenly it is Black's king that is in
tDxb4 leaves an equal position. Fi­ danger!
nally, 16 iLb3 is met by 16 . . . tDc6, pro­ 1 8 . . . �xh 2 1 9 5lxc6 .l::!. a c8 20 �xf6
tecting the bishop, when 17 tDxb4 �xf4+ 21 �c2 g6 22 .l:!.df 1 �h 2+ 23
tDxb4 18 Mfe 1 Mc6! (the point of .l::!. g 2 �h3 24 fxg6 fxg 6
1S . . . Mc8) 19 iLa3 as ! , intending . . . �f7,
is fine for Black. I feel that the onus is
on White to demonstrate more than
just sufficient play for the pawn.
1 1 5l b 5+ tLlc6 1 2 5lf4 !

25 .i:!xg6+ hxg6 26 �xg 6+ �h8 27

.l::!. h 1 !
Winning the queen. The game is
over.
27 . . . .i:!f2+ 28 �b3 .i::!. b 8+ 29 �a3 1 -0
This prevents the capture 1 2 . . . dxc3
due to 13 tDxc6 cxb2+ 14 tDxb4+! , So it seems as if, despite these two
when, thanks t o the great strength of reverses, Shirov's 9 . . . iLb4 may well be
the discovered check, White wins the just about playable. Let us now take a
whole house! look at the more restrained continua­
1 2 . . . 0-0 1 3 tLlxc6 �xf4 14 �xd4 ! tion 9 . . . ctJc6.
5lxc3+ 1 5 bxc3 �g 5 1 6 f4 �xg 2 1 7
O-O-O ! bxc6 1 8 .i:! hg 1 ! Came 28
l I Iescas-Gelfand
Dos Hermanas 1 996

1 d4 d 5 2 c4 c6 3 tLlf3 tLlf6 4 tLlc3

dxc4 5 a4 5lf5 6 tLle5 e6 7 f3 c 5 8
e4 cxd4 9 exf5 tLlc6

see follo wing diagram

The older and more solid move.

1 0 tLlxc6 bxc6 11 fxe6 fxe6 12
5lxc4

56
 Th e N e w M a in L in e : B la c k c o u n t e ra t t a c k s

gxbS �xbS 20 93 lLa3 2 1 g b 1 .!:!.xb 1

22 <;i>xb 1 lL c 5 23 lLd3 h6 24 c4
lL b4 25 cxd 5 Y2 Y2 -

Game 29
Van der Sterren-Petu rsson
San Bernardino Open 1 992

1 d4 d 5 2 c4 c6 3 ttJc3 ttJf6 4 ttJf3

dxc4 5 a4 lLf5 6 ttJe5 e6 7 f3 c5 S
e4 cxd4 9 exf5 ttJc6 1 0 ttJxc6 bxc6
11 fxe6 fxe6 1 2 lLxc4 dxc3 13
1 2 �e2 i s considered i n Game 30, bxc3
while more adventurous players may A different move-order that should
like to consider the alternative 12 ltJa2 be met by 13 . . . �xd1+ 14 <;t>xd1 <;t>d7,
d3, keeping the piece and hoping to with the same ideas as in Illescas­
unravel later with g2-g3, �g2 and 0-0, Gelfand above.
although Black's counterplay is very 1 3 . . . '>il\Va 5 ? 1 4 '>il\Ve2 ! !
dangerous!
12 . . . dxc3 1 3 '>il\VxdS+ <;i>xdS 1 4 bxc3

This rook sacrifice is a magnificent

White is a bit better in this ending,
since he has a slightly better pawn
structure (fewer pawn islands) and the
two bishops, but Black's pieces are
active.
14 .':tJd 5 1 5 <;i>d2 lL d 6 1 6 <;i>c2 <;i>d 7 !
. .
The king protects both weak
pawns and helps to cover the only
open file on the board: the b-file.
1 7 lLd2 .§.hfS 1 S gab 1 gabS 19
concept!
1 4 . . . �xc3+ 1 5 <;i>f 1 �xa 1 1 6 '>il\Vxe6+
<;i>dS
16 . . . �e7 loses to 17 �xc6+ <;t>f8 1 8
�xa8 ctJe8 19 <;t>e2 �xa4 20 �d5 ctJd6
2 1 �d3 with a crushing attack
(Petursson) .
1 7 <;i>e2 ! !
This quiet move, allowing the rook
to j oin in the attack, justifies White's
brilliant idea.

57
Th e Sla v
1 7 . . : �xa4 1 8 .l:!.d 1 + �xd 1 + 1 9 �xd 1
� e 5 20 �f7 ge8 2 1 �xg 7 ctJd 7 22
iLf7 .l:!.f8 23 iLe6 ctJf6 24 �b7 .l:!.e8
25 �xa8+ � e 7 26 iLf4+ 1 -0
Came 30
Topalov-G elfand
Dos Hermanas 1 996
1 d4 d 5 2 e4 e6 3 ctJf3 ctJf6 4 ctJe3
dxe4 5 a4 iLf5 6 ctJe5 e6 7 f3 e 5 8
e4 exd4 9 exf5 ctJe6 1 0 ctJxe6 bxe6
1 1 fxe6 fxe6 1 2 �e2 ! ?
An interesting novelty that aims for
a slightly different endgame.
1 2 . . . dxe3 1 3 �xe6+ �e7 1 4 iLxe4
�xe6+ 15 �xe6 exb2 16 iLxb2
iL b4+ 1 7 �e2 �e7 1 8 �e4
58
This ending is somewhat more
awkward for Black since his king is a
little more open. Nonetheless, it is
surprising how quickly his position
goes downhill.
1 8 . . . .l:!.hd 8 ? !
Black must try and activate his
knight: ls . .ebds! is stronger, when 19
.
�xg7 .l:IhgS ! regains the g2-pawn. Af­
ter 1 9 .l:Iac 1 , 19 . . . .l:IheS is best, prevent­
ing 20 �xds cxds 2 1 .l:Ic7 + due to
2 1 . . .�d6+, a discovered check that
wins the rook!
1 9 :l:!hd 1 :l:!ab8 20 g3
This places the g-pawn on a pro­
tected square and thus prevents the
black knight from moving. Black
now has serious problems which he is
unable to overcome.
20 . . . h 5 2 1 gd4 �a5 22 �a3+ �e8
23 :l:!ad 1 .l:!.xd4 24 .l:!.xd4 .l:!.b 1 25
�d3 ge 1 + 26 �f2 �f7 27 �e5 :l:!a 1
28 �e4+ �e8 29 �d3 �f7 30 �xa 7
.l:!.a2+ 3 1 �f 1 :l:!a 1 + 32 �g2 ga2+
33 �h3
N ow White is just winning.
33 . . . ga3 34 f4 �e 1 35 �c5 .l:!.e3 36
�e4+ �e8 3 7 � b4 Be 1 3 8 �xe 1
:l:!xe 1 39 a5 Ba 1 40 a6 �e7 4 1 gd2
1 -0
 Th e N e w M a in L in e : Bla c k c o u n t e ra t t a c k s

Both 6 . . . e 6 7 f3 �b4 and 6 . . . e 6 7 f3

c5 are popular counterattacking sys­
tems at all levels of play. However,
Black can also play more slowly, aim­
ing to break out from a cramped posi­
tIOn.

Game 3 1
Kramni k-Short
Novgorod 1 994

tiJf3 d 5 2 d4 tlJf6 3 c4 dxc4 4

tZJc3 c6 5 a4 .iLf5 6 tlJe5 tlJbd7 7
tilxc4 tlJb6 8 tlJe5 a 5
Preventing a4-as-a6, breaking up
the black queenside. 8 . . . e6 is consid­
ered in the next game, while 8 . . .'�Jbd7,
still seeking the exchange of knights,
was crushed by 9 iVb3 ! ctJxe5 10 dxes
CLlg4 1 1 iVxb7 ctJxes 12 f4 ctJg6 13 e4
_�d7 14 fs ctJes 15 �f4 f6 16 �xes
[xeS 17 .s,dl in Kasparov-Timman,
Riga 1995 .
The standard plan of development:
the queen supports e3-e4 from e2,
leaving the rook to cover d4 from d l .
1 4 . . . . tlJfd 7 1 5 tlJd3 ! ?
1 5 ctJxd7 ctJxd7 16 e4 is also slightly
better for White.
1 5 . . . �e7 1 6 e4 e 5 1 7 d 5 gfd 8 1 8
.iLe 3? !
A mistake, allowing Black to
weaken the white queenside. 18 �d2
would have kept an edge according to
Kramnik.
1 8 . . . .iL x c 3 1 9 b x c 3 cxd 5 20 exd 5
tlJc4 ! ?

9 g3!?
A novelty. 9 f3 ctJfd7 10 ctJxd7
CLlxd7 1 1 e4 �g6 12 �e3 e6 13 �c4
�b4 14 0-0 is normal, with a slight I think that Short may have missed
advantage for White. White's next, but this is actually a
9 . . . e6 1 0 .iLg2 .iLb4 1 1 0-0 0-0 1 2 good move!
e3 h6 1 3 �e 2 ! .il.. h 7 1 4 J::i: d 1 ! 21 .iLxh6 gxh6 22 �g4+ �g 5 23

59
Th e Sla v
�xc4 �ac8 24 � b 5 l::t x c3
Black has good counterplay in this
murky position.
2 5 ttJe 1 b 6 26 d 6 e4 2 7 � d 5 �f6 28
J:!ad 1 e3 29 fxe3 �xe3 3 0 ttJd3 �c3
31 ttJf4 1Le4 32 �5d2 �c5 3 3 �xc 5
ttJ x c 5 34 1Lxe4 .l:!.xe4 3 5 ttJ d 5 <Jif8
3 6 ttJxb6 ttJxa4 37 ttJxa4 .6!xa4 38
J:! d 5
White has a tiny edge in the end­
game but he is unable to make any­
thing of it.
38 . . . l:i.d7 39 <Jig2 .l::!. a 2+ 40 'it> h 3 a4
4 1 J:! a 5 a3 42 g4 'it>g7 43 <Ji g 3 <Jig6
44 h 3 <Jig7 4 5 'it>h4 'it>g6 46 .l:!.a8
<Ji h 7 % - %
Game 32
Ruzele-Thorsteins
Lyon (European Club Cup) 1994
1 d4 d 5 2 c4 c6 3 ttJc3 ttJf6 4 ttJf3
dxc4 5 a4 1Lf5 6 ttJe5 ttJ b d 7 7 ttJxc4
ttJb6 8 ttJe5 e6 ! ?
9 f3
A sensible reply. 9 as ttJbds
(threatening . . . ttJb4 or . . . .\tb4) 10 a6!?
is consistent but very risky, while 9 g3
is met by 9 . . . .\tb4 10 .\tg2 ttJe4 when
the natural 1 1 .\td2 loses to 1 1 . . .ttJxf2!
60
1 2 �xf2 ik'xd4+ and . . . ik'xeS.
9 . . . a 5 1 0 e4
10 g4 ttJfdS! 1 1 h4 (1 1 gxfS ik'h4+ 12
�d2 ik'f4+ wins back the piece) l 1 . . .f6
1 2 gxfS fxeS (Ruzele) is extremely un­
clear.
1 0 . . . 1Lg6 1 1 1Le3 1Lb4 1 2 1Le2 O-O ? !
1 2 ... ttJfd7 is better since 13 ttJxg6
hxg6 gives Black play on the h-file,
while 13 ttJxd7 ttJxd7 transposes to
S . . . aS 9 f3 .
1 3 0-0 ttJfd 7
Now White can take the two bish­
ops.
1 4 ttJxg 6 ! hxg6 1 5 <Jih 1 �e7 16
1Lg 1 �fd8 1 7 �b3
White has effortlessly obtained a
wonderful version of the old main
line with 6 e3.
1 7 . . .c5 18 ttJ a 2 ! cxd4 19 ttJxb4
�xb4 20 �xb4 axb4 2 1 a5 d 3 22
1Lxd3 ttJe5 23 1Lb5 ttJbc4 24 f4 ttJd6
25 1Le2 ttJc6 26 1Lb6 .l:!.e8 27 e5
ttJc8 28 1Lf3 ttJxb6 29 axb6 ttJe7 30
1Lxb7 �ab8 3 1 �a7 ttJc8 3 2 1Lxc8
�exc8 3 3 �c7 nxc7 34 bxc7 .l:!.c8
35 .l:!.c 1 <Jif8 36 <Jig 1 <Jie7 3 7 <Jif2
<Jid7 38 <Jie3 �xc7 39 l::t x c7+ <Jixc7
40 <Jid4 <Jib6 41 <Jic4 1 -0
Model strategy by White.
 Th e Ne w M a in L in e : Bla c k c o u n t e ra t t a c k s

S u m m a ry

6 . . . e6 7 f3 cS should definitely be studied by Black players. I particularly like

Shirov's handling of the line with 8 e4 cxd4 9 exfS �b4. 9 . . .':tJc6 is for the
calmer players amongst you who don't mind taking on a slightly worse end­
ing. Probably it is important to choose the right opponent: 9 . . .':tJc6 will be
ideal against an impatient attacking player, while 9 . . �b4 would unsettle a
.

more positionally inclined player.

If you prefer the more solid 6 . . . ctJbd7 7 ctJxc4 ctJb6, and don't mind the
slightly cramped positions that arise from this line, then Thorsteins's 8 . . . e6
looks like a good move-order, since it avoids Kramnik's 9 g3 .

1 d4 d 5 2 c4 c6 3 ctJf3 ctJf6 4 ctJc3 dxc4 5 a4 iLf5 6 ctJe5

6 . . . e6
6 . . . ctJbd7 7 ctJxc4 ctJb6 8 ctJeS (D)
8 . . . aS - game 3 1
8 ... e6 game 32
-

7 f3 c 5 8 e4 cxd4 9 exf5 (D)

9 �xc4 game 25
-

9 . . . ctJc6
9 . . . �b4 10 �xc4 'iVd6 1 1 �bs + ctJc6
12 ctJc4 game 26
-

12 �f4 game 27
-

10 ctJxc6 bxc6 1 1 fxe6 fxe6 (D) 1 2 iLxc4

12 'iVe2 game 30
-

12 . . . dxc3
13 'iVe2 game 28
-

13 bxc3 game 29
-

8 ctJe5 9 exf5 l ' . . . fxe 6

61
 CHA PTER FIVE

The S m yslov Variation

1 d4 d 5 2 c4 c6 3 ttJf3 ttJf6 4 ttJc3 Answer: Smyslov's style as Black is

dxc4 5 a4 ttJa6 perfectly reflected in this system: he is
The move 5 . . . ctJa6 was originally an prepared to accept a slight space dis­
idea of Emanuel Lasker; and it was advantage and will just place his pieces
later taken up and played with success on good squares where they coordi­
by another World Champion Vassily nate well with each other. Since
Smyslov. Recently Ivanchuk and Black's position is very solid, the op­
Short have also used it to good effect. ponent will not be able to launch a
Question 1: 5 . . ctJa6 looks a little
. sudden attack and Smyslov will pa­
strange. What does it do? tiently unravel, gaining space little by
Answer: 5 a4 (preventing . . . b7-b5) little until he frees himself.
has weakened the b4-square. In the
main lines with 5 . . . itf5 6 e3, Black Game 33
puts a bishop on b4 and develops his Ivanchuk -Smyslov
queen's knight to d7; here, Black Tallinn (rapidplay) 1 996
leaves his bishop on e7 to keep b4 free
for the knight. Black's light-squared 1 d4 d 5 2 ttJf3 ttJf6 3 c4 c6 4 ttJc3
bishop will now go to g4 to put pres­ dxc4 5 a4 ttJa6 6 e4
sure on the d4-square. Very straightforward play.
Question 2: So what difference does
this make?
Answer: In the 6 e3 lines, Black's
bishops on f5 and b4 combine to pre­
vent White from easily achieving e3-
e4. In this line, Black exerts virtually
no pressure on e4, and very little on
d4, which means that White pretty
much has the centre to himself.
Question 3: Well that doesn't sound
very promising for Black, does it?
What am I supposed to do as Black?

62

6 . . . i!.. g 4 7 i!.. x c4 i!.. x f3
7 . . . e6 is dealt with in the next game.
8 gxf3 e6 9 i!.. x a6 bxa6
Black has no problems here. First,
he has exchanged two sets of minor
pieces, and such exchanges always
help the player with less space, since it
means that there are fewer pieces in a
confined area; second, Black's doubled
a-pawns give Black the b-file on which
to activate his major pieces and attack
the vulnerable white queenside. Black
would be much less active if his a6-
pawn were on b7! Certainly Ivanchuk
is happy to exchange queens and es­
cape with a draw.
1 0 \We2 a 5 1 1 \Wc4 Mc8 1 2 Mg 1 g6
1 3 r;t>f 1 i!.. g 7 1 4 �c5 �b6 1 5 �xb6
axb6 1 6 i!.. e 3 0-0 1 7 MC 1 tiJ d 7 Y:z - Y:z
If Black delays exchanging on f3 , a
more complex situation arises, as we
shall see in the next game.
Game 34
N ovikov-Greta rsson
Berlin Open 1995
1 d4 d5 2 c4 c6 3 tiJf3 tiJf6 4 tiJc3
dxc4 5 a4 tiJa6 6 e4 i!.. g 4 7 i!.. x c4 e6
Th e S m y s / o v Va ria tio n
A different approach which leads to
a complicated middlegame.
8 i!.. e 3 tiJb4 9 a 5
A typical idea from White, aiming
to prevent either . . :iVaS, activating the
black queen, or . . . a7-aS, cementing the
knight on b4. Black has to be a little
careful that this knight, protected
only by the bishop on e7, does not get
cut off from the rest of his army.
9 . . . i!.. e 7
9 . . . �xf3 , forcing 10 gxf3 (10 'iVxf3
ctJc2+) , was still possible but Black
prefers natural development.
1 0 i!.. e 2 0-0 1 1 0-0 b 5 !
This far from obvious move is the
black plan for survival in these mid-
dlegames.
63
Th e Sla v
Question 4: What is the pomt of
this move?
A nswer: The main idea is that Black
gains just a little more space for his
pieces - remember what I said about
Smyslov patiently improving his posi­
tion, taking extra territory little by
little. The other point is that Black
would like to strike at the white cen­
tre with . . . c6-c5, but first he needs a
reasonable square for his queen: he
can't put it on the c- or d-files, since
after . . . c5xd4 these files will be opened
and the queen will be in the firing line
of white rooks on d1 and c 1 . 1 1 . . .b5
frees b7 for the black queen, where it
is absolutely safe. A nice bonus is that
after . . . c6-c5, the black queen will join
with the knight on f6 in attacking the
e4-pawn.
1 2 �b3 �c7 1 3 Rfc 1 �b7 1 4 � g 5
R f d 8 1 5 .i1Lxf6 gxf6
A forced recapture as 15 . . . 1lLxf6
loses a piece to 16 iVxb4. The weaken­
ing to Black's kingside is not too seri­
ous, however, since White has no
pieces in that area. Moreover, without
his dark-squared bishop, White loses a
lot of control over the central dark
squares, which means that d4 is
64
weaker and hence it is easier for Black
to achieve the . . . c6-c5 break.
1 6 ttJe 1 1£. x e 2 1 7 ttJxe2 J:t a c 8 1 8
z::!. c 3 c 5 ! 1 9 dxc5 ttJa6 ! !
A very neat idea. 19 . . . 1lLxc5 would
have lost to 20 's'xc5 's'xc5 2 1 iVxb4.
20 ttJd3 ttJxc5 2 1 ttJxc5 axc5 22
axc5 �xc5 23 �c3 �xe4 24 �xf6
�d 5 25 ttJc3 �d4 26 �f3 �e 5
Black's control of the dark squares
gives him good chances.
27 g3 b 4 2 8 Rd 1 �xd 1 + 2 9 ttJxd 1
�d4 30 �d 3 �g7 3 1 b3 h6 32 a6
�c5 33 �f 1 �f6 %-%
This game is a model illustration of
Black's middlegame strategy in this
vanatlOn.
N ow we move on to look at 6 e3.
Game 35
Benz-Gretarsson
Oberwart Open 1996
1 d4 d5 2 ttJf3 ttJf6 3 c4 c6 4 ttJc3
dxc4 5 a4 ttJa6 6 e3
The most solid option and proba­
bly the best move. White does not
give Black the chance to double his f­
pawns with . . . 1lLxf3 .

6 . . . 1Lg4 7 1Lxc4 e6 8 h 3 1L h 5 9 0-0
iLlb4 1 0 �e2 !
A typical manoeuvre in queen's
pawn openings: the queen moves to
e2, supporting the e4 push, while the
rook is played to d 1 , supporting the
d4-pawn and discouraging . . . c6-cS due
to the opposition of the rook to the
black queen on d8 .
1 0 . . . 1Le7 1 1 J:::!. d 1 0-0 1 2 g4 1L g 6 1 3
e4 c5 ! ?
An unusually active move at this
stage of the Smyslov variation, but
Helgi Gretarsson, a fanatic of this
variation, has an interesting idea in
mind. The more restrained 13 . . .':tJd7 is
considered in the next two games.
14 d 5 ! ?
Th e S m y s l o v Va ria tio n
A very sharp reply, but I would be
intrigued to discover what Helgi had
in mind against the ECO recommen­
dation of 14 iLf4! (taking c7 away
from the queen) 14 . . . 'iYaS 15 ctJd2!,
intending ctJb3 to harass the queen
some more. 1s . . . cxd4 16 ctJb3 'iYb6 1 7
a s i s not nice for Black and the reck­
less lS . . . ctJc2 loses a piece to 16 ctJb3
'iYb4 17 ctJa2 'iYxa4 18 ctJc3 ! (18 'iYxc2
iLxe4 causes some problems) 18 . . . 'iYb4
19 Ra4 'iYb6 20 'iYxc2.
1 4 . . . exd 5 ! 1 5 e 5 d4 ! ? 1 6 exf6 1Lxf6
1 7 1Lf4
Question 5: What is going on?
A nswer: As compensation for the
piece, Black has two pawns, a strong
centre and a tempo on the queen with
. . . Re8 . Unfortunately, I don't think
that this is quite enough; and this is
almost entirely due to the bad placing
of the bishop on f6, which takes away
a brilliant square for the black queen.
Perhaps Black could try 1 S . . . Re8 ! ? to
meet 16 exf6 (16 iLbs ctJc6) with ei­
ther 16 . . . iLd6 or 16 . . . iLf8 , intending
to recapture on f6 with the queen,
though I would be the first to admit
that it all looks a bit speculative!
1 7 . . . J:::!. e 8 1 8 �f 1 a6
65
Th e Sla v

To prevent LLlb5 . 8 h 3 iL h 5 9 0-0 ctJb4 1 0 �e2 iLe7

1 1 tId 1 0-0 1 2 g4 iLg6 1 3 e4 ctJd7 ! ?

1 9 .a.d2? !
White starts to go wrong around
here and drifts very quickly into a lost
position. 19 LLla2! LLld5 (19 . . . LLlc6 20
�e l ! beginning to exchange pieces) 20
ilg3 , intending �e l , would have
given White the better chances.
1 9 . . . '>&d 7 20 �g 2? '>&c6 21 ctJb5
axb5 22 iL x b 5 '>&e4 23 iLxe8 '>&xf4
24 .a.e 1 h6 25 iL b 5 iLe4 26 .a.xe4
�xe4 2 7 iLc4 �f4 28 b3 .)de8 29
�g 3 '>&xg 3+ 3 0 fxg 3 ctJc6 31 iLd5
ctJa5 3 2 b4 .a.d8 3 3 iLe4 cxb4 34
ctJe 1 b3 3 5 iLd3 iL g 5 36 tIb2 tIc8
3 7 �f 1 g 6 3 8 �e2 iLc 1 39 tIb 1 b2
40 ctJf3 ctJb3 0 - 1
Black anticipates the threat of llJeS
and f2-f4, intending f4-f5 to trap the
bishop on g6.
1 4 ctJe5 ! ?
A very double-edged decision.
White allows his central pawns to be
doubled, but also frees the f-pawn to
advance.
1 4 . . . ctJxe 5 1 5 dxe5 �a5 1 6 f4 .l::!. a d8
1 7 iLe3 h 6 !

The main line for Black i s consid­

ered in the next game, probably the
finest blindfold game ever played. I
wish I could play this well in normal
chess!

Game 36 Making an escape square for the

K ra m n i k-Ivanchu k bishop.
Monte Carlo (blindfold) 1 996 1 8 �g2 iL h 7 1 9 .a.xd8 .a.xd8 20 .a.d 1
g5!
1 ctJf3 d 5 2 d4 ctJf6 3 c4 c6 4 ctJc3 2 0 . . . a 6 i s considered i n the next
dxc4 5 a4 ctJa6 6 e3 iLg4 7 iLxc4 e6 game.

66

2 1 :i::!. x d8+ �xd8 22 �d 2 gxf4 23
ixf4 iL b 6 24 iL b 5 ! !
24 �xh6 allows 24 . . . iVxeS, so
White sacrifices a bishop to keep the
queen boxed in on as .
24 . . . cxb5 25 iLxh6 � c 5 26 �d7
ig6 2 7 �c8+ �h7 28 iLg5
With his threat of �f6 and iVhs
mate, White just seems to be winning,
but now it is Black's turn to sacrifice a
piece, this time to free his queen.
28 . . . tLl d 5 ! ! 29 �xc5
Not 29 exdS when 29 . . iVb4! . Kramnik obviously believes in this
launches a powerful counterattack.
29 . . . tLlxc3 30 �xc3 �xa4 3 1 �g3
�xe4 3 2 iLf6 b4 3 3 �c8 �e 1 + 34
�f4 �f2+ 35 � g 5 �d2+ 36 �h4
�h6+ 37 �g3 �e3+ Yz - Yz
Th e S m y s l o v Va ria tio n
Black must give perpetual check
due to the threat of mate on hS.
Game 37
Kramni k-Short
Moscow {Intel Grand Prix} 1996
1 tLlf3 d 5 2 d4 c6 3 c4 tLlf6 4 tLlc3
dxc4 5 a4 tLla6 6 e3 iLg4 7 iLxc4 e6
8 h 3 iLh5 9 0-0 tLl b4 1 0 �e2 �e7
1 1 :i::!. d 1 0-0 1 2 g4 iLg6 1 3 e4 tLld7
1 4 tLle5
continuation for White, but it seems a
little hasty to me. Since Black is
threatening little in the centre, a sen­
sible move like 14 �f4, taking c7
away from the black queen and
67
Th e Sla v

intending perhaps h3-h4-hS, makes 'iVxe 5 30 'iVb6 ctJd4 3 1 .1Ld 1 .1L g 8 !

more appeal to me.
1 4 . . . ctJxe5 1 5 dxe5 �a5 1 6 f4 h 6
1 7 �g2 J::r a d8 1 8 .1Le3 .1L h 7 1 9 J::r x d8
J::r x d8 20 t1.d 1 a 6 !
The new idea, activating Black's
queenside majority.

3 2 .i.f3 f6 ! 3 3 ctJ e 2 ctJxe2 34 .i.xe2

�xe4+ 3 5 .i.f3 �c2+ 36 �f2 'iVxf2+
37 �xf2 a5 38 fxe6 .i.xe6 !

2 1 .1L b 3 �h8 !
So that the bishop on h7 can reacti­
vate itself by means of . . . iLgS , . . .f7-f6
and . . . iLf7! This position would not
be to everyone's taste, but Short wins
a mce game.
22 J::r d 2 b 5 23 axb5 cxb5 24 f5
ctJc6!

Yes it's free! White will not be able

to cope with the two potential outside
passed pawns.
39 �e3 b4 40 .1Ld 1 f5 4 1 �f4 fxg4
42 hxg4 �g8 43 �e5 �f7 44 �d6
�f6 45 b3 g5 46 �c5 �e5 4 7 �b5
�d4 48 �xa5 �c3 49 �a4 .1L d 5 0 - 1

Kramnik's approach with CDeS is

Stressing the new weakness on eS.
25 J::r x d8+ � x d 8 26 .i.f4 .i. c 5 2 7
.i.e3 '�b6 28 .1Lxc5 ·�xc 5 29 'iVf2
rather impatient. A quieter method is
demonstrated in the following game,
the last of a match between France's
13-year-old star Etienne Bacrot and
ex-World Champion Vassily Smyslov,

68
 Th e S m y s / o v Va ria tio n

whose name this variation bears. The
match score was a rather crushing 5- 1 ,
and one player was made t o look
vastly inferior in the endgame. But
not the player one might have ex­
pected!
Game 38
Bacrot-S myslov
Albert (sixth match game) 1 996
f4-fs or h3-h4-hs with a clear advan­
tage. Black must play . . . ctJd7 to pre­
vent ctJeS, either before or after . . . b7-
bS, with a typicaI S . . . ctJa6 position.
1 2 . . J�� c 8 ? ! 1 3 .:iLb3 c 5 ?
This is excessively active from
Black at this early stage.
1 4 CL:l b 5 !

1 d4 d 5 2 c4 c6 3 CL:lc3 CL:lf6 4 CL:lf3

dxc4 5 a4 CL:la6 6 e3 .:iLg4 7 .:iLxc4 e6
8 0-0 .:iLe7 9 �e2 CL:lb4 1 0 J::i. d 1 0-0
1 1 h3 .:iLh5 1 2 a 5 ! ?

This fine move threatens ctJxa7 and

d4xcS followed by ctJd6.
1 4 . . . .:iLxf3 1 5 gxf3 a6 1 6 dxc5 CL:lbd5
1 7 CL:ld6 .:iLxd6 1 8 cxd6 �xd 6 1 9
ga4 .:tIc5 20 f4 �c6 2 1 .:iLd2 �b5 22
�xb 5 l::!. x b5

We have already seen White's idea

in Novikov-Gretarsson. Our first
thought should be therefore to im­
plement the plan of . . . b7-bS, followed
by the transfer of the queen to b7:
12 bs 13 �b3 "Wic7 14 e4 (14 g4 �g6
. . .

15 ctJeS! ? , intending a quick h3-h4-hS

trapping the bishop) 14 . . . "Wib7. It is
obvious that White is better prepared
for his opponent's plan than in the
above game. The rook covers d4 from
d 1 , while the queen on e2 both pro­ The ending is very nasty for Black,
tects e4 and attacks bS, making . . . c6-cS since his knights have no outposts.
more difficult to achieve. After 15 g4 23 ga3 CL:le4 24 .:iLe 1 CL:ldf6 25 J::i. c 1
itg6 16 ttJeS White intends either f2- &i.d8 2 6 f 3 CL:ld6 2 7 .:tId 1 CL:lfe8 28

69
Th e Sla v

�a4 gd 5 29 gxd 5 exd 5 30 gd3 37 �f2 CiJxc7 38 gb3 CiJe6 39

CiJc7 31 � b4 CiJ d b 5 3 2 � c 5 f5 3 3 .a.xb7+ �f8 40 .a.b8+ �e7 41 .a.b7+
�b6 gd6 34 e4 ! .a.c7 42 .a.xc7+ CiJxc7 43 �c6 d4 44
b4 1 -0

This allows White to make a mas­

sive stride forwards with his e-pawn as A fine vlCtOry for the young
34 . . . dxe4 loses a piece to 35 Mxd6. Frenchman, crowning an amazing
34 . . . �f7 3 5 e5 .a.c6 36 �xc7 .a.c 1 + match result.

70
 Th e Sm y s / o v Va ria tio n

S u m m a ry

Theoretically, S . . . ctJa6 is doing well for Black and if a system has been played
by Smyslov, Ivanchuk and Short then it must have some merit! If you don't
mind playing slightly cramped positions, then it could be the system for you.
6 e3 is the most critical test.

1 d4 d 5 2 c4 c6 3 ctJf3 ctJf6 4 ctJc3 dxc4 5 a4 ctJa6

6 e4 (D)
6 e3 iLg4 7 iLxc4 e6 8 h3 iLh5 9 0-0 ctJb4 10 'Wie2 iLe7 1 1 Md1 0-0
12 g4 iLg6 13 e4 (D)
1 3 . . . cS game 35
-

13 . . . ctJd7 14 ctJeS ctJxe5 15 dxeS 'Wia5 16 f4 Mad8 17 iLe3 h6

18 �g2 iLh7 19 Mxd8 Mxd8 20 Md1
20 . . . g5 game 36
-

20 . . . a6 game 37
-

12 as game 38
-

6 . . . ilLg4 7 ilLxc4 (D)

7 iLxf3 - game 33
. . .

7 . . . e6 game 34
-

6 e4 1 3 e4 7 ilLxc4

71
 CHA PTER SIX

The Bronstein Variation

1 d4 d 5 2 c4 c6 3 tLlf3 tLlf6 4 tLlc3 the kingside with g2-g4 and develop

dxc4 5 a4 jLg4
Like Smyslov's 5 .. .'�Ja6, 5 . . . iLg4 is a
variation that has been around for a
long time without ever gaining wide­
spread popularity.
Question 1 : What is the point of
5 . . . iLg4?
A nswer: 5 . . . iLg4 looks to delay e3-
e4 by putting pressure on the d4-
pawn. White should avoid 6 e4 as af­
ter 6 . . . e6 7 iLxc4 iLb4 (threatening
. . . ctJxe4) , he has problems holding his
centre. Therefore 6 ctJe5, gaining a
tempo on the bishop, is almost always
played. Now after 6 . . . iLh5 White still
cannot play 7 e4, as this would allow
7 . iLxd l !
. .
the bishop to g2. This variation is so
complicated, however, that a whole
book would be needed to explain its
ramifications! I will do my best, but
I'm afraid you'll only get a brief
taster!
For the less savage, there is also the
quiet 7 g3 , planning a fianchetto.
Game 39
Kram nik-Damljanovic
Moscow Olympiad 1994
1 d4 d 5 2 tLlf3 tLlf6 3 c4 c6 4 tLlc3
dxc4 5 a4 jLg4 6 tLle5 jLh5 7 f3
tLlfd7

Question 2 : What plans does White

have?
A nswer: Since White can take the
c4-pawn at his leisure, there is no need
for him to hurry with ctJxc4. His two
most dangerous plans both aim to
exploit the slightly precarious posi­
tion of the bishop on h5:
a) 7 f3 , which blocks the h5-dl di­
agonal and threatens to achieve e2-e4,
while supporting g2-g4 iLg6, h2-h4,
intending to trap the bishop. Question 3: This is a strange­
b) 7 h3, intending to gain space on looking move. What does it do?

72

A nswer: Black must challenge the
knight on eS or his bishop will be
caught by g2-g4 and h2-h4. Now after
8 g4, 8 . . . �g6 9 h4 4JxeS ! 10 dxe5
'i'xdl+ 1 1 �xdl hs is fine for Black.
However, another point is to give
Black an unexpected opportunity to
continue his pressure against the d4-
pawn.
8 ttJxc4 e 5 !
And this i s it! The knight's move
from f6 has freed the h4-d8 diagonal
for the queen, giving Black the oppor­
tunity to exploit the slight weakening
on the e l-h4 diagonal created by 7 £3 .
9 ttJe4
The old move and a very sensible
one. White cannot play 9 dxeS as
9 ... iVh4+! wins the knight on c4,
while 9 4JxeS 4Jxe5 10 dxeS 4J d7 1 1
f4 JLb4 is a very risky pawn grab.
Black has a substantial lead in devel­
opment and will follow up with
. . . 'i'e7 and . . .f7-f6 or . . . g7-gS, opening
further lines. 9 g3 is considered in
Games 4 1 and 42.
Question 4: What does 9 4J e4 do?
Answer: Black's pieces are a little
strange at the moment: his king's
knight is on d7, which is the queen's
Th e B r o n s t e in Va ria tio n
knight's natural square, while his
bishop is biting granite on hs. 9 4J e4
seeks to exploit this temporary confu­
sion by threatening a devastating
check on d6, and also shields the
knight on c4 along the fourth rank; so
10 dxeS is now a threat as . . . iVh4+ will
no longer achieve anything.
9 . . . SL b4+ 10 SLd2 Wife7 11 SLxb4
Wifxb4+ 1 2 Wifd2 Wifxd2+
Forced, as 12 . . .iVxc4 loses to 1 3
4J d6+.
1 3 �xd 2 exd4 1 4 iLled6+ �e 7
The alternative, the enterpnslllg
14 . . . �d8, is considered in the next
game.
1 5 iLlf5+ �f6 1 6 iLlxd4 J:;!d8
We have reached an ending, but the
tactical complications continue for
some while yet.
1 7 �c3 iLlc5 1 8 e4 ! B.xd4 1 9 �xd4
iLlb3+ 20 �c3 iLlxa 1 21 .iLe2 �e 7
22 J::i. x a 1 iLl d 7 23 b4 !
Question 5: Isn't this just a boring,
equal ending?
A nswer: Unfortunately for Black,
no. White does enjoy a definite edge
here, and it all boils down to that
wayward bishop on hs. First, Black is
going to have to spend a tempo with
73
Th e Sla v

. . . f7-f6 to bring it back into play; and stopping the knight from activating
second, if it was still on cS , Black via eS . Black obviously felt very un­
wouldn't have such an annoying comfortable round here, since he
weakness on b7! The white knight is starts just moving his knight around
excellently placed on c4, as it can at­ for no reason.
tack b7 via d6 (with the help of a rook
on dl) or as .
Question 6: You mean Black is
lost?!
A nswer: No, not at all. He only has
one real weakness, so he should be
able to defend, but it isn't really that
much fun.

31 . . . iLlb6 3 2 iLl c 5 iLld7 3 3 iLlb3 96

34 iLld4 iLlb6 3 5 �c4 J:!.b7 3 6 �b3
�d7 3 7 ga5 !

23 . . . f6 2 4 gd 1 iLl b 6
Kramnik suggests that 24 ... Mbs IS

more solid.
2 5 iLl a 5 !
Exchanging knights would greatly
simplify Black's defensive task. Now
White forces unpleasant weaknesses
in the black queenside.
25 . . . iLlxa4+ 26 �b3 iLl b 6 27 iLlxb7
�f7+ 28 �c3
The trade of the a4-pawn for the
b7-pawn has been profitable for
White, as now he has two targets: a7
and c6.
28 . . . gb8 29 �a6 �e8 30 J:!.a 1 iLld7
31 f4 !
Gaining space on the kingside and
Threatening a breakthrough with
e4-eS.
3 7 . . . iLla8? 38 3L a 4 J:!.c 7 39 lIc5 iLlb6
1 -0
And Black lost on time in this
hopeless position: he is just going to
lose his c-pawn.
Let us now take a look at 14 . . . �d8
instead of 14 . . . �e7.

74
 Th e B r o n s t e in Va ria tio n

Game 40
Schandorff-Hellsten
Copenhagen 1996

1 d4 d 5 2 c4 c6 3 ctJf3 ctJf6 4 ctJc3

dxc4 5 a4 �g4 6 ctJe5 � h 5 7 f3
IiJfd 7 8 ctJxc4 e 5 9 ctJe4 � b4+ 1 0
�d2 �e7 1 1 � x b4 �x b4+ 1 2 �d 2
�xd2+ 1 3 '.t>xd2 exd4 1 4 ctJed6+
�d8 ! ?

1 9 '.t>c3 �he8 ! 20 ctJac4 .!;l;ab8 2 1

� d 3 ? ctJxd3 2 2 '.t>xd3 ctJc5+ 2 3 '.t>e2
f5! 24 '.t>f 1 f4 2 5 ctJd 1 ctJb3 0 - 1
An amazingly quick defeat for
White!

Game 41
Epishin-Pomes
Manresa 1995

1 d4 d 5 2 c4 c6 3 ctJf3 ctJf6 4 ctJc3

An interesting idea: in the ending dxc4 5 a4 �g4 6 ctJe5 � h 5 7 f3
above, Black would have loved to ctJfd 7 8 ctJxc4 e 5 9 g 3
have his king on c7 protecting the
weak b-pawn! One drawback is that
White can take on b7 with check, but
at least the position is unbalanced,
unlike the safe edge which White eas­
ily obtained in the game Kramnik­
Damljanovic.
1 5 ctJxb7+
15 g4!? JlLg6 16 f4 f6 17 f5 JlLeS 1S
tLlxb7+ c/£c7 19 tbba5 does not really
improve the white position, but in­
stead 15 h4!? (threatening g2-g4 and
h4-h5) 1 5 . . .f6 16 g4 JlLeS 17 tbf5!?, at­ This has been the most popular
tacking g7 and d4, is an interesting choice recently. White prevents
attempt. . . . �h4+ and threatens d4xe5 .
1 5 . . '.t> c 7
. 16 ctJba5 ctJa6 17 e4 9 . . . � b4
dxe3+ 1 8 ctJxe3 ctJ b4 For 9 . . .f6 see Game 43.

75
Th e Sla v

1 0 dxe5 0-0 1 1 ilL h 3 ! ? is ttJh6+ gxh6 19 �xe6 �f7 20 f5 h5

White now threatens e5-e6. 21 0-0 would have led to a murky po-
1 1 .. .'�e7 1 2 f4 slttOn.
1 7 ilL e 3
17 'iVe4!, intending ttJh6+, was bet­
ter.
1 7 . . . LiJa6 1 8 0-0 ilLf7 1 9 LiJe4 It>h8
20 gad 1 �c7 21 b3 ilLf8 22 �b2
LiJec5 23 LiJxf6 ilLxb3 24 !:!.xd8 !:!.xd8
25 LiJg4 ilLd5 26 LiJd4 LiJxa4? ! 27
�a 1 b 5 ?

12 �f4 is considered in the next

game.
1 2 . . . .l:!.d8 1 3 �c2 f6
Absolutely necessary in order for
Black to free himself.
1 4 e6
14 exf6 ttJxf6 gives Black some
counterplay for the pawn.
1 4 . . . LiJc5 1 5 LiJe3 Black is determined to ignore his
Or 15 f5 Md4 16 ttJe3 ttJba6 17 0-0 kingside defences. The punishment is
MadS with counterplay (Epishin) . swift.
1 5 . . . LiJxe6 1 6 LiJf5 28 f5 ilL b4 29 LiJe6 ! ilLxe6 3 0 fxe6
ilLc3 3 1 �b 1 �e7 32 LiJ h 6 !
The pawn capture 32 ... gxh6 loses t o
33 Mf7.
32 . . . .l:!.f8 33 LiJf7+ It>g8 34 ilLf5 ':!:!'xf7
35 ilLxh7+ It>h8 36 ':!:!'xf7 �e8 37
�f5 1 -0

Game 42
Parker-Hellsten
Copenhagen 1996

1 d4 d 5 2 c4 c6 3 LiJf3 LiJf6 4 LiJc3

1 6 . . . �d 7 ? ! dxc4 5 a4 ilLg4 6 LiJe5 ilLh5 7 f3
A mistake according t o Epishin. LiJfd 7 8 LiJxc4 e 5 9 g 3 ilL b4 1 0 dxe5
The alternative 1 6 . . :�eS 1 7 'iVe4 ttJa6 0-0 1 1 ilLh3 �e7 1 2 ilLf4

76

A very interesting idea. 12 . . . g5, to
chase the bishop away, is simply met
by 13 �d2 (13 �xd7 'bxd7 14 �xg5
'iVxg5 15 'MUxd7 b5 16 f4 is very messy)
13 . . . 'bxe5 14 'bxe5 'MUxe5 15 'be4
Jtxd2+ 16 'MUxd2 g4 17 �g2 with a
better position for White.
1 2 . . Rd8 1 3 �c2 �g6 1 4 �f5
.
14 e4 is very risky but not easy to
refute. For example, 14 . . . 'MUc5 15 'bd6
fLlxe5 16 'bxb7 (forking queen and
rook) 16 . . . 'bxf3+ 17 �f1 'MUc4+ 1 8
'iVe2 seems fine for White, while
14 . . . b5 15 axb5 cxb5 16 'be3 'bxe5 17
0-0 is also difficult to judge.
1 4 . . . ctJb6 1 5 �xg6 hxg6 1 6 ctJxb6
axb6 1 7 h4
Preventing . . . g7-g5 .
Th e B r o n s t e in Va ria tio n
1 7 . . . Ile8
17 .. .lbd7 fails to 18 �g5 ! 'MUxe5
(18 .. .f6 19 exf6 gxf6 20 'MUxg6+) 19
�xd8 'MUxg3+ 20 �f1 Mxd8 21 'be4,
when Black has insufficient compen­
sation for the exchange.
18 0-0 �c5+ 19 <;t>g2 �xc3 20
�xc3 �xc3 21 bxc3 ctJd7
Black just manages to hold the en­
suing endgame, but the whole line
seems extremely uncomfortable for
him.
22 gfb 1 ga6 23 a 5 gea8 24 gd 1
ctJf8 25 gab 1 b5 26 c4 bxc4 27
gxb7 I!xa 5 28 Il b4 R d 5 29 R c 1
ctJ d 7 30 Rbxc4 ctJxe5 3 1 �xe5 Ilxe5
3 2 e4 ga2+ 3 3 <;t>h3 f5 34 gxc6
fxe4 3 5 fxe4 Ilxe4 36 1::!. x g6 <;t>h7 37
gg 5 P.e7 38 Id.h 5+ '/z - '/z
Game 43
Dautov-Ni kolic
Ter Apel 1994
1 d4 d 5 2 c4 c6 3 ctJf3 ctJf6 4 ctJc3
dxc4 5 a4 �g4 6 ctJe5 � h 5 7 f3
ctJfd 7 8 ctJxc4 e 5 9 g 3 f6
A very solid continuation, just pro­
tecting the e5-pawn.
77
Th e Sla v
1 0 dxe5 ctJxe5 1 1 \'i{{x d8+ <;t>xd8 1 2
ctJxe5 fxe5 1 3 iL g 5+ <;t> c 7 1 4 0-0-0
ctJd7 1 5 iL h 3 iLe8 1 6 iLe3 !
This very nice move prepares f3-f4,
breaking the position open.
1 6 . . . ctJc5 1 7 f4 iLd7 1 8 iLxd7 ctJxd7
1 9 f5
19 ctJe4 Me8 20 f5 was even more
accurate. The ending is basically very
pleasant for White.
1 9 . . . iLe7 20 g4 h 6 21 ctJe4 ctJf6 22
ctJxf6 iLxf6 2 3 h4 iLe 7 24 iL f2 gad8
25 iLg3 iLf6 26 <;t>c2 .!cl.xd 1 27 J::l: x d 1
gg8 28 g 5 h x g 5 29 hxg 5 iLxg 5 30
iLxe5+ <;t>c8 31 gg 1 ge8 3 2 iLxg7
l:;l;.xe2+ 3 3 �d3 l:;l;.d2+ 34 <;t>c3 iLe3
3 5 gg3 iLf4 3 6 gf3 gg2 3 7 gxf4
gxg7 38 f6 g g 8 39 <;t>d4 <;t>d7 40
78
<;t>e5 J::l: e 8+ 41 �f5 ge2 42 b4 �e8
43 gh4 J::l: f 2+ 44 �e6 ge2+ 45 �d6
a5 46 bxa5 <;t>f7 47 gh7+ <;t>xf6 48
gxb7 ne4 49 �xc6 gxa4 50 gb5
1 -0
This is all very sensible. However,
White has another rather crazy idea.
Game 44
Shirov- N i kolic
Wijk aa n Zee 1993
1 d4 d 5 2 c4 c6 3 ctJf3 ctJf6 4 ctJc3
dxc4 5 a4 iLg4 6 ctJe5 iL h 5 7 f3
ctJfd 7 8 ctJxc4 e 5 9 e4
And this is it! This very natural
move was completely ignored until
recently.
9 . . . \'i{{ h 4+ 1 0 g 3
The wacky 10 �e2 is considered in
the next game.
1 0 . . : 'II¥f 6 !
Having softened up the kingside
with . . . "iVh4+, the queen retreats to f6,
where it helps attack f3 with the
bishop on h5 , and d4 with the pawn
on e5.
1 1 dxe 5 ! �xf3 1 2 ctJd6+ ! <;t> d 8
Amazingly, 12 ... �xd6 loses to 1 3
 Th e B r o n s t e in Va ria tio n

'iVxd6 i¥xh l 1 4 �,g5 (threatening a5 25 h4 b6 26 �c2? !

'iVe7+ mate) 14 . . .f6 15 exf6 gxf6 16
'iVe6+ WdS (16 . . . WfS 1 7 ilh6+ mate)
17 ilxf6+ lLlxf6 i s i¥xf6+ Wc7 19
'iVe5+! followed by i¥xh5 or i¥xhs
with a crushing position, as pointed
out by Ivan Sokolov.
1 3 �xf3 �xf3 1 4 ctJxf7+ �e8 1 5
e6 ! ?
The first new move o f the game! In
I.Sokolov-Lautier, Belgrade 199 1 ,
White had played 1 5 CLlxhS but
15 . . . CLlxe5 ! (Shirov) 16 ilf4 CLlSd7 is
nice for Black, as after 17 fIg l , Either 26 Wd2 or 26 g5 would have
1 7 . . . ilc5 ! makes sure the rook does kept a slight advantage for White ac­
not escape. cording to Shirov.
1 5 . . . ctJ c 5 1 6 � c 4 � x h 1 1 7 ctJxh8 26 . . . ctJca6 27 ctJe4 �a7 28 ctJg 5+
JLxe4 ! 1 8 � g 5 �d5 29 ctJhf7 Ide7 3 0 gh8 ctJd7 3 1
�xh 7 ctJe5 3 2 ctJxe5 M,xh7 3 3 ctJxh 7
�xe5 34 ctJf8 I!>f4 3 5 ctJxg6+ �xg4
36 ctJe5+ �xh4 37 ctJxc6 �g5 Y2 - Y2

Game 45
Nesterov-Imanaliev
Bishkek Zonal 1 993

1 d4 d 5 2 c4 c 6 3 ctJf3 ctJf6 4 ctJc3

dxc4 5 a4 �g4 6 ctJe5 � h 5 7 f3
ctJfd 7 8 ctJxc4 e 5 9 e4 �h4+ 1 0
�e 2 ! ?
i s b4 seems to win a piece, but
ls . . . ild3 ! (Shirov) 19 e7 il.xc4
(19 . . . ilxe7 20 ilf7+ followed by bxc5
keeps the fun going) 20 exfSi¥ + �xfS
21 bxc5 WgS favours Black.
1 8 . . . �f5 1 9 0-0-0 �e7 20 �xe7
ri;xe7 21 �f 1 g 6
A slight error according t o Shirov.
21 . . . r�e6 22 ilxe6 Wxe6 23 fIfS as is
suggested instead, but this also seems
quite nice for White.
22 g4 �xe6 23 �xe6 �xe6 24 Idf8

79
Th e Sla v
A magnificent idea, the tactical just­
ification of which lies in my all-time
favourite opening trap!
1 0 . . . exd4
Tempting but not the best. Black
should react more calmly with
10 . . . jLb4, intending . . . 0-0.
1 1 �xd4 3L c 5 1 2 lZJd6+! �f8
12 . . . c,t>e7 loses to 13 ctJfS+ while
12 . . . c,t>ds is met by 13 ctJxb7+.
1 3 �xg 7+ ! !
1 3 . . . �xg7 1 4 1ZJf5+
Regaining the sacrificed queen and
winning a pawn, with a good position
to boot!
1 4 . . . �f6 1 5 lZJxh4 lZJa6 1 6 3Lh6 lZJe5
1 7 g4 lZJxg4 1 8 fxg4 3Lxg4+ 1 9 �d2
�ad8+ 20 � c 2 lZJ b4+ 21 �b3 3Le6+
2 2 3Lc4 3Lxc4+ 23 �xc4 .l:!.d4+ 24
�xc 5 lZJc2 2 5 3Lg7+ 1 -0
Finally, White can try the develop­
ing move 9 jLe3 .
Question 8: This looks very strange,
doesn't it?
A nswer: With 9 jLe3 , White de­
fends the d4-pawn and prepares to
meet 9 . . :�h4+ with the simple 10
jLf2 . However, 1 0 dxeS is still not a
threat due to 10 . . :iVh4+, winning the
80
knight on c4. The move is extremely
aggressive: White will expand on the
kingside with g2-g4 and h2-h4 and try
to win Black's light-squared bishop,
while rapid queenside castling is also
on the agenda.
Game 46
I . Sokolov- Hellsten
Malmo 1 995
1 d4 d5 2 c4 c6 3 lZJc3 lZJf6 4 lZJf3
dxc4 5 a4 3Lg4 6 lZJe5 3L h 5 7 f3
lZJfd 7 8 lZJxc4 e5 9 3Le3 3Lg6
Question 9: This looks odd as well!
A nswer: This is a typical idea in this
line. Black realises that the bishop is
doing nothing on hS, where it merely
bites against the pawn on f3 . There­
fore, he moves it to a more active di­
agonal, delaying the decision of which
piece to put on b4: the bishop on f8
or the knight on bS, via a6.
1 0 h4!
This not only aims to harass the
bishop on g6 with hS , but also threat­
ens d4xeS by removing Black's re­
source of . . . iVh4+.
1 0 . . . 3Le7 1 1 h 5 ! 3Lf5
1 1 . . .�h4+ is met simply by 12 jLf2

�xf2+ 1 3 'it'xf2 i,f5 14 liJd6+, win­
nmg a pIece.
1 2 dxe5 0-0 1 3 h 6 ! lLla6 1 4 hxg7
l:!.e8 1 5 g4 XL g 6 1 6 f4 !
Quite amazing! White is only mov­
ing pawns, but Black just seems to be
helpless!
1 6 . . . lLlb4 17 Uc 1 lLld5 18
Ibxe3 1 9 lLlxe3 XL h4+ 20 �f 1 XLg3
21 f5 �b6 2 2 fxg6 hxg6 23 �d2
�xe 5 24 lLled 5 1 -0
24 . . . cxd5 loses to either 25 Mh8+
\t>xg7 26 �h6+ 'it'f6 27 liJxd5+, fork­
ing king and queen, or instead 25
tt'lxd5 �d4 26 Mh8+ 'it'xg7 27 �h6+
mate! A game of astonishing ferocity,
even by Ivan Sokolov's remarkable
standards!
Th e B r o n s t e in Va ria tio n
Game 47 .
K rasen kov-Sapis
Polish Championship 1995
1 lLlf3 d 5 2 d4 c6 3 c4 lLlf6 4 lLlc3
dxc4 5 a4 XLg4 6 lLle5 XLh5 7 f3!
lLlfd 7 8 lLlxc4 e 5 9 XLe3 XLb4
A more natural developing move
than 9 . . . i,g6.
1 0 g4 XLg6 1 1 dxe 5 !
This is possible now, since
l 1 . . .�h4+ no longer hits the knight
on c4!
1 1 . . . 0 -0 ! ?
XLg2
This i s a new idea. The alternative
1 1 . . . �e7+ was played in the original
game Granda Zuniga-Nikolic, Biel
Interzonal 1993, and now 12 liJd6+
i,xd6 13 �xd6 would have given
White a safe edge due to his two bish­
ops and the weakness of Black's dark
squares. Instead in the game, White
went for the crazy complications of
12 f4 �h4+ 13 i,f2 �xg4 14 �b3
liJa6 15 liJd6+ i,xd6 16 �xb7 liJb4 17
�xa8+ �b8 which is just unfathom­
able.
1 2 h4 h6 1 3 h 5 XL h 7 1 4 g 5 ! ? hxg 5
1 5 h6 g 6
81
Th e Sla v
Question 1 0: Why is White playing
so aggressively? He was a pawn up
with more space; why did he sacrifice
his g-pawn?
A nswer: White has played some
very strange moves in the opening.
First, he put his bishop on e3 in front
of the e-pawn, blocking in his light­
squared bishop . Then, instead of de­
veloping his pieces, he advanced his
kingside pawns in order to chase
Black's light-squared bishop. White
has won a pawn, but his pawn ad­
vances have left many weak squares in
his position. For example, if Black
could get a rook to dS and then play
. . . ctJc5, aiming for the weak b3-square,
then White's position would become
critical. White has raised the stakes
with his risky opening play - he must
continue in the same aggressive man­
ner or Black will develop and exploit
White's weaknesses.
1 6 Rg 1 'J}f/e7 1 7 Rxg 5 :i:!.dS 1 S 'J}f/c2
b5?
A tactical miscalculation. Chekhov
recommends lS . . . 'i'e6! 19 'i'e4 ctJa6!
(intending . . . ctJdc5) 20 'i'h4 b5!, when
I think that Black has good counter­
chances.
82
1 9 axb5 cxb5 20 'J}f/e4 ! bxc4 2 1
'J}f/xaS ttJxe5 22 'J}f/xa 7
White is winning now due to his
large material advantage.
22 . . . ttJbd7 23 '>jj'd 4 f6 24 Rg3 ctJc5
25 .l'::!. a S ! l:baS 26 '>jj' d 5+ ttJ e 6 27
'J}f/xaS+ ttJfS 2S Wf2 iLd6 29 ttJe4
iLb4 30 iLd4 WhS 3 1 f4 ttJed7 3 2
Re3 '>jj'f 7 33 ttJ g 5 '>jj' g S 3 4 ReS 1 -0
And now things get even more
complicated. Let us take a look at 7
h3 .
Game 48
Gelfand - N i kolic
Manila Interzonal 1990
1 d4 d5 2 ttJf3 ttJf6 3 c4 c6 4 ttJc3
dxc4 5 a4 iLg4 6 ttJe5 iL h 5 7 h3
ttJa6 S g4 iLg6 9 iLg2
I find it hard to recommend this
line to players of either colour, unless
they have six months in which to ana­
lyse the mind-boggling complications!
9 e3 is discussed in the next game.
9 . . . ttJb4 1 0 0-0 iLc2 1 1 'J}f/d2 iLb3
Question 7: Why is Black doing
this?
A nswer: If Black were to play nor-
 Th e B r o n s t e in Va ria tio n

mally, then White's space advantage iVxg5, which looks like good com­
would guarantee him a substantial pensatIon to me.
advantage. Black therefore keeps the
pawn and challenges White to make
something of his lead in development.

1 8 �e4 �d8 1 9 f4 �e7 20 e3 0-0

21 �e2 c 5 22 dxc5 �xc 5 23 � d 2
Black does not really have enough
1 2 a5 for the exchange.
Threatening a5-a6 to break up the
black queenside. 12 tL'le4, attacking the
knight on b4, is the other major con­
tinuation, aiming for the attractive
trap 12 . . . tL'lxe4 13 iVxb4 tL'ld6 14
'i'xb7! ! , when 14 . . . tL'lxb7 loses to 15
c�oxc6+.
1 2 . . . a6
Stopping a5-a6, but in fact this may
not be so dangerous: 12 . . . e6 13 a6 iVc7
14 axb7 iVxb7 15 g5 tL'lfd5 16 e4 tL'lb6
17 d5 looks impressive, but 17 . . . :l':rd8
18 tL'lxc6 tL'lxc6 19 dxc6 iVxc6 20 e5 2 3 . . . CLl d 5 2 4 �h 1 � d 7 25 l::i. a c 1 �a4
4Jd5 21 :l':rxa7 �c5, with . . . 0-0 to fol­ 26 �f3 c3 27 bxc3 �xa 5 28 f5 �c4
low, was nice for Black in the game 29 fxe6 ! liJe7 30 �xh7+ �xh 7 3 1
Moreno-Rogers, Manila Olympiad �e4+ liJg6 3 2 �xc4 fxe6 3 3 �g4
1992. l::i. x f 1 + 34 .ti.xf 1 iVb5 35 c4 �c6+ 36
1 3 liJa4 e6 1 4 g 5 liJ d 7 1 5 liJxd 7 �g 1 �g8 37 �f3 �d 7 3 8 l::i. f 2 a 5 3 9
�xd 7 1 6 tLlb6 �d8 1 7 tLlxa8 �e4 tLle7 4 0 �xa5 tLlf5 4 1 � d 2 1 -0

see follo wing diagram

If White wishes to duck the critical
1 7 . . . �xa 8 ? lines that we saw in the previous
Ftacnik suggests 1 7 ... tL'lc2 1 8 :l':rb 1 game, he can play 9 e3 instead of 9
'i'xa8 19 e3 �b4 20 iVe2 iVd8 2 1 �d2 �g2.

83
Th e Sla v

iLlc2!? 15 �xc2 �xd4+ 16 �g2 �xc4

Came 49 and 17 f5 (Rogers) .
Klarenbeek-Rogers
Dutch Team Championship 1 996

1 d4 e6 2 e4 d 5 3 tZlf3 tZlf6 4 tZle3

dxe4 5 a4 ]Lg4 6 tZle5 ]Lh5 7 h 3
tZl a 6 8 g4 ]L g 6 9 e 3
A quieter attempt, making sure that
White regains the pawn.
see follo wing diagram

9 . . . tZlb4 1 0 ]Lxe4 e6
1O . . . 'Llc2+ loses to 11 �xc2 �xc2
12 �xf7 mate, but 1o . . . 'Lld7 1 1 iLlxg6 14 �e2 �xd 7 15 tZla2 �e4 16
hxg6 12 �f3, intending �f1-g2, is tZlxb4? ! ]Lxb4 17 ]Ld2 0-0 18
more normal. White has a small ad­ ]Lxb4? !
vantage here due to his slight space White is just playing for a draw,
advantage and bishop pair. but he is doing this badly. The text
makes the a-pawn very weak.
1 8 . . . axb4 1 9 b3 Ra 5 ! 20 Rad 1 b5
21 axb5 exb5 22 ]Ld3 ]L b 7 !
Black is not going to exchange this
bishop, while the as-h 1 diagonal is so
tempting!
23 \t>h2 �d 5 24 %:!.b 1 Re8 2 5 Rfd 1
g5!
An unexpected and really strong
move.
26 :!:!.f 1
26 fxg5 loses to 26 . . . �d6 (Rogers),
1 1 0-0 a5 1 2 f4 tZld7 1 3 tZlxd 7 as 27 �g1 �g3+ is terminal.
Rogers recommends instead 13 e4 26 . . . Re3 27 Rb2 gxf4 28 Rxf4 Ra1
�h4 14 �g2 with a slight advantage 29 \t>g3 Rg 1 + 30 \t>h4 h 6 0 - 1
for White, but with the threat of 15 And White lost o n time i n this
iLlf3 , intending f4-f5, this looks horri­ hopeless position.
ble for Black.
1 3 . . . ]L e 2 ! Came 50
We have seen this before. Black Leitao-Beliavsky
makes sure that the bishop does not Erevan Olympiad 1 996
get shut in behind the e4-pawn.
13 . . . �xd7 is strongly met by 14 e4! 1 d4 d 5 2 e4 e6 3 tZlf3 tZlf6 4 ctJc3

84
 Th e B r o n s t e in Va ria tio n

dxe4 5 a4 �g4 6 tLle5 � h 5 7 g 3

The quiet option. White ignores
the bishop on h5 and just develops
normally.
7 . . . e6 8 � g 2 � b4 9 tLlxe4 tLl d 5
Attacking c3 .
1 0 'i'b3 0-0 1 1 �d2
11 0-0 fails to 1 1 . . .itxc3 12 bxc3
Jtxe2, winning a pawn.
1 1 . . a 5 1 2 e4 tLl b 6 !
.

An important and typical manoeu­

vre. Black exchanges a pair of knights,
relieving his slightly cramped position
while attacking d4.
1 3 tLlxb6
This rather helps Black. Beliavsky
suggests 1 3 ite3 instead.
1 3 . . . 'i'x b6 1 4 �e3 e 5 ! 1 5 d5 tLl d 7
1 6 0-0 tLle5 1 7 h 3 tLlf3+ 1 8 �h 1
itJd4
Clearly Black has now taken over
the initiative. The bishop on h5 has
suddenly become a major player, sup­
porting the incursion of the black
knight into the vulnerable kingside
light squares.
1 9 'i'e4 tLle2 20 '!:!'ae 1 tLlxe3 2 1 fxe3
'i'd6 22 �h2 .!:!.ae8 23 .l:!.ee 1 �g6 24
'!:!'e2 .l:!.e7 25 tLlb5 'i'e5 26 �f3 g d 7
2 7 �g2 �h8 28 tLle3 .l:!. f d 8 29 gd 1
f6 30 ged 2 �f7 3 1 .!:!.d3 h 5 32 'i'b5
e4 3 3 'i'xe4 h4 34 tLle2 � h 7 3 5 gd4
ge7 36 'i'd3 '!:!'de8 3 7 gxh4 '!:!'e2 3 8
d6 '!:!' x b 2 39 d 7 .l:!.d8 40 �h 1 � h 5 4 1
�xh 5 'i'xh 5 42 e5+ f5 43 tLlg 1 � e 5
4 4 .!:!. g 4 .l:!. x d 7 45 'i'xd 7 fxg4 4 6
'i' d 3 + �h6 47 'i'e3 g 3 4 8 'i'xb2
'i'xd 1 49 �g2 'i'xa4 50 'i'xb 7 'i'e2+
51 �xg3 �xe3 0 - 1
A fine game b y Beliavsky.

85
Th e Sla v

S u m m a ry

I cannot really recommend the 5 . . . itg4 line for Black, not because it is a par­
ticularly bad line, but simply because unless you have loads of time for de­
tailed analysis, you won't be able to feel comfortable playing it. There are
many theoretical problems to solve: 7 f3 ttJfd7 8 ttJxc4 e5 9 ttJe4 gives White a
safe endgame edge, while 9 g3 is also dangerous. Even the crazy 9 e4 and 9
ite3 pose difficult problems! By contrast 7 h3 gives Black too many counter­
chances, while 7 g3 is a little tame.

1 d4 d 5 2 c4 c6 3 ctJf3 ctJf6 4 ctJc3 dxc4 5 a4 �g4 6 ctJe5 � h 5 7 f3

7 h3 ttJa6 8 g4 �g6 (D)
9 itg2 - game 48
9 e3 game 49 -

7 g3 - game 50
7 . . . ctJfd 7 8 ctJxc4 e 5 (D) 9 ctJe4
9 g3
9 . . . � b4 10 dxe5 0-0 1 1 ith3 Wlic7
12 f4 game 41 -

12 itf4 - game 42
9 .. .f6 game 43 -

9 e4 Wlih4 +
10 g3 - game 44
10 �e2 game 45 -

9 ite3
9 . . . itg6 game 46 -

9 j�ob4 - game 47
..

9 . . . � b4+ 1 0 �d2 Wii e 7 1 1 �xb4 Wiix b4+ 1 2 Wii d 2 Wiix d2 1 3 �xd2 exd4
1 4 ctJed6+ (D)
14 . . . �e7 - game 39
14 . . . �d8 game 40-

8 . . . �g6 8 . . . e5 14 ctJed6+

86
 CHA PTER SEVEN

The 4 . . . a6 Slav :
White plays 5 e3

1 d4 d 5 2 c4 c6 3 lL:lf3 lL:lf6 4 lL:lc3

a6
In the next two chapters we shall
look at the move-order 1 d4 d5 2 c4 c6
3 ctJf3 ctJf6 4 ctJc3 a6.
Question 1 : What is the point of
4 .. a6?
.

Answer: 4 . . . a6 allows the bishop to

develop outside the pawn chain to g4
or f5, since after . . . iH5 , 'i"b3 attacking
b7 can be met by . . . b7-b5, advancing
the b7-pawn to a safe square, or
... J.:!a7!, an ugly looking but brilliant
thought of Julian Hodgson's.
Question 2: Sounds great. Any
drawbacks?
A nswer: You had to ask. Black is
placing a lot of pawns on light
squares , so he can often suffer from
weak dark squares.
Game 5 1
O i l -Anand
Biel Interzonal 1993
1 d4 d 5 2 c4 c6 3 lL:lf3 ttJf6 4 lL:lc3
a6 5 e3
White protects the c4-pawn and
prepares to develop his kingside.
White's numerous alternatives here
are discussed in the next chapter.
5 . . . b5
Black would like to develop his
bishop outside the pawn chain by
playing 5 . . . �g4, for example, but 6
'i"b3 is strong, as 6 . . . b5 7 cxd5 cxd5 8
a4! breaks up the black queenside.
Black must be able to meet a2-a4 with
. . .b5-b4; after . . . b5xa4, his a-pawn be­
comes very weak. In fact Black could
play an interesting tactical idea here:
after 8 . . . �xf3 9 gxf3 he can try
9 . . . b4!? , so that after 10 'i"xb4, 1 0 . . . e5!
attacks the white queen. 1 1 'i"b3 exd4
1 2 exd4 ctJc6 13 �e3 �b4 gives com­
pensation for the pawn due to
White's weakened pawn structure,
but 1 1 'i"b7! gains a tempo by attack­
ing the rook on a8 : 1 1 . . .ctJbd7 1 2 dxe5
ctJxe5 13 �xa6, when Black has

87
Th e Sla v

enormous problems on the light For 8 gxf3, see the next game.
squares. I'm sure that Julian Hodgson 8 . . . e6
would suggest 6 . . . Ma7(!) , but after 7 This was a novelty at the time of
ctJe5 (threatening 8 ctJxg4 ctJxg4 9 the game, as black players had been
cxd5, winning a pawn) 7 . . . e6 (as usual experimenting with the violent 8 ... eS,
in the Slav Black does not mind to exploit the absence of the queen
swapping off his light-squared bishop from the queenside and the slight
for White's knight) 8 f3 ! 11Lh5 9 g4 weakness of the knight on c3 (it is no
11Lg6 10 h4! , White's threat of h4-h5, longer protected by a pawn on b2) . I
trapping the bishop, forces Black to prefer Anand's simple move, which
play the disastrous 1 O ... h6, when 1 1 carries the same threats but without
ctJxg6 fxg6 1 2 �c2 is just winning for the risk.
White. 9 :iL.d2 :iL.b4 1 0 �d 1
The inclusion of 5 . . . b5 6 b3 takes
the b3-square away from the white
queen, allowing Black to develop his
light-squared bishop in greater com­
fort.
6 b3
The exchange 6 cxd5 is considered
in Game 55.
6 . . . :iL.g4 7 h 3

So that 10 . . . �a5 can be met by 1 1

�c2, but this move i s a little meek. A
few years ago, I played 10 11Ld3 �aS
1 1 Me l ! ? 11Lxc3 12 MXc3 �xa2 13 �d1
against Jon Levitt, sacrificing a pawn
in order to gain the advantage of two
bishops against two knights. In fact I
won a nice game after 13 . . . 0-0 14 0-0
�a3 15 �a1 ! �xa1 16 Mxa1 Ma7!? 17
A very natural reaction, putting the cxb5 cxb5 (hoping for 1 8 11Lxb5 ctJe4,
question to the bishop. 7 11Le2 is dealt swapping off one of my bishops) 18
with in Games 53 and 54. Mc2 ! , allowing my dark-squared
7 . . . :iL.xf3 ! bishop to activate itself via b4. With
7 . . . 11Lh5 8 g4 11Lg6 9 ctJe5 , intending control of the c-file and Black's pas·
h2-h4, is rather awkward, as we have sive pieces, I quickly gained a decisive
seen. advantage. Black should have played
8 �xf3 . .. ctJe4 at some point before Mc2 in

88
 Th e 4 . . . a 6 Sla v : Wh i t e p la y s 5 e 3

order to force the exchange of one of

White's bishops, but White has rea­
sonable compensation for the pawn.
1 0 . . . 0-0 1 1 �e2 bxc4 ! 1 2 bxc4 c 5 !
This i s a typical freeing manoeuvre
for Black. The immediate 1 1 . . .c5
would of course have lost a pawn to
12 cxb5, so Black first exchanges on
c4 and then breaks in the centre.
1 3 dxc 5 ?
This is a serious mistake, and after
Anand's superb play it almost looks
like the losing move. White had to 17 �a4 �ad8 18 �e1 tDd4! 19
play for equality with 13 cxd5 cxd4 14 �xb4?
exd4 1i.xc3 15 1i.xc3 tLlxd5 . Retreating with 19 1i.d1 was the
1 3 . . d4! 1 4 exd4 �xd4
. only (but rather miserable) way to
avoid material loss.
1 9 . . . tDxe2+ 20 tDxe2
Or 20 �h l :d3 !
20 . . :�xa 1 2 1 tDc3 �c 1 22 �a5 �f4
23 �xa6 l:l:a8 24 �d6 �xc4 0 - 1
White had had enough. This game
is a really impressive demolition job
by Anand.

Let us now see what happens if

Black recaptures on f3 with the pawn
instead of the queen.
White has a worse pawn structure
and real tactical problems, as Black's Game 52
pieces quickly become amazingly ac­ Van der Sterren-Shi rov
tive. Biel Interzonal 1993
1 5 �c2
15 iVc 1 (to avoid Black gammg a 1 d4 d 5 2 c4 c6 3 tDf3 tDf6 4 tDc3
tempo on the queen after . . . tLlc6-d4) a6 5 e3 b 5 6 b3 �g4 7 h 3 �xf3 8
15 . . J::t d S ! prevents White from cas­ gxf3 ! ?
tling. The recapture with the queen
1 5 . tDc6
. . leaves White vulnerable to a quick
15 .. J::r d S is well met by 16 :d1 ! . . . 1i.b4. This recapture aims to suffo­
1 6 0-0 �e 5 ! cate Black by playing f3-f4
Freeing the d4-square for the (preventing Black's . . . e7-e5 break) and
knight. c4-c5 (preventing the . . . c6-c5 break) .

89
Th e Sla v

5 liJf3 b 5 6 b3 �g4 7 �e2

8 . . . liJ b d 7 !
Threatening . . . c7-eS .
9 f4 bxc4 1 0 bxc4 dxc4!
By taking the c-pawn, Black pre­
vents c4-cS . Now White cannot stop
Black from playing . . . c6-cS himself,
and the game soon fizzles out.
1 1 �xc4 e6
Peter Wells and Glenn Flear are
well known for their deep knowledge
of Slav systems, so this game is espe·
cially interesting.
7 . . . e6 8 0-0 ..IiLd6? !
An inaccuracy that has unpleasant
consequences. When White plays
LiJeS, attacking c6, Black needs to be
able to exchange it as quickly as pos·
sible for one of his own knights.
Therefore Black should either play
8 . . LiJbd7 or 8 . . �e7 (to meet 9 4Je5
. .

with 9 . . . �xe2 10 ¥llix e2 and then

1O . . . LiJfd7!) . In the game, White gets a
grip on the dark squares and wins in
model fashion.
9 h 3 �h5 1 0 liJe5 �xe2 1 1 liJxe2!
0-0 1 2 liJf4 Wlic7 1 3 liJfd 3 ! liJbd7 1 4
�b2 Wlib7 1 5 '!:!c 1 �ac8 1 6 Re2
1 2 �d2 ..IiL b4 1 3 0-0 0-0 1 4 liJe4 a5 .!:!fd8 1 7 Wlif3 �f8 1 8 J:!fc 1 !
1 5 ..IiLxb4 axb4 1 6 liJxf6+ liJxf6 1 7 White's pieces are coordinating
Wlid3 liJ d 5 1 8 �fc l .!:!a5 1 9 J:!ab 1 beautifully.
�d 6 Y:z - Y:z
see follo wing diagram

Game 53 1 8 . . . bxc4 1 9 bxc4 liJxe5 20 liJxe5

Wells-Flear �d6 21 liJd3 liJe4 22 �e2 Wlib8 23
Oakham+1994 liJc5 �xc5 24 dxc5 Wlib7 2 5 cxd 5
�xd5 26 ..IiLxg7 �xg7 27 Wlig4+ LtJg5
1 d4 d5 2 c4 c6 3 liJc3 liJf6 4 e3 a6 28 e4 J::!. e 5 29 f4 J::!. x e4 30 Wlixg 5+

90

Wf8 3 1 g d 2 �e7 32 �h6+ �g8 3 3
l:!: c 3 �h8 34 gg3 gg8 35 gd7
�xe5+ 3 6 � h 2 1 -0
Game 54
K rasenkov-Epishin
Bmo 1 994
1 d4 LZlf6 2 e4 e6 3 LZlf3 d 5 4 LZle3
a6 5 e3 b 5 6 b3 JiLg4 7 JiLe2 e6 8 h3
Jih5 9 0-0 LZl bd 7 ! 1 0 LZle5 JiLxe2 1 1
CLJxe2 LZlxe5 1 2 dxe5 LZld7
13 exd 5 exd 5 14 JiLb2 JiLe7 1 5 LZld4
�b6 1 6 l:i.e 1 0-0 1 7 ge6
17 CDc6 �cS is equal according to
Epishin. Here White's control of the
c-file looks impressive, but with some
fine moves, Epishin gradually pushes
Th e 4 . . . a 6 Sla v : Wh i t e p la y s 5 e 3
White back.
1 7 . . . �b7 1 8 �e2 ga 7 ! !
I f White now defends the eS-pawn
with 19 f4, then 19 . . .tDb S ! 20 l:\,c3 b4!
21 l:\,d3 l:\,cs (Epishin) gains the c-file
for Black.
1 9 z:!. e 1 LZlxe 5 ! 20 LZlxe6 �xe6 2 1
LZlxf8 �xe2 2 2 gxe2 LZle4! ! 2 3 bxe4
dxe4
The white knight is trapped. Black
regains his piece and his queenside
pawns prove to be far too much for
his opponent to cope with.
24 JiLd4 ge7 2 5 LZlxh 7 �xh7 26 JiLb6
ge6 27 JiLa5 JiLa3 ! 28 �f 1 b4 29
�e2 ge5 30 JiLb6 g b 5 31 JiLd4 b3
32 gxe4 b2 3 3 JiLxb2 gxb2+ 34 �f3
JiLb4 0- 1
91
Th e Sla v

Instead of 6 b3, sometimes White �e2 e5 33 h3 e4 V2 - V2

plays 6 cxd5 .

Game 55
Karpov-Short
Dortmund 1 995

1 d4 d5 2 c4 c6 3 tZJc3 tZJf6 4 e3 a6
5 tZJf3 b5 6 cxd 5 cxd 5 7 tZJe5

In the next game we see a tricky al­

ternative move-order from White.

Game 56
Sadler-Hodgson
Hastings 1 995/96

1 d4 d 5 2 c4 c6 3 tZJc3
A different plan: White prevents Of course playing 3 ttJc3 first gives
the black bishop from developing Black the extra possibility of playing
outside the pawn chain and tries to 3 . . . dxc4 (see Chapter 10) .
prove that Black's queenside is weak. 3 . . . tZJf6 4 e 3 a 6 5 �c2 ! ?
7 . . . e6 8 .1i.d2 .1i.e 7 9 .1i.e2 0-0 1 0 0-0
.1i.b 7 1 1 tZJd3
A typical manoeuvre by White,
bringing the knight in contact with
the weak c5-square.
1 1 . . . tZJ b d 7 1 2 b4 tZJ b 6 !
White has weak squares too!
1 3 a 4 tZJe4!

see follo wing diagram

1 4 axb5 tZJxc3 1 5 .i,xc3 axb5 1 6

tZJc5 .i,c6 1 7 .1lxa8 �xa8 1 8 .1i.d3
�a2 1 9 � h 5 g 6 20 �e5 �a7 21
ga 1 �b8 22 �x b8 gxb8 23 ga5
.1i.d8 24 .1i.e1 ga8 2 5 !!xa8 tZJxa8 26
g4 tZJb6 2 7 f3 tZJc4 28 .i,f2 .i,g5 29
f4 .i,e7 3 0 � g 2 f6 31 �f3 �f7 3 2
5 ttJf3 would o f course simply
transpose to the games we have al­
ready seen in this chapter. On c2 the
white queen prevents Black from
developing his bishop to f5, which

92

suggests that it would be natural for
Black to put his bishop on g4 instead.
However, after 5 . . . b5 6 b3 �g4,
White's idea is to play 7 tiJge2, and if
7 . . . tiJbd7 then 8 h3 �h5 9 tiJf4,
picking up the bishop pair.
Question 3: I thought you said that
Black wanted to exchange off his
light-squared bishop for White's
knight! Aren't you contradicting
yourself?
A nswer: It is a conflict of ideas - as
Black you say, ' The bishop on c8 was
my problem piece and I'm glad I've
exchanged it,' whereas with White
you say 'Yes, I've won the bishop
pair!' Frankly I would be happy to
play either colour! It is clear, how­
ever, that in comparison with the line
5 tiJf3 b5 6 b3 �g4 7 h3 �xf3 , White
has gained the two bishops at a much
lower cost: he has not had to either
weaken his kingside pawn structure,
or misplace his queen on the kingside.
So basically White has got a good ver­
sion of this typical sort of position.
And that is the point of waiting with
5 'lic2.
A similar idea for White is 5 �d3 ,
preventing . . . �f5 . Personally, I would
grab this opportunity to transpose
into a Queen's Gambit Accepted with
5 . . . dxc4 6 �xc4 e6 7 tiJf3 c5, but I
know that not everyone feels the
same way! 5 . . . �g4!? 6 'lib3 Ma7 is
possible, however, as neither 7 f3 dxc4
(7 �h5 8 cxd5 cxd5 9 g4 �g6 10
. . .
�xg6 hxg6 11 g5 tiJh5 12 tiJge2 [12
tLlxd5 tiJg3 !J 1 2 ... e6 1 3 f4 is better for
White due to the offside knight on
h5) 8 �xc4 �h5, intending . . . tiJbd7
and . . . e7-e5, nor 7 cxd5 cxd5 8 f3
Th e 4 . . . a 6 Sla v : Wh i t e p la y s 5 e 3
�c8 !?, with . . . tiJc6 and . . . e7-e5 to fol­
low, is advantageous for White.
Right, back to the game!
5 . . . e6 6 iDf3 e 5 !
A very imaginative idea. Black to­
tally changes his plan; he no longer
plays for . . . b7-b5, but strikes in the
centre with . . . c6-c5 .
Question 4: Hasn't Black just
wasted a move, since he's played first
. . . c7-c6 and now . . . c6-c5?
A nswer: This is true of course.
White is playing the variation 1 d4
tiJf6 2 c4 e6 3 tiJf3 c5 4 e3 d5 5 tiJc3
a6, with the move 'Iic2 added in for
free. Black's contention is that this
variation is not advantageous for
White normally, and that the move
'lic2 does not make any difference in
White's favour.
I felt that the best way to try to
make use of 'lic2 was to play 7 cxd5,
so that after 7 . . . exd5 I could attack a
clear target on d5 by bringing my
rook to d 1 , exploiting the fact that
my queen has already vacated this
square. Moreover, my queen could be
very useful on c2 to attack a bishop
on c5 after d4xc5 �xc5.
7 exd 5 exd 5 8 �e2 iDe6 9 0-0 �e6 !
93
Th e Sla v
The start of a superb tactical plan.
Normal development could have eas­
ily ended in disaster. For example, if
9 . . . iLe7 10 Md1 0-0 then 1 1 dxcS iLxcs
12 ttJxds wins a pawn, since the
bishop on cS hangs. Therefore Black
begins a complicated tactical manoeu­
vre that seeks to exploit the exposed
position of the queen on c2 .
1 0 gd 1 ctJ b4 ! 1 1 �d 2
1 1 'iVb 1 'iV cS , threatening . . . iLfS,
could be embarrassing.
1 1 . . . ctJe4! 1 2 ctJxe4 dxe4
White's pieces are not coordinating
well, and if White were to continue
routinely with 13 ttJeS, then 13 . . . cxd4
14 exd4 MCS , threatening . . . ttJc2,
would be extremely annoying. Instead
94
of this defensive course, White plays
for the initiative.
1 3 a 3 ! exf3 1 4 iLxf3 ctJc6 !
Julian did not like the look of
14 . . . ttJdS 15 dxcs ttJc7 16 iLxb7 with
three pawns and an initiative for the
piece.
1 5 d5 ctJe5 1 6 dxe6 ctJxf3+ 1 7 gxf3
fxe6 1 8 b3 �xd 2 1 9 iLxd2
Black's queenside pawn maJonty
even gives him a slight pull, but after a
few adventures the game was eventu­
ally drawn.
1 9 . . . �d 7 ! 20 iLc3+ �c6 21 ga2
iLe7 22 gad2 gad8 23 gxd8 gxd8
24 gxd8 iLxd8 2 5 iLxg7 c4!
Sacrificing a pawn to activate the
king.
26 bxc4 �c5 2 7 f4 �xc4 28 e4
iLc7 29 f5 exf5 30 exf5 � d 5 3 1 a4
b 5 32 axb5 axb5 33 iLc3 iLe5 34
iLd2 �e4 35 �f 1 �xf5 36 h3 iLf4
37 iLc3 �e4 38 �e2 iLd6 39 iLd2
�d4 40 iLe3+ �c4 41 iLd2 iLb4 42
iLxb4 !
I spent a while just checking that
the pawn ending was drawn. Remem­
ber that 4 1 . . .b4 just leads to a draw
after 42 iLxb4 as Black has the wrong­
coloured bishop for the rook's pawn!

42 . . . <;t>xb4 43 <;t>d3 <;t>a3 44 <;t>c2 b4
45 <;t>b 1 <;t>b3 46 h4 <;t>c3 47 h5 � - �
Since 47 . . . h6 48 f4 �d4 49 �b2
�e4 50 �b3 �xf4 51 �xb4 �g4 52
cj;c3 �xh5 53 �d2 �g4 54 �e1 �f3
55 �f1 draws Gust) for White.
Another tricky move for White is 5
a4 .
Gam? 57
Atali k-Miles
Hastings 1995/96
1 d4 d 5 2 c4 c6 3 ttJc3 ttJf6 4 e3 a6
5 a4
An unusual move in this position
when White has already committed
himself to e2-e3, shutting in his dark­
squared bishop.
5 . . . �f5 6 �b3 ga7 7 a 5 !
Suddenly, t o his horror, Miles real­
ised that after the natural 7 . . . e6, 8
'{!Vb6! is extremely strong, as after the
forced 8 . . . 'i'xb6 9 axb6 �a8 10 c5 ,
intending b2-b4 and b4-b5 breaking
through, White has a magnificent end­
ing. Tony, practical as ever, just
played a few necessary defensive
moves and got on with the game!
Th e 4 . . . a 6 Sla v : Wh i t e p la y s 5 e 3
7 . . . �d 7 S ttJf3 e 6 9 ttJe5 �cS 1 0 f3
ttJfd7 1 1 ttJxd 7 ttJxd7
Also possible was 1 1 . . .'i'xd7! ?
1 2 cxd 5 cxd 5 1 3 e4 dxe4 1 4 fxe4
�g6 1 5 e 5 ! §Le7 1 6 §Le2 0-0 1 7 0-0
gaS 1 S §Le3 .l:!.bS 1 9 �a4 �dS 20
.l:!.ad 1 MCS 2 1 d5 �c5 22 §Lf2 �xf2+
2 3 gxf2 ttJc5 24 �b4 �g 5 25 dxe6
fxe6 2 6 gxfS+ <;t>xfS 27 �d4 <;t>gS
2S �c4 Wie7 2 9 b4 ttJ d 7 30 §Lb3
ttJfS 31 �d6 �eS 3 2 §La4 �f7 3 3
Mf 1 �f5 3 4 ttJe4 �g 6 3 5 ttJ c 5 h 6 3 6
�d 1 �e4 3 7 B f 2 §Ld 5 3S §L c 2 Wig 5
39 h3 �e3 40 �e7 <;t>hS 4 1 �f7
ttJh7 42 �f4 �e 1 + 43 <;t> h 2 .l:!.gS 44
�g3 �xb4 45 ttJd3 �a3 46 ttJf4
�xg 3+ 47 <;t>xg3 gdS 4S ttJg6+ <;t>gS
49 ttJe 7 + <;t>hS 50 ttJg6+ <;t>gS 5 1
95
Th e Sla v
ClJe7+ �h8 52 ClJg6+ %-%
After 5 a4 Black can simply play
5 . . . e6, aiming to put a bishop on the
hole on b4, but this leads more to a
Semi-Slav type of position, so for Slav
devotees, I offer a few other ideas:
a) 5 . . . g6 is interesting, leading to a
sort of Schlecter Slav (see Chapter 1 1) .
96
b) 5 . . . 'ii c 7 is not as stupid as it
seems: Black defends b7 in advance. If
6 ct'Jf3 then 6 . . . �g4 7 'ii b 3 �xf3 8
gxf3 e6!
c) 5 . . . Ma7!? with the same idea. If 6
as then maybe 6 . . . �e6!? 7 ct'Jf3 (7 'ii b 3
dxc4!) 7 . . . dxc4 8 ct'Jg5 �g4 9 f3 �h5
with a completely unclear position.
 Th e 4 . . . a 6 Sla v : Wh i t e p la y s 5 e 3

S u m m ary

In general Black is doing fine in these lines, but since the 4 . . . a6 Slav is such a
recent development, there is still scope for improvements for both colours. 4
Qjf3 a6 5 e3 bS 6 b3 oltg4 7 h3 �xf3 8 'i'xf3 e6 9 �d3 �b4 10 �d2, as in
Sadler-Levitt, is worth further tests, and the game Sadler-Hodgson is certainly
crazy enough to be worth analysing!

1 d4 d5 2 c4 c6 3 tUf3 tUf6 4 tUc3

4 e3 a6 (D)
5 'i'c2 game 56 -

5 a4 game 57 -

4 . . . a6 5 e3 b 5 6 b3
6 cxdS game 55
-

6 . . . iLg4 (D) 7 h3
7 �e2 e6
8 0-0 - game 53
8 h3 game 54 -

7 . . . iLxf3 (D)
8 'i'xf3 game 5 1-

8 gxf3 game 52
-

4 . . . a6 6 . . . iLg4 7 . . . iLxf3

97
 The 4 . . . a6 Slav:
Aggressive o ptions for White

1 d4 d 5 2 c4 c6 3 lLlf3 lLlf6 4 lLlc3 possible to achieve if Black can organ­

a6 ise his pieces properly. White has
In this chapter we deal with the closed the centre, so Black has more
more attacking systems against the time to achieve his plan, as he never
4 . . . a6 Slav, in which White delays e2- has to worry about a sudden central
e3 in order to develop the bishop on breakthrough. Moreover, Black can
c 1 to an active square outside the spare a lot more pieces than usual to
pawn chain. We shall first examine S prepare . . . e7-eS, since he doesn't need
cS, which has virtually become the any to defend his own centre.
main line.
Question 1 : S cS looks like a begin­ Game 58
ner's move. Isn't it bad to release the K ra m n i k-Shirov
tension in the centre so early? Vienna 1 996
A nswer: White is aiming for a 'big
clamp' on the centre: S cS stops Black 1 lLlf3 d 5 2 d4 c6 3 c4 lLlf6 4 lLlc3
from breaking with . . . c6-cS, and �f4 a6 5 c 5 .iLf5 6 �b3 .l:!.a7 !
will prevent Black from achieving
. . . e7-eS . White will then either launch
a queenside offensive with b2-b4, a2-
a4 and b4-bS, or he will organise a
central break with e2-e4. Overall, this
is a very ambitious plan.
The queenside break . . . b7-b6 is not
great for Black, since cSxb6, . . . ik'xb6
leaves him with weak dark squares on
cS, b6 and as, and a backward c-pawn.
Question 2: Black can't play on the
queenside, and he can't get in any of
his breaks. What can he do? I couldn't help laughing when Jul­
A nswer: . . . c6-cS is not on the ian Hodgson told me that he played
agenda, but . . . e7-eS should not be im- . . . a7-a6 in order to facilitate . . . .l::!. a 7,

98
 Th e 4 . . . a 6 Sla v : A g g r e s s i v e o p t io n s fo r Wh i t e

but the idea is a good one. It really is White positionally. White had to an­
not clear whether the queen on b3 is ticipate . . . e7-eS and either prevent it
any better placed than the rook on a7: with 17 eS or play 17 iYc4, intending
on b3 the queen attacks very little and 17 . . . eS 1S dS .
gets in the way of one of White's 1 7 . . . e 5 ! 1 8 �c4 �f6 1 9 dxe5 lLlxe5
main plans, the queenside pawn storm
with b4, a4 and bs. If the queen
moves away from b3, then the rook
can simply return to as !
The more conventional 6 . :�cS 1S.

discussed in the following game.

7 .ltf4 lLl b d 7 8 h 3 h6 9 e3 g 5 !

The pos1tlOn is now horrible for

White: Black can just gang up on his
weak pawns.
20 �d4 gaa8 !
Having done its job, the rook
comes back into play.
2 1 � e 3 J::!. a d8 2 2 .lte2 � e 7 23 J:Ihe 1
This move shows a good under­ lLld7 24 �d4+ lLlf6 25 �e3 �e5 26
standing of the position. Black leaves .ltf3 h 5 !
his weakened queenside alone and Preparing . . . gS-g4, driving away a
grabs space on the kingside, where defender of e4.
White has little going for him. The 2 7 a3 gfe8 28 nxd8 nxd8 29 .ltd 1
move . . . g7-gS also allows the bishop J::!. d 4 30 .ltc2 J::!. c4 3 1 Wb 1
to come to g7, supporting the . . . e7-eS
break.
1 0 .lth2 .lt g 7 1 1 lLle5 0-0 1 2 f3
Unwilling to let Black have things
his own way, White prepares the e2-
e4 break. This raises the stakes, as
... e7-eS will become doubly effective
against an expanded white centre.
1 2 . . . lLlxe5 1 3 .ltxe5 lLl d 7 1 4 .ltxg7
�xg 7 1 5 e4 dxe4 1 6 fxe4 .ltg6 1 7
O-O-O?
A careless move that condemns

99
Th e Sla v

3 1 . . . 1:!.xc5 S . . . e6 9 ctJe5 3Le7 1 0 g4 3Lg6 1 1 e3

The first pawn falls . . . 3LdS 1 2 3Le2 ctJxe5 1 3 3Lxe5 3Lc7
3 2 �d2 1:!.c4 3 3 �dS ctJxe4 3 4 3Lxe4 14 ctJa4
3Lxe4+ 35 ctJxe4 .!:!xe4
. . . and now the second.
0- 1
A really good exposition of the
ideas behind the 4 . . . a6 Slav.

Game 59
I . S okolov-Shirov
Erevan Olympiad 1 996

1 d4 d 5 2 c4 c6 3 ctJf3 ctJf6 4 ctJc3

a6 5 c 5 3Lf5 6 �b3 �cS ! ?
This is a normal plan, trying to
gain the two bishops by exchanging
the knight for the dark-squared
bishop on ds . However, there is a
tactical problem on this occasion.
1 4 . . . 3La5+! 1 5 ctJc3
Or 15 Wfl CDe4 16 Wg2 f6, intend-
ing . . . e6-e5 .
1 5 . . 0-0 1 6 h4 h6 1 7 0-0-0 ctJd7 1 8
.

3Ld6 geS 1 9 ghg 1 3L c 7 2 0 h 5 3Lh7

21 g 5 .1Lxd6 22 cxd6 <t>hS

I'm not sure why Shirov did not

want to repeat his previous game.
Perhaps it was nothing objective, just
the desire not to be too predictable.
The idea behind this move is ex­
tremely neat: Black will develop nor-
mally with . . . e7-e6, . . . CDbd7, . . . :ie7
and . . 0-0, and will then prepare . . . e6-
.

e5 by means of . . . MeS and . . . :idS-c7,

making use of the fact that the queen
has vacated dS !
7 3Lf4 ctJ b d 7 S h 3 ! ?
Black i s going t o castle kingside, so
White wants to have his pawn storm
ready on that side of the board.
I think that Black IS Just better
here: White is having to make all sort
of positional concessions and his at­
tack just isn't getting anywhere.

100
 Th e 4 . .
2 3 gxh6 g x h 6 2 4 e 4 dxe4 2 5 tLia4
�d8 26 tLic5 tLixc5 2 7 dxc5 �f6 28
�c3 �xc3+ 29 bxc3 Uab8 30 gd4
a 5 31 �d 1 b6 3 2 cxb6 gxb6 33
jLc2 .!:!.d8 34 d7 .!:!.b5 3 5 .!:!.gd 1 '!:!'xh 5
Black wins yet another pawn.
Though Sokolov fights hard, the re­
sult is never in doubt.
36 c4 �g7 37 �b2 �f8 3 8 .ita4
nc5 3 9 \t>c3 �e7 40 .!:!.d6 e3 4 1
fxe3 �e4 42 .!:!.6d4 .!:!. e 5 4 3 gf 1 �g2
44 gg 1 gxe3+ 45 gd3 .!:!.xd3+ 46
\ii x d3 '!:!'xd 7 + 47 �e3 �h3 48 �xc6
nc7 49 � b 5 e 5 50 Uh 1 .ite6 5 1
\ii d 3
51 . . . �f6 52 .!:!.xh6+ �g 5 53 gh8
\ii f4 54 .!:!.h4+ �g4 55 gh8 f5 56
ne8 g c 5 5 7 a4 �f3 58 �a6 �f2 59
. a 6 Sla v : A g g r e s s i v e o p tio n s fo r Wh i t e
� b 5 �e2+ 60 �c3 f4 61 gd8 �f3
6 2 .l:!.d2+ �e 1 0 - 1
Obviously Black is rather happier
than White in this line at the mo­
ment!
Another very popular idea for
White is to play 5 a4.
Question 4: What is the point of
this move?
A nswer: 5 a4 merely aims to pre­
vent Black's idea of . . . b7-b5, and force
him to find another plan. Though it
severely weakens the b4-square, this
move does give White the possibility
of a4-a5, cramping the black queen­
side.
Game 60
Pushkov-Epishin
Russian Championship 1 995
1 tLif3 tLif6 2 c4 c6 3 tLic3 d 5 4 d4
a6 5 a4 e6
The most natural and overwhelm­
ingly the most popular choice. The
game loses a little of its Slav character
I'm afraid, but Black does get to oc­
cupy that weak b4-square.
6 �g5
101
Th e Sla v
The fianchetto 6 g3 is dealt with in
the following game.
6 . . . tiJ b d 7 7 e3 iLe7 8 iLd3 0-0 9 0-0
dxc4 1 0 iLxc4 c 5 1 1 a 5 cxd4 1 2
exd4 b 5 1 3 axb6 tiJxb6
Black has already equalised.
1 4 iLe2 iL b 7 1 5 tiJe5 h 6 1 6 iLh4
tiJfd 7 1 7 iLxe7 �xe7 1 8 tiJxd 7 �xd 7
1 9 tiJ a 4 tiJxa4 20 l::l x a4 l::l f d8 2 1
�a 1 �c6 2 2 iLf3 �b6 2 3 iLxb7
�xb 7 24 �c 1 �d6 2 5 h 3 gad8 26
gcc4 e 5 !
Black wins a pawn, but can't quite
convert it into a win.
27 gab4 �e4 28 d5 �xd 5 29 �a3
gg6 30 .i:tg4 l:I d d 6 3 1 gb8+ �h7 3 2
� b 3 �xb3 3 3 g x b 3 gd 1 + 34 � h 2
g f 6 3 5 f3 g d 2 3 6 h4 g 6 3 7 ga4
1 02
�g7 38 �ba3 .i:txb2 39 �xa6 .i:tf4 40
�6a4 �xa4 4 1 nxa4 nd2 42 �g 3
.i:td3 43 na5 Wf6 44 na6+ �f5 45
ga7 �e6 46 .l:!a6+ .i:td6 47 ga5 f5
48 �a8 Wd5 49 �f2 nb6 50 l:IdS+
�e6 51 �g3 nb1 52 h 5 gxh 5 53
l:Ih8 h4+ 54 �xh4 �d5 5 5 XlfS f4
56 .i:td8+ �e6 57 ne8+ �f6 58 .l:!fS+
�e7 59 �h8 gg 1 60 nh7+ �f6 6 1
l:Ixh6+ � g 7 62 n e 6 .l::i. x g2 6 3 gxe5
�f6 64 J::!. a 5 .l::!. g 1 % - %
Let us see what happens if White
opts to fianchetto .
Game 61
Cifuentes-Dreev
Wijk aan Zee 1 995
1 c4 c6 2 d4 d 5 3 tiJc3 tiJf6 4 CUf3
a6 5 a4 e6 6 g3 tiJbd7 7 iLg2 iLb4 8
0-0 0-0 9 �b3 a5 1 0 gd 1 b6
White will find it hard to achieve
e2-e4, since Black has his bishop en­
trenched on b4 and can always play
. . . �xc3 , removing a defender of e4.
Without this idea, however, White
may struggle to find a plan.
1 1 tiJe5 tiJxe 5 1 2 dxe5 tiJd7 1 3 cxd 5
exd 5 1 4 iLf4 �e7 1 5 e4 d4! 16
 Th e 4 . .
gxd4 lZJxe 5 1 7 J£. x e 5 \'ixe5
Black's two bishops and queenside
pawn majority give him a clear advan­
tage.
1 8 .l:!.d 2 J£.e6 1 9 \'i c2 gad8 20 gad 1
gxd2 2 1 gxd2 b 5 22 h 3 h 5 23 h4
bxa4 24 \'ixa4 J£.xc3 2 5 bxe3 \'ixe3
26 ge2 \'ib4 2 7 \'ixe6 a4 28 \'ie3
Vi1Vb3 29 �f 1 ge8 3 0 \'ix b3 axb3 3 1
gd2 g e 2 3 2 gd8+ � h 7 3 3 e 5 J£.f5
0-1
This does all seem fine for Black
but if you cannot live happily with­
out developing your bishop outside
the pawn chain, then S . . . �fs 6 iVb3
Ma7 seems interesting. After 7 as,
threatening 8 iVb6, 7 . . . dxc4 8 iVxc4
CLlbd7 is not so stupid. For example, 9
�gS h6 1 0 i.. xf6 (10 i.. h 4 gS!? 1 1 i.. g 3
i.. g7 seems fine for Black) 10 . . . exf6
(Please don't even think of 10 . . . CLlxf6,
allowing 1 1 iV cS ! , winning a piece) 1 1
e4 i.. g4 (1 1 . . .i.. h 7!?; 1 1 . . .i.. e 6!?) with a
very murky position.
S i.. g S is a sharp continuation that
is only occasionally seen. White de­
velops h is bishop to its most aggres­
sive square and is willing to sacrifice
the pawn on c4 for the chance of a
. a 6 Sla v : A g g r e s s i v e o p tio n s fo r Wh i t e
swift attack.
Game 62
Ward-Levitt
British Championship 1 995
1 d4 d 5 2 e4 e6 3 lZJe3 lZJf6 4 lZJf3
a6 5 J£.g5 dxe4
6 a4
Also possible is the immediate 6
e4!?
6 . . . h 6 7 J£. h 4 lZJd5 8 e 4 lZJ x e 3 9
bxc3 b 5
Black has gone about this in a
strange way. S . . . Lbe4 6 i.. h4 (6 h4 has
been played but I don't believe it!
Even 6 . . . h6 is not stupid, when I don't
see the value of h2-h4!) 6 . . . Lbxc3 7
bxc3 dxc4 8 e4 bs is more natural if
Black wants this type of position,
since the omission of a2-a4 and h7-h6
must help Black a little. After S . . . dxc4
6 a4 another idea is to play 6 . . . �fS (a
more natural Slav move) , aiming for a
little trap: 7 e3 Lbe4!? 8 Lbxe4 i.. x e4 9
i.. x c4 i.. xf3 ! 1 0 iVxf3 �aS+! 1 1 �e2
�xgS 12 �xf7+ �d8 , when White
does not have sufficient compensation
for the piece.
1 0 lZJe5 \'ie7 1 1 J£.g3 \'ib7 1 2 g b 1
1 03
Th e Sla v

Wii a 7 1 3 Wii f 3 e6 1 4 3l.e2 g6 1 5 Wiif 6

After this freeing break, Black has

The opening has not been a success no problems.
for Black, but somehow he hangs on 11 tiJxd 7 tiJxd7 1 2 3l.e3 3l.e 7 13
and almost turns the tables com­ dxc5 tiJxc5 1 4 Wiix d8+ gxd8 1 5 gd 1
pletely. 0-0 1 6 a3 gxd 1 + 1 7 �xd 1 M.c8 1 8
1 5 . . . g h 7 1 6 �f3 3l. b 7 1 7 0-0 tiJd7 3l.e2 �f8 1 9 �c2 tiJa4 20 3l.d2 3l.d6
1 8 tiJxd7 �xd7 1 9 �f4 �e8 20 21 h 3 �e7 22 Bd 1 tiJxc3 23 3l.xc3
axb5 axb5 21 ga 1 ! Wii b 6 22 :!::i. x a8+ b4 24 axb4 3l.xb4 25 gd3 a5 26
3l.xa8 23 g a 1 Wii b 7 24 d5 cxd 5 25 �b3 3l.xc3 2 7 Bxc3 .l:i.xc3+ 28 �xc3
exd 5 3l.e7 26 �e3 3l. g 5 2 7 f4 3l.e7 �d6
28 3l.f3 Wii c 8 29 d 6 3l.xf3 30 Wiix f3
�c5+ 31 3l.f2 Wiix d6 3 2 .!:!.a8+ 3l.d8
3 3 ga7 Wii d 5 34 Wiix d 5 exd 5 35 g b 7
h 5 36 �xb5 f5 3 7 M.xd 5 .!:!. d 7 38
ge5+ �f7 3 9 g 3 3l. f6 40 M.c5 V2 - V2

Next, the adventurous 5 CLJeS . I

don't like this move at all for White
and in the next game Shirov neutral­
ises it very easily with simple devel­
opment.

Game 63 You might have expected the play­

Beliavsky-Shirov ers to have called it a day here, espe­
Erevan Olympiad 1 996 cially when you consider that this was
the last round of a gruelling Olym­
1 d4 d5 2 c4 c6 3 tiJf3 tiJf6 4 tiJc3 piad for both players: Beliavsky was
a6 5 tiJe5 tiJ b d 7 6 3l.f4 dxc4! 7 playing his 14th game on board 1 and
tiJxc4 b5 8 tiJe5 3l.b 7 9 e4 e6 1 0 f3 Shirov his 1 3th!
c5! 29 �d4 e 5+ 30 �c3 �c5 31 3l.c4 f6

1 04
 Th e 4 . . . a 6 Sla v : A g g r e s s i v e o p tio n s fo r Wh i t e

32 h4 �c6 3 3 �g8 h 6 3 4 � c 4 � d 7
3 5 h 5 f5 3 6 e x f 5 � x f 5 3 7 �d3 �e6
38 �e4 �f7 3 9 g4 �c4 40 �b7
� b 5 41 �a8 �c4 42 �e4 �d5 43
�xd 5 �xd 5 44 b3 e4 45 f4 e3 46
�d3 e 2 4 7 �xe2 �e4 48 g5 �xf4
49 gxh6 gxh6 50 � d 3 �g4 5 1 �c4
�xh 5 5 2 �b5 �g4 5 3 �xa 5 h5 54
b4 h4 55 b5 h3 56 b6 h2 57 b7
h 1 � 58 b8� �a 1 + 59 �b6 �b 1 +
60 rtJc7 �x b8+ 6 1 �xb8 V2 - V2
In the end even these two battlers
cannot continue the game! A really 6 1Lg 5 ? !
impressive display of fighting spirit 6 cxd5 cxd5 7 iLg5 iLe7 S e 3 h 6 9
from both players. iLxf6 iLxf6 10 iLd3 ctJc6 1 1 0-0 0-0 12
Mac 1 was played in Piket-Shirov,
The final idea to be considered is 5 Aruba 1995, and now Shirov suggests
'i'b3 , preventing the development of 12 . . . iLd7, as 13 �xb7 ctJa5 14 �b4
the light-squared bishop by attacking �e7 snares the queen.
b7. The advantage of this move is that 6 . . . dxc4 7 �xc4 b 5 8 �d 3 c 5 9 a4
Black cannot counter in normal Slav cxd4! 1 0 LiJxd4 b4 1 1 LiJe4 1Lb7 1 2
fashion, but the Semi-Slav approach is �xf6 gxf6 1 3 ld.d 1 1Le7 1 4 �f3 b 3 !
easy to understand, and nice for
Black.

Game 64
Lautier-Ba reev
Linares 1994

1 d4 d 5 2 c4 c6 3 LiJf3 LiJf6 4 LiJc3

a6 5 �b3 e6
The most popular move for Black.
5 . . . dxc4 6 �xc4 iLf5 7 g3 is the 4 �b3
dxc4 5 �xc4 iLf5 line with the extra
moves 4 ctJc3 a6, which favour White To meet 1 5 ctJxe6 with 1 5 ... �a5+!
more than Black, while 5 . . . Ma7 6 iLf4! 1 5 e3 �a5+ 1 6 �e2
is awkward. I briefly considered 5 . . . b5 Watch this king!
6 cxd5 cxd5 7 a4 b4!? S �xb4 ctJc6 9 1 6 . . . rtJf8 1 7 �f4 �e5 1 8 f3 �xf4
'i'b3 Mbs 10 �dl iLf5, intending 1 9 exf4 �g7 20 f5 LiJc6 21 fxe6
. . . ctJc6-b4, but 1 1 ctJh4 chases away LiJxd4+ 22 .i;l;,xd4 ghd8 23 gd7 gxd 7
the bishop and just leaves White a 24 exd 7 J:i.d8 25 rtJe3 f5 26 LiJg3
pawn up. iLc5+ 27 �f4 !

1 05
Th e Sla v
27 . . . 1Ld6+ 28 �xf5 gxd 7 29 ctJe4
1Lc7 30 1Lc4 g e 7 3 1 ctJc3 ge5+ 32
�g4 � h 6 33 f4 ge3 34 1L d 5 f5+ 3 5
�xf5 1Lc8+ 3 6 �f6 1Lxf4 3 7 1Lxb3
1Lg5+ 38 �f7 1Lg4 3 9 �f8 ! ! 1Le7+
40 �g8 ! !
The white king completes a re-
markable J ourney: �e2-e3-f4xfs-
1 06
g4xfS-f6-f7-f8-g8 ! !
40 . . . 1Lb4 4 1 1Lf7 1Lxc3 4 2 bxc3
gxc3 43 ge 1 .l:i.c2 44 1Ld5 .l:i.d2 45
1Le4 1Lh5 46 gf 1 1Lg6 47 1Lb7 gd4
48 a 5 na4 49 gf6 gxa5 50 nxa6
g h 5 5 1 h3 ge5 52 ga8 � g 5 53 1Lf3
h 5 12 - 12
A brilliant game!
 Th e 4 . . . a 6 Sla v : A g g r e s s i v e o p t io n s fo r Wh i t e

S u m m a ry

If you had asked me a few months ago, I would have told you that Black had
a few problems in the 4 . . . a6 Slav, but now I am not so sure! 5 cS should be
avoided for the time being, while 5 a4, 5 ctJeS and 5 'iYb3 don't really seem to
promise a great deal, though 5 a4 and 5 'iYb3 can be good weapons if you
know that your opponent does not like to play systems with . . . e7-e6. 5 �gS is
aggressive and deserves further tests.

1 d4 d 5 2 c4 c6 3 CLJf3 CLJf6 4 CLJc3 a6

5 c 5 (D)
5 a4 e6 (D)
6 �gS - game 60
6 g3 game 61
-

5 �gS game 62
-

5 ctJeS game 63
-

5 'iYb3 game 64
-

5 . . . jLf5 6 �b3 (D)

6 . . . .l:"!.a7 game 58
-

6 . . . 'iYc8 game 59
-

5 c5 5 . . . e6 6 �b 3

107
 CHA PTER NINE
The Exchange Variation
1 d4 d 5 2 c4 c6 3 cxd 5 cxd 5
The Exchange variation is often
dismissed as just a dead draw, but sev­
eral top players, Yusupov and Vaiser
in particular, play it to win. What do
they see in this variation?
Although the pawn structure is
symmetrical, there is the open c-file to
play with and we know from experi­
ence the annoying pressure that
White can exert on b7. White also has
the advantage of moving first in this
position. Indeed, if Black does not
know what he is doing, his position
can very easily become highly un­
pleasant.
Question 1 : What plans has Black
tried in this position?
A nswer: Black has tried two ap­
proaches:
a) To put the bishop on c8 outside
the pawn chain, on either f5 or g4.
b) To unbalance the pawn structure
with . . . ctJe4xc3 .
Question 2: And what does White
do?
A nswer: White also has two ideas:
He can put his king's knight on either
f3 or e2 (after �d3) .
Question 3: What is the difference
between them?
1 08
A nswer:
a) ctJf3 allows White to increase his
pressure on the c-file by playing ctJe5,
attacking a black knight on c6. If
Black exchanges knights with . . . ctJxe5,
then White has removed the main
barrier to invasion on the c-file.
b) With �d3 and ctJge2 White aims,
not to attack on the c-file, but rather
to keep Black passive by preventing
the light-squared bishop from devel­
oping outside the pawn chain. The
white bishop on d3 controls f5, while
the absence of a knight from f3 means
that . . . �g4 can be met by f2-f3 .
Game 65
M i l ov-Sadler
Isle ofMan 1994
1 d4 d 5 2 c4 c6 3 cxd 5 cxd 5 4 iLf4
White must be wary of move­
orders here as 4 ctJc3 e5 !? 5 dxe5 d4 6
ctJe4 ctJc6 7 ctJf3 �f5 8 ctJg3 �g6 9 a3
�c5 10 "i'b3 ctJge7, as in Tozer-Levitt,
London (Lloyds Bank) 1993, gives
Black good counterplay for the pawn.
4 . . . lLlc6 5 e3 lLlf6 6 lLlc3
White could try 6 �d3 ! ? here to
prevent 6 . . . �f5. Black's best is 6 . . . i,g4
7 ctJe2 �h5 ! , intending . . . �g6 to swap

off bishops. 7 �b3 ctJaS ! 8 �bs+ �d7
also poses few problems.
6 . . �f5
. little carried away. The positional idea
The old main line. The modern
6 . . a6 is considered in the next game.
. tension in the centre very early, Black
7 ]L b 5 e6 8 tZJf3
4 ctJf3 ctJf6 S ctJc3 ctJc6 6 �f4 �fS 7
e3 e6 8 �bs is the normal way to
reach this position. White is threaten­
ing ctJeS, ganging up on the knight on
c6.
8 . . . tZJd7
Breaking the pin on the knight and
thus dealing with the annoying threat
of ctJeS .
9 0-0 �e7 1 0 �b3 g 5 ! ?
Bold but probably not the best.
1O . . . ,Sc8 1 1 �xc6 bxc6 12 �b7 as is
Th e Ex c h a n g e Va ria tio n
the main line, which does not seem
bad for Black, but at the board I got a
is correct: since White has released the
feels much more able to take action
on the wings. Indeed this idea was
seen to great effect in Kramnik-Shirov
from Chapter 8, but in this case
White has the open c-file to help his
queenside play.
1 1 �g3 h5 1 2 h3 g4 1 3 hxg4 hxg4
1 4 tZJd2 �h4 1 5 tZJe2 �a5 ! ?
S o far so good. With my last move,
attacking the knight on d2, I was hop­
ing for the reply 16 'sfd 1 , when
16 . . . �c2! 17 �xc2 �xb5 is reasonable
for Black.
1 6 e4 ! !
After a great deal of thought, my
opponent found a very strong idea,
blowing the centre open.
1 6 . . . �xd 2 1 7 exf5 ]Lxg3 1 8 tZJxg3
�xd4 1 9 t!.fe 1 �d8 20 fxe6 fxe6 2 1
l:!xe6
Well, this is not great for Black, but
at least I'm still alive! My opponent
was already in serious time-trouble,
and after the game I was really upset
that I had made things so easy for him
1 09
Th e Sla v

in the rest of the game. a6 5 cxd5 cxd5 6 .tf4 ttJc6.

21 . . . �c7 22 �d 1 4:Jc5 23 !!xd4 Question 4: Why is . . . a7-a6 useful in

4:Jxb3 24 �xd 5 4:Jbd4 2 5 �e4 4:Jxb5
26 �x b 5 �ad8 2 7 .l::i. e 1 � d 2 28 4:Je4
.l::i. e 8 29 �e3 �c2 30 a4 .l::i. e 7 3 1 4:Jf6
I!.xe3 32 4:Jd5+ �c8 3 3 4:Jxe3 :!:!.d2
34 4:Jx94 �d4 3 5 4:Je5 :!:!.xa4 36
4:Jxc6 bxc6 3 7 �b3 �c7 38 f3 c 5 39
�f2 �c6 40 94 c4 41 �b8 a 5 42
�e3 �c7 43 �b5 �c6 44 :!:!.b8 Wc7
45 .l:!a8 I!.a2 46 95 �b7 47 96! c3
48 9 7 c 2
Everything else is also hopeless, e.g.
4S ... cxb2 49 .l:IbS+ �xbS 50 gSiV +
�a7 5 1 'i¥xa2 and wins.
49 �d2 1 -0
Now let us take a look at the fash­
ionable 6 . . . a6.
Game 66
H odgson-Sadler
Ischia 1 996
1 c4 c6 2 4:Jf3 d5 3 cxd 5 cxd 5 4 d4
4:Jf6 5 4:Jc3 4:Jc6 6 iL.f4 a6
The modern main line, which will
be of particular interest to 4 . . . a6 Slav
players, as this line can occur by
transposition after 3 ttJf3 ttJf6 4 ttJc3
this position?
A nswer: 6 . . . a6 is a constructive
waiting move: Black keeps the white
pieces from occupying the b5 square,
which means he no longer has to
worry about .tb5, pinning his knight
on c6.
Question 5: What if White just
plays 7 e3?
A nswer: Then Black plays 7 . . . �g4 8
h3 .txf3 9 iVxf3 e6. This is another
example of Black giving up his light­
squared bishop for White's king's
knight in the Slav. In this case, since
White has played cxd5 so early, Black
has been able to play his knight to c6,
its most natural and best square. The
manoeuvre . . . .tg4xf3 also removes
the attacking idea ttJe5 .
Question 6: So what does White do?
A nswer: The only way that White
can go for advantage is to avoid play­
ing e2-e3 too early and thus sidestep
. . . .txf3 .
7 4:Je 5 ! ? e6
Black does not fear S ttJxc6 bxc6, as
White is not sufficiently active to pre­
vent Black from freeing himself with

1 10
 Th e Ex c h a n g e Va ria tio n

. . . c6-cS .
8 e3 CLlxe5 9 iLxe5 iLd7 !

An important new idea. 9 . . . bs was
also possible, but that does give White
a bit of a target to attack on the
queenside. 9 . . . �d7 develops the bish­
op to the as-h 1 diagonal (preventing
White from ever achieving e3-e4) ,
while still allowing Black to cover the
cS-square with the advance . . . b7-b6 if
necessary.
1 0 iLd3 iLc6 1 1 �f3 CLld7 1 2 iLg3
Y:z - Y:z
A real no-holds-barred classic! In
mitigation, England were playing
Switzerland that day in 'Euro 96' so
we did have other things on our
minds. (Although after watching the
match for half an hour we began to
think that even our game might have
been more exciting!)
7 . . . iLf5
7 . . . �g4 S ctJ eS ! (the point) is annoy­
ing, but now S ctJ eS is met by s . . . ncs .
8 e3 e6 9 LDe5 LDxe5 1 0 iLxe5 CLld7
1 1 iLg3 .!:!.c8 1 2 iLd3 iLxd3 1 3 �xd 3
iL e 7 1 4 0-0 0-0 1 5 '!::!' c 2 �a 5 Y:z - Y:z
Despairing o f making anything
against this plan of . . . �g4xf3 , white
players turned to 'Plan B ' : �d3 and
ctJe2.
Game 68
Yusupov-Sh i rov
Zurich 19941%
1 d4 d 5 2 c4 c6 3 CLlc3 CLlf6 4 cxd 5
cxd 5 5 iLf4 CLlc6 6 e3 iLg4

Game 67
Andersson-Epishin
Ter Apel 1995

1 CLlf3 af6 2 c4 c6 3 d4 d5 4 cxd 5

cxd 5 5 LDc3 CLlc6 6 iLf4 a6 7 .1:Ic 1
Here we see another attempt to de­
lay e2-e3 .

1 1 1
Th e Sla v

Question 7: Why not 6 . . . a6 here?

A nswer: After 6 . . . a6 White can play
7 i,d3 i,g4 8 tLJge2 (but not 8 f3 �hs
and Black can exchange bishops
with . . . i,g6) .
Question 8: What's the problem?
Can't I just go 8 . . i,xe2?
.

A nswer: You certainly can, but in

comparison with the line 6 tLJf3 a6 7
e3 i,g4 8 h3 i,xf3 9 'iVxf3 , we see that
Black has an inferior version: White
has played i,d3 instead of h2-h3 , and
his queen is better placed on e2 than Also playable is 8 . . . g6! ? , intending
on f3 . This position is perfectly play­ 9 . . i,g7 to support the . . . e7-eS break.
.

able for Black, but there just seems no 9 Si.g3

point in going in for a worse version The aggressive 9 g4 is considered in
of something if there is no real need the next game, while 9 tLJge2 allows
to. Black to win the bishop pair with
Question 9: What is the point of the 9 . . . tLJhS .
6 . . . i,g4 7 f3 i,d7 manoeuvre? 9 . . . Si.e 7 1 0 tLJge2 0-0 1 1 0-0 a6
A nswer: By forcing f2-f3 , Black Black's main counterplay is to ex­
weakens a dark square on e3 in pand on the queenside, placing his
White's position. Moreover, Black has knight on the c4 outpost and advanc­
more chance of achieving the . . . e7-eS ing the a- and b-pawns.
break, since the white knight can no 1 2 tLJc 1 ! ? tLJa5 1 3 tLJ b 3 !
longer go to f3 . Finally, the bishop is
not too badly placed on d7, as it sup­
ports Black's counterplay with . . . b7-
bS.
Question 1 0: Wait a minute, if I get
this via a transposition from the . . . a7-
a6 Slav, then I won't be able to play it
in this way will I?
A nswer: Don't worry! White can't
use this subtle move-order if he trans­
poses to the Exchange variation via
the 4 . . . a6 Slav, since he will either
have shut in his dark-squared bishop An imaginative idea: on b3, the
(3 tLJc3 tLJf6 4 e3 a6) or will have al­ knight protects d4, allowing White to
ready committed his knight to f3 (3 expand in the centre with e3-e4, while
tLJf3 tLJf6 4 tLJc3 a6) . eyeing the cS-square, which will be
7 f3 Si. d 7 8 Si. d 3 e6 weakened by . . . b7-bs. The one

1 12
 Th e Ex c h a n g e Va ria tio n

problem is that the knight will be in

the line of fire when Black plays . . . a6-
as-a4.
1 3 . . . lLle4 1 4 �e2 J:i.e8 1 5 e4 b5 1 6
e5 lLlh 5 ! 1 7 f4 lLlxg3 1 8 hxg3 a 5 1 9
2Lxh7+ �xh7 20 �h5+ �g8 2 1 �f2
f5 Y, - y,
After 22 �h l White will force per­
petual with iVhs+ and iVhs+.

Game 69
Vaiser-Nalbandian
Erevan Open 1 996 Although White's position is a little
loose, his extra pawn will count in the
1 d4 d 5 2 e4 e6 3 lLlc3 lLlf6 4 exd 5 end.
exd 5 5 .itf4 lLle6 6 e3 .itg4 7 f3 2Ld7 20 . . . �b6 21 lLld2 0-0 2 2 g5 h5 23
8 .itd3 e 6 9 g4 �e3 a 5 24 a 3 2Lb7 2 5 .l:!.b1 �a6 26
ge7 2Le6 27 .l:!.e1 lhe7 28 .itxe7
.it b 5 29 lLlg3 ge8 3 0 �b2 g 6 3 1
lLlge4 a4 3 2 l:!e3 2Le6 3 3 .itg3 .it d 5
3 4 .l:!.xe8+ �xe8 3 5 lLle3 .ite6 3 6 e4
�a6

This aggressive thrust is V aiser' s

trademark: White advances on the
kingside, while Black counter-attacks
on the queenside.
9 . . . a6
9 . . . i, b4 is interesting, so that after 37 d5 .ite5+ 38 �g2 exd 5 39 exd5
10 a3 i,aS 1 1 b4, the bishop returns to 2Lxd 5 40 lLlde4 .itxe4 4 1 lLlxe4 iU8
c7 and supports the central break 42 �d4 �e2+ 43 .itf2 f5 44 gxf6
. . . e6-eS . �b5 45 lLlg 5 lLle5 46 �e4 .ith6 47
1 0 g e 1 h 6 1 1 h4 ge8 1 2 lLlge2 lLla5 f7+ 1 -0
1 3 �f2 lLl e4 1 4 ge2 b 5 1 5 lLlb 1
2Le6 1 6 �e 1 .ite7 1 7 b3 lLld7 1 8 Finally, we take a look at the most
bxe4 bxe4 1 9 2Lxe4 dxe4 20 .l:!.xe4 recent attempt from Black, 6 .. .cDe4.

1 13
Th e Sla v

This game brought the whole idea to

popular attention.

Game 70
Portisch-Kra m n i k
Biel lnterzonal 1993

1 d4 d 5 2 c4 c6 3 cxd 5 cx d 5 4 ct:Jf3
ct:Jf6 5 ct:Jc3 ct:Jc6 6 �f4 ct:Je4 ! ?

The key to this concept. Black an­

ticipates that the centre will open
with c3-c4 or e3-e4, and places his
bishop on an influential diagonal.
9 .lte2
9 c4 �g7 10 cxd5 'i'xd5 is a posi­
tion from the 4 �g5 Griinfeld (1 d4
lZJf6 2 c4 g6 3 lZJc3 d5 4 �g5 lZJe4 5
�f4 lZJxc3 6 bxc3 �g7 7 e3 c5 8 cxds
An idea o f Iosif Dorfman. cxd4 9 cxd4 'i'xd5 , etc.) , which is fine
Question 1 1 : What is the point of for Black, while Kramnik shows that
the . . . lZJe4xc3 manoeuvre? 9 �d3 �g7 10 0-0 0-0 1 1 e4 �g4 12 h3
A nswer: With . . . lZJe4xc3 , Black �xf3 1 3 'i'xf3 e5! 14 dxe5 (14 exdS
closes the c-file, blocking one of lZJxd4! 1 5 cxd4 exf4 equalises) 14 . . . d4!,
White's major assets. Moreover, he intending . . . lZJxe5, is also fine. Finally,
unbalances the pawn structure, isolat­ 9 lZJe5 is dealt with in the next game.
ing the white a-pawn but giving 9 . . . �g7 1 0 0-0 0-0 1 1 c4 dxc4 1 2
White the chance to play a pawn to c4 �xc4 �f5
for the second time!
Question 12: What if White had
played one of the subtle move-orders,
such as 3 cxd5 cxd5 4 �f4 lZJc6 5 e3
lZJf6 6 lZJc3?
A nswer: Well, 6 . . . lZJe4 is still possi­
ble and leads to unclear play after 7
lZJxe4 (or 7 lZJge2! ? , intending f2-f3 to
recapture on c3 with a knight)
7 . . . dxe4, intending a quick . . "e7-e5 and
. . . �b4+.
7 e3 ct:Jxc3 8 bxc3 g 6 !

1 14
 Th e Ex c h a n g e Va ria tio n

Kramnik gives 12 . . . �g4 13 h3 �xf3

14 �xf3 es 1 5 dxes �xes 16 �xes
�xes as equal: 17 �xb7 �xh2+ 1 8
�xh2 �h4+ 19 �gl 'iVxc4 keeps the
balance. He also suggests 12 . . . a6!?,
intending ... b7-bs.
1 3 Rc 1 l;l c 8 1 4 � e 2 a6 1 5 h 3 ?
1 5 ds bs 16 �b3 leads to a slight
advantage for White according to
Kramnik.
1 5 . . . ttJa5 1 6 Ji.d3 Ji.xd3 1 7 �xd3
�d 7 1 8 .l::!. c 3 b 5 1 9 Rfc 1 ttJc4
A very impressive demonstration
of Black's strategy.

Game 71
Cifuentes-Leyva
Cien/uegos 1 996

1 d4 c5 2 c3 cxd4 3 cxd4 d 5 4 ttJf3

20 �e2 ttJb6 21 'f1.c7 �e6 22 Ji. g 5
4Jd 5 ! 23 'f1. 7 c 5 h 6 2 4 Ji.h4 b4 ! 25
�b2 ttJc3 !
I like the way in which Black has
moved his outpost from c4 to c3 !
26 .l::!. x c8 Rxc8 27 \t>h 1
27 'iVxb4 ctJe2+ wins.
27 . . . \t> h 7 ! 28 'f1.a 1
Or 28 'iVxb4 ctJxa2!
28 . . . a5 29 �b3 �xb3 30 axb3 g 5
3 1 Ji. g 3 a4
ttJc6 5 ttJc3 ttJf6 6 Ji.f4 ttJe4
After a weird transposition, we are
back to the main position.
7 e3 ttJxc3 8 bxc3 g 6 9 ttJe5 Ji. g 7 ?
A careless error. As Lalic has
shown, Black can equalise with
9 . . . 'iVas! 10 �b3 �g7 1 1 �bs �xes!
12 �xes 0-0.
1 0 ttJxc6 bxc6 1 1 �a4 !

see follo wing diagram

32 ttJd2 a3 33 l;l c 1 e5 34 d 5 a2 3 5
'f1.a 1 e 4 3 6 d 6 'f1. a 8 3 7 ttJc4 ttJb5 3 8
.Jt e 5 ttJxd6 39 l:!xa2 .l::!. x a2 4 0 Ji.xd6
gxf2 41 Ji.xb4 l:!f 1 + 42 \t>h2 l;lb 1
0-1 A very awkward move: Black

1 15
Th e Sla v
suddenly finds himself in big trouble.
This is a good illustration of what can
happen to Black if he does not take
enough care.
1 1 . . . SL d 7 1 2 iLa6 ! e5 1 3 �a3 exd4
1 4 exd4 0-0 1 5 0-0 iLe8
Black has to shed a pawn to meet
the threat of iLb7, winning the ex­
change.
1 6 iLxe8 .!:!:xe8 1 7 VJlixa7 f6 1 8 a4 e 5
1 9 SL g 3 � a 8 20 VJlie5 �a5 2 1 VJlie3
1 16
�e8 22 VJlib3 �b8 23 VJlia2 �h8 24
.!:!:fd 1 VJlia8 25 h4 h 5 26 '!:!: d e 1 exd4
27 exd4 .!:!:e4 28 �e2 � h 7 29 J::t d 1
�a7 3 0 iLe7 .!:!:a6 3 1 a 5 .!:!:xd4
3 2 SLb6
Now the inevitable advance of the
white a-pawn swiftly proves to be
decisive.
32 . . . .!:!:xd 1 + 33 .!:!:xd 1 VJlib7 34 �c5
�a8 3 5 �xd 5 VJlixd 5 36 �xd 5 f5 37
�d8 .!:!:a6 38 �d7 �g8 39 �a7 1 -0
 Th e Ex c h a n g e Va ria tio n

S u m m ary

All three major continuations of the Exchange variation seem to be doing fine
for Black. Personally I would recommend either 6 . . . a6 or 6 . . . .iiJ5 , as 6 . . .liJe4
leads to positions that are more typical of the Griinfeld than the Slav.

1 d4 d5 2 c4 c6 3 cxd 5 cxd 5

4 11. f4 CLlc6 5 e3
5 liJf3 liJf6 6 liJc3
6 . . . a6 (D)
7 liJe5 game 66
-

7 �c 1 game 67
-

6 . . . liJe4 7 e3 liJxc3 8 bxc3 g6 (D)

9 �e2 game 70
-

9 liJe5 game 71
-

5 . . . CLlf6 6 CLlc3 1Lg4

6 . . . �f5 game 65
-

7 f3 lL d 7 8 lL d 3 e6 (D)
9 �g3 game 68
-

9 g4 game 69
-

6. . . a 6 B . . . g6 B . . . e6

1 17
 CHA PTER TEN

Move-Orders an d
T ranspos itions

1 d4 d5 2 c4 c6 cxd5 cxd5 5 iYb3 , 5 . . :Vllic 7 is possible

This chapter is extremely impor­
tant both for white players, and for
black players who wish to play lines
with . . . d5xc4. Until recently, it was
thought that White had no need to be
too accurate with his move-order.
However, due to the efforts of Ivan
Sokolov, this is no longer true.
We shall first consider 3 e3 . With
this move White aims to remove the
force from . . . d5xc4 (White can simply
recapture with the bishop and has no
need to play a2-a4, preventing . . . b7-
b5) , and thus to prevent Black from
entering the Slav. Usually, the game
continues 3 . . .cbf6 4 e3 e6, leading to
the Semi-Slav, a very interesting open­
ing but not the one we want to play.
The following game shows the way to
meet 3 e3 .
Game 72
K rasen kov-I . S o kolov
Malmo 1 995
1 d4 d 5 2 c4 c6 3 e3 jLf5 !
I n the introduction, I said that
. . . �f5 can only be good for Black if
he can defend the b-pawn with his
queen. Sokolov noticed that after 4
as 6 iYxd5 loses to 6 . . . iYxc 1+. 3 . . . ilfs
is not the end of the world for White,
but he has only very small chances of
gaining an advantage once Black has
developed his queen's bishop outside
the pawn chain, while White has shut
his inside.
4 �b3 �c7 5 cxd 5 cxd 5 6 CiJc3
6 ,�b5+ is considered in the next
game.
6 . . . e6 7 jLd2 CiJc6 8 CiJf3 CiJf6 9 jLe2
a6 1 0 0-0 jLd6 1 1 gfc 1 0-0
Black has developed very naturally,
and has equalised comfortably.
1 2 a3 �e7 1 3 jLe 1 h6
Ivan Sokolov suggests that
13 . . . MadS, aiming for . . . e6-e5, was
most accurate.

1 18

1 4 ctJa2 gac8 1 5 ctJb4 ctJa5 1 6 �a2
iLc7 1 7 ctJe5 ctJe4 1 8 ctJbd3
Somewhere around here, White of­
fered a draw, but Black was having
none of it!
1 8 . . .f6 1 9 iL b4 ctJd6 20 ctJf4 gfd 8 !
21 g4 iL e 4 22 .l:!. x c 7 ? �xc7 23
ctJxe6 � e 7 2 4 ctJ x d 8 fxe 5
2 5 iL d 3 ?
After the natural 25 iiLxa5, winning
a piece, 25 . . . nc2 26 iiLdl �h4 gives
Black a very dangerous attack: 27
iiLxc2 �xg4+ 28 �f1 �g2+!
(28 . . . iiLxc2 29 �xd5+!) 29 �e2 �g4+
30 �e l �gl+ leads to a draw by per­
petual, while Ivan Sokolov also men­
tions 29 . . . iiLxc2!? when, with threats
of . . . t2Je4 or . . . t2Jc4 and . . . �g4+, the
M o v e - O r d e rs a n d Tr a n sp o s i tio n s
attack continues despite Black's deficit
of a rook. However, White had to try
this as 25 iiLd3 ? , trying to deflect the
bishop on e4 from its protection of
the d5-pawn, simply loses a piece.
Sokolov finishes very efficiently.
2 5 . . :�xd8 26 dxe5 ctJdc4 27 iLxe4
dxe4 28 .l:!.c 1 r;t>h8 29 �b 1 ctJxe5 30
.l:!.xc8 �xc8 3 1 �xe4 ctJac6 32 h 3
ctJxb4 33 axb4 ctJ c 6 3 4 h4 � d 8 3 5
h 5 �e7 36 �f 5 �xb4 3 7 �c8+ r;t> h 7
38 ,, 5 hxg5 39 �f 5+ r;t>g8 40 �e6+
r;t>f8 41 �f5+ r;t>e8 42 �g6+ r;t>d8 43
�xg 7 �g4+ 44 r;t>f 1 �xh 5 4 5 �x b7
�h 1 + 0-1
46 �e2 t2Jd4+ wins the queen on
a8 . A very important game: Sokolov
won his last four games in the tour­
nament to pip Krasenkov by half a
point!
Game 73
N ogueiras-I . Sokolov
Erevan Olympiad 1 996
1 d4 d 5 2 c4 c6 3 e3 2Lf5 4 �b3
�c7 5 cxd 5 cxd 5 6 2L b 5+ ctJd 7 ! ?
7 ctJc3 e 6 8 2Ld2 ctJe7 ! ?
An interesting development o f the
knight.
1 19
Th e Sla v
9 J:rc 1 ctJc6 1 0 ctJa4 J::t c 8 1 1 ctJe2
iLe 7 1 2 0-0 0-0 1 3 ctJf4 ctJdb8 ! ?
There is no way that White is going
to get through on c6!
1 4 iL d 3 iLxd 3 1 5 ctJxd3 iLd6 1 6 93
�e7 1 7 ctJ c 3 a 6 1 8 ctJa4 �c7 1 9
Mc2 Y2 - Y2
Exchanges will follow on the open
c-file.
The next move that we shall exam­
ine is 3 tLlc3, which has exactly the
same idea as 3 e3: after 3 ... CLJf6, 4 e3
would prevent . . . dSxc4 and lead to the
highly theoretical pastures of the
Semi-Slav after 4 ... e6. This was the
main move-order for several years
until that man Ivan Sokolov again got
cracking.
Game 74 ·
LautieH .So kolov ·
Groningen 1995
1 d4 d 5 2 c4 c6 3 ctJc3 dxc4!
Once White has already committed
his knight to c3, the sequence . . . b7-b5,
a2-a4 (to regain the pawn) ... b5-b4
gains a tempo on the knight on c3,
which must then move to an inferior
1 20
square, away from the centre.
4 e4
The most natural continuation:
White gets in his desired e2-e4 as soon
as possible. If White plays 4 a4 in 0[­
der to prevent . . . b7-b5, then Black can
exploit the fact that White has not
played CLJf3 by playing 4 ... e5! 5 e3 (5
dxe5 �xd1 + is very nice for Black)
5 . . . exd4 6 exd4 �e6! Instead 4 e3 b5 5
a4 b4 6 CLJe4!? is the critical test of this
idea, when 6 ... �dS (not 6 . . . �a6 7
CLJc5!; but 6 . . . CLJf6!? 7 CLJxf6 + exf6 8
�xc4 �d6 9 'Wic2 0-0 1 0 �d3 g6 led
to a quick draw in Epishin-Sakaev, 5t
Petersburg 1997) 7 CLJg3 CLJf6 (7 ... e 5!?)
8 CLJf3 �a6 (8 ... h5!?) is very unclear.
4 . . . b5 5 a4 b4 6 ctJa2 ctJf6 7 f3
Advancing with 7 e5 is possible, al­
though 7 ... CLJdS 8 �xc4 e6 9 CLJf3 ile7
10 �d2 as 1 1 CLJc1 CLJdl 12 CLJb3 ilb7
13 'Wie2 c5! was absolutely fine for
Black in Yakovich-Sadler, European
Team Championship, Pula 1997.
7 . . . e 5 8 iLxc4?
White sacrifices a pawn, but he
must have missed something, as he
never gets anything for it. The correct
move is 8 dxe5, which is considered in
the next game.

8 . . .'�Vxd4 9 �c2
9 'i'b3 , attacking f7, is strongly met
by 9 . . . iilcS ! 10 iilxf7+ cJ;; e 7, when
Black's threat of . . . 'i'f2+ gives him an
overwhelming position.
9 . . . it c 5 1 0 it g 5 ita6 ! 1 1 itb3 ite7
1 2 liJe2 �b6 1 3 liJg3 h 6 1 4 itd2 g6
1 5 � c 1 liJfd 7 1 6 liJf 1 liJ c 5 !
White's position is truly horrible.
1 7 a 5 �b7 1 8 liJxb4 liJxb3 1 9 �xb3
c5!
This wins a piece due to the pin on
the knight.
20 �d 5 cxb4 2 1 �xe 5 f6 0 - 1
White players will obviously not
want to repeat this experience. A re­
cent game has shown a more interest-
M o v e - O r d e rs a n d Tra n sp o s i t io n s
ing path for White.
Game 75
Hjartarson-Gulko
Reykjavik 1 996
1 d4 d5 2 c4 c6 3 liJc3 dxc4 4 e4
b5 5 a4 b4 6 liJa2 liJf6 7 f3 e5 8
dxe 5 ! �xd 1 + 9 �xd 1 liJfd 7 1 0 e6 ! ?
Hjartarson suggests that 10 f4 iila6
1 1 iile3 f6 is unclear, but 10 . . . CLJcS
may be stronger, meeting 1 1 iilxc4 by
1 1 . . .iila6! 12 iilxa6 CLJbxa6 with ideas
of . . . CLJb3 and . . . CLJxa4 as well as
. . . CLJxe4.
1 0 . . . fxe6 1 1 itxc4
White has the more attractive pawn
structure and hence the better long­
term prospects, but Black's piece ac­
tivity and the slightly open position
of the white king should give him
equally good chances.
1 1 . . . ita6
Also possible was l 1 . . .CLJeS! ?
1 2 itxa6 liJxa6 1 3 ite3 it c 5 1 4 �e2
see follo wing diagram
1 4 . . 0-0-0?
.
Hjartarson says that Black should
have taken this opportunity to swap
12 1
Th e Sla v
off bishops by 14 . . . ii,xe3 1 5 �xe3
0-0-0 16 ctJh3 ctJe5 17 Mac 1 with mu­
tual chances. In the game, Black fails
to make the most of his chances and
falls into an unpleasant ending.
1 5 .l1i. g 5 ! ctJf6 1 6 b3 .l1i.d4 1 7 n c 1
�b7 1 8 ctJ h 3 e 5 1 9 �hd 1 c5 20 .l1i.e3
.l1i.xe3 21 nxd8 �xd8 22 �xe3 h 6 23
ctJf2 gd6 24 gc2 ctJd7 25 ctJc 1 h5
26 ctJfd3 g5 27 ctJ b 2 g4 28 ctJcd 3 !
gxf3 29 gxf3 g g 6 30 ctJc4!
White's pieces are ideally placed,
and Black can no longer hold his
pawn weaknesses.
30 . . . ng 1 31 nb2 ctJc7 3 2 ctJcxe5
ctJxe5 3 3 ctJxe5 ne 1 + 34 �f2 gh 1
3 5 �g2 nc 1 36 f4 ctJe6 37 f5 ctJg5
38 f6 �c7 3 9 f7 ctJ h 7 40 ctJg6 .!;i.e 1
1 22
4 1 e5 �c6 42 �f2 �e4 43 ne2 ng4
44 e6 1 -0
Nice technique from Hjartarson,
and a critical new idea for White.
In view of Ivan Sokolov's success
with 3 . . . dxc4 against 3 ctJc3 , you may
be wondering whether Black can play
3 . . . dxc4 against 3 ctJf3 but, amongst
others, the following game has put the
line under a cloud. It is so unbalanced,
however, that I would not be at all
surprised if there is a sneaky resource
for Black!
Game 76
Miles-Hodgson
Hastings 1 995/96
1 ctJf3 d 5 2 d4 c6 3 c4 dxc4
3 ctJf3 ii,f5 4 cxd5 cxd5 5 'i'b3 'i'c7
is tactically possible as 6 'i'xd5 allows
mate after 6 . . . 'i'xc 1+. The crucial dif­
ference with the 3 e3 line is that
White has not blocked in his dark­
squared bishop, which means that
after 6 ctJc3 e6, 7 ii,f4! is extremely
nasty: 7 . . . 'i'xf4 loses the rook on as
after 8 'i'xb7, while 7 . . . 'i'b6 8 'i'xb6
axb6 9 e3 , intending ii,b5+, �e2 and
then Mhc 1 to invade on the c-file,
gives Black a very depressing ending
to defend.
4 e3 b 5 5 a4!
see fol/o wing diagram
The main difference between 3 ctJc3
and 3 ctJf3 is that Black cannot ad­
vance his queenside pawns with
tempo against 3 ctJf3 . This gives
White plenty of time to undermine
them.

5 . . . e6 6 axb5 cx b 5 7 b3 iL b4+ 8
iLd2 iLxd 2+ 9 CLlbxd 2 a5 1 0 bxc4 b4
A very confusing situation: Black
has two passed queenside pawns while
White has more central control. The
essential conflict is whether Black can
get his pawns moving or whether
White can blockade them so that they
will become weak.
1 1 CLle 5 !
Threatening 12 ttJxf7! 'it>xf7 1 3
"Wf3+, winning the rook o n a8 .
1 1 . . . CLlf6 1 2 �a4+ !
This is the key idea, forcing Black's
pieces into a nasty tangle. 12 . . . ttJbd7
loses a pawn to 13 ttJc6, while
12 . . . �d7 1 3 ttJxd7 is exactly what
White wants: 13 . . . ttJbxd7 14 �e2 0-0
M o v e - O r d e rs a n d Tra n sp o s i tio n s
1 5 0-0 "¥IlIc7 16 �f3 ! (16 ttJb3? ttJe4!,
aiming for c3 , is to be avoided at all
costs) 16 . . . �a7 17 cS is very nice for
White.
1 2 . . . CLlfd7 1 3 c 5 ! 0-0 1 4 CLlec4!
Exchanges help Black free himself.
1 4 . . . CLlf6 1 5 iLe2 CLl d 5 1 6 iLf3 iLd7
1 7 �c2 iLb5 1 8 CLld6 iLa6 1 9 CLl2c4
ttJc6 20 0-0 Wifc7 21 Bfc 1 iLxc4 22
CLlxc4 CLlce7 23 �b3 Ba7 24 g 3 !
R b 8 2 5 �g 2 !
Very instructive: Black's pawns are
going nowhere, so White quietly im­
proves his position, removing the
possibility of back-rank mates and
making sure that if Black does queen,
it won't be with check.
25 . . . h6 26 Rc2 CL'lc3?
1 23
Th e Sla v

Allowing a combination, but Black b3 0-0 1 2 e4?

was struggling to find anything to do.
27 nxa 5 ! �xa 5 28 tLlxa5 �xa 5 29
�xc3 bxc3 3 0 �xb8+ \t>h7 3 1 �b7
f5 3 2 �xe 7 c 2 3 3 �xe6 c 1 � 34
�xf5+ g 6 35 �f7+ \t>h8 3 6 �f8+
\t>h7 37 .li d 5 !
Mate cannot be averted.
3 7 . . . �h 1 + ! ? 3 8 �xh 1 1 -0

The next line is 3 tLlf3 tLlf6 4 e3. I

am surprised at how popular this
variation is, as it really doesn't prom­
ise White very much at all, and his
results have not really been that good.
Game 77
Kozu l - l il escas
Erevan Olympiad 1 996
1 d4 d 5 2 c4 c6 3 tLlf3 tLlf6 4 e3
.lif 5 !
I don't think that this needs any
comment!
1 2 . . . tLlc5 ! ! 1 3 dxc5 dxe4 ! 1 4 �xd8
�fxd8
The point: Black regains his piece
with a vastly superior position, as the
knight on f3 is tied to the defence of
the bishop on d2.
1 5 tLla4 exf3 1 6 �fd 1 �d3 1 7 itc3
:ad8 1 8 ne1 :8d7 1 9 :ac 1 fxg2
20 tLlb2 �f3 21 \t>xg2 z:i.f5 2 2 t2Ja4
�d3 23 .l:!.cd 1 .l:i:df3 24 �e2 h 5 ! 25
h3 g 5 !
This kingside advance finishes off
the game.
26 b4 axb4 27 axb4 g4 28 hxg4
tLlxg4

5 .lid 3
5 cxd5 cxd5 6 iVb3 is the other way
to play, leading to positions very simi­
lar to Game 72.
5 . . . .lixd3 6 �xd3 e 6 7 0-0 tLlbd7 8 Now f2 must fall.
tLlc3 1I.. b 4 9 1I.. d 2 a5 1 0 a3 .lie7 1 1 29 .l:i:h 1 :xf2+ 30 nxf2 tLlxf2 3 1

1 24

J:!.a 1 h4 3 2 tZ:lb6 h3+ 33 �h2 J::!. g 5 34
J:!.g 1 .l:!.xg 1 3 5 �xg 1 tZ:lg4 36 tZ:lc8
�f8 0 - 1
And finally, 4 'iVb3 and 4 'iVc2. In
both case White's queen protects c4
and prevents the light-squared bishop
from developing safely: 4 'iVb3 attacks
b7, while 4 'iVc2 covers the f5-square.
However, these moves do nothing to
further White's development, while
exposing the white queen to attack by
Black's minor pieces. Although nei­
ther line promises much, they are
both popular with positional players
who wish to avoid any sharp options.
Recently, black players have been
trying a Semi-Slav approach, 4 . . . e6 5
g3 dxc4 6 'iVxc4 b5 7 'iVc2 j,b7, with
. . .tDbd7 and . . . c6-c5 to follow, but as
befits a book on the Slav, I will con­
centrate on our beloved plan of de­
veloping the light-squared bishop out­
side the pawn chain.
Game 78
Akopian-Shirov
Wijk aan Zee Open 1993
1 d4 d 5 2 tZ:lf3 tZ:lf6 3 c4 c6 4 �b3
dxc4
The most solid and reliable option.
Black diverts the queen from its attack
on b7, thereby allowing the light­
squared bishop to develop to f5. I
suppose that 4 . . . a6 is possible here, as
5 j,f4 (preventing . . . l:!.a7!) 5 . . . b5 6
cxd5 cxd5 7 a4 b4! 8 CtJbd2 (8 'iVxb4
e5! wins a piece) 8 . . . CtJc6 9 l:!.c 1 j,b7
10 CtJe5 (to remove the knight on c6,
which both guards b4 and blocks the
c-file) is met by 10 . . . CtJa5! (10 . . . CtJxd4
M o v e - O rd e rs a n d Tra n sp o s i tio n s
1 1 'iVxb4 attacks b7 and d4) 1 1 'iVxb4
e6 12 'iVc3 l:!.c8 ! , winning the queen, as
13 'iVd3 allows 1 3 . . . l:!.xc1+ mate! In-
stead of 6 . . . cxd5, 6 . . . CtJxd5! ? (hitting
the bishop on f4) 7 j,g3 e6!?, intend­
ing a quick . . . c6-c5, is also interesting.
Note that 5 CtJc3 transposes to 4 CtJf3
a6 5 'iVb3 .
5 �xc4 .itf5 6 tZ:lc3 tZ:lbd7 7 g 3 e6 8
.itg2 .ite7 9 0-0 0-0 1 0 e 3
Be warned! The 'natural' 10 l:!.d1
loses the exchange to 10 ... j,c2!, as 1 1
l:!.d2 allows 1 1 . . .CtJb6! , trapping the
queen!
1 0 . . . tZ:le4 1 1 �e2
A novelty. 1 1 CtJd2 is the old move,
when theory recommends 1 1 . . .CtJxd2
12 j,xd2 e5! 13 e4 cxd4 14 exf5 dxc3
15 j,xc3 j,f6! 16 l:!.ad 1 ? ! j,xc3 1 7
bxc3 'iVc7 with equality. White must
consider 16 j,b4!?, retaining the
bishop pair.
1 1 . . . tZ:lxc3 1 2 bxc3 .ite4 !
A typical manoeuvre, preventing
e3-e4 and neutralising White's bishop
on g2.
1 3 c4 c5 1 4 ld:.d 1 wtic7
see fol/o wing diagram
1 5 .itb2 tZ:l b6 ! ?
1 25
Th e Sla v
The start of an eccentric plan from
Shirov. 1 5 . . . MfdS , intending 16 . . . MacS
and then . . . ct:Jd7-bs-c6 to pressurise c4
and d4, was also interesting.
1 6 gac 1 jfLf6 1 7 jfLf 1 !
To chase the bishop from e4 with­
out allowing the exchange of bishops.
1 7 . . . ltJa4 ! ? 1 8 jfLa 1 .l::i. f d8 1 9 ltJd2
jfLc6 20 ltJ b 3 b 6 21 d5
21 . . . jfLxa 1 2 2 dxc6 jfLf6
ltJc3 24 .l::i. x d8+ gxd8 2 5 �c2 gd6
26 .!de 1 gxc6 27 jfLxc6 �xc6 28
ltJd2 a6 29 ltJ b 1 ltJe4 30 h4 b 5 3 1
cx b 5 axb5 3 2 ltJ d 2 ltJc3 3 3 ltJ b 1
%-%
ltJe4 3 4 ltJ d 2 ltJc3
I am a little surprised that White
accepted the draw here. Although his
opponent has good counterplay,
126
White i s material up, after all!
Game 79
Razuvaev-Sturua
Erevan Open 1 996
1 d4 ltJf6 2 c4 c6 3 ltJf3 d 5 4 �c2
g6
A very sensible idea. Black allows
his bishop to develop to g7 whilst
supporting . . . �cS-f5 , attacking the
queen. 4 . . . a6 5 �f4 b5?! 6 cxb5 cxb5 7
ct:Jbd2! (intending ct:Jb3) is best
avoided, as White's pieces are well
placed to exploit the queenside dark­
square weaknesses, so Black should
try 5 . . . dxc4 6 'iVxc4 e6 7 e3 b5 S 'iVc2
�b7, intending . . . ct:Jbs-d7 and . . . c6-c5.
Finally, 4 . . . dxc4 5 'iVxc4 transposes to
the previous game.
5 �f4 �g7
23 jfLg2 5 . . . ct:Ja6!? is considered in the next
game and 5 . . . dxc4 is also good: 6
'iVxc4 �g7 7 e3 (7 ct:Jc3 0-0 S e4 b5
leads to a typical Griinfeld position
which, though reasonable for Black,
may not appeal to pure Slav players)
7 . . . 0-0 S �e2 �e6 9 iVc1 ct:Jbd7 10 0-0
c5! 1 1 ct:Jc3 ct:Jd5 ! 12 Mdl MCS 13 ct:Jxd5
�xd5 14 dxc5 ct:Jxc5 15 �c4 ct:Jd3 ! 16

g,xd3 g,xc4 1 7 'iVd2 e6 gave Black no
problems in Goldin-Yusupov, Tilburg
1992. However, s . . . iiJs 6 'iVb3 'iVb6 7
cS 'iVxb3 S axb3 is a touch better for
White, and not very exciting for
Black.
6 e3 0-0 7 tLlc3 ilLe6
S tLlg 5
S 'iVb3 dxc4 9 �xc4 (9 'iVxb7 ctJdS ! ? ,
intending to trap the queen i n the
corner after 10 'iVxaS 'iVb6, is ex­
tremely murky but not worse for
Black) 9 . . . �xc4 10 'Vi'xc4 ctJds 1 1 �g3
ctJa6, intending . . . ctJdS-b6 and . . . c6-cS
is about equal.
S . . . ilLf5 9 �b3 �b6 1 0 ilLe2 tLlbd7
1 1 tLlf3
The endings with 1 1 cS 'iVxb3 12
M o v e - O r d e rs a n d Tra n sp o s i tio n s
axb3 are improved for Black, as the
white knight is misplaced on gS,
which makes . . . e7-eS easier to achieve.
1 1 . . . ilLg4 1 2 cxd 5 tLlxd 5 1 3 tLlxd 5
cxd5 1 4 �xb6 tLlxb6 Y2 - Y2
And finally, an amazing sacrificial
line in this most solid of openings!
Game 80
Alburt-S habalov
USA Championship 1996
1 d4 d 5 2 c4 c6 3 tLlf3 tLlf6 4 �c2
g 6 5 ilLf4 tLla6 ! ? 6 e3 ilLf5 7 �b3
tLlb4 ! !
A typically inventive idea of Ku­
preichik.
S �xb4 e 5 9 �x b7
9 cS cxf4 10 exf4 b6 1 1 ctJeS bxcs 12
'iVb7 �d7 13 ctJxd7 ctJxd7 14 'iVxc6
g,cS 15 'iVxds cxd4 16 �bs �b4+
(16 . . . g,c1+ 17 �d2 g,xh 1 1S 'Vi'eS+!
'iV e7 19 'Vi'xhS+ 'Vi'fs 20 'Vi' eS+ leads to
a draw by repetition) gave Black a
powerful lmtlatlve in Epishin­
Kupreichik, Russia 1989.
9 . g b S 1 0 �xc6+ ilLd7
. . 11 �xf 6 !
�xf6 1 2 ilL x e 5 � b 6 1 3 b3 ilL b4+ 1 4
tLlbd2 0-0 1 5 ilLxbS J::!. x bS
127
Th e Sla v

wmmng.
1 6 . . . 'ii' a 5 1 7 �c4 �g4 1 8 0-0 �xd2
White has some pawns, but Black
has the big guys!
1 9 CLJe5 �f5 20 a3 'ii' c 3 21 d 6 �g7
22 CLJxf7 .l::!. b 6 23 e4 �c8 24 .l:i:fd 1
�f4 25 e 5

1 6 cxd 5
A strange move to play since
Shabalov had already won a convinc­
ing game in this line. 16 i,d3 is the
theoretical recommendation, meeting
16 . . . i,g4 with 17 �e2 ik'a5
(threatening . . . i,b4xd2) 18 �hd l ,
which i s a bit of a mess. 17 O-O!? seems 25 . . Jbb3 26 � x b 3 'ii'x b3 27 CLJd8
more natural, to counter 17 . . . ik'a5 'ii' b 6 28 g 3 'ii' x d8 29 g xf4 'ii' h 4 30
with 18 h3 ! i,h5 (18 . . . i,xd2 19 hxg4 f3 'ii'x f4 3 1 �f2 �b7 3 2 .l:i:d3 �a6
i,c3 20 �ac 1 i,b2 2 1 �c2 ik'xa2 22 3 3 d7 'ii'x h2+ 34 �e3 'ii' h 6+ 35 �e2
4::l d2! [intending �b l] 22 . . . �xb3 23 'ii' h 4 3 6 �d2 �f7 37 d5 �e7 38 e6
4::lx b3 ik'xb3 24 �b l ik'xd3 25 �cxb2 'ii'f 2+ 39 �c3 'ii' c 5+ 40 �d2 �xd3
dxc4 26 �c 1 , intending �bc2 with an 41 .l::!. c 1 'ii' d 4 42 .l::!. c 8 � b 5+ 43 �c2
advantage) 19 a3 ! i,xd2 20 b4! i,xb4 �xd7 0- 1
21 axb4 ik'xb4 22 �ab l ik'f8 23 g4! , A fascinating game.

1 28
 M o v e - O rd e rs a n d Tra n sp o s i tio n s

S u m mary

3 e3 JLfs equalises for Black; 3 CDc3 dxc4 is a very interesting line and only
Hjartarson-Gulko (Game 7S) is a possible attempt to play with White; 3 CDf3
dxc4 is very unbalanced but seems to be good for White; 3 CDf3 CDf6 4 e3 JLfs
is nothing for White; while 3 CDf3 CDf6 4 'iVc2 and 4 'iVb3 are also nothing spe­
cial. Hence 3 CDf3 CDf6 4 CDc3 dxc4 is the most accurate order for both sides.

1 d4 d 5 2 c4 c6

3 e3
3 CDc3 dxc4 4 e4 bs S a4 b4 6 CDa2 CDf6 7 f3 eS (D)
8 JLxc4 game 74 -

8 dxeS game 75 -

3 CDf3
3 . . . dxc4 game 76-

3 . . . CDf6
4 e3 JLfs game 77
-

4 'iVb3 game 78
-

4 'iVc2 g6 S JLf4 (D)

S . . . JLg7 game 79
-

S . . . CDa6 game 80
-

3 . . . i1Lf5 4 �b3 �c7 5 cxd 5 cxd 5 (D)

6 CDc3 game 72
-

6 JLbS + game 73 -

7 . . . e5 5 �f4 5 . . . cxd5

1 29
 CHA PTER ELEVEN

Odd s and Ends

1 d4 d5 2 c4 c6
In this chapter, we take a brief look
at all the sidelines that I couldn't fit
into the main chapters!

A) T h e Winawer Cou nter­

G a m b i t : 3 CL'lc3 e5

This gambit was all the rage four or

five years ago, but the following game
somewhat dampened the ardour of
the black players.
7 . . . \'l;Vd 6
Game 81 Black has also tried 7 . . . cS, but after
. Kasparov-N ikolic 8 CL'lf3 Vic7 9 e3 dxe3 1 0 �xe3 CL'lf6 1 1
Manila Olympiad 1 992 �d3 White i s slightly better due t o his
lead in development.
1 d4 d 5 2 c4 c 6 3 ltJc3 e 5 4 dxe5 8 ltJf3 ltJf6 9 \'l;Vc2 �e7 1 0 O-O-O!
d4 5 ltJe4 \'l;Va5+ 6 � d 2 ! 0-0
This natural move had been practi­
cally ignored prior to this game (6
CL'ld2 had been the main line) .
6 . . . . \'l;Vxe 5 7 ltJg 3 !
This move looks obvious, but it
was Kasparov's new idea. Of course,
the genius is not in the move itself,
but in the astonishing attacking plan
that flows from it. It is not easy to
guess that White is gearing up for
kings ide attack!

1 30

1 1 e3
Azmaiparashvili, a long-time ana­
lyst for Kasparov, played 1 1 jLc3
agamst Eslon in Seville 1994.
1 1 . . .dxc3 !? 12 �xd6 cxb2+ is given as
unclear by ' Gazza' , but 'Azmai'
claims an advantage after 1 3 'iVxb2
jLxd6 14 e4 (threatening e4-e5)
14 . . . jLf4+ 15 'It>b 1 . This assessment is
objectively correct, as White's mate­
rial advantage, coupled with the threat
of e4-e5, driving away the knight on
f6 and exposing the bishop on f4 to
attack by White's pieces ('iVb2-d4,
ctJg3-h5) , should tell in the end. How­
ever, in a practical game White's dark­
square weaknesses and exposed king
give Black definite counter-chances.
Eslon played 1 1 . . .'iVf4+, unpinning,
and White sacrificed a pawn with 12
e3 dxe3 13 fxe3 (13 �d4!? looks very
strong, meeting 13 . . . 'iVh6 with 14
�h4! 'iVg6 1 5 jLd3 ! , trapping the
queen, and 13 . . . 'iVc7 with 14 jLd3 ! ? or
just 14 fxe3 with very dangerous at­
tacking play) 1 3 . . . 'iVxe3+ 14 'It>b 1 ctJa6
15 ctJd4 (aiming to put a knight on f5
and threatening �de 1 , winning the
bishop on e7) .
Now after 1 5 . . . jLb4 16 ctJdf5 jLxf5
O d ds a n d En ds
17 ctJxf5 'iVe6 18 jLd3 White had a
superb attacking position for the
pawn. This looks very smooth and, in
his annotations, Azmaiparashvili
comments that if Black tries to im­
prove with 15 . . . ctJb4 (instead of
. . . jLb4) then 16 'iVb3 c5 17 �e 1 wins
after 17 . . . 'iVgS 18 �xe7 cxd4 19 jLxb4,
but in fact 16 . . . jLcS! is extremely good
for Black, as 17 jLxb4 'iVxb3 18 ctJxb3
jLxb4 and 17 �e 1 'iVf4! (threatening
. . .jLxd4) 18 jLxb4 'iVxd4 both leave
Black simply a pawn up . The position
is extremely risky for Black, of
course, and I would not recommend
this sort of position against a good
attacking player, but Black's tactical
chances must not be underestimated.
1 1 . . . dxe3 1 2 fxe 3 !
This is the key to the whole idea as
12 jLxe3 'iVc7 promises nothing for
White. The text allows the bishop on
d2 to move with tempo to the attack­
ing diagonal a1-h8.
1 2 . . .'�c7
12 . . . ctJa6 is considered in the next
game.
1 3 SLc3 SLg4? ! 1 4 SLd3 tLlbd7 1 5
SLf5 !
Naturally, White wants to occupy
131
Th e Sla v
the f5-square with his knight.
1 5 . . . 1l.xf5 1 6 CLlxf5 ZUeS 1 7 CLlxg7 ! !
�xg7 1 S �f5 CLlfS 1 9 h4!
In order to chase away the knight
on fs if it should surface on g6.
1 9 . . . h 6 20 g4? !
20 �g4! iDg6 2 1 h5 was even
stronger according to Kasparov.
20 . . :�cS 21 �xcS .!:!:axcS 22 g 5 !
The ending i s still very unpleasant
for Black, and Kasparov powers
through with his customary energy.
22 . . . CLlSd7 23 e4 �cdS 24 � d f 1 �fS
25 gxf6 1l.xf6 26 e5 1l. g 7 27 .i:i.hg 1
c5 2S �c2 .!:!:e6 29 .!:!:g4 1l.hS 30 b4
b6 3 1 bxc 5 bxc5 32 .!:!:b 1 �a6 3 3
.!:!:b2 1l. g 7 34 �b7 .i:i.xa2+ 3 5 � b 3
1;!a6 3 6 e6 .l::!. x e6 3 7 .!:!: x g 7 1 -0
1 32
If Black wishes to play this line, he
needs to find an improvement on
12 . . . �c7.
Game 82
Rogozenko-Bets
Moldovan Championship 1 994
1 d4 d 5 2 c4 c6 3 CLlc3 e 5 4 dxe5
d4 5 CLle4 �a5+ 6 1l.d2 �xe 5 7 CLlg3
CLlf6 S CLlf3 �d 6 9 �c2 1l.e7 10
0-0-0 0-0 1 1 e3 dxe3 1 2 fxe3
Following Kasparov's example.
1 2 . . . CLla6
12 . . . .sdS is possible, preventing 13
Jtc3 as 13 . . . �xd1+ 14 �xd1 .sxd1+ 15
'ii;>x d1 exchanges queens, destroying
White's attacking possibilities.
1 3 1l.c3 �c7
Since this turns out so badly, Black
must consider 1 3 . . . �e6, keeping the
queen close to the kingside to help
with defence, while threatening
. . . iDb4 and . . . iDg4 as well as . . . �xe3+.
14 a 3 !
see follo wing diagram
Preventing 14 . . . iDb4, activating the
knight.
1 4 . . . CLlg4 ! ? 1 5 '!:!:e 1 1l.f6

This manoeuvre, exchanging the
dangerous bishop on c3, briefly gave
black players hope in this line.
1 6 h 3 ! jLxc3 1 7 hxg4
Obviously 17 . . . .txe l loses to 1 8
iYxh7+ mate.
1 7 . . . h 6 1 8 �xc3 �xg3 1 9 g 5 ! hxg 5
20 jL d 3 �d8 2 1 l:!.h7 f6 22 l:Ieh 1
�c7 23 l2Jxg 5 �f8 24 c 5 ! jLg4 25
�xf6+! 1 -0
25 . . . gxf6 26 �h8+ cJi;e7 27 �lh7+ is
mate!
B) The S c h leeter S l a v :
3 CD c 3 CD f 6 4 e3 9 6
Question 1 : What is the idea behind
this system?
O dds a n d En ds
A nswer: This is another Smyslov
favourite. Black accepts a slight space
disadvantage and develops his king's
bishop on g7, avoiding . . .e7-e6 in or­
der to retain the option of developing
his bishop on c8 outside the pawn
chain if he wishes.
Game 83
G u l ko-Salov
Reykjavik (World Cup) 1991
1 l2Jf3 d 5 2 c4 c6 3 e3 l2Jf6 4 l2Jc3
g 6 5 d4 jLg7 6 jLe2 0-0 7 0-0 b6
A solid move, developing the
bishop to b7 to support Black's cen­
tre. The more active 7 . . . dxc4 8 .txc4
.tg4 is dealt with in the next game.
8 cxd 5 cxd 5 9 l2Je 5 !
When the central pawn structure
becomes fixed, the first side to gain a
central space advantage will be able to
claim the initiative.
9 . . . jLb7 1 0 jLd2 l2Jfd 7 ? !
10 . . . CDc6 was stronger, meeting 1 1
f4 with 1 1 . . .CDe8 ! , intending to de­
velop the knight to d6, when Black
only stands a little worse.
1 1 f4 ! f6 1 2 l2Jf3 ! ?
1 2 CDd3 i s also possible. Black has
1 33
Th e Sla v
not developed his pieces harmoni­
ously: the king's knight stops the
queen's knight from developing to d7
and has no moves of its own.
1 2 . . JH7 1 3 iLd3 CL:Jf8
1 4 f5? !
A little hasty according to Gulko,
who prefers 14 g4! e6 15 f5 .
1 4 . . . gxf5 1 5 iLxf5 e6 1 6 iLd3 CL:Je6
1 7 CL:Je2 �d6 1 8 CL:Jg3 CL:Jg6 1 9 CL:Jh5
iLh8 20 W'e2 gaf8 21 a 3 e 5 22 dxe5
fxe 5 23 CL:Jg5 gxf 1 + 24 .i:l:.xf 1 iLe8?
2 5 .i:l:.xf8+ W'xf8 26 iLe4!
A very nice move: 26 . . . dxe4 loses to
27 "ik'c4+.
26 . . . CL:Jee7 2 7 iLb4! iLb7 28 W'g4
\�i"c8 29 W'f3 �f8 3 0 �h3 �e8 3 1
CL:Jf6+ �g7 32 W'xh 7+ �xf6 33
1 34
iLxe7+ CL:Jxe7 34 W'h6+ CL:Jg6 35
W'xg6+ � e 7 36 W' h 7 + �d6 3 7 CL:Jf7+
rt;e7 3 8 iLf5 1 -0
Game 84
D reev-Piket
Dortmund 1994
1 d4 CL:Jf6 2 e4 e6 3 CL:Jf3 d 5 4 e3 g6
5 CL:Je3 iLg7 6 iLe2
White can try 6 �d3 (preventing
. . . �f5) 6 . . 0-0 7 h3 (preventing . . . il,g4)
.
if he wants to prevent the plan in the
game, although 7 . . . c5 is an interesting
reply. The game transposes to a quiet
variation of the Griinfeld, in which
White has played the useful, though
hardly earth-shattering, extra move
h2-h3 .
6 . . . 0-0 7 0-0 dxe4 ! ? 8 jLxe4 iLg4
This is a much more active idea:
Black will follow up with . . . 4Jbd7 and
a quick . . . e7-e5.
9 h 3 iLxf3 1 0 W'xf3 CL:Jbd7 1 1 iLb3
The critical line is 1 1 gd1 e5 12 d5
e4!? 13 4Jxe4 4Jxe4 14 "ik'xe4 4Jb6 15
�b3 (threatening d5xc6; Bareev sug­
gests 15 gb 1 !?, protecting b2 and seek­
ing to avoid the time-loss with �c4-
b3xd5 as in the game) 15 . . . cxd5 16
 O dds a n d En ds

,�xds CiJxds 17 Mxds 'i'b6. ity and gains space.

1 8 '¥fie2 tLl b 6 ! 1 9 gxd8 gxd8 20 ge 1
tLlc4 !
Imperceptibly, White has drifted
into big trouble: Black controls the
central dark squares and White's
queenside is an easy target.

Question 2: What is going on here?

A nswer: Black has sacrificed a pawn
for a lead in development and pres­
sure against the b2-pawn. However,
this play is sufficient to regain the
pawn, but not to achieve complete
equality: i s 'i'd3 MadS 19 e4! itd4 20 2 1 jLc 1 tLle6 22 '¥fif 1 '¥fic5 23 jLb3
Wh l itxf2 2 1 ith6 Mxds 22 exdS Mds tLld2 24 jLxd2 gxd2 2 5 jLxe6 fxe6
23 Mdl 'i'd6 24 'i'c3 , intending 'i'g7+ 26 e 5 gxb2 27 tLle4 '¥fixe 5 28 '¥fid3
mate, when White's passed d-pawn IWd 5 29 IWg3 jLe5 3 0 f4 jLd4+ 3 1
and Black's weak kingside gave White �h2 c 5 3 2 \\!,Vh4 gxa2 3 3 \\!'ve7 h 5 !
a slight advantage in Bareev-Kramnik,
Novgorod 1994.
1 1 . . . e 5 1 2 gd 1 '¥fi e 7 1 3 e4? ! exd4
1 4 gxd4 gad8 1 5 jLe3 tLlc5 1 6 jLc2
tLlfd7 1 7 gdd 1 b 5 !

3 4 \\!'ve8+ � g 7 3 5 \\!'ve7+ � h 6 3 6 h4
IWf5 3 7 tLlg3 \\!'vxf4 38 gxe6 g a 1
Here White lost on time.

What I like about the Schleeter Slav

Black activates his queenside major- is its flexibility: on move 7, Black has

1 35
Th e Sla v

a huge range of plans. We have seen:

a) 7 . . . b6 reinforcing the centre by
fianchettoing the other bishop; and
b) 7 . . . dxc4 8 i1Lxc4 i1Lg4 to break
quickly in the centre by means of
. . . tLlbd7, . . . i1Lxf3 and . . . e7-e5; but
Black can also try
c) 7 . . . a6 to expand on the queenside
with . . . b7-b5, before or after . . . d5xc4;
and
d) 7 . . . tLle4!? to unbalance the pawn
structure with . . . tLlxc3 . Bates-Sadler,
British Championship 1995, contin­ 5 . . . b5 6 a4 b4 7 CLJb 1
ued 8 'iVb3 b6 9 tLlxe4 (9 cxd5 tLlxc3 ! 7 tLla2, attacking the b4-pawn, re­
10 bxc3 cxd5 intending . . . tLlc6-a5, hit­ gains the pawn by force, but seriously
ting the queen) 9 . . . dxe4 10 tLld2 f5 1 1 misplaces the knight. The text is more
f3 e5 1 2 dxe5 exf3 ! 1 3 tLlxf3 tLld7, re­ ambitious.
gaining the pawn with a better pawn 7 . . . 1La6 8 1Le2
structure; 8 tLlbd2 regains the pawn, but after
. . . and the list continues! There 8 . . . c3 9 bxc3 i1Lxf1 10 tLlxfl bxc3
really is huge scope for personal ideas. Black will break with . . . c6-c5 and
equalise.
C) 3 tLlf3 tLlf6 4 tLlc3 dxc4 8 . . . c 5 ! 9 0-0 CLJc6 1 0 dxc5 CLJ a 5 ! 1 1
5 e3 CLJbd2 e6 1 2 b3 1Lxc5 1 3 1Lb2 c3 1 4
1Lxa6 cxb2 1 5 1L b 5+ cI;e7
This was one of my favourites when I
was little, but I'm not quite sure why
I decided to play it against Bareev! It is
actually not a very promlSlng con­
tinuation.

Game 85
Sadler-Bareev
Hastings 1992/93

1 d4 d 5 2 c4 c6 3 CLJf3 CLJf6 4 CLJc3

dxc4 5 e3
The king is very safe in the centre.
see fol/o wing diagram
Bareev rapidly outplays me, but
Trying to manage without the around the time control begins to ask
standard 5 a4, which prevents the ad­ too much of his slight advantage.
vance . . . b7-b5. 1 6 .i:t b 1 a6 1 7 1Le2 �b6 1 8 .i:txb2

1 36

l2Jd 5 1 9 l2Jc4 l2Jxc4 20 .ltxc4 l2Jc3
21 �c2 �c7 2 2 �c 1 Ithd8 23 .l::i. d 2
a 5 2 4 � c 2 g 6 2 5 .l::i. c 1 .ltd6 26 g 3
.ite5 27 .itf 1 h 5 28 .it c 4 �c 5 29 h 4
.lt f 6 30 .itf 1 l:! a c 8 3 1 .l::i. x d8 l:!xd8 3 2
l2Jd2 .l::i. c 8 3 3 l2J c 4 g 5 34 hxg 5 �xg 5
3 5 � h 7 � h 8 3 6 � d 3 .i:!. d 8 3 7 �h7
.l::i. h 8 3 8 �d3 �c5 39 J:!c2 l:!d8 40
�h7 h4 41 gxh4 Rh8 42 �d3 lad8
43 �h 7 .l::i. h 8 44 �d 3 .i:!.g8+ 45 .ltg2
.l::i. d 8 46 �h7 .l::i. h 8 47 �d3 :ad8 48
�h7 l2Jxa4?
Too risky. After the tactics, Black
only just manages to hold on to the
draw.
49 .l::i. a 2 ! l2J b 2 50 l2Jxa5 �c 1 + 5 1
c;t>h2 lah8 5 2 l2Jc6+ �xc6 5 3 l:!a7+
c;t>d6 54 �xh 8 .itxh8 5 5 .ltxc6 c;t>xc6
5 6 .l::i. x f7 l2Jd3 5 7 h5 l2Jc5 58 h6
l2Jxb3 5 9 Itf8 .ite5+ 60 f4 .itb2 6 1
e4 l2Jc5 6 2 e 5 l2Jd7 6 3 h 7 l2Jxe5 64
fxe 5 .itxe5+ 65 c;t>g2 b3 66 h 8�
.itxh8 67 J:!xh8 c;t>c5 68 c;t>f2 c;t>c4 69
c;t>e2 b 2 7 0 lab8 c;t>c3 71 c;t>d 1 e5
YZ - YZ
D) T h e S l a v G a m bit :
3 tL'lf3 tL'lf6 4 tL'lc3 dxc4
5 e4 b 5 6 e 5 tL'ld5 7 a4 e6
O d ds a n d En ds
As far as I am concerned, 5 e4 just
loses a pawn, but some die-hards just
keep on playing it! Foremost amongst
them is the attacking Swedish player
Tiger Hillarp-Persson. So here is one
of his typical efforts.
Game 86
Hillarp-Persson - Acs
Budapest 1996
1 c4 c6 2 l2Jf3 l2Jf6 3 d4 d 5 4 l2Jc3
dxc4 5 e4 b 5 6 e 5 l2Jd 5 7 a4 e6 8
l2Jg 5 ! ?
The modern line. 8 axb5 tL'lxc3 9
bxc3 cxb5 10 tL'lg5 .i,b7 1 1 'i'h5 g6 1 2
'i'g4 .i,e7 i s the old continuation,
when Black will follow up with
. . . tL'ld7-b6-d5, while White plays for
tricks!
8 . . . .lte7 9 h4 ! ? h6 1 0 l2Jge4 b4
10 . . . .i,a6 immediately is interesting,
to avoid weakening the queenside too
early. The text leads to absolute
chaos, though I think that Black is
fine.
1 1 l2J b 1 .lta6 1 2 �g4 g6 1 3 l2Jbd2
c3 14 l2Jc4 .itxc4 1 5 .itxc4 a 5 1 6 h 5
g 5 1 7 0-0 l2Jd7 1 8 bxc3 l2J7b6 1 9
.ltb3 l2Jxc3 20 l2Jxc3 bxc3 2 1 �e4
137
Th e Sla v
tiJ d 5 22 f4 gb8 23 ll.a2 gb4 24
fxg 5 �b6 ! 25 ll.xd 5 exd 5 26 �f3
IWxd4+ 27 ll.e3 �xe 5 28 IWxf7+
28 . . . �d8 29 gae 1 l:!e4 30 �b6+
� d 7 3 1 l:!xe4 �xe4 3 2 ll.e5 J:!. h 7 33
�f2 ll.xg 5 34 J:!.e1 �f 5 3 5 �e2 e5
3 6 �b5+ �e8 3 7 �e6+ ge7 38
IWa8+ � d 7 39 IWxd 5+ �e8 40 �a8+
�d7 41 gd 1 + ll.d2 42 �d5+ �e8
43 gf 1 .liLf4 44 �g8+ �d7 4 5 �d5+
1h - 1h
An amazing game!
E) 3 tLl f 3 tLl f 6 4 tLl c 3 dxc4
5 a4 �f5 6 tLlh4 ! ?
This i s an interesting little idea that
has even been tried by Kasparov.
1 38
Here 6 . . . �c8 !? 7 tLl£3 takes us back to
square one, but 7 e3 e5 8 ilxc4 exd4 9
exd4 ile7 10 0-0 0-0 1 1 h3 tLla6!? 12
ct'lf3 ct'lc7!, intending ... ile6, leads to
an unclear position. In the next game
we see Black allowing the exchange
on f5.
Game 87
Savchenko-N inov
Cappelle La Grande Open 1 994
1 d4 d 5 2 e4 e6 3 tiJf3 tiJf6 4 tiJe3
dxe4 5 a4 .liLf5 6 tiJh4 e6 7 tiJxf5
exf5 8 e3 .liLd6
This cannot be bad, but I wonder
whether Black cannot develop more
effectively. White's basic idea is to
play iVf3 , h2-h3 and g2-g4, removing
the f5-pawn and thus undermining
Black's control of e4. In order to pre­
vent this plan, I would therefore sug­
gest protecting the f5-pawn with
. . . iVd7, putting the bishop sensibly on
e7 and then developing the queen's
knight to b4 via a6. This line is of
course very similar to Yusupov­
Kramnik from the 6 e3 main lines,
but since White has taken on £5 so
early, allowing Black to prevent e3-e4

with a pawn on fS , rather than his
pieces, Black has more flexibility with
his piece placement: S . . . aS 9 iLxc4
CUa6 10 'i'f3 'i'd7 1 1 h3 CUb4 12 0-0
,YLe7 is fine for Black.
9 1Lxe4 0-0 1 0 0-0 CL.lbd7 1 1 a 5 a6
1 2 h 3 h5 1 3 �f3 g 6 1 4 e4! CL.lxe4
1 5 CL.lxe4 fxe4 1 6 �xe4
The move . . . h7-hs has rather weak­
ened the black kingside and now
'i'xg6+ is threatened.
1 6 . . . wg7 1 7 1Ld2 CL.lf6 1 8 �f3 CL.l d 5
1 9 gfe 1 1L e 7 20 1Lxd 5 exd 5 2 1 1L b4
1L d 6 22 � e 5 �xe5 23 dxe5 �d7 24
gad 1 gad8 25 gd4 gfe8 26 Uxe8
�xe8 27 �e3 �xe3 28 fxe3
White's space advantage gives him a
very pleasant rook ending.
O d ds a n d En ds
28 . . . Wf6 29 gb4 gd7 30 gb6+ We5
31 Wf2 ge7 3 2 b4 we4 3 3 Ud6 h4
34 We2 g 5 3 5 Wd2 ge7 36 e6 bxe6
3 7 gxe6 ga7 38 gd6 gb7 39 gb6
ga7 40 gb8 ge7 4 1 b 5 axb5 42
gxb5 ga7 43 Ub4+ We5 44 £a4
Wd6 45 a6 We5 46 We3 f 5 47 ga5+
Wb6 48 gxd 5 Wxa6 49 gxf5 £g7 50
Wd3 Wb6 51 We2 g4 5 2 gh 5 gxh3
5 3 gxh3 £a7 54 Wd3 We6 5 5 gxh4
Wd5 56 Uh5+ We6 5 7 We4 ga2 58
gh6+ Wf7 59 Wf4 1 -0
F) 3 iLf4
With 3 iLf4 White intends e2-e3,
when he will have safely developed
his queen's bishop outside the pawn
chain.
Game 88
Psakhis-Sadler
Megeve (peA rapidplay) 1 994
1 d4 d 5 2 e4 e6 3 �f4 dxe4!
Basically winning a pawn.
4 CL.le3 e6 5 e3 b5 6 �f3 �a5 7 g4
b4 8 CL.le4 b3+ 9 CL.le3 1Lb4 1 0 CL.lge2
�xa2 !
Now after 11 Mxa2 bxa2 the pawn
1 39
Th e Sla v

queens! ttJa4 26 �e 1 �xc3+ 27 � O .:gxg7

1 1 g c 1 �xb 2 1 2 e4 ttJf6 1 3 g 5 e 5 28 �xe6+ fxe6 29 �e4 exd 5 30
1 4 g x f 6 exf4 1 5 �xf4 � a 3 1 6 �e3 �e6+ �b8 3 1 ttJg3 �c2 3 2 �g2
b2 1 7 .:g b 1 �a2 �e4+ !
Threatening 1 8 . . . 'iVxb l+.
1 8 �d 1 �b3+ 1 9 �d2 ttJa6 20 fxg 7
gg8 2 1 �h6 �e6 2 2 � h 3 ttJ c 5 !

Making things safe!

33 �xe4 dxe4 34 .:ghd 1 gxd 1 35
gxd 1 �e5 0 - 1
23 �xh 7 0-0-0 2 4 d 5 cxd 5 2 5 exd 5 Not a good advert for 3 �f4!

1 40
 O d ds a n d En ds

S u m m a ry

The Schleeter variation is a reasonable alternative to 3 . . . dxc4 after 3 CDc3 .

However, note that Black can only play this line after e2-e3 by White, as 3
CDf3 CDf6 4 CDc3 g6 5 cxds cxdS 6 itf4! leads to a superior version of the Ex­
change variation, which is rather depressing for Black.
White players looking for an offbeat alternative might care to examine 6
CDh4!?

1 d4 d 5 2 c 4 c 6

3 i2J c 3
3 CDf3 CDf6
4 CDc3 dxc4 (D)
5 e3 game 85
-

5 e4 game 86
-

5 a4 itfS 6 CDh4 game 87

-

3 itf4 dxc4 - game 88

3 . . .e5
3 . . . CDf6 4 e3 g6 5 CDc3 itg7 6 ite2 0-0 7 0-0 (D)
7 . . . b6 - game 83
7 . . . dxc4 game 84
-

4 dxe5 d4 5 i2Je4 'iVa5 + 6 3Ld2 'iVxe 5 7 i2Jg3 i2Jf6 8 i2Jf3 'iVd6 9 'iVc2 3Le7
1 0 0-0-0 0-0 1 1 e 3 dxe3 1 2 fxe3 (D)
1 2 . . . iVc7 - game 81
1 2 . . . CDa6 - game 82

4 . . . dxc4 7 0-0 1 2 fxe 3

1 41
 INDEX OF COMPLETE GAMES

Adianto-Kramnik, London (Intel Grand Prix) 1994 49

Akopian-Shirov, Wijk aan Zee Open 1993 125
Alburt-Shabalov, USA Championship 1996 127
Andersson-Epishin, Ter Apel 1 995 111
Atalik-Miles, Hastings 1 995/96 95
Bacrot-Smyslov, A lbert {sixth match game} 1 996 69
Beliavsky-Shirov, Erevan Olympiad 1996 1 04
Benz-Gretarsson, Oberwart Open 1 996 64
Cifuentes-Dreev, Wijk aan Zee 1 995 1 02
Cifuentes-Leyva, Cienfuegos 1 996 115
Dautov-Nikolic, Ter Apel 1 994 77
Dreev-Piket, Dortmund 1 994 134
Ehlvest-Schwartzman, New York Open 1996 23
Epishin-Pomes, Manresa 1 995 75
Gelfand-Lautier, Zurich 1 994 30
Gelfand-Nikolic, Manila Interzonal 1 990 82
Gelfand-Shirov, Dortmund 1 996 55
Gofshtein-Sadler, Ischia 1 996 27
Gulko-Salov, Reykjavik (World Cup) 1991 133
Hillarp-Persson - Acs, Budapest 1996 137
Hjartarson-Gulko, Reykjavik 1 996 121
Hodgson-Sadler, Ischia 1 996 110
Hiibner-Beliavsky, Munich 1 994 24
Illescas-Gelfand, Dos Hermanas 1 996 56
Ivanchuk-Bareev, Dortmund 1 995 18
Ivanchuk-Lautier, Linares 1994 22
Ivanchuk-Smyslov, Tallinn (rapidplay) 1996 62
Karpov-Hj artarson, Tilburg 1988 45
Karpov-Kramnik, Monte Carlo (blindfold) 1995 16

1 42
 In dex o f C o mp l e t e G a m e s

Karpov-Short, Dortmund 1995 92

Kasparov-Bareev, Novgorod 1994 35
Kasparov-Nikolic, Manila Olympiad 1992 130
Kasparov-Shirov, Dos Hermanas 1996 54
Khalifman-Georgiev.Kir, Elenite 1994 35
Klarenbeek-Rogers, Dutch Team Championship 1996 84
Kozul-Illescas, Erevan Olympiad 1996 124
Kramnik-Damljanovic, Moscow Olympiad 1994 72
Kramnik-Ivanchuk, Linares 1994 47
Kramnik-Ivanchuk, Monte Carlo (blindfold) 1996 66
Kramnik-Lautier, Linares 1994 43
Kramnik-Shirov, Dortmund 1996 47
Kramnik-Shirov, Vienna 1996 98
Kramnik-Short, Moscow (Intel Grand Prix) 1996 67
Kramnik-Short, Novgorod 1994 59
Krasenkov-Epishin, Bmo 1994 91
Krasenkov-Sapis, Polish Championship 1995 81
Krasenkov-Sokolov.I, Malmo 1995 118
Lalic-Sadler, Hastings 1995/96 39
Lautier-Bareev, Linares 1994 105
Lautier-Sokolov.l, Groningen 1995 120
Leitao-Beliavsky, Erevan Olympiad 1996 84
Miles-Hodgson, Hastings 1995/96 1 22
Milov-Sadler, Isle ofMan 1994 108
Nesterov-Imanaliev, Bishkek Zonal 1993 79
Nogueiras-Sokolov.l, Erevan Olympiad 1996 119
Novikov-Gretarsson, Berlin Open 1995 63
Oll-Anand, Biel Interzonal 1993 87
Parker-Hellsten, Copenhagen 1996 76
Piket-Ge1fand, Wijk aan Zee 1996 53
Portisch-Kramnik, Biel Interzonal 1993 1 14
Psakhis-Sadler, Megeve (PCA rapidplay) 1994 139
Pushkov-Epishin, Russian Championship 1995 101
Razuvaev-Sturua, Erevan Open 1996 1 26
Richardson-Sadler, Islington Open 1995 13
Rogozenko-Bets, Moldovan Championship 1994 132
Ruze1e-Thorsteins, Lyon (European Club Cup) 1994 60
Sadler-Bareev, Hastings 1992/93 136
Sadler-Ferguson, British Championship 1996 42
Sadler-Hodgson, Hastings 1995/96 92
Sadler-Miles, British Championship 1998 34
Savchenko-Ninov, Cappelle la Grande Open 1994 138

1 43
Th e Sla v

Schandorff-Hellsten, Copenhagen 1996 75

Shirov-Bareev, Biel 1 991 46
Shirov-Nikolic, Wijk aan Zee 1 993 78
Sokolov.I-Bareev, Leon 1 995 50
Sokolov.I-Hellsten, Malmo 1 995 80
Sokolov.I-Shirov, Erevan Olympiad 1996 1 00
Strauss.D-Lakdawala, USA 1992 36
Topalov-Gelfand, Belgrade 1 995 21
Topalov-Gelfand, Dos Hermanas 1996 58
Vaiser-Nalbandian, Erevan Open 1996 1 13
Van der Sterren-Petursson, San Bernardino Open 1992 57
Van der Sterren-Shirov, Biel lnterzonal 1993 89
Ward-Levitt, British Championship 1995 103
Wells-Flear, Oakham 1 994 90
Xu Jun-Akopian, Moscow Olympiad 1994 32
Yusupov-Kramnik, Riga 1 995 17
Yusupov-Shirov, Zurich 1 994 111

1 44