1 e4 e6 2 d4 d5 3 'Lid2

The French Tarrasch

BATSFORD CHESS OPENING GUIDES

Other titles in this series include:

0 7134 8456 x Budapest Gambit Bogdan Lalic

0 7134 845 1 9 King's Gambit Neil McDonald
0 7134 8466 7 Scotch Game Peter Wells
0 7134 8471 3 Spanish Exchange Andrew Kinsman

For further details for Batsford chess titles, please write to Batsford Chess
Books, 583 Fulham Road, London SW6 5BY.
 Batsford Chess Opening Guides

The French Tarrasch

John Emms

B. T .Batsford Ltd, London

First published 1998
Copyright© 1998 John Emms

ISBN O 7134 8461 6

British Library Cataloguing-in-Publication Data.

A catalogue record for this book is
available from the British Library.

All rights reserved. No part of this book may be

reproduced, by any means, without prior permission
of the publisher.

The Batsford Chess Opening Guides were designed

and developed by First Rank Publishing
Typeset and edited by First Rank Publishing, Brighton
Printed in Great Britain by
Redwood Books, Trowbridge, Wilts
for the publishers,
B. T. Batsford Ltd,
583 Fulham Road,
London SW6 SBY

Chess set used in cover photograph courtesy of the

London Chess Centre
Author photograph by Claire Smith

A BA TSFORD CHESS BOOK

General Manager: David Cummings
Advisors: Mark Dvoretsky, Raymond Keene OBE,
Daniel King, Jon Speelman, Chris Ward
 CONTENTS I

Bibliography 8
Introduction 9

Part One: Systems with 3 tbf6

. . .

1 Main Line with 1 1. . .0-0 16

2 Main Line with 1 1 ...'i!fc7 and 1 1 ...'i'b6 34
3 Early Deviations from the Main Line 50
4 The Big Pawn Front with 5 f4 65

Part Two: Systems with 3 . cs 4 exdS 'i!ixdS

. .

5 Main Line with 10 tbxd4 73

6 Early Deviations from the Main Line 86

Part Three: Systems with 3 . . . cs 4 exdS exdS

7 Main Line with 6 i.bS 99
8 Fifth Move Alternatives for White and Black 1 12

Part Four. Other Systems

9 3 . .. cS 4 tbgf3 125
10 Third Move Alternatives for Black 133

Index of Games 143

 BIBLIOGRAPHY I

Books
Encyclopaedia of Chess Openings vol.C, Sahovski Informator 1997
Batsford Chess Openings 2, Kasparov & Keene (Batsford, 1989)
The Complete French, Psakhis (Batsford, 1992)
Winning with the French, Uhlmann (Batsford, 199 1)
Mastering the French, McDonald & Harley (Batsford, 1997)
Play the French, Watson (Cadogan, 1984 and 1986)
The Modern French Tarrasch, Gufeld (Cadogan, 1996)

Periodicals
Informator
ChessBase Magazine
New In Chess Yearbook
British Chess Magazine
Chess Monthly
 INTRODUCTION
'The problem with playing tbc3
against the French, is that Black just
plays the Winawer, doubles your c­
pawns, gangs up and wins your a­
pawn, regroups and defends for a bit
and then promotes his a-pawn. It's just
too easy for Black! ' So complained
one of my students when he realised
that his next opponent was a hardened
French Defence player. Well, I must
admit to feeling a mixture of smugness
and relief at the knowledge that, as an
avid Tarrasch devotee, I never have to
worry about such things. I briefly
showed him a few lines and he was
struck by how simple it all looked.
Often when studying a new open­
ing, it can look rather foreign to begin
with, but this was not the case here.
Perhaps, having played the Tarrasch
for virtually all my chessplaying life, I
am a little biased, but it is true that the
Tarrasch is really quite simple to play
for both White and Black. Of course,
the real secret is to know how to play
it well, but that's what this book is all
about.
I
The German grandmaster Siegbert
Tarrasch, one of the world's leading
chess authorities at the end of the 19th
century, introduced this variation into
tournament practice in 1890. Since
then it has always been a popular way
for White and it is still regarded as the
soundest approach against the French.
In particular, the Tarrasch has al­
ways appealed to players who prefer
to start with a sound positional base,
and because of this it is unlikely to
ever go out of fashion. Its most fa­
mous devotee was Anatoly Karpov,
who did much to elevate the opening's
status in the early seventies, when he
used it continuously on his way to the
World Championship. More recent
exponents include England's World
Championship semi-finalist Michael
Adams, together with the young Rus­
sian stars Peter Svidler (currently
ranked in the world's top ten) and
Sergei Rublevsky. To this list we can
add a whole host of other grandmas­
ters, much too numerous to name in­
dividually here!
9
The Fre n c h Ta rra s c h

Positional and Tactical Themes

Before we begin studying the nitty­
gritty of the actual theory, it is impor­
tant to take a brief look at some of the
positional and tactical themes associ­
ated with the Tarrasch. As the two
distinct main lines with 3 ...lbf6 and
3 ... cS constitute at least 90% of the
material in this book, I've decided to
focus on those two moves here, al­
though you will appreciate that some
ideas do overlap with 3 ... a6, 3 ...lbc6
and 3 ...dxe4. In view of the argument above you
will not be surprised to find out that
3... l2Jf6 in the first diagram the main theoreti­
cal move is 12 i.f4!, offering the trade.
The exchange of dark-squared In the second case White's play is
bishops more subtle. 12 i.f4 is impossible as it
One of White's principal goals in loses a piece, but White can carry out
many ...lbf6 variations is to achieve the exchange with 12 i.h4, intending
the exchange of the dark-squared i.g3.
bishops. The pawn structure dictates
that White has more control of the The fight for the dark squares
dark squares, and a trade of these
bishops normally only emphasises
White's authority. Take the following
two positions.

This diagram shows an ideal situation

for White. Three sets of minor pieces
have been exchanged, leaving White
with the classic ' good knight versus
see following diagram
bad bishop' situation. Added to this,
White's rooks sit very nicely on the
open c-file and semi-open e-file. White

10

can choose between attacking Black's
weak backward e6-pawn, or piling up
on the c-file. In contrast, Black has no
constructive plan and can only sit and
wait, while White leisurely builds up
the pressure. Notice that all of White's
pieces stand on dark squares.
Wells-Fries Nielsen
Copenhagen Open 1995
Here's another depressing scenario
for Black. Again the dark-squared
bishops have been exchanged early in
the game. Black has failed to achieve
any counterplay and now faces a grim
defence. The black e-pawn is ex­
tremely weak, while White can also
consider a direct attack on the black
king starting with 27 hS. Indeed this
was Wells's choice in the actual game
(see Game 20).
Marjanovic-Timman
Sarajevo 1984
see following diagram
Here's an altogether different situa­
tion. White has exchanged off his
dark-squared bishop, but only for a
knight. Now the dark squares are with
In troduction
Black. His queen is doing a marvellous
job on g7, where it protects the black
kingside, attacks the weak d4-pawn
and prepares a pawn assault starting
with 35 ... g4! See Game 14 for the rest
of this very instructive encounter.
The .lhfl exchange sacrifice
..
In order to achieve counterplay
against the solid white position, Black
sometimes employs a common ex­
change sacrifice which seriously weak­
ens the pawn cover around the white
king, while also eliminating a defender
of the d4-pawn.
The main option here is 14 .. .�f3
15 gxf3 tbgS, when a very unclear po­
sition arises. See Chapter 1, Games 1-4
for a detailed discussion of this line.
11
The Fre nch Tarra s c h

The ... e6-e5 break has constant problems with his d- and
In many Tarrasch positions the move b-pawns.
... e6-e5 enables Black to free his pieces
and thereby solve most of his prob­
lems immediately.

Bialas-Uhlmann
Leipzig 1951

Dismantling the white centre

In the early f2-f4 lines White erects a
powerful looking pawn centre, which
can often be used to mercilessly crush
Black into submission. Because of this,
Black players often employ quite vio­
After 14 ...e5! 15 dxe5 tbxe5 16 tbxe5 lent means to destroy White's pride
i.xe5 Black has a very active position, and joy. Here are a couple of typical
with open lines on which to attack the examples.
white king. The rest of this game is a
graphic illustration of Black's chances
here: 17 i.c3 i.xh2+! 18 @xh2 tbg4+
19 @g3 �xf2 20 i.d2 tbe3 21 �xf2
"ifd6+ 22 @f3 i.g4+ 23 @xe3 "ife5+
and White resigned as it is mate next
move.

Emms-Poulton
Hastings 1997
see fol/o.ving diagram

Here's an example of Black failing to
solve his problems, even after achiev­
ing the desired advance. After 14 ... e5
15 dxe5 tbxe5 16 tbxe5 "ifxe5 17 "ifb3
White retains a more comfortable po­
sition. In this case Black has no attack­
ing chances against the white king, but
In this position Black has the ex­
traordinary move 10 ... g5!, which has
indeed become the main line here. See
Games 32-33 for more details on this
variation.
Here's an even more extreme ex­
ample.

12

Normal moves would consign
Black to a terribly passive position.
The brutal solution is 13 ...tbdxe5!? 14
dxe5 tbxe5, which gives Black two
central pawns and active play for the
piece. Note that 15 tbxe5 allows an
immediate catastrophe with 15 ...�f2
mate.
3 ... c5
The ...'iixd5 structure
Here is the typical pawn structure
arising from the 3 ... c5 4 exd5 'ii'xd5 of
the Tarrasch. Black's e-pawn gives him
more control over the centre and the
island of four pawns provide a valu­
able shield for the black king. This
I ntro duc ti on
protection is often vital in repelling
early onslaughts from White, who will
be attempting to utilise his superior
development and activity. Black often
has to play the early middlegame very
accurately and must be particularly
careful not to allow a combinational
breakthrough.
Adams-Luther
Oakham 1990
Here the casual 14 ...tbd7 was pun­
ished by 15 �xe6! and already Black
has a lost position, since 15 ... fxe6 loses
to 16 tbxe6 'ii'b 6 (or 16 ... i.h2+ 17 'it>hl
'iie 5 18 tbxf8+ 'it>xf8 19 'iih5) 17
tt'ixf8+ 'it>xf8 18 'iid 5.
The ... 'i!:Yxd5 lines often liven up
when Black chooses the ambitious
option of castling queenside. Take the
following position, for example.
see following diagram
Black's bishops, knight and queen
point menacingly towards the white
king, which is enough to give White a
lot of headaches.
Black also has some chances of suc­
cess in an endgame. White owns a use­
ful queenside pawn majority, but the
13
The F re n c h Tarra s c h

importance of this is often exagger- while having a few guidelines for these
ated. Black also holds some trump sorts of positions.
cards.

Here we see the IQP as a real

Here is an example of a very suc­
cessful ..."ifxdS endgame for Black.
Hennigan-Levitt
British Championship 1989
strength. Note that it supports the
crucial squares e4 and c4, which can be
used as outposts for the black knights.
As always, piece activity is paramount.
In this position Black's minor pieces
are superior to their white counter­
parts and this is enough to give Black
the advantage. Notice that if all the
minor pieces were taken off the board,
then the edge would swing to White,
as the weakness of the d-pawn would
become more significant. So we can
say in general that exchanges of minor
pieces normally benefit the side bat­
tling against the IQP.

In the ending Black's extra central

pawn can be useful in supporting a
knight in the centre. Note that Black's
kingside pawn phalanx is looking very
impressive indeed.

The isolated queen's pawn

After 3 . . cS 4 exdS exdS some sort of
.

isolated queen's pawn (IQP) position

is usually reached, and it is worth-

14

In this position Black's strategy has
failed. Two sets of minor pieces have
been eliminated and White has the e4-
and c4-squares firmly under his con­
trol. More importantly, however, the
d-pawn is chronically weak and is
ready to be picked off at any moment.
Part of White's strategy against the
IQP is to exchange the right pieces. In
I n t ro duc tion
the above diagram there are three pos­
sibilities, but only one gives White
chances of an advantage. White can
exchange a minor piece with 1 i.xc6
bxc6. However, that change in the
pawn structure favours Black, as the
dS-pawn is no longer isolated. An­
other idea is 1 i.xe7, but after
1...i.xe7 Black gets control of the dark
squares and his bishop can be used to
support a later . .. d5-d4. The best solu­
tion is 1 i.h4!, intending i.g3, offering
a trade of those dark-squared bishops,
which would eliminate Black's best
piece and increase White's control
over the dark squares. Throughout the
book you'll find me harping on con­
tinuously about the importance of
these colour complexes. I apologise in
advance, but their importance can
never be underestimated!
15
I CHAPTER ONE I
3 '2Jf6:
0-0
. . .

Main Line with 11 ...

1 e4 e6 2 d4 d5 3 ll'id2 ll'if6 4 e5 although there are alternatives (see

ll'ifd7 5 .id3 c5 6 c3 ll'ic6 7 ll'ie2 Game 9). Finally we consider the rarer
cxd4 8 cxd4 f6 9 exf6 ll'ixf6 1 0 ll'if3 12 i.gS, which often just transposes to
.id6 1 1 0-0 0-0 lines involving an earlier ...'i!lc7 or
The line starting with 1 1.. .0-0 is ...'i!lb6, in Game 10.
perhaps the most important variation
stemming from 3 ... 4Jf6. Black allows Gamel
his opponent to play the positionally Rublevsky-Gleizerov
desirable 12 i.f4 in the hope that the USSR 1991
temporary 'looseness' in the white
position will allow him to create seri­ 1 e4 e6 2 d4 d5 3 ll'id2 ll'if6 4 e5
ous counterattacking chances. ll'ifd7 5 .id3 c5 6 c3 ll'ic6 7 ll'ie2
Most of the games in this chapter cxd4 8 cxd4 f6 9 exf6 ll'ixf6 10 0-0
continue 12 i.f4 i.xf4 13 tbxf4 tbe4!?, .id6 1 1 ll'if3 0-0 1 2 .if4 .ixf4 1 3
when Black immediately starts opera­ ll'ixf4 lbe4 1 4 ll'ie2
tions on the half-open f-file. Games 1-4
are particularly important, as here we
will cover the notorious exchange sac­
rifice (14 tbe2 a.xf3), which has
breathed new life into Black's posi­
tion. Some White players have been
scared off by the practical difficulties
of facing the exchange sacrifice, and in
Games 5-8 we shall look at other ways
for White to play the position after
13 ...tbe4. This knight move continues
to be the popular choice at all levels,

16

1 4 . . .�xf3! ?
14 ...tbg5 is hardly ever seen, but it
may well be a playable alternative to
the exchange sacrifice. 15 tbxg5 'ifxg5
16 'ifc l transposes to Game 5, while
15 tbe5! is an obvious attempt to claim
a more concrete edge. After 15 ... tbxe5
16 dxe5 White can try to exploit
Black's offside knight, although it is
not all one-way traffic: 16 ...'ifb6! 17
'ifc2 i.d7! 1 8 h4 �ac8 19 'ifhl seems
to win a pawn for White, but after
either 19 . ..tbf7!? 20 i.xh7+ 'it>h8,
when the e5-pawn hangs, or 19 . ..tbe4
20 i.xe4 dxe4 21 'ifxe4 i.c6, Black
obtains substantial counterplay.
1 5 gxf3 lbg5 1 6 'it>h 1
Seemingly leaving the f3-pawn en
prise, but White has some tricks based
on i.xh7+. The major alternative 16 f4
will be discussed in Games 3 and 4.
1 6 . . . e5!
It is imperative for Black to release
his light-squared bishop immediately.
16 ...tbxf3 17 i.xh7+ 'it>xh7 18 'ii'd3+
and 19 'ifxf3 just favours White.
1 7 dxe5 ll'ixf3
17 ...tbxe5 is an underrated alterna­
tive to the text. White has two ways to
combat this idea, but neither has pro­
duced particularly good results:
a) 18 f4? i.g4! 19 f3 i.xf3+ 20 �xf3
tbgxf3 2 1 fxe5 'ifh4 22 'it>g2 �f8 23 h3
'ii'g5+ 24 'it>h l 'ifxe5 25 i.xh7+ 'it>h8
26 tbf4 'ifxf4 27 'ifc2 'it'g3 28 i.f5 tbd4
29 'ifc5 'iff3+ 30 'it>h2 'ii'f2+ 3 1 'it>h l
l!xf5 and White resigned in Donovan­
Kinsman, London (Lloyds Bank Mas­
ters) 1994. An easy point for Kinsman,
who had had the good fortune of play­
ing exactly the same variation two
years earlier!
3 . . . l'Df6: Main Line with 1 1 . . . 0-0
b) 18 tbgl 'iff6 1 9 i.e2 i.d7!
(19 ...i.e6 20 'ifd4 with the idea of f3-f4
gives White a clear plus) 20 'ifd2 (20
'ifxdS+ 'it>h8 looks critical; Black's
compensation for the material is based
on ideas involving ...i.c6) 20 ... tbe6 21
.i::I.fe l i.c6 22 i.fl tbf4 2 3 �adl? and
here in Emms-LB.Hansen, Gausdal
1992, we both missed 23 ... tbxf3! 24
tbxf3 d4, when Black has a winning
position.
1 8 ll'ig 1
18 i.xh7+ will be studied in Game
2.
1 8 . . .ll'ifxe5
A characteristic posltlon for this
line. In compensation for the ex­
change, Black has a menacing passed d­
pawn and White's king is not alto­
gether safe. Objectively Black proba­
bly has enough, and in any case his
position is easier to play.
1 9 .ic2
Other tries include:
a) 19 i.e2 i.f5 20 'ifd2 d4 2 1 �fel
'ifd5+ 22 f3 d3 23 i.fl l:td8 24 b3 tbg6
25 l:tacl tbce5 26 �c3 'ifd4 27 �eel??
'ifxgl+! 0-1 Lhagvasuren-Gleizerov,
Cheliabinsk 199 1. White has to stay
very alert in this variation! ·
17
The Fre n c h Ta rra s c h

b) 19 f4 tbg4 (19 ... tbxd3!? also de­ <it>xa2 5 1 !Ixb4 <it>al 52 <it>d2 h2 53 !Ih4
serves attention) 20 'ifd2 d4 2 1 i.e4 <it>b2 54 !Ixh2 a2 and finally the draw
tbe3 22 l:.f3 i.g4 23 �f2 'iff6? (I like becomes inevitable.
23 ... i.fS! here, as after 24 i.xfs 'ifdS+
25 tbf3 tbxfS the exchange of bishops
emphasises the weaknesses around the
white king) 24 l:.cl �f8 25 tbe2 'ife6
26 tbg3 'ifh6 27 tbfl ttJfs 28 <it>g1 'ifh4
29 'ifd3 'ifd8? 30 h3! 1-0 Hunt­
Hagarova, Yerevan Women's Olym­
piad 1996.
1 9 . . .d4 20 .ie4 .ie6 21 �c 1 •h4 22
f3 �dB 23 �e 1 •ta

Game 2
Kramnik-Ulibin
USSR Championship 1991

Black has completed his develop­
ment and his pieces stand harmoni­
ously. All in all, he has a very reason­
able position.
24 b3 .id5 25 b4! d3 26 .ixd5+
�xd5 27 •b3 •t1 28 �ed 1 g5! 29
�c5 �xc5 30 bxc5 lt:Jd4 31 "ii'xf7+
�xf7 32 lt:Jh3! lt:Jdxf3 33 lt:Jf2 d2 34
�g2 �e6 35 lt:Je4 �d5 36 lt:Jxd2
lt:Jxd2 37 �xd2+ �xc5 Yz Yz -
1 e4 e6 2 d4 d5 3 lt:Jd2 lt:Jf6 4 e5
lt:Jfd7 5 c3 c5 6 .id3 lt:Jc6 7 lt:Je2
cxd4 8 cxd4 f6 9 exf6 lt:Jxf6 1 0 lt:Jf3
.id6 1 1 0-0 0-0 1 2 .if4 .ixf4 1 3
lt:Jxf4 lt:Je4 1 4 lt:Je2 �xf3 !? 1 5 gxf3
lt:Jg5 1 6 'iii>h 1 e5! 1 7 dxe5 lt:Jxf3 1 8
.ixh7+ 'it>h8!
18 ... @xh7 19 'ifd3+ <it>g8 20 'ifxf3
tbxeS 21 'it'g3 is good for White.
1 9 lt:Jg 1

Here Rublevsky gives 38 �f2 �d4

39 �fs g4 40 �hs bs 41 !Ixh7 as as
leading to a draw. Perhaps this should
be expanded a little further. For ex­
ample, 42 <it>g3 a4 43 �b7 <it>c4 44 <it>f4
tbf3! 45 h3 gxh3! 46 Wxf3 b4 47 <it>e3
<it>c3 48 �c7+ <it>b2 49 �a7 a3 50 !Ib7

18
 3 . . . l'Df6: Main L ine with 1 1 . . . 0-0

1 9 ...lbcd4! couple of lines:

Other moves do not quite hit the
mark:
a) 19 ...tbxgl? 20 'ifh5 and wins.
b) 19 ...tbcxe5? 20 i.e4! also wins for
White.
c) 19 ... tbfxe5 doesn't lose immedi­
ately, but after 20 i.c2 White still has
a clear plus.
20 ll'ixf3
20 l:.e 1 i.g4 2 1 !Ie3 'ifh4! gives
Black a clear advantage according to
Holmov. 22 tbxf3 'ifxf2 23 'ifxd4!
i.xf3+ 24 !Ixf3 'ifxd4 bears out his b21) 23 ... !Ixe5? 24 hxg4 l:.xe3
opinion. However, White does have (24 .. .'i'f6!? is an improvement, but
another important alternative to 20 White is still better after 25 �g2 llxe3
tbxf3 with 20 i.d3 i.g4 21 'it'cl !? 26 fxe3 'it'g5 27 :Xf3 tbxf3 28 tbxf3
a) 2 1...tbxgl and now: 'ifxg4+ 29 'it>f2) 25 fxe3 'iff6 26 tbxf3
al) 22 �xgl? tbf3+ 23 'it>hl tbxe5 tbxf3 27 g5! 'ifxb2 28 l:.xf3 'ifxal+ 29
with a dangerous attack. l:.fl and White wins.
a2) 22 'iff4! {this clever move refutes b22) The calm 23 .... i.h5! maintains
2 1...tbxgl) 22 ... i.e2 23 !Ixgl i.xd3 24 all of Black's attacking chances, e.g. 24
'ifxd4 i.e4+ 25 !Ig2 !Ic8 26 !Idl and tbxf3 tbxf3 25 i.c2 l:.xe5 26 'ifd3 g6
the extra pawn ensures White a plus. and the threat of ... 'iff4 is looming.
b) 2 1...'iff8! 22 'ife3 and now: 20 . . ..ig4 21 lbxd4 .ixd 1 22 l:f.axd 1
b l) 22 ... g5?! 23 tbxf3 hf3+ 24 'it>gl 'it>xh7
'ifh6 25 e6! 'iff6 (or 25 ... 'ifh3 26 'ife5+
'it>g8 27 'ifxg5+ Wh8 28 'ife5+ 'it>g8 29
'ifg3+) 26 l::ta el tbc6 27 i.b 1 g4 28
'ifc3 tbd4 29 'ifc7 i.e4 30 i.xe4 1-0
was played in Johnson-Du Pree, corre­
spondence 1993.
b2) Johnson gives 22 ...l:.e8 as an ob­
vious improvement, yet still assesses
the position as clearly better for
White. In fact, Black seems to have a
dangerous initiative.
see following diagram
This is a crucial position for the as­
The threat is ... !Ixe5 followed by sessment of the whole 14 ...l:.xf3 line .
... �5, so White has to react quickly White has obtained a rook, knight and
with 23 h3 and now I've looked at a dangerous passed pawn for the queen.

19
The Fren c h Ta rra s c h

However, White's king is still a little ual check.

airy and Black can use this to his ad­ 27 �h3+ 'iti>g6 28 �g3+ 'iti>h5 v.i-v.i
vantage in the ensuing tactics. White can draw with 29 �h3+ or
23 f4 else allow Black the perpetual with 29
Supporting the passed pawn. 23 �xcl 'ifxcl+ 30 .l::I.g l 'ife3! 31 fxg7
�d3!?, with the idea of !Ig3 and �fgl, 'ife4+ 32 �g2 'ife l+.
is another way to play the position.
23 . . .•b6 24 f5 Game]
24 �d2!? may be a more serious Antonov-Bal inov
winning try. Bode-Voigt, Hamburg Djuni 1987
1992, continued 24 ...'ifg6 25 .i::I.e l 'i!:Yh6
26 l:tf2 'ifh4 27 !Iffl �h8 28 e6 g6?! 29 1 e4 e6 2 d4 d5 3 lbd2 lbf6 4 e5
f5! gxf5? (29 ...'ifxd4 30 f6 'ifc5 31 e7 lbfd7 5 .id3 c5 6 c3 lbc6 7 lbe2
d4!, when Black has ideas of a perpet­ cxd4 8 cxd4 f6 9 exf6 lbxf6 1 0 lbf3
ual check, is the last chance) 30 tllxf5 .id6 1 1 0-0 0-0 1 2 .if4 .ixf4 1 3
'iff6 and here 3 1 !If3, threatening lbxf4 lbe4 1 4 lbe2 �xf3!? 1 5 gxf3
!Ih3+, is a killer. lbg5 1 6 f4
24 . . .•xb2! The only real alternative to 16 'it>h 1.
Not so much grabbing a pawn, as 1 6 . . .lbf3+ ! ?
paving the way for the rook to travel
to cl.
2 5 lld3 �ca 2 6 f6
26 �h3+?! 'it>g8 27 f6 fails to
27 ...'ifxd4! 28 f7+ 'it>f8 29 !Ih8+ 'it>e7,
when 30 f8'if+ l:txe8 is slightly better
for Black and 30 �xc8 runs into
30 ...'ife4+ 3 1 'it'gl 'it'g4+ 32 'it>h l
'i'xc8.
26 . . .�c 1

This move adds more fuel to the

fire. The safer 16 ... tll h 3+ is discussed
in Game 4.
1 7 'iti>g2 •h4 1 8 'iti>xf3 •h3+ 1 9 lbg3
e5 20 'iti>e3
The major alternative to Black's
threat of . . .i.g4 is 20 f5, when after
20 ... e4+ we have a further split:
see following diagram

Now everything leads to a perpet- a) In Emms-Kinsman, Wuppertal

20

(rapidplay) 1993, I tried 2 1 'lt>e3 'ifh6+
22 f4 exd3 23 'ifbs, but after 23 ... 'ifxhS
24 ltJxhS .5lxf5 25 ltJg3 .l::!.e 8+ 26 'lt>d2
.5le4 Black certainly had enough play
for the exchange. In hindsight 23 'ifhS
looks to be too accommodating and
should be replaced with 23 'ifb3. Even
so, it is quite clear that there is no easy
solution for White.
b) 2 1 .5lxe4 dxe4+ 22 'lt>xe4? (22 'lt>e3
is safer) 22 ....txfS+! 23 ltJxfS l:te8+ 24
'lt>f4 'ifg2 25 ttJg3 ttJb4! 26 'ifb3+
ttJdS+ 27 'lt>g4 'lt>h8 28 .l::!.ae l l::.f 8!, and
White's king has slipped into a mating
net. Novotny-Nilsson, European
Team Correspondence Championship
199 1, concluded 29 h4 l:I£4+ 30 'lt>hs
.l::!.fS+ 3 1 'lt>g4 g6 32 .l::!.e8+ 'lt>g7 33
'ifxb7+ .l::!.f7 34 'ifxf7+ 'lt>xf7 35 .l::!.e S
h5+ 36 'lt>gS 'iff3 37 'lt>h6 'iff4+ 0-1, as
after 38 .l::!.g S 'ifxd4 39 .l::!.xg6 ttJf4 the
rook goes or it is mate.
20 . . . exf4+ 2 1 'iti>d2
2 1 'lt>xf4 'ifh6+ 22 'lt>f3 .5lh3 23 .tf5
.l::!.f8 occurred in De Souza-Ferraro,
Zurich 1990, a game that White went
onto win. Instead of 23 ... :f8, Black
should play 23 ...'iff6!, hitting the d4-
pawn. Then 24 %:!.e l ltJxd4+ 25 'ifxd4
'ifxd4 26 .5lxh3 .l::!.f8+ and 24 'ifd3 ltJe7!
3 . . . 4J f6 : Main L ine with 1 1 . . . 0-0
(24...ttJxd4+ 25 'ifxd4 'ifxd4 26 .5lxh3
isn't so good now, as f2 is protected)
25 .l::!.fe 1 ltJxf5 26 ltJxf5 .5lxf5 are both
very uncomfortable for White.
2 1 . . . fxg3 22 hxg3 'il'h6+ 23 f4 'il'f6!
24 'il'a4 i..d7 25 llae 1 llc8 26 i..b5
a5!
It seems that White is on the verge
of consolidation, but Black manages to
obtain just enough play to force a
draw. 26 . aS prepares ... ttJb4, discon­
. .
necting White's queen from the de­
fence of the vital d-pawn.
27 l:tc 1
After 27 a3 White gets hit with
27... ttJb4! (anyway) as 28 .5lxd7 'ifxd4+
29 'lt>e2 'ifd3+ 30 'lt>f2 .l::!.c2+ 3 1 'ifxc2
ltJxc2 is good for Black.
27 . . . lleS! 28 llfe 1 tLlb4! 29 J:te5
29 .l::!.xe8+ .5lxe8 30 .5lxe8 'ifxd4+ 3 1
'lt>e2 'ifxb2+ 3 2 'lt>dl 'ifd4+ is also a
draw by perpetual check.
29 . . . llxe5 30 dxe5 'il'b6 3 1 i..xd7
'il'f2+ 32 'iti>d 1 'il'f1 + 1h -1h
Game 4
Komarov-Barsov
Reims 1994
1 e4 e6 2 d4 d 5 3 tLld2 tLlf6 4 i..d 3
21
The Fre n c h Ta rra s c h
c 5 5 e 5 tLlfd7 6 c 3 tlJc6 7 tLle2 cxd4
8 cxd4 f6 9 exf6 tLlxf6 1 0 tLlf3 i..d6
1 1 0-0 0-0 1 2 i..f4 i..xf4 1 3 tLlxf4
tlJe4 1 4 tLle2 llxf3 ! ? 1 5 gxf3 tlJg5 1 6
f4 tLlh3+
1 7 'iti>h1
17 'lt>g2 was thought to be bad due
to 17... ttJxf4+ 18 ltJxf4 °ifgS+. After 19
'lt>h l (19 'lt>f3? ltJxd4+ 20 'lt>e3 "ifeS+ is
very good for Black) 19 ... "ifxf4 20
"ifhS, however, White can drum up a
powerful initiative before Black has a
chance to co-ordinate, e.g. 20... g6 21
.l::!. g l ltJxd4!? (very provocative, but
21 ...ltJe7 22 °if gS! "ifxgS 23 .l::!.xgS ..5ld7
24 %:!.cl ttJc6 2S .tbs .l::!.f8 26 'lt>g2 .l::!.f4
27 l:tdl gives White a pleasant end­
game) 22 ..5lxg6 hxg6 23 "ifxg6+ 'lt>f8.
22
Here White's attack seems to be on
the verge of succeeding, but Black has
some surprising counter-chances as
shown, for example, by the following
crazy line: 24 °ifg8+ 'lt>e7 2S .l::!.g7+ 'lt>d6
26 .l::!.f7 ttJfs 27 "ifd8+ 'lt>c6 28 .l::!.c7+
'lt>bs 29 .l::!.xc8. Now it is Black's turn
to bat! After 29 ..."iff3+ 30 'lt>gl ltJh4 3 1
"ifd7+ 'lt>a6 White has an immediate
draw with 32 "ifa4+ 'lt>b6 33 "ifb4+ but
nothing better. The winning attempt
32 "ifxe6+!? b6 33 °ifg8 fails to
33 ..."ifh3 34 °ifg3 tiJf3+ 3S "ifxf3 "ifxf3
36 .l::!.xa8, when it is Black's turn to
deliver perpetual with 36 ...°ifg4+.
In view of the unclear nature of
these variations, Black most practical
choice against 17 'lt>g2 is, however,
17 ..."ifh4. After 18 .tbs ..5ld7 Black
will follow up with ....l::!.f8 with clear
compensation for the exchange.
1 7 . ..'il'h4 1 8 'il'd2 tLlxf2+ 1 9 'iti>g2
tLlxd3 20 'il'xd3 i..d 7
Taking stock, we see that White has
withstood his opponent's initial on­
slaught, but there is still plenty to
fight for. Black has a pawn for the ex­
change and White still has weaknesses
on f4 and d4. All in all, the position is
very finely balanced.
 3 . . . li:if6: Main L i ne with 1 1 . . . 0-0

2 1 h3!? tLlxc2 40 llc4 tLla3 41 llc7+ 'iii>g 6 42

Making a slightly safer haven for llxb7 a5 43 llb6+ ..ti>g5 44 lle6 ..ib5
the king on h2. In contrast 2 1 f5? 45 lle5+ ..ti>f6 46 l:th5 '1ti>g6 4 7 llc5
failed to hit the mark in Dobosz­ a4 48 tlJe4 ..id7 49 tLlc3 tLlc2
Boehsmueller, Loosdorf 1993, which
concluded 21. ..exf5 (why not?) 22 lbf4
.l::!.e8 23 "ifb3 .l::!.e4 24 "ifxd5+ .ll e6 25
"ifd6 .ll c4 26 .l::!.f2 :Xd4 27 "ifc7 °if g4+
28 'lt>hl .l::!.xf4 29 "ifc8+ 'lt>f7 30 "ifxb7+
lbe7 3 1 .l::!.xf4 "ifxf4 and the threat of
32 ....ll d 5+ cannot be parried by 32
.l::!. dl as 32 ... ..td5+ 33 .l::!.xd5 "iffl is
mate.
21 . . . ..ie8 ?!
This allows White to open up the
position for his two rooks. 21.. .:f8
would have been much more natural. 50 tLlxa4! ..ixa4 51 llc4 1 -0
22 f5! e5
In his notes to the game Komarov Game 5
gives 22 ...exf5 23 "ifxf5 .llg6 24 "ifxd5+ Dvoirys-Ulibin
'lt>h8 25 .l::!.f4! "ife7 26 lbg3 as winning Cheliabinsk 1990
for White and I see no reason to disa­
gree with his assessment. 1 e4 e6 2 d4 d5 3 tLld2 tLlf6 4 e5
23 dxe5 tLlxe5 24 'ii'd4 lLlfd7 5 c3 c5 6 ..id3 tLlc6 7 tLle2
24 "ifxd5+ also looks good, although cxd4 8 cxd4 f6 9 exf6 �xf6 1 0 0-0
after 24 ...lbf7 Black can hope for ..id6 1 1 tLlf3 0-0 1 2 ..if4 ..ixf4 1 3
counterplay based on threats along the tLlxf4 tlJe4 1 4 'ii'c 1
a8-h 1 diagonal.
24 . . .'ii'g 5+ 25 tLlg3 tLlc6 26 'ii'f4 'ii'f6
27 llae 1 ! 'ii'x b2+
Grabbing the bait, but 27....llf7 was
wiser, restricting White to a small
edge.
28 llf2 'ii'f6 29 lle6 'ii'f7 30 lld6
tlJa5 3 1 'ii'g 5 tlJc4 32 lle6 d4 33 f6!
h6 34 'ii'x g7+ 'ii'x g7 35 fxg7 tLle3+
36 '1ti>h2 'ittx g7 37 lld6
The d-pawn drops and White's
rooks rule in the open position. The
rest of the game is quite painful for A safe move. The queen keeps an
Black. eye on the important dark squares.
37 . . .llc8 38 llxd4 llc2 39 llxc2 White now plans lie 1 to kick the

23
The Frenc h Ta rra s c h
black knight from e4.
1 4 . . . tlJg5!
A positionally sound idea. The ex­
change of these knights eliminates a
defender of the d4-pawn, while also
lessening White's control over eS. In
contrast the developing 14 ....5ld7 turns
out to be too slow after lS :e l! ltJgS
16 ltJxgS 'ifxgS 17 ltJxe6 'ifxcl lS
:axcl .5lxe6 19 :xe6 ltJxd4 20 :el
and White retains a clear advantage in
the endgame.
1 5 tlJxg5 'il'xg5 1 6 tLle2
For 16 .5lxh7 +!? see Game 6.
1 6 ...'il'f6 1 7 'il'e3
1 7 . . . i.. d 7
17 ... eS lS dxeS ltJxeS, immediately
releasing the tension in the centre, is
the principal alternative. Practical re­
sults have favoured White, although
with exact play Black may be able to
equalise. Tiviakov-Maksimovic, Cheli­
abinsk 1990, continued 19 l::.a dl ltJg4
20 'ifcS .5le6 2 1 'ifd4 ttJes 22 .tbl
.l::!. aeS 23 ltJc3 .l::!. dS 24 .l::!.fe l ltJc6 2S
'ifxf6 .l::!.xf6 26 l:td2 a6 27 .l::!.e dl ltJe7 2S
ltJe4! .l::!.h 6 29 ltJgS and Black was al­
ready in trouble. Maksimovic suggests
19 ...ltJc4!? as an improvement. It is
true that after 20 .5lxc4 dxc4 21 ltJc3
24
.5lg4 22 f3 .tfs Black seems to be rid of
most of his problems.
1 8 J:tad 1
1 S .tbs is another idea, hoping to
exchange on c6 and reach a classic
'good knight versus bad bishop' sce­
nario. However, Black can thwart this
attempt with the strange looking
lS ...ttJbS!?, when White may have
nothing better than to retreat with 19
.5ld3. Instead Heinemann-Kinsman,
German Bundesliga 1997, continued
lS ....l::!.acS 19 l:.ac l (the immediate 19
.5lxc6 .l::!.xc6 20 ltJc3 initiates White's
plan, but after 20 ... .l::!. c 4 21 .l::!.a dl .l::!. fcS
Black can obtain counterplay with
...b7-bS-b4) 19 ... ttJbs 20 .td3 ttJc6 (of
course this jig with bishop and knight
isn't necessary. White can reach the
same position with the simple lS .l::!.ac 1
.l::!.acS) 2 1 l::.c 3 ltJe7 22 .l::!. fcl .l::!.xc3 23
:xc3 .5lc6 24 :ct ltJcS 2S ltJc3 ttJd6 26
�el :es 27 .l::!.e2 a6 2S f4 ltJc4 29 'iff2
ms 30 b3 ttJd6 3 1 g3 l::.cS 32 .l::!.c2 and
here Black should have played
32 ... .teS, not fearing 33 fS as 33 ... .5lf7
34 fxe6 'ifxf2+ 3S 'ittxf2 .5lxe6 is com­
pletely equal.
1 8 . . .J:tacS 1 9 i.. b 1 .tea 20 h3 i..f7
Not 20... .5lg6 2 1 .5lxg6 and Black
has to capture with the pawn, as
21.. .'ifxg6 runs into 22 ltJf4.
2 1 tLlf4 l:tfe8 22 b3 e5 23 dxe5 .:txe5
24 'il'c 1 d4 25 tLld3
The ... e6-eS break has been achieved
at a favourable moment, so that Black
has enough piece activity to support
the passed d-pawn. Notice that 2S
:xd4? ltJxd4 26 'ifxcS+ l:teS loses the
knight on f4.
25 . . .l:tg5 ! ?
The simple 2S ....l::!. SeS would have

been enough to maintain the balance,
but Black decides to go 'all in' with a
kingside attack.
26 llde 1 i..d 5 27 f3 .!If8 28 .:tf2
llg3?! 29 '1ti>h2 'ii'h4? 30 'ii'c5 lld8
Around here Ulibin must have real­
ised that his kingside offensive had
been too ambitious and is forced into
an ignominious retreat. The critical
line is 30 ... .5lxf3, but after 3 1 .l::!.xf3
:fxf3 32 .l::!.e8+ @f7 33 "ii f8+ @g6 34
ltJf4+ @h6 35 .l::!.e6+ l:tg6 36 l:.xg6+
hxg6 37 "ii h8+ @gs 38 ltJe6+ White
wins comfortably.
3 1 llfe2 i..f7 32 lle4 'ii'g 5 33 ll 1 e2?
The computer program Fritz im­
mediately spotted 33 "ii c 2! .tds 34
"ii f2, when the black rook is severely
embarrassed!
33 . . . 'ii'xc5 34 l2Jxc5 .:tg5! 35 l2Je6
The theme of trapping pieces con­
tinues. 35 ttJxb7 .l::!.d7 36 f4 .l::!. b s 37 l:tc2
.tds wins for Black.
35 . . . i.. x e6 36 llxe6 .:tc5
37 i..e 4??
The final and decisive blunder.
37 . . . d3 38 lld2 ..ti>f7!
Now it is White's rook that is
trapped! Ulibin's endgame technique
proved to be extremely efficient.
3 . . . li:i f6: M a in Line with 1 1 . . . 0-0
39 llxc6 bxc6 40 llxd3 llxd3 41
i..x d3 J:tc3 42 i..c4+ 'itf 6 43 'iti>g3 g5
44 a4 llc2 45 a5 llb2 46 a6 c5 4 7
i..d5 ..ti>e5 48 i.. f 7 h6 49 h4
Or 49 .5lc4 h5 50 'itih2 h4 and the
black king will march into the dark
squares.
49 . . . gxh4+ 50 '1ti>h3 '1ti>f4 51 i..d5 J:ta2
52 i..c4 lla 1 53 '1ti>h2
53 . . . h3! 0-1
Game 6
Svidler-Ulibin
Russ ian Championship 1994
1 e4 e6 2 d4 d5 3 l2Jd2 l2Jf6 4 e5
l2Jfd7 5 c3 c5 6 .td3 l2Jc6 7 l2Je2
cxd4 8 cxd4 f6 9 exf6 l2Jxf6 1 0 0-0
i..d6 1 1 l2Jf3 0-0 1 2 .tf4 i..xf4 1 3
l2Jxf4 l2Je4 1 4 'ii'c 1 l2Jg5 1 5 l2Jxg5
'ii'x g5 1 6 i..x h7+!?
see following diagram
This surprising tactic (possibly a
Fritz invention!) nets White a rook
and two pawns for two minor pieces.
Theoretically speaking, however, this
material imbalance should be okay for
Black and practical results have re­
flected this.
25
The Fre n c h Ta rra s c h
16 . . . '1ti>xh7 1 7 lLixe6 'il'f5
The main alternative is to head into
the endgame immediately with
17 ...'ifxc l 18 lLixf8+ 'lt>g8 19 .l::!.axcl
'lt>xf8. Holzke-Kinsman, German
Bundesliga 1995, continued 20 .l::!.fdl
..te6 2 1 .l::!.d2 nes 22 f3 'lt>f7 23 h4! 'lt>f6
24 'itif2 gS 25 hxgS+ 'lt>xgS 26 .l::!.e 1 and
the connected pawns on the kingside
gave White an edge, although the
game was eventually drawn. Kinsman
suggests that Black should aim to play
... g7-g5 before White has a chance to
play h2-h4, thus making it more diffi­
cult for White to utilise the pawn ma­
jority. So 22 ... gS is more accurate than
22 ... 'lt>f7; and likewise 22 h4! may be
stronger than 22 .l::!.d2. Taking this to
the very limit, it is possible that Black
should consider 20... gS!?
1 8 lLixf8+ 'il'xf8 1 9 'il'g5 'il'f5
Black has also secured reasonable re­
sults with 19 ... ..tfs. Emms-Vareille,
Paris 1994, continued 20 .�.fel .l::!.d8 2 1
.l::!.e 3 .l::!.d 6! (the rook patrols the third
rank, along which it can repeatedly
annoy the white queen) 22 .l::!.f3 .l::!.f6 23
:te l 'itig8 24 �g3 ..te4 and Black had
equalised. Notice that 25 f3 can be
answered with 25 . .. lLixd4!, as 26 fxe4?
26
lLie2+ forces mate.
20 'il'xf5+ ..ixf5 21 llfe 1 ..ti>g6
It is only good for Black to snatch
the d4-pawn if he can retain his own d­
pawn. Here 2 1 ...lLixd4 only succeeds
in opening paths for the white rooks
after 22 :adl lLic2 23 .l::!.e S.
22 f3 ..ti>f6
22... lLixd4? still fails due to 23 .l::!.adl
lLic2 24 .l::!.e 2 with threats of g2-g4 and
.l::!.xdS. Now, however, Black is threat­
ening to grab the pawn.
23 .:tad 1 g5!
We touched upon the importance
of this move earlier. The lunge h2-h4
has been prevented.
24 g4 ..id7! 25 �2 .:th8! 26 ..ti>g3
tlJe7!
Heading for f4. It is now White
who must play accurately to maintain
the balance.
27 lle5 tlJg6! 28 llxd5 ..ie6 29 .:tb5!
Not 29 .l::!.d6? 'lt>e7, trapping the
rook.
29 . . .tLlf4
30 d5!
It is a good time to bail out. If
White continues in pawn grabbing
fashion with 30 .l::!.xb7 then 30 ...l::th3+
3 1 'itif2 ..tds! 32 .l::!.xa7 .l::!. xh2+ 33 'lt>el

.l::!.e2+ 34 @fl :xb2 and it is Black who
is doing the pushing.
30 . ..l:th3+ 3 1 �2 .txd5 1h -1h
32 :dxdS ttJxdS 33 !hds .l::!.xh2+ 34
'lt>g3 l::.xb2 leads to a totally drawn
rook and pawn endgame.
Game l
Henao-Sequera
M erida 1992
1 e4 e6 2 d4 d5 3 l2Jd2 l2Jf6 4 e5
l2Jfd7 5 c3 c5 6 .td3 l2Jc6 7 l2Je 2
cxd4 8 cxd4 f6 9 exf6 l2Jxf6 1 0 l2Jf3
.td6 1 1 0-0 0-0 1 2 .tf4 ..ixf4 1 3
l2Jxf4 l2Je4 1 4 l2Jh5
The knight heads for g3, where it
will confront its black counterpart on
e4. This is a perfectly logical, albeit
rather time-consuming plan.
1 4 . . . g6 1 5 l2Jg3 l2Jxg3 1 6 hxg3
Now White threatens .tbs, ex­
changing on c6 and establishing a mas­
sive bind on the dark squares. Need­
less to say, Black has to remain very
active.
1 6 . . .'il'b6! 1 7 'il'a4 a6
17 ...'ifxb2!? is meant to be unplay­
able due to 18 .tbs, threatening to
trap the black queen with %:!.ab 1 and
3 . . . fiJf6: Main L in e with 1 1 . . . 0-0
.l::!. fcl . This is far from the end, how­
ever, as Black can play 18 ...:xf3 19
gxf3 ltJxd4 with a very messy position.
18 llad 1 ..id7 19 'il'a3 l2Jxd4 20
l2Jxd4 'il'xd4 2 1 ..ixg6 'il'f6 22 ..ib 1
..ib5 23 ..id3
23 . . . ..ic6!
Ulibin's suggested improvement
over 23 ... .5lxd3, which he labels as a
' strategic howler'. Spasov-Ulibin, Tun­
ja 1989, continued 24 .l::!.xd3 :ac8 2S
.l::!.f3 'ifg7 26 %:!.e l :c6 27 'ifb4 bs 28 b3
.l::!. fc8 29 :fe3 l:Ie8 30 .l::!.e S and Black's
shaky king position had become the
overriding feature of the position. The
idea behind 23 ... .5lc6 is to preserve the
bishop in order to support the central
pawn mass.
24 'il'b3 '1ti>h8 25 ..ib1 e5 26 .:td2 llf7
27 lle 1 llg8 28 'il'c3 .J:!.g5 29 .tc2 h5
In contrast with the Spasov-Ulibin
game, the principal features here are
Black's centre and active pieces. Black
now has no problems at all.
30 ..ib3 d4 31 'il'c4 l:te7 32 'il'c5
.:teg7 33 .td1 h4 34 g4 h3 35 gxh3
see following diagram
35 . . .llxg4+ 36 hxg4?
After 36 .5lxg4 'iff3 37 @fl Black
27
The Fre n c h Tarra s c h

has only the choice of perpetuals with

37 ... .5lb5 38 'ifxb5 'ifhl+ 39 'lt>e2 'ii'e4+
40 'itidl 'ifb 1 + 4 1 'lt>e2 'ife4+ or
37 ... :xg4 38 hxg4 'ifh3+ 39 'lt>e2 'iff3+
40 'lt>fl . Now White is forced to give
up the queen.

1 4 . . . g5
The most ambitious reply to 14 g3,
but perhaps not the best. Other ideas
include:
a) 14 ...'ii'f6 (perhaps the most testing
response) 15 h4 h6 16 .5lxe4 dxe4 17
36 . . .'il'h4 37 'il'xc6 bxc6 38 llxe5 ttJe5 l:.d8 18 ttJxc6 bxc6 19 'ifc2 g 5 20
llh7 39 .J:!.e8+ ..ti>g7 40 ..if3 'il'h2+ 41 hxg5 hxg5 2 1 ltJe2 'iff3 22 'ii'd2? 'lt>f7
�1 'il'f4 23 'ifxg5 :h8 24 'iff4+ 'lt>e7 25 'ifxf3
A vigorous technical phase lies exf3 26 ltJf4 .l::!.h6 27 ttJd3 and White
ahead, but White's resistance was not resigned in Kotronias-Ulibin,
up to much. Chalkidiki 1992. Ulibin gives 22 ltJc3
42 lld3 'il'c 1 + 43 l:te 1 'il'c4 44 lled1 as an improvement for White and as­
c5 45 b3 'il'b5 46 a4 'il'b6 47 .:te 1 sesses 22 ... .5la6 23 :fe l .td3 24 'ifd2
'ill'd6 48 lle8 'ii'g6 49 .:te4 .:th8 50 'iff5 as unclear. There is plenty of
b4? l:!.e8! 0-1 scope for home analysis on this posi­
After 51 �xe8 'ifxd3+ 52 'lt>g2 cxb4 tion!
the two passed pawns prove to be de­ b) 14 ...ltJg5 15 ltJe5 ltJxe5 16 dxe5
c1S1ve. l::.xf4!? 17 gxf4 ltJh3+ 18 'lt>h 1 ltJxf4 19
.l::!. g l .5ld7 20 'it'g4 'iff8 21 .l::!.g3 .5lc6 22
.l::!.dl d4+ 23 'lt>gl 'iff7 and here the
Game 8 players agreed a draw in Geller-Zi.iger,
Stigkin-Basin Bern 1988, just when the position was
St Petersbu rg 1992 getting interesting. Clearly Black has
substantial compensation for the ex­
1 e4 e6 2 d4 d5 3 tLld2 tLlf6 4 e5 change in the final position, so White
tLlfd7 5 ..id3 c 5 6 c3 tlJc6 7 tLle2 should look for improvements earlier
cxd4 8 cxd4 f6 9 exf6 tLlxf6 1 0 0-0 on .
..id6 1 1 tLlf3 0-0 1 2 ..if4 .txf4 1 3 1 5 tLlh5 e5 1 6 tlJxe5! tlJxd4
tLlxf4 tlJe4 1 4 g3 16 ... .5lh3 falls short after 17 .5lxe4

28

.5lxf1 18 .5lxh7+! 'itixh7 19 'ifc2+ 'itih8
20 ltJg6+ 'lt>g8 2 1 ltJxf8 with a winning
position, while 16 ...ttJxe5 17 dxe5 .5lh3
18 .5lxe4 is no better.
1 7 'il'a4!
Ulibin has suggested 17 f3, but after
17 ...'ifb6 1 8 fxe4 .5lh3! 19 .l::!.xf8+ lhf8
Black has a very dangerous attack for
the piece. For example, 20 'lt>h 1 .l::!.f2,
when the threat of ....5lg2+ is ex­
tremely difficult to meet.
1 7 . . .'il'b6 1 8 llad 1 l2Jxf2 1 9 llxf2
l2Je2+ 20 'it>h 1 'il'xf2
20 ...l:txf2 leads to disaster after 2 1
'ife8+ .l::!.f8 2 2 .5lxh7+ 'lt>xh7 2 3 'ife7+
'lt>h6 24 'ifxf8+ 'lt>xh5 25 'ife8+ 'lt>h6 26
'ifh8 mate.
21 'il'c 2!
The final word in this tactical se­
quence. White restores some material
balance and can look forward to a fa­
vourable endgame.
21 . . . l2Jxg3+
Other moves lose quickly:
a) 21.. . .tf5 22 'ifc7!
b) 2 1 . . .. .5lh3 22 .5lxh7+ 'lt>h8 23
ltJg6+ 'itixh7 24 ltJxf8+ 'lt>g8 25 'ifh7+
'lt>xf8 26 'ifg7+ 'itie8 27 ltJf6+ 'itid8 28
:xd5+ 'itic8 29 'iff8+ 'lt>c7 30 'ifd6+
'lt>c8 3 1 'ifd8 mate.
3 . . . lll f 6: Main L in e with 1 1 . . . 0-0
2 2 l2Jxg3 'il'xc2 2 3 ..ixc2
Materially speaking, Black is doing
okay, but White's active pieces and his
opponent's weak pawns add up to an
edge for White.
23 . .. ..ie6 24 ..ti>g2 llad8 25 ..ib3 ..ti>g7
26 l:td4 h6 27 h4 gxh4 28 l2Jh5 'it>g8
29 .:txh4 a5 30 lld4?!
Throwing away the advantage. 30
ttJf4 was the right course.
30 . . . b5?
Missing a chance. After 30 ...l:tf5 3 1
.l::!.g4+ .l::!.g5 Black is not worse, but now
White simply rounds up the d-pawn.
31 l2Jg6 llf7 32 l2Jgf4! 1 -0
In fact Black can still struggle on
with 32 ....5lf5 33 .5lxd5 'lt>h7, but per­
haps Basin didn't fancy being tortured.
Game 9
Alvarado-Matamoros
Telde 1993
1 e4 e6 2 d4 d5 3 l2Jd2 l2Jf6 4 e5
l2Jfd7 5 ..id3 c5 6 c3 l2Jc6 7 l2Je2
cxd4 8 cxd4 f6 9 exf6 l2Jxf6 1 0 l2Jf3
..id6 1 1 0-0 0-0 1 2 ..if4 ..ixf4 1 3
l2Jxf4 'ill'd6
A useful alternative to the main
line, especially if Black wants to avoid
29
Th e Fre n c h Tarra s c h
the reams o f theory that follow after
13 ...ltJe4. Note that 13 ...'ifc7 14 g3
transposes to Chapter 2, Game 15.
However, 1 3 ... .id7, while looking like
a perfectly useful move, is in fact too
pedestrian for the demands of this po­
sition. After 14 l1e1 White can claim a
comfortable edge.
1 4 g3
Not 14 'ifd2?? ltJe4 and White drops
a piece, but 14 ltJe2 can be considered.
1 4 . . . l2Jg4
1 5 .txh7+ ! ?
A very tempting idea, but Black has
some deeply hidden resources! It is
safer to play 15 'ifd2, when we have
two main alternatives:
a) 15 ... eS 16 dxeS 'ii'h6 17 h4 was
clearly better for White in Zapata­
Matamoros, Seville 1992.
b) 15 ... .id7 is more subtle, reserving
the inevitable ... e6-e5 until an oppor­
tune moment arises. Now 16 .l::!.fel
l::.ae8 17 h 3 e S! carries more punch, as
White has weakened himself with h2-
h3. For example, 18 dxeS ltJgxeS 19
ltJxeS ltJxeS 20 .ie4 d4 21 .ixb7?!
.ic6! and Black has excellent play
along the a8-h 1 diagonal. White can
also opt for the waiting 17 .l::!.adl
30
(instead of 17 h3) . Erald-Kobalija,
World U-18 Championship, Cala
Galdana 1996, continued 17 ... eS!? 18
dxeS ltJcxeS 19 .ie4 'lith8 2 0 ltJxeS
.l::!.xeS 21 .ixdS .l::!.xel+ 22 'ifxel l::.xf4
and now instead of 23 .if3? ltJeS
White has to take the bait with 23
gxf4! 'ifxf4. It looks as if there is no
way out, but Fritz has pointed out the
brilliant defence 24 .ic6!!, when
White defends after 24 ...'ifxh2+ 25
@fl bxc6 26 .l::!.xd7 'ifh 1 + 27 'lt>e2
'ife4+ 28 'lt>d2 'iff4+ 29 'ii'e 3!
1 5 . . .'it>xh7 16 l2Jg5+ '1ti>g8 1 7 'il'xg4
e5 1 8 dxe5
Another Matamoros encounter
went 18 'ifh4 'ifh6 1 9 ltJxdS 'ii'xh4 20
gxh4 exd4 21 .l::!.adl .ifs 22 ltJc7 .l::!. ac8
23 ltJce6 .ixe6 24 ltJxe6 l:If6 25 ltJxd4
ltJxd4 26 .l::!.xd4 �c2 27 .l::!.b4 .l::!.g6+ 28
'lt>hl l::.b 6 29 l1xb6 1h-1h Godena­
Matamoros, Cannes 1996. In the final
position Black immediately regains a
pawn, when White's wrecked pawns
don't offer any winning chances.
1 8 . . .'il'h6 1 9 'il'h5 l:txf4 20 'il'xh6
gxh6 21 gxf4 hxg5 22 fxg5 ..if5!
This is much stronger than
22 ...ltJxeS, which allows White to roll
the passed pawns with 23 f4 ltJc4 24 fS.

After 24 ... tDe3 25 .l::!.f 3! tDxfs 26 .l::!.afl
lbd6 2l :f8+ 'lt>gl 28 h4 White has a
winning position. The text move
22 .....tfs leads to a very unusual posi­
tion. White has four connected passed
pawns, but they are all well blockaded
and in fact it is Black's lone passed d­
pawn that wins the day.
23 f4 �7 24 llfd 1 ..ti>e6 25 l:td2
llh8! 26 llc1 llh4 27 l:tf2 d4 28 ..ti>f1
ttJb4 29 llc 7 tLld3 30 llxb 7
Giving up an exchange, but after 30
l:If3 tDxf4 White's position collapses.
30 . . .tLlxf2 3 1 ..ti>xf2 .:txf4+ 32 ..ti>e 1
'1ti>xe5 33 llxa7 '1ti>e4 34 lle7+ ..ti>d3
35 lla7 ..ti>e3 36 b4 d3 0-1
The d-pawn will soon advance to
promotion.
Game 10
Rozentalis-Hergott
Montreal 1995
1 e4 e6 2 d4 d5 3 tLld2 tLlf6 4 e5
lLlfd7 5 i..d 3 c5 6 c3 tlJc6 7 tLle2
cxd4 8 cxd4 f6 9 exf6 tLlxf6 10 tLlf3
..td6 1 1 0-0 0-0 1 2 i.. g 5
This is probably the most signifi­
cant alternative to 12 .tf4, although it
does contain many transpositional
3 . . . ll'if6: Main L i n e with 1 1 . . . 0-0
possibilities. Other ideas include:
a) 12 h3 (this quiet line should cause
Black no problems whatsoever)
12 ... .tdl 13 .5le3 'ii'el 14 1:%.cl a6 1 5 a3
'lt>h8 16 %:!.el .5le8 ll t2lg3 tDhs 18
tbxhs .5lxh5 19 g4 .5le8 20 'lt>g2 'ii'f6 2 1
.5lg5 'iifl 2 2 .th 1 lbxd4! and Black had
won a pawn in Zaichik-Panchenko,
Lvov 198l, as 23 lbxd4 'ii'xf2+ 24 'lt>h l
'ifh2 is mate.
b) 12 lbc3 a6 (12 ... 'ifb6 transposes to
Chapter 2, Games 16-18) 13 .5le3 (13
.5lg5!?) 13 ... ..tdl 14 %:!.cl 'ii'el (naturally
cl and b6 are also possible squares for
the queen, but here Black opts for the
kingside) 15 l:e l lbg4 16 .5lg5 'ii'fl ll
h3 h6!? 18 .5lh4 tDf6 19 .5lg3 .5lxg3 20
fxg3 'iie l 2 1 a3 'ifd6 22 'lt>h2 .5le8 23
'ife2 .tbs 24 'ifxe6+ 'ii'xe6 25 l:Ixe6
.5lxf3 26 gxf3 tDxd4 2l %:!.el l;tfl with
an equal endgame in Rovid-Ulibin,
Cappelle la Grande 1994.
c) After the tricky 12 %:!.e l!? Black's
best bet looks to be a transposition to
Chapter 2 with either 12 ...'ifb6 or
12 ...'ifcl, as both 12 . . ...tdl and
12 ...'ife8 may be answered by 13 .tf4,
which gives White an edge.
1 2 . . . 'il'e8
The only other independent idea is
12 ... ..tdl!?, keeping Black's options
open. After 13 .5lh4 .5le8 14 'ifb 1 h6
15 .5lg6 .5lxg6 16 'ii'xg6 'ii'e8 ll 'ifxe8
:axe8 18 .5lg3 lbe4 19 .5lxd6 tDxd6 20
.l::!.fdl lbe4 21 t2lg3 lbxg3 22 hxg3 .l::!.c8
23 .l::!. d2 @fl 24 %:!.el 'litel the players
agreed a draw in T.Horvath­
Kindermann, Germany 1996. Once
again Black also has 12 ... 'i'b6 and
12 ...'ifcl.
1 3 i..f4!
A clever switchback! Rozentalis be-
31
The Fre n c h Ta rra s c h

lieves that the black queen is worse off

on e8 than on d8 and after this game it
is difficult to argue with his assess­
ment.
1 3 . . . ..ixf4 1 4 l2Jxf4 l2Je4
Or 14 ... eS 15 dxeS tDxeS 16 .ie2
tDf7 17 l:Ie l 'ifd8 18 'ifd4 and White's
control over d4 and pressure against
the d-pawn was sufficient for an ad­
vantage in Brodsky-Lipka, Karvina
1992.
1 5 'il'c 1
1 9 d5
This is stronger than 19 'ifxe6+
'lt>h8!, when 20 d5 fails to 20 ....lhf4!
and 20 'ifxe8 .l::!.axe8 gives Black plenty
of play.
1 9 . . . l2Jxd5 20 'il'xe6+ .:tf7! 21 'il'xe8+
llxe8 22 l2Jxd5 ..ixd5 23 l2Jd4 .J:!.d7
24 l2Jf5 .J:!.e2 25 b3 ..ie6?! 26 l2Jg3
llc2 27 llfe 1 �7 28 lle3 lldd2?

1 5 . . . ..id7
Aggressive measures only backfire
for Black. For example, 15 ... gS 16
.ixe4 .l:r.xf4 17 .ixh7+ 'lt>xh7 18 lDxgS+
wins, while 16 ...gxf4 17 .id3 leaves
Black's pawn structure in a real mess.
Black's 'extra move' ... 'ife8 also means
that the positionally desirable ...lDgS is
impossible. All in all, it would appear
that White can retain a significant After 28 ....idS! Black's active pieces
edge. ensure good drawing chances.
1 6 'il'e3 l2Jb4 1 7 ..ixe4?! 29 l2Je4! l:te2 30 l2Jg5+ \tie 7 3 1 l:Z.xe2
After this move Black obtains some l:!.xe2 32 ..ti>f 1 ..ig4
unexpected counterplay. Rozentalis After 32 ...l::.c2 33 %:!.e l or 32 ... .l::!.eS 33
gives 17 tDeS tDxd3 18 tDfxd3 .ibS 19 tDxe6 'lt>xe6 34 l:Ie 1 White can achieve
.l:r. fel .ixd3 20 tDxd3 as an improve­ a won king and pawn endgame. Now,
ment. however, Black loses the exchange.
1 7 . ..dxe4 1 8 'il'xe·4 ..ic6! 33 h3 lle5 34 f4 1 -0

32
 3 . . . li:Jf6: Main Line with 1 1 . . . 0-0

Summary
Despite rigorous tests, the 12 .if4 .ixf4 13 ltJxf4 ltJe4 14 ltJe2 l:txf3 exchange sac­
rifice continues to annoy White, although it is clear that both players need to
know their stuff before they venture into these critical lines. White players
wanting an easier life will probably prefer 14 °ifcl, which is certainly a super­
solid option. It will be interesting to see whether 14 ltJhS or 14 g3 become more
popular.

1 e4 e6 2 d4 d5 3 l2Jd2 l2Jf6 4 e5 l2Jfd7 5 i..d3 c5 6 c3 l2Jc6 7 l2Je2 cxd4 8

cxd4 f6 9 exf6 l2Jxf6 1 0 l2Jf3 i..d6 1 1 0-0 0-0

1 2 i..f4
12 .igS - Game10
1 2 ... hf4 1 3 llJxf4 llJe4 (D)
13 ..."ifd6 - Game9
1 4 l2Je2
14 °if cl ltJgS 15 ltJxgS °ifxgS (D)
16 ltJe2 Game5 -

16 .ixh7+ Game6 -

14 ttJhs Game7
-

14 g3 Game8
-

1 4.. .llxf3 1 5 gxf3 l2J g5 1 6 'iti>h1

16 f4 (DJ
16 ...ttJf3+ Game3 -

16 ...ttJh3+ Game4 -

1 6 . . . e5 1 7 dxe5 l2Jxf3 1 8 l2Jg 1

18 .ixh7+ Game2 -

1 8 .. llJfxeS
. Game1
-

13...llJe4 15. . ."ikxgS 16 f4

33
 CHAPTER TWO I
3 ...'2Jf6: Main Line
with 11...'i/c7 and 11 ...'ii'b 6

1 e4 e6 2 d4 d5 3 l2Jd2 l2Jf6 4 e5 1 1...0-0. However, there was no con­

l2Jfd7 5 i..d 3 c5 6 c3 l2Jc6 7 l2Je2 crete reason for its decline in popular­
cxd4 8 cxd4 f6 9 exf6 l2Jxf6 1 0 l2Jf3 ity, and it remains a fundamentally
i..d6 , , 0-0 sound variation.
In this chapter we shall look at lines
involving an early ...'i!lc7 or ... 'i!lb6, Game 1 1
both of which discourage White from Emms-Menoni
playing the desirable ..5lf4. In contrast Montecatini Te rme 1996
to the previous chapter, play tends to
be more positional, with both colours 1 e4 e6 2 d4 d5 3 l2Jd2 l2Jf6 4 i..d 3
battling for control over the all­ c5 5 e5 l2Jfd7 6 c3 l2Jc6 7 l2Je2 cxd4
important dark squares in the centre. 8 cxd4 f6 9 exf6 l2Jxf6 1 0 l2Jf3 i..d6
Games 1 1-15 deal with 1 1 ...'i!lc7, 1 1 0-0 -.c7 1 2 i.. g 5
which directly prevents 12 ..5lf4. This The most popular move. 1 2 g3 is
line reached its height of popularity in considered in Game 15.
the 1980s, when it was used success­
fully by top players such as Jan Tim­
man, Mikhail Gurevich and Sergei
Dolmatov. Although new methods
have been discovered for White that
have rendered 1 1 ...'i!lc7 less fashion­
able than 1 1 ...0-0, it remains a very
important line.
In Games 16-20 we look at
1 1...'i!lb6. Originally this was the most
usual move, but it has been gradually
superseded by 1 1 ...'i!lc7 and then

34
 3 . . .
1 2 . . 0 -0 1 3 i.. h 4
. .:tae8 20 h3 .:te 7
Again this is the most fashionable
move. 13 %:!.cl and 13 ltJc3 a6 are the
subjects of Games 13 and 14 respec­
tively.
1 3 . . .e 5
We shall look at 13 ...ttJhS in Game
12.
14 dxe5 l2Jxe5 1 5 l2Jxe5 i.. x e5 1 6
.tg3 i.. x g3 1 7 l2Jxg3
In practice White has scored very
well from this position. Indeed, in
searching through over twenty games
I was unable to find a Black victory.
Although his position looks quite
sound, Black has no real attacking
prospects; White has no weaknesses
whatsoever and this makes it difficult
for Black to achieve any counterplay
along the f-file. Furthermore, the d­
pawn often becomes a liability, while
Black's king also sometimes suffers
from its slight lack of cover. The real
problem is that White's edge often
persists well into the endgame. From a
purely theoretical standpoint Back
must be quite close to equality, but it
is not surprising that many Black ex­
perts avoid this position.
1 7 . . . 'il'b6 1 8 'il'd2 i..d 7 1 9 .!Iad 1
ll'i f6: Main Line with 1 1 . . . 1kc7 a n d 1 1 . . . 1k b 6
20... .ibS 21 'iib4 .ixd3 22 'iixb6
axb6 23 .l::!.xd3 still leaves White with
an edge, due to the pawn weaknesses
in the black camp.
2 1 llfe 1 .:tfe8 22 .:txe7 .!Ixe7 23 i..c2
.tb5 24 'il'g5 h6 25 'il'd2 a5 26 .tb1
.:td7
If 26 ... d4 then not 27 'iixd4 'iixd4
28 lhd4 %:!.el+, but 27 .ifs! and Black
has problems defending the d4-pawn.
27 l2Jf5 l2Je4 28 'il'd4 'il'e6
28 .. .'iixd4 29 ltJxd4 .ia6 30 ttJb3
wins a pawn for White, as e4 and a5
are attacked.
29 l2Je3 i..c6 30 a3 'il'f6 3 1 f3 'il'xd4
32 llxd4 l2Jf6 33 �2
White's agreeable posltlon contin­
ues into the endgame. White has a
simple plan of advancing the kingside
pawn majority to gain more space,
while the dS-pawn remains a perma­
nent worry for Black.
33 ...�7 34 h4 h5 35 ..ti>g3
35 g4! is more direct, e.g. 35 ... hxg4
36 fxg4 ltJe4+ 37 .ixe4 dxe4 38 .l::!.xd7+
.ixd7 39 'lt>g3 .ic6 40 'lt>f4.
35 . . .l2Je8 36 l2Jf5 g6 37 l2Je3 l2Jc7 38
ftd2 'itf6 39 'iti>f2 l2Je6 40 g4 l2Jc5?!
40... gS! would have offered good
35
The Fre n c h Ta rra s c h
drawing chances.
41 g5+ 'i;f7 42 ..ia2! 'i;e6 43 b4
axb4 44 axb4 l2Ja6
45 b5! ..ixb5 46 l2Jxd5 !
46 .l::!.xd5 l:txd5 47 .5lxd5+ @e7 48
.5lxb7 also maintains reasonable win­
ning chances, but I was tempted by a
mating net.
46 . . . 'i;f5 47 llb2!? ..ic6 48 l2Jf6 llg7
49 lle2 l2Jc5?
49 ...@f4 was the last chance. Now
the web closes around the black king.
50 'i;g3 ! ..ib5 5 1 ..ib 1 + ..id3
52 f4!
Needless to say, I was very satisfied
with the conclusion of this game, but
in fact it turns out White has an alter­
native finish with 52 .5lxd3+ ltJxd3 53
36
ltJg4! (53 ... hxg4 54 fxg4 mate is quite
visual) 53 ... .l::!.h7 54 :es and ltJe3 mate.
52 . . .l2Je4+ 53 l2Jxe4! ..ixe2 54 l2Jc5+
1 -0
It is mate next move.
Game 12
Yemelin-Kostenko
S zeged 1994
1 e4 e6 2 d4 d5 3 l2Jd2 l2Jf6 4 e5
l2Jfd7 5 c3 c5 6 .td3 l2Jc6 7 l2Je2
cxd4 8 cxd4 f6 9 exf6 l2Jxf6 1 0 l2Jf3
..id6 1 1 0-0 'il'c7 1 2 ..ig5 0-0 1 3
.th4 l2Jh5
The major alternative to 13 ... e5, al­
though the uncommon 13 ... .5ld7 also
looks quite playable.
1 4 'il'c2
The most questioning reply. Other
moves do nothing to prevent Black's
standard plan of ... g7-g6 and .. . °ifg7.
1 4 . . . h6
The natural 14 ...g6 falls short after
15 .5lxg6! and now:
a) 15 ... hxg6 16 "ifxg6+ ltJg7 17 ltJg5
and wins.
b) 15 .. Jlxf3 16 .txh5 .5lxh2+ 17
@h l .l::!.f5 18 .5lg4 .l::!.f8 19 g3 and the
bishop is lost.
 3 . . . ll'i f6: Main Line with 1 1 . . . 1kc7 a n d 1 1 . . . 1kb6
1 5 i.. g 6 l2Jf4
After this move White tends to ob­
tain a clear advantage, so perhaps
Black players should look into
15 ...llxf3!?, which hasn't been wit­
nessed much but is a trickier move.
After 16 gxf3 .5lxh2+ 17 'ifi>hl ltJf4 18
ltJg3 Black has two ways to play:
a) Akopian gives 18 ...ttJxg6 19 'ifxg6
.5lxg3 20 .5lxg3 'iff7 2 1 'ifxf7+ 'iti>xf7 22
nfdl with only a minute advantage
for White in the endgame. Indeed
Black's position does seem quite diffi­
cult to breach.
b) Altgelt-Stamenkovic, Sofia 1994,
saw instead 18 ... .5lxg3 19 .5lxg3 .5ld7
20 'ifd2 l:.f8 with some compensation
for the exchange. Indeed, Black went
on to win this game.
1 6 l2Jxf4 i..xf4 1 7 .:tad 1
Black's predicament lies with the
shaky light squares around his king.
White has a very simple plan of
switching the roles of bishop and
queen along the b 1-h7 diagonal and
playing for checkmate. It is difficult to
find a suitable antidote to this idea.
1 7 . . . i..d7 1 8 'il'e2! a6 1 9 .:tfe 1 l2Jd8
20 l2Je5 i.. x e5 21 dxe5 llc8 22 b3
i..e8 23 i.. b 1 g5
Extremely committal, but Black
feels obliged to cover h7.
24 i..g 3 'il'g7 25 'il'd3 l:!.f7 26 'il'd2
.:tfc7
see following diagram
2 7 f4!
The start of a decisive breakthrough
27 . . .i.. g 6 28 f5! exf5 29 'il'xd5+ '1ti>h7
30 e6 f4 31 e7 llxe7 32 i..xg6+
..ti>xg6 33 'il'd6+ 'il'f6 34 llxe7 fxg3
35 'il'xf6+ ..ti>xf6 36 lle8 1 -0
Game 13
Smagin-Dolmatov
USSR Championship 1986
1 e4 e6 2 d4 d5 3 l2Jd2 l2Jf6 4 e5
l2Jfd7 5 i..d3 c5 6 c3 l2Jc6 7 l2Je2
cxd4 8 cxd4 f6 9 exf6 l2Jxf6 1 0 0-0
i..d6 1 1 l2Jf3 'il'c7 1 2 i..g5 0-0 1 3
llc1
An alternative way to proceed. The
objective is to dissuade ... e6-e5 before
proceeding with the .5lh4-g3 plan. Of
course the tempo spent achieving this
idea gives Black more time to start a
counterattack.
1 3 . . .l2Jg4!
Cutting across White's main goal.
Now 14 h3 allows 14 ...nxf3!, when 15
gxf3 ltJh2 and 15 hxg4 .l::!.f7 are both
fine for Black.
1 4 l2Jg3 g6 1 5 l2Jh4
15 tiJd2!? is a fairly new move of
Tiviakov's which certainly deserves
attention. After 15 ...ttJf6 16 ttJb3 'it'g7
17 'ifd2 .5lb4 18 'ife3 ltJg4 19 'ife2 ltJh6
20 .tbs .5ld7 2 1 .5lxc6 bxc6 22 ttJcS
.5lxc5 23 .5lxh6 'ifxh6 24 :XcS Black's
bad bishop condemned him to a pas-
37
The F re n c h Tarra s c h
sive position in Tiviakov-Komarov,
Kherson 199 1.
1 5 . . .e 5
Also possible is 1 5...ttJf6 16 'ifd2
and then:
a) 16 ... 'ifb6? walks into a nasty at­
tack with 17 ..5lh6 'ifxd4 18 ttJf3 'ifb6
19 'it'g5! ltJe4 20 ltJxe4 ..5lf4 2 1 ltJf6+
:xf6 22 'ifxf6 ..5lxh6 23 ..5lxg6 hxg6 24
'ifxg6+ ..5lg7 25 ltJg5 'ifd8 26 'iff7+
'lt>h8 27 'ifh5+ 'lt>g8 28 .l::!. c3 ttJe5 29 f4
1-0, Geller-Ravi, Coimbatore 1987.
b) 16 ...ttJg4!? 17 'if dl ttJf6 18 'ifd2
ltJg4 19 'ife2 (19 'ifdl ftJf6 would be a
threefold repetition) 19 ...ttJf6 20 ..5le3
(20 ..5lh6! .l::!.f7 2 1 ..5le3 is stronger, as it
prevents the ...'ifg7 plan) 20 ...'ifg7! 21
f4? (2 1 ltJf3 is equal) 2 1 . ..'ifh6!, win­
ning a pawn in Emms-Klingelhofer,
Hamburg 1992.
1 6 ..ie2 l2Jf6 1 7 dxe5 ..ixe5 18 b4
..if4 1 9 ..ixf4 'il'xf4 20 b5
see following diagram
20 . . .l2Jd4!
A big improvement over 20 ... ttJb4
21 ltJxg6! hxg6 22 a3, winning back
the knight with an excellent position
in Geller-Dolmatov, Russia 1985. Af­
ter 20 ... ttJd4 White's advantage is
38
small enough for Black not to have
any real problems.
21 l2Jf3 l2Jxe2+ 22 l2Jxe2 'il'd6 23
l2Jed4 ..id7 24 l:l'.e1 a6 25 a4 axb5 26
axb5 .:tae8 27 'il'b3 .:txe 1 + 28 l:l'.xe 1
.:tea 29 l:l'.xe8+ ..ixe8 30 h3 ..id7 3 1
'ill'e 3 l2Je4 Yz - Yz
Game 14
Marjanovic-Timman
Sarajevo 1984
1 e4 e6 2 d4 d5 3 l2Jd2 l2Jf6 4 e5
l2Jfd7 5 c3 c5 6 ..id3 l2Jc6 7 l2Je2
cxd4 8 cxd4 f6 9 exf6 l2Jxf6 10 0-0
..id6 1 1 l2Jf3 'il'c7 1 2 l2Jc3
The old line. The white knight
heads for the queenside, gaining a
tempo because of the threat of ttJb5.
White's plans include l:tc l and ltJa4-c5,
which would put the knight on a very
useful square. The flip side is that
Black has a freer role to operate on the
kingside, now that the squares f4 and
g3 have less protection.
1 2 . . . a6 1 3 ..ig5 0-0
This position also often arises from
the alternative move order 12 ..5lg5 0-0
13 ltJc3 a6.
1 4 ..ih4 l2Jh5
 3 . . .
1 5 .:tc 1
1 5 'ifc2 can b e answered with
15 ... g6! here, as 16 .5lxg6? runs into the
surprising 16 ... ltJf4! 17 .5ld3 ltJb4 and
Black wins a piece. This is one of the
disadvantages of 12 ltJc3.
1 5 . . . g6 1 6 i.. b 1
16 ttJa4 .td7 17 ttJcs lhf3 ! 18 'ifxf3
ltJxd4 19 'it'g4 .5lxc5 occurred in Griin­
feld-Shachar, Israeli Championship
1994. Here White's best is 20 'ifxd4,
but after 20 ... 'ifxh2+! 21 'lt>xh2 .5lxd4
Black is certainly no worse.
1 6 . . .'il'g7 1 7 'it>h 1 ? !
This seems a little slow. White can
also try:
a) 17 ttJa4 .l::!.xf3!? (17...l:tb8!?) 18 gxf3
.td7 19 ttJcs .l::!. f8 20 .5lg3 .5lxc5 21
dxcS 'ifxb2 2 2 l:te l 'iff6 2 3 'lt>hl ltJd4
24 .5le5 'ifxf3+ 25 'ifxf3 ttJxf3 was
clearly better for Black in Womacka­
Pahtz, East German Championship
1985, but I think that White's play can
be improved upon.
b) 17 .5lg5!?, returning to the c1-h6
diagonal, is an interesting idea. The
point is that with the knight on hS,
the bishop has no real future on h4, so
it returns to its original diagonal. I
believe that Karpov was the first to
4J f6 : Main L ine with 1 1 . . . • c l and 1 1 . . . • b 6
use this plan, in a game against Jona­
than Mestel. A more recent example
went 17 ....5ld7 18 %:!.e l l:tae8 19 ltJa4
.5lf4 20 .5lxf4 :xf4 2 1 ltJcS! ltJxd4 22
ttJes .tbs 23 a4 b6 24 axbS bxcS 25
bxa6 'ifa7 26 ttJd3 and White was bet­
ter in Unzicker-Rozentalis, Germany
1995.
1 7 . . . i..d 7 18 lle1 llf7 1 9 lbe2 '1ti>h8
20 a3 llaf8
While White is groping for an active
plan, his opponent can steadily im­
prove his position before expanding
on the kingside. Already Black's posi­
tion is easier to play.
2 1 llc3 h6 22 i..g3
An admission of defeat. White ex­
changes his dark-squared bishop, but
only gets a knight in return - a poor
swap.
22 . . . lbxg3+ 23 lbxg3 llf4 24 lbe2
ll4f6 25 lbg3 i..e8 26 llce3 i..f7 27
..ti>g 1 i..g8 28 'il'd3 i..f4 29 ll3e2 'il'f7
30 b4 g5 31 h3 i..c7 32 lbf 1 i..b6 33
lb 1 d2 'il'g7 34 lbb3 .:tf4 35 lld 1 g4!
This thrust represents the culmina­
tion of Black's strategy. The d4-pawn
is very sensitive, while Black also
dominates proceedings on the king­
side.
39
The Fre n c h Tarra s c h

arise after 15 b3 (15 'ifd2!?) 15 ... 'lt>h8

16 ltJgS eS! 17 ttJxh7 ttJxh7 18 'ifhs e4
19 ltJg6+ 'lt>g8 20 'ifxdS+ IU7 21 .5lc4
ttJd8 22 ltJe7+ 'lt>f8. Here White can
take a draw with 23 ltJg6+. Instead in
Dvoirys-M.Gurevich, USSR Champi­
onship 1986, White played on with 23
ltJxc8, but after 23 ....l::!.xc8 24 'ifxe4
ltJgS 25 'it'g4 tiJf3+ 26 'lt>g2 'ifc6 27 d5
'ifh6 28 h4 .l::!.b8 matters were still far
from clear, and Black eventually won
the game.
36 l2Jh4
After 36 hxg4 .l::!.xg4 37 g3 eS
White's position falls apart, as 38 dxeS
allows 38 ....l::!.xg3+.
36 . . .l2Je7 37 g3?
White chooses the path of least re­
sistance. 37 ltJg6+ ltJxg6 38 'ifxg6
would have offered more hope.
37 . . . gxh3 38 'it>h 1 ? llxh4 39 f4
Or 39 gxh4 'ifg2 mate.
39 . . . llg4 40 �h2 l2Jf5 0-1
A model game from Timman.
1 4 llc 1 l2Jh5 !? 1 5 l2Jg5?!
Game 15 This attempt to complicate only
Kr .Georgiev-Dolmatov gives Black the chances. It would have
Sofia 1985 been better to play 1 5 .5lxd6 'ifxd6 16
ltJc3 g6 and now:
1 e4 e6 2 d4 d5 3 l2Jd2 l2Jf6 4 e5
l2Jfd7 5 i..d 3 c 5 6 c3 l2Jc6 7 l2Je2
cxd4 8 cxd4 f6 9 exf6 l2Jxf6 1 0 0-0
i..d6 1 1 l2Jf3 "iic 7 1 2 g3
This introduces another procedure
for exchanging the dark-squared bish­
ops.
, 2 . . . 0-0 , 3 i..f4

see following diagram

, 3 . . ..td7
13 ....txf4 14 ttJxf4 'ifb6!? is a differ­
ent idea. Great complications may a) 17 lle 1 a6 18 ltJeS ltJxeS 19 .l::!.xeS

40
 3 . . .
'i'Vb6 2 0 :c 2 lbg7! (the knight per­
forms a very useful defensive role
here) 2 1 'ii'g4 .tbs 22 lbxbS axbS 23
b3 :ac8 24 :ce2 :c3 2S :se3 b4 26
h4 'ii'd8 27 'ii'gS 'ii'f6 28 'ii'xf6 :xf6 29
.tbs :c7 30 'it>g2 'it>f7 31 :d2 'it>e7
with an equal position in Dvoirys­
Dolmatov, USSR Championship
1986.
b) Psakhis recommends 17 .tbs,
giving White a small, secure advan­
tage, but the sacrifice 17 ... :Xf3! is very
unclear. Indeed, after 18 'ii'xf3 lbxd4
19 'ii'd3 lbxbS 20 lbxbS 'ii'a6 2 1 a4
'ii'xa4 I prefer Black, who has two
pawns and some light-squared control
for the exchange. In Malevinsky­
Komarov, Leningrad 1989, White
tried 1 8 i.xc6, only to be shocked by
18 ... :xc3! 19 i.xd7 :xc l 20 i.xe6+
'ii'xe6 2 1 'ii'xc 1 :cs and Black already
had a winning position.
1 5 . . . h6 1 6 iLxd6 'ii'xd6 1 7 ll'ih7 l:l.f3!
1 8 iLb1
Or 1 8 'it>g2 :fs! 19 lbc3 g6 and the
knight on h7 is dead and buried.
1 8 . . . e5 1 9 l:l.xc6 bxc6 20 dxe5 'ii'x e5
21 ll'id4 lU7 22 l:l.e 1
22 . . .il.g4! 23 f3 'ii'd 6! 24 fxg4 ll'ixg3
25 ll'ie6 l:.e8 26 il.g6 l:l.xe6 27 iLxf7+
lll f6 : Main Line with 1 1 . . . 1ic7 a n d 1 1 . . . 1i b 6
�xf7 28 l:l.xe6 'ii'xe6 29 hxg3 �98
30 'ii'd 3 'ii'xg4
The knight can be picked up at
Black's leisure, while the most White
can hope for in return is a pawn.
Dolmatov, renowned for his immacu­
late technique, doesn't let it slip.
31 �f2 'ii'e 6 32 'ii'c2 g6 33 'ii'c5
'ii'f7+ 34 �g2 �xh7 35 'ii'xc6 h5 36
b4 d4 37 a4 'ii'e7 38 b5 d3 39 'ii'd5
'ii'e 2+ 40 �h3 'ii'f 1 + 41 �h2 'ii'f2+
42 �h 1 d2 0-1
Game 16
Nijboer-Farago
Dieren Open 1988
1 e4 e6 2 d4 d5 3 ll'id2 tl'if6 4 e5
ll'ifd7 5 c3 c5 6 il.d3 ll'ic6 7 ll'ie2
cxd4 8 cxd4 'ii'b6 9 ll'if3 f6 1 0 exf6
ll'ixf6 1 1 0-0 il.d6
see following diagram
This is the standard starting posi­
tion of the 1 1... 'i'Vb6 variation. Now
12 lbc3 is White's most popular move,
while 12 lbf4 and 12 b3 are dealt with
in Games 19 and 20.
1 2 ll'ic3 0-0 1 3 il.g5
13 i.e3 is seen in Game 18.
41
The Fre n c h Ta rra s c h
, 3 . . . il.d7
13 ... 'it>h8!? is the subject of the next
game.
1 4 l:l.e1
Alternatively White can reach this
position via the 13 :el .td7 14 .tgS
move order.
1 4 . . .ltJg4
The other key line is 14 ...'it>h8,
which has actually been scoring quite
well for Black. White can answer
with:
a) 15 lba4 (this is common, though
I'm not sure why) 15 ...'ii'a S! (I prefer
this to 15 ...'ii'c 7, which, however, is
also fully playable) and now:
al) 16 a3 lbg4 (16 .. .lbxd4 17 lbxd4
.txa4 1 8 b3 'ii'c 3 19 i.c2 .td7 20 lbxe6
42
is equal) 17 i.c2 eS 1 8 dxeS lbcxeS 19
lbxeS i.xeS 20 :xeS lbxeS 2 1 i.e7
.txa4 22 .txa4 :f4 with a clear advan­
tage to Black in Marciano-Jonsson,
Reykjavik 1993.
a2) 16 .td2 'ii'd8 17 :ct lbg4 18
.tgS? 'ii'b 8! 19 .th4 (19 h3 .th2+ 20
'it>fl lbxf2!! 21 'it>xf2 lbxd4 is an acute
example of Black's attacking potential
in this line) 19 ... .txh2+ 20 'it>fl 'ii'f4, as
in Fernandez Garcia-Illescas, Spain
1986.
b) 15 lbeS!? (I believe that this
causes Black more problems):
bl) 15 ... .te8 16 lbxc6 'ii'xc6 and
now 17 :ct can be answered with
17 ... .ths 18 'ii'd2 lbg4, so perhaps
White should play the prophylactic 17
h3.
b2) 15 ... 'ii'xd4!? 16 lbbS 'ii'xeS 17
:xeS i.xeS grants Black compensation
for the queen in Tolnai-Ziiger, Buda­
pest 1988, although after 18 f4 (instead
of Tolnai's 1 8 lbc3) I still prefer
White.
1 5 iLh4 ll'ih6 1 6 il.g3 iLe 7 1 7 ll'ia4
'ii'a 5 1 8 iLc2 ll'if5
After 18 ...:xf3?! 19 gxf3 :f8 20 a3
lbfs 2 1 .txfs :Xfs 22 b4 'ii'd8 23 bs
lbaS 24 lbcS Black had no compensa­
tion for the exchange in Howell-Ryan,
Isle of Man 1994. The American IM
John Watson, who has done much to
popularise Black's cause in his excel­
lent Plcry the French books, suggests
18 ... .te8 19 :xe6 .tbs, but here White
should play 20 a3 (not 20 'ii'd3 lbfS 2 1
lbe5 lbcxd4) when I don't believe that
Black has enough compensation for
the pawn.
1 9 a3 ll'ixg3
19 ...:ae8 20 b4 'ii'd8 21 lbcS gives
 3 . . .
White a clear edge, while 1 9...b 6 has
been suggested by Watson, though I
don't trust that at all after 20 'ii'd3
(threatening b2-b4) as 20 ... lbcxd4 fails
to 2 1 lbxd4 i.xa4 22 lbxf5 :Xf5 23
i.xa4 'ii'xa4 24 :xe6 :fl 25 :xe7
:xe7 26 'ii'xdS+.
20 hx93 'iic7 21 l:.c 1 !
Despite the absence of the dark­
squared bishop, White still retains
control of many of the important dark
squares. In particular, both rooks and
both knights are actively posted.
Overall White holds a significant plus.
2 1 . . . 96 22 b4 a6 23 tllc 5 l:.f6 24
ila4 b5 25 il.b3 iLxc5 26 l:l.xc5 'iib6
27 tlle 5 tllx e5 28 dxe5 l:.f7 29 'iid4
'iia7 30 l:l.e3 a5 3 1 bxa5 'iix a5
tll f6: Main Line with 1 1 . . . • c l and 1 1 . . . •b6
32 l:l.xd 5!? exd5 33 iLxd5 'iia 7!
Other moves lose. Nijboer gives the
following lines:
a) 33 ... i.e8 34 e6 l::t e7 35 :f3!, in­
tending :xf8+ and 'ii'h8 mate.
b) 33 ...i.f5 34 g4.
c) 33 ...l::td8 34 i.xf7+ 'it>xf7 35 l:r.d3.
34 iLxf7+ �xf7 35 'iid6 l:.e8?
35 ...'it>g8 isn't mentioned by Ni­
jboer and seems like a tougher de­
fence. After 36 e6 i.e8 37 :£3 h6 38
39 ...'ii'xa3 40 'ii'xa3 (40 'ii'f6 'ii'a l +)
l::tf8+ 'it>h7 39 e7 (intending 'ii'f6)
40 ...:xa3 41 :Xe8 :al+ 42 'it>h2 :el
Black can hold the rook and pawn
endgame. After 35 ...:e8? it is all quite
straightforward for White.
36 l:.c3 .Uc8
37 'iif6+ �98 38 l:txc8+ iLxc8 39
'iid8+ �f7 40 'iix c8 'iixa3 41 'iid7+
�8 42 �h2! 'iib2 43 f4 'iie2 44 94
'iic4 45 'iid8+ �7 46 'iif6+ �98 47
e6 'iic7 48 e7 'iic8 49 f5 'iib8+ 50
�h3 'iie8 51 'iie6+ �h8 52 f6 1 -0
Game 1 7
Majer-Ruzicka
Correspondence 1993
1 e4 e6 2 d4 d5 3 tlld 2 tllf6 4 e5
43
The Fre n c h Ta rra s c h
tllfd7 5 il.d3 c 5 6 c 3 tll c6 7 tll e2
cxd4 8 cxd4 f6 9 exf6 tll xf6 1 0 tll f3
il.d6 1 1 0-0 0-0 1 2 il.g5 'ii'b6 1 3
tll c3
This position could also be reached
via the move order 1 1 ...'ii'b 6 12 lbc3
0-0 13 .tg5.
1 3 . . .�hB ! ?
This has some advantages over the
often-played 13 ....td7, as Black retains
the option of transferring the queen to
the kingside via c7-f7. Note that
13 ...'ii' xb2 is a little too ambitious: 14
lbb5! leaves the black queen struggling
to find a retreat.
1 4 tll a4 'ii'c7 1 5 l:.c1
1 5 . . . 'ii'f7
15 ...lbg4!? is another enticing alter­
native, leading to extremely intense
positions. After 16 h3 lbh2 we have:
a) 17 lbe5!? i.xe5 18 dxe5 lbxfl 19
'ii'h 5 h6 20 .tf6 gxf6 2 1 'ii'xh6+ 'it>g8
22 exf6 :xf6 23 'ii'xf6 'ii'h2+ 24 'it>xfl
'ii'h l+ 25 'it>e2 'ii'xcl 26 f4 'ii'g l 27
.th7+ lh-lh Akopian-Ulibin, Minsk
1990. A crazy miniature.
b) 17 lbxh2 .txh2+ 18 'it>h l .tf4! 19
'ii'h 5 g6 20 .txf4 :xf4 2 1 i.xg6 'ii'e7 22
.td3 :h4 23 'i' e2 e S and Black had an
unexpected attack for the pawn in
44
Kosashvili-Ulibin, Santiago 1990. The
game concluded 24 'ii'e3 i.g4 25 lbc3
l::tg8 26 'it>h2 e4 27 i.e2 .txh3 28 gxh3
'ii'd6+ 29 'it>hl 'ii'd7 30 'iii>h2 'ii'd6+ 3 1
'it>hl 'ii'e6 3 2 'iii>h2 lbe5 3 3 f4 exf3 34
:xf3 lbg4+ and White was forced to
resign.
1 6 tllc 3?!
I don't like this move at all. What's
the point of sticking the knight on a4,
only for it to return, tail between its
legs, a few moves later? Here are two
other possibilities:
a) 16 .th4!? is positionally well mo­
tivated, but after 16 ...'ii'h5 17 .tg3
.txg3 either recapture allows 1 8 ...lbg4
with annoying pressure on the king­
side.
b) 16 :el looks quite natural. The
game Andrade-Pettinger, Rimavska
Sobota 1996, continued 16 ...'ii'h5 17
h3 e5, when instead of 18 dxe5 White
should play 18 l:.xc6!, as 1 8 ... bxc6 19
dxe5 .tb4 20 exf6 i.xe 1 21 fxg7+
'it>xg7 22 'ii'xel clearly favours White.
17 ...lbxd4 is stronger, and after 18
lbxd4 'ii'xg5 19 lbxe6 i.xe6 2 0 :xe6
:ad8 we reach a level position.
1 6 . . .'ii'h 5 1 7 iLxf6
17 h3? lbxd4 18 lbxd4 'ii' xg5 led to a
quick Black victory in Hossain­
Wiedenkeller, London 1987. 17 .txf6
avoids this trick but renounces any
claim for an advantage. Indeed, I al­
ready like the look of Black's position.
The rest of the game is very one-sided
and sees White putting up very little
resistance to the inevitable kingside
attack.
1 7 . . .l:.xf6 1 8 h3 il.d7 1 9 iLe2 l:.af8
20 tlle5 'ii'e8 2 1 tllxd7 'ii'xd7 22
ilg4? 'ii'f7 23 iLf3
 3. . .
2 3 . . . l:l.g6! 2 4 il.g4 h5! 0-1
25 i.xh5 'ii' f4 26 g3 :xg3+ 27 fxg3
'ii'xg3+ 28 'it>hl 'ii'h 2 is mate.
Game 18
Sowray-B. Martin
London 1996
1 e4 e6 2 d4 d5 3 tll d2 tllf6 4 e5
tllfd7 5 c3 c 5 6 il.d3 tll c 6 7 tlle2
cxd4 8 cxd4 'ii'b6 9 tllf3 f6 1 0 exf6
tllxf6 1 1 0-0 il.d6 1 2 tllc3 0-0 1 3
ii.. e3 il.d7
13 ... 'ii' xb2 is again met by 14 c!bb5.
14 a3 ilea
Note that 14 ...'ii'x b2 drops the
queen immediately to 15 c!ba4!
1 5 tll g5
li:J f6 : Main Line with 1 1 . . . "ii c l a n d 1 1 . . . "ii b 6
Many moves have been played in
this position, but no route to an ad­
vantage has been found for White, e.g.
a) 15 'ii'c 2 c!be7 16 h3 i.g6 17 l::lfe l
l::lae8 18 i.g5 i.xd3 19 'i'xd3 c!bg6 20
'ii'd2 c!bh5 21 c!ba4 'ii'b5 and the ex­
change of light-squared bishops had
eased Black's position in Grinfeld­
N.Pert, Hastings 1996.
b) 15 'ii'd2 i.h5 16 c!be5 c!bg4 17
c!bxg4 i.xg4 18 :ac l 'ii'd8 19 c!be2 i.f5
20 i.xf5 :Xf5 21 c!bg3 :£7 22 l:r.fe 1
'ii'b6 23 :c3 .U.af8 and once again
Black had no problems in I.Gurevich­
Gdanski, Santiago 1990.
c) 15 c!be5 i.xe5 16 dxe5 'ii'xb2 17
exf6 'ii'xc3 18 fxg7 'ii'xg7 leaves White
with insufficient play for the pawn.
1 5 . . .tlle7 1 6 tlla4
This gives Black a comfortable
game, but it is difficult to suggest an
improvement. After 16 h3 i.c7 17
'ii'c2 h6 18 c!bf3 c!bh5 the knight heads
for f4 and Black can once more look
to the future with confidence.
1 6 . . . iLxa4 17 'ii'xa4 h6!
17...c!bf5 has also been played, but I
like the forcing nature of this move.
1 8 tllxe6 iLxh2+ 1 9 �xh2 'ii'xe6
White has the token advantage of
45
The Fre n c h Ta rra s c h
the bishop pair, but it soon becomes
obvious that the most important as­
pect of the position is his shaky king.
20 l:tae 1 tllg4+ 21 �g1 'ii'd 6 22 g3
l:l.f3 23 �g2
23 .tf4 :xf4! 24 gxf4 'ii'xf4 25 'it>g2
:f8 is the end.
23 . . .l:Iaf8 24 'ii'b4 'ii'f6 25 iLe2
25 . . . l:l.xg3+! ! 26 �xg3 tllf5+ 27 �g2
27 'iii>xg4 'ii'h4+ 28 'iii> f 3 'ii' e 4 is mate.
27 . . . 'ii'g 6 28 iLf3
Other moves also wind up in
checkmate, e.g. 28 :g1 lbgxe3+ 29
'it>h2 'ii'f6 30 fxe3 'ii'h4+ 3 1 'it>g2 lbxe3
and 28 .txg4 'ii'xg4+ 29 'it>h2 lbh4 30
:g1 lbf3+ 3 1 'it>hl 'ii'h3 .
28 . . . tllh4+ 29 �g3 l:txf3+ 30 �xh4
'ii'f6+ 0-1
46
31 'it>xg4 'ii'fS+ 32 'it>h4 :h3 is mate.
A convincing display from Martin.
Game 19
Muratov-Bata
Correspondence 1967
1 e4 e6 2 d4 d5 3 tlld 2 tll f6 4 e5
tllfd7 5 il.d3 c5 6 c3 tllc6 7 tlle2
'ii'b6 8 tllf3 cxd4 9 cxd4 f6 1 0 exf6
tllxf6 1 1 0-0 il.d6 1 2 tllf4
Initially a critical choice, this line
has virtually disappeared from tour­
nament chess, as it merely seems to
give Black all the chances.
1 2 . . . 0-0 1 3 l:l.e 1 il.d7! 1 4 tllxe6 l:.fe8
1 5 iL f5 il.b4 1 6 iLd2 iLxd2 1 7 'ii'xd 2
tlle 7!
 3. . .
1 8 g4?
This makes matters worse. It would
have been better to play either 1 8
lbxg7!? o r 1 8 .th3:
a) 18 lbxg7!? Wxg7 19 'ii'g5+ lbg6 20
.txd7 lbxd7 2 1 h4 h6 22 'ii'xd5 lbf6 23
'ii'f5 .Uad8 and Black's extra piece is
worth more than the three pawns.
b) 1 8 .th3 lbe4 19 :xe4 dxe4 20
lbfg5 lbd5 2 1 lbxg7 .txh3 22 lbxe8
.tf5 23 lbf7 e3 24 fxe3 :xe8 25 lbe5
'ii'h6, with advantage to Black in
Suster-Prokopp, correspondence 1984.
18 . . . �xf5 1 9 gxf5 �e4 20 'ii'e3
iLxe6 21 fxe6 l:l.xe6
The airy kingside gives Black a deci­
sive advantage. The rest is very pain­
ful, but for one side only!
22 l:l.e2 l:.f8 23 �e5 'ii'd8 24 f3 �g5
25 f4 l:l.xf4! 26 l:l.c2 h6 27 l:l.ac1 'ii'f6
28 �h1 'ii'f5 29 l:.c3 l:.xe5! 30 dxe5
d4 3 1 l:.c8+ �h7 32 'ii'g3 l:tg4 0-1
Game 20
Wells-Fries Nielsen
Copenhagen Open 1995
1 e4 e6 2 d4 d5 3 �d2 �f6 4 e5
�fd7 5 c3 c5 6 il.d3 �c6 7 �e2
cxd4 8 cxd4 'ii'b6 9 �f3 f6 1 0 exf6
tll f6: Main Line with 1 1 . . . 1ic7 and 1 1 . . . 1i b 6
�xf6 1 1 0-0 il.d6 1 2 b3
Seemingly planning a fianchetto,
but in fact White has a different
scheme in mind. The immediate 12
.tf4 is also possible, when after
12 ....txf4 13 lbxf4 'ii'xb2 14 lbg5!?
White has some compensation for the
pawn.
1 2 . . . 0-0 1 3 iLf4!? iLxf4
13 ...lbxd4 is a playable alternative,
e.g. 14 lbfxd4 e5 15 i.e3 exd4 16 .txd4
'ii'd8 17 lbg3 .td7 and Black was okay
in Wahls-Zhang Gong, Beijing 1993.
14 �xf4 il.d7
14 ... @h8 and 14 ...g6!?, both renew­
ing the threat against the d4-pawn,
also merit consideration. 14 ... lbxd4?!,
however, cannot be recommended.
After 15 lbxd4 e5 16 lbxdS Black has
two captures but neither lead to equal­
ity:
a) 16 ... lbxd5 17 'ii'h 5! lbf6 18 'ii'xe5
and White was simply a clear pawn up
in Emms-Kinsman, England 4NCL
1995.
b) 16 ...'ii'xd4 17 i.c4! 'ii'xdl
(17 ...@h8 18 'ii'xd4 exd4 19 lbc7 :b8
20 .Ufdl wins a pawn; and although
17 ... .te6!? is stronger, after 18 lbe7+
Wf7 19 i.xe6+ Wxe6 20 'ii'xd4 exd4 2 1
47
The Fre n c h Ta rra s c h
:fel+ 'it>f7 2 2 lbfS! White is still bet­
ter) 1 8 :fxdl 'it>h8 19 l:le l bS?
(19 ...lbxdS 20 i.xdS is clearly better
for White) 20 lbc7 bxc4 21 lbxa8 i.d7
22 lbc7 :cs 23 lba6 i.bs 24 lbb4 as 2S
l:txeS 1-0 Emms-M.Anderton, Maid­
stone 199S.
1 5 l:.c1 !
An important move. The bishop on
d3 will have to move so that the queen
can support the d4-pawn, and since bl
is the natural retreat square, it is wise
to activate the queen's rook first.
1 5 . . . 96
Two other moves have also been
witnessed from this position:
a) 15 ...lbe4 16 g3 lUS!? 17 lbh4 :£7
18 lbf3 nfs 19 lbh4 :fl 20 lbf3 and
the players agreed a draw in Wahls­
Kindermann , Dudweiler 1996. 17
l:te l!?, hassling the knight on e4, looks
48
more critical.
b) 1S ...:ae8 16 i.bl lbxd4 17 'ii'xd4
'ii'xd4 18 lbxd4 es 19 lbfe2 exd4 20
lbxd4 with a small edge for White in
Rausis-Farago, Germany 1996.
1 6 ilb1 tll e 4 1 7 g3 l:.aca 1a .Ue 1
tlld6
This retreat represents a concession,
but 18 ...gS 19 i.xe4 gxf4 20 i.b l fxg3
2 1 hxg3 looks better for White, as the
black king remains unsafe.
1 9 tlle5 tllx e5 20 l:.xe5 l:If6 2 1 'ii'd 2
l:tcta 22 h4!
White has established an ideal posi­
tion.
22 . . .tllf7 23 l:.e2 ilb5 24 l:l.e3 'ii'd6
25 l:l.ce 1 tllda 26 l:l.e5 ilea 27 h5
'ii'd7 2a hxg6 hxg6 29 l:.g5 l:l.xf4 30
gxf4 .Uf6 31 'ii'e3 ilf7 32 �g2 tll c6
33 l:l.h 1 ilea 34 l:.h3 'ii' g7 35 f5 ilf7
36 fxg6 1 -0
 3 . . . tll f6: Main Line with 1 1 . . . 1i c7 a n d 1 1 . . . 1i b 6

Summary
The plan of i.g5-h4-g3, in conjunction with leaving the knight on e2, continues
to be the most dangerous plan against 1 1 ...'ii'c 7. Black has to play extremely care­
fully to avoid drifting into a lifeless position in this line. On the other hand it is
still unclear how White should proceed against 1 1 ...'ii'b 6. Many different ways
have been tried but Black seems to have enough resources in each variation.

1 e4 e6 2 d4 d5 3 ll'id2 ll'if6 4 e5 ll'ifd7 5 il.d3 cs 6 c3 ll'ic6 7 ll'ie2 cxd4 8 cxd4 f6

9 exf6 tl'ixf6 1 0 ll'if3 il.d6 1 1 0-0

1 1 . .'ii'c7
.

1 1...'ii'b6
12 ltJc3 0-0 (D)
13 .tgS
13 ....td7 Game16 -

13 ... @hs Game17 -

13 i.e3 Game18
-

12 ltJf4 Game19-

12 b3 Game20
-

12 il.g5 (D)
12 ltJc3 Game14
-

12 g3 Game15
-

1 2 . . . 0-0 13 iLh4 (D)

13 l:tc l Game13
-

1 3 ... es
13 ... ltJhS Game12
-

14 dxe5 - Game11

12... 0-0 1 2 ilg5 13 ilh4

49
 CHAPTER THREE
3 '2Jf6: Early Deviations
. . .
from the Main Line
1 e 4 e6 2 d 4 d5 3 tlld 2 tllf6
Thus far we have studied the most
popular and important variations of
the 3 ...ll'if6 Tarrasch; the main lines
that arise after 4 eS ll'ifd7 5 .td3 cS 6
c3 ll'ic6 7 ll'ie2 cxd4 8 cxd4 f6 9 exf6
ll'ixf6 10 ll'if3 .td6 1 1 0-0. In this chap­
ter we will be looking at a number of
attractive options for both White and
Black, before the main position is
reached. In Games 2 1 and 22 we con­
sider 9 ...'ii'xf6, which only came to
prominence on the early 1990s. Ad­
herents such as Knaak and Gleizerov
have injected some important new
ideas into Black's play, which has ele­
vated the line from the assessment 'a
bit dodgy' to 'quite playable'. Con­
versely 9 ll'if4 (Game 23) was an ex­
tremely important line during the
mid-eighties but it is hardly ever seen
now, since White has not found a way
to obtain any advantage. In Games 24-
26 we consider some other deviations
by Black from the main line, namely
8 ... tt:'ib6, 7...'ilfb6 (which usually trans­
poses to Games 16-20 from Chapter 2
50
I
after 8 ll'if3 cxd4 9 cxd4 f6 10 exf6
ll'ixf6 1 1 0-0 .td6) and 6 ... b6. Games
27 and 28 deal with 7 tt:'igf3, which can
lead to a pawn sacrifice, although
Black also has other respectable ways
to play the position. 5 tt:'igf3 cS 6 c4?!
cannot be recommended, as we see in
Game 29. Finally, 4 ...ll'ie4 doesn't
have a good reputation, but are things
really so clear? We take a close look at
this underrated move in Games 30 and
31.
Game 21
Korneev-Knaak
Bad Worishofen Open 1992
1 e4 e6 2 d4 d5 3 tlld 2 tllf6 4 e5
tllfd7 5 il.d3 c5 6 c3 tllc 6 7 tlle 2
cxd4 8 cxd4 f6 9 exf6 'ii'xf6 10 tllf3
see following diagram
1 0 . . .il.d6
10 ... .tb4+, offering the exchange of
bishops, is the old continuation. After
1 1 .td2 .txd2+ 12 'ii'xd2 0-0 13 0-0 eS
14 dxeS tt:'idxeS 15 ll'ixeS 'ii'xeS we
 3 . . .
reach a position similar to the one dis­
cussed in Chapter 2. Theoretically
White maintains an infinitesimal edge
and Black's position is solid, but
slightly depressing. Note that if Black
wishes to play with ... .td6, then
10 ... h6 is the most accurate move or­
der, preventing White's option in the
next note.
1 1 0-0
Here 1 1 .tgs 'ii'f7 12 0-0 0-0 13
.th4, threatening lbgS and also intend­
ing the positionally desirable .tg3,
gives White a traditional plus. Hence
the growing popularity of the move
10... h6.
1 1 . . . h6 1 2 iLc2
The alternative 12 .th 1 is the sub­
ject of Game 22.
1 2 ... 0-0 1 3 'ii'd3
White's play looks very natural, but
in fact it founders on an original de­
fensive manoeuvre.
1 3 . . . l:.dS !
Black's fortifying strategy includes
...lbf8, covering the weak light squares
around the king. This will be followed
by ... i.d7-e8-g6! (or hS), after which
Black's pieces will co-ordinate per­
fectly. Simple stuff, but in this game
tll f6 : E arly De viati o n s fro m t h e Main L ine
this strategy proves to be mightily
effective.
1 4 a3
14 'ii'h7+ Wf7 is just a waste of time.
The queen would soon have to beat a
hasty retreat after ... lbf8.
1 4 . . .tllfB 1 5 b4 il.d7 1 6 iLb2 b5 1 7
iLc3 ilea 1 8 'ii'e3 il.96
The plan has been completed and
Black stands well. Now White chooses
not to exchange, but this merely leaves
Black with two good bishops against
two bad ones.
1 9 il.b3 tlld7 20 l:tae1 .Ue8 2 1 tllg3
iLf4 22 'ii'e2 a6 23 tlle5 tlldxe5 24
dxe5 'ii'g5 25 iLb2 %lad8
An uncomplicated plan. The d­
pawn will roll up the board. White's
next move is a sure sign that things are
slipping.
26 tllh 1 iLh5 27 f3 d4 28 tllt2 'ii'h4
29 tllh3
29 g3 .txg3! 30 hxg3 'ii'xg3+ 3 1 Whl
.txf3+ is the end.
29 . . .iLe3+ 30 �h 1 tllxe5 3 1 'ii'c 2
iLxf3 32 gxf3 'ii'xh3 33 'ii'e4 tll f7 34
f4 tlld6 35 'ii'e 5 tll c4 36 iLxc4 bxc4
37 l:.g 1 il.xg 1 38 %lxg 1 'ii'f3+ 39 l:.g2
'ii'd 1 + 40 l:.g 1 'ii'f3+ 41 l:.g2 l:.e7 42
'ii'f6
51
The Fre n c h Tarra s c h
4 2 . . .'ii'x g2+ 0-1
A powerful game by Knaak.
Game 22
Todorovic-Dimitrov
Belgrade 1995
1 e4 e6 2 d4 d5 3 lLld2 lLlf6 4 e5
lLlfd7 5 c3 c5 6 il.d3 lLlc6 7 lLle2
cxd4 8 cxd4 f6 9 exf6 'ii'xf6 1 0 0-0
An interesting move order, which
eliminates ... .tb4+ options, though
after Black's response we arrive at the
same position as the previous game.
Given that Black's king is a little ex­
posed and he still needs time to de­
velop, grabbing the pawn on d4 is too
risky. After 10 ... lbxd4 1 1 lbxd4 'ii'xd4
12 lbf3 'ii'f6 13 .tgs 'ii'f7 14 'ii'c 2!
White has a very strong initiative. For
example, 14 ... g6 15 :ac t 'ii'g7 16 :fe l
'it>f7 17 'ii'e2 lbcS 18 b4 1-0 was the
brutal finish to Cvachoucek-Iveges,
correspondence 1977.
1 0 . . . il.d6 1 1 lLlf3 h6 1 2 iLb1 ! ?
In Emms-De Francesco, Bad
Worishofen 1996, I tried 12 .tc2 0-0 13
.te3 and after 13 ... lbe7!? (13 ...:d8 is
more consistent, when I was planning
14 'ii'd2, in order to exchange bishops
52
with .tf4) 14 'ii'd3 eS 15 lbxeS (15
dxeS lbxeS 16 lbxeS .tfS!) 15 ... lbxeS
16 dxeS i.xeS 17 lbg3! I liked my posi­
tion. My opponent now slid rapidly
downhill and after 15 ... lbfS? 18
'ii'xdS+ i.e6 19 'ii'e4 .txb2? (19 ...:ac8
20 lbxfs .txfs 21 'ii'xfs 'ii'xfs 22 .txfs
l:txfS 23 :act :xct 24 l::lxc t is clearly
better for White) 20 :abl .tes 21
lbhS he was forced to resign.
1 2 . . . 0-0 1 3 'ii'd 3
A similar idea to the previous game,
except with the bishop on b 1. Because
of this White need not worry so much
about the move ... lbb4.
1 3 . . .lidS 1 4 g3 lLlf8?!
This actually proves to be too slow
here. The disadvantage of 12 .th 1 is
that it disconnects the white rooks.
Black needs to exploit this by violent
means, so I suggest that 14 ... eS may be
the answer to Black's problems. In­
deed, in Baron Rodriguez-Shaked,
Cala Galdana 1996, Black equalised
after 15 dxeS lbdxeS 16 lbxeS i.xeS 17
lbf4 lbd4 18 'ii'h7+ 'it>f7 19 'ii'g6+ ( 19
i.g6+ 'it>e7 20 .td3 .txf4 21 .txf4
lbe2+! 22 i.xe2 .tfS! traps the queen)
19 ...'ii'xg6 20 .txg6+ 'it>f6 21 .td3 .tfs.
1 5 iLf4
 3 . . .
15 . . . e5 16 dxe5 iLxe5 17 tllxe5
tllxe5 1 8 iLxe5 'ii' xe5 1 9 iLc2 iLe6
20 tll f4 iL f7 2 1 l:l.fe 1 'ii'x b2?
This runs into an irresistible attack.
2 1...'ii' d6 would have restricted
White's plus.
22 l:.ab1 'ifxa2 23 l:.e 7! l:l.ab8?!
23 ....Ue8 also loses after 24 l:txf7!
'it>xf7 2S .tb3 'ii'aS 26 .txdS+. The last
chance to continue the game was
23 ... l:.ac8 24 :bxb7 'ii'xc2 2S 'ii'xc2
:xc2 26 �xf7, which still looks very
good for White.
24 'ii'f5 d4
25 l:.xf7! 1 -0
Both 2S ...'ii'xf7 26 i.b3 and 2S ... d3
26 l:.xg7+! 'it>xg7 27 lbhs+ 'it>g8 28
'ii'g4+ 'it>f7 29 .tb3+ are winning for
White.
Game 23
Pirrot-Hertneck
Ge rman Bundesliga 1990
1 e4 e6 2 d4 d5 3 tlld2 tllf6 4 e5
tllfd7 5 il.d3 c5 6 c3 tllc6 7 tll e 2
cxd4 8 cxd4 f6 9 tll f4 !?
see following diagram
9 . . . tll xd4 1 0 'ii'h 5+ �e7 1 1 exf6+
ltJ f6 : Ea rly De viations fro m t h e Main L ine
This is more accurate than 1 1 lbg6+
hxg6 12 exf6+ 'it>xf6! (12 ...lbxf6 trans­
poses to the main game) 13 'ii'xh8 'it>f7
14 0-0 lbcS! and Black is better, e.g. lS
'ii'h3 es 16 'ii'e3 lbxd3 17 lbf3 lbxf3+
18 'ii'xf3+ .tfs 19 g4 e4 20 'ii'g2 .te6 21
.te3 lbeS 2 2 f4 lbf3+ 23 :Xf3 exf3 24
'ii'xf3 'it>g8 2S .td4 .Uc8 0-1 Comben­
Short, Brighton Zonal 1984.
1 1 . . .tllxf6 1 2 tllg6+ hxg6 13 'ii'xh8
�7 1 4 0-0
14 f4, attempting to halt Black's
surge in the centre, has been tried in a
bid to inject some new life into this
variation. Nevertheless, with direct
play Black can still drum up an initia­
tive. The game Radlovacki-Rasidovic,
Yugoslavian Championship 199 1, con­
tinued 14 ...lbc6 lS 'ii'h4 es 16 0-0 e4 17
.te2 lbd4 18 .tdl lbfs 19 'ii'e l .tcs+
20 'it>hl 'ii'h 8 21 lbb3 .tb6 22 g4 lbh6
23 gS lbhg4 24 'ii'g3
see following diagram
'ii'h4 27 :g3 (or 27 gxf 6 :h8 with the
24 ...lbxh2!! 2S 'ii'xh2 .th3 26 :gl
decisive threat of 28 ... i.g2+ 29 .Uxg2
'ii'e l+) 27 ....Uh8 28 'ii'xh3 'ii'xg3! 0-1.
A well-worn alternative line is 1 4
'ii'h4, where no real improvements
53
Th e Fre n c h Ta rra s c h

have been found for White on the

stem game Van der Wiel-Timman,
Brussels 1986, which continued 14 ... eS
15 lbf3 lbxf3+ 16 gxf3 .tfs 17 .txfS
gxfS 1 8 .tgS 'ii'aS+ 19 'it>fl g6! (with
the idea of ... .tg7 and ... :h8) 20 .txf6
'ii'a6+ 2 1 'it>g2 'ii'xf6 22 'ii'xf6+ 'it>xf6 23
:ac t .td6. Here Black is doing well
and Timman went on to win the
game.

21 . . . il.g4!! 0-1
After 22 l:txg4 Black mates with
22 ...'ii'f l+ 23 :gl lbg3+! 24 hxg3 'ii'h3 .

Game 24
Lputian-Agzamov
USSR Championship 1985

1 d4 e6 2 e4 d5 3 tlld2 tllf6 4 e5
tllfd7 5 il.d3 c5 6 c3 tllc6 7 tlle2
1 4 . . . e5 1 5 tllf3 tllxf3+ 1 6 gxf3 tllh5 cxd4 8 cxd4 tll b6
1 7 il.xg6+
Drastic measures are required, oth­
erwise Black will consolidate with
...'ii'f6.
1 7 . . . �xg6 1 8 �h1 'ii'h 4!
Black does best to give back the
piece. 1 8 ... lbf6 allows 19 :gl+ 'it>f7 20
:Xg7+ 'it>e6 21 .tgS .te7 22 l:.xe7+!
'ii'xe7 23 l:.cl!
1 9 'ii'xf8 �h7 ?!
It is better to play 19 ...'ii'h 3 20 :gl+
'it>h7 21 'ii'a3 .tfs 22 :g2 d4, when
Black had good compensation in 9 0-0
Doric-D'Amore, Formia 1995. After Here White can also take immediate
the text move, 19 ... 'iti>h7, White should steps on the queenside with 9 a3 aS 10
play 20 'ii'f7! b3. At first sight these moves seem a
20 l:l.g 1 ? ! 'ii' xf2 2 1 'ii'f 7?? little strange, but White is taking pro­
One move too late. This allows a phylactic measures. Black's... a5-a4 will
very pleasing finish. now be answered by b3-b4, establish-

54
 3 . . . liJ f6 : Ea rly De viations fro m t h e Main Line

ing a fair share of space. I've had nu­ is mate.

merous experiences with this line: 26 fxg6 'ii'xd4 27 'ii'e 2 il.d7
a) 10 ....tdl 1 1 0-0 .tel 12 f4 f5 13
exf6 .txf6 14 lbf3 a4 15 lbe5 (the logi­
cal 15 b4 is also good for White)
15 ... axb3 16 'ii'xb3 0-0 ll i.e3 with a
clear plus for White in Emms­
H venkilde, Copenhagen 1992.
b) 10 ... a4 1 1 b4 .tdl 12 0-0 .tel 13
f4 g6! (this is stronger than 13 .. .f5) 14
.tb2 (14 g4!? is critical, although after
14 ...h5 15 f5 exf5 16 gxf5 .txf5 ll
i.xf5 gxf5 1 8 l:.xf5 'iWdl, White's ex­
posed king offers Black definite
counter-chances) 14 ...h5 15 @h 1 'it>f8 Now 28 'i'xh5 'i'h4 allows Black to
16 .Uc l 'it>gl 1 l .Uf3 'ii'e8 and Black had hang on. Instead Lputian produced an
equalised in Emms-Knott, British inspired queen sacrifice which com­
Championship 1996. pletely ties up his opponent.
9 . . . il.d7 1 o f4 h5 1 1 tll f 3 :ca 1 2 a3 28 iLf5! il.b5 29 iLxc8 ! ! iLxe2 30
a5 1 3 tll c3 tll a 7 14 h3 tllc4 l:.xe2 iLe5 31 il.d7 tll c6 32 iLxc6
Once more 14 ...g6 looks like the bxc6 33 l:Z.fe 1 'ii'd 3 34 l:Z.xe5
most solid defence. Now White exer­
cises a thematic break.
1 5 f5! 'ii'b 6 1 6 �h 1 iLe7
Grabbing the b-pawn with
16 ...lbxb2 allows ll .txb2 'ii'xb2 18
lba4 .txa4 19 'ii'xa4+ lbc6 2 0 fxe6 fxe6
21 i.g6+ 'it>dl 22 .Ufb l, when White's
initiative is becoming menacing.
1 7 'ii'e 1 ! ?
Naturally White can defend the b­
pawn, but Lputian has something
much more enticing in mind.
1 7 . . .tll x b2 1 8 iLxb2 'ii'xb2 1 9 tllxd5! 34 . . .'ii'xg6 35 l:.e8+ �xf7 36 l:. 1 e7+
exd5 20 e6 iLc6 �6 37 l:l.e6+ �5 38 l:Z.xg6 �xg6 39
Or 20 ...fxe6 21 fxe6 .tc6 22 .tg6+ l:Z.e6+ 1 -0
when 22 ...'it>f8 23 lbh4+ 'it>g8 24 i.f7+
'it>hl 25 'ii'e 5! and 22 ...'it>d8 23 'ii'xaS+ Game 25
are winning for White. Menke-Wilde
21 exf7+ �8 22 tlle5 l:.h6 23 'ii'e3 Correspondence 1957
'ii'b 6 24 l:.ae 1 iLf6 25 tll g6+ l:.xg6
Naturally not 25 . 'itafl, as 26 'ii'e 6
. . 1 e4 e6 2 d4 d5 3 tll d 2 tllf6 4 e5

55
Th e Fre n c h Ta rra s c h
tlltd7 5 il.d3 c5 6 c 3 tllc6 7 tll e2
'ii'b 6 8 tllt3 cxd4 9 cxd4 f6
Usual and best here is 10 exf6 ll'ixf6
1 1 0-0 .td6, transposing to the
1 1...'ii' b6 lines from the previous chap­
ter.
1 0 tllf 4?!
After 10 0-0 Black should just go
ahead and grab the pawn. Pirrot­
Kindermann, Bad Worishof en 1992,
went 10 ... fxe5 1 1 dxe5 tt:'idxe5 12 tt:'ixe5
ll'ixe5 13 ll'if4 tt:'ixd3 14 tt:'ixd3 .td6 15
'ii'h 5+ g6 16 'ii'h 6 'ii'd4 17 .tf4 .tfs 1s
'ii'h3 e5! (18 ... .tg7 19 :ac l 0-0 20 :c7
gives White a dangerous attack) 19
i.xe5 'ii'xe5 20 'ii'xc8+ :xc8 2 1 ll'ixe5
.tg7 22 ll'id3 'it>d7 23 :fcl �c4 24 b3
:c6 and Black's bishop and better
king ensured an edge in the endgame.
1 o . fxe5 1 1 dxe5
. .
Or 1 1 tt:'ixe6 e4 12 .tf4 tt:'ide5! 13
dxe5 .txe6 14 tt:'ig5 .tfs 15 .te2 .tb4+
with a clear advantage to Black.
1 1 . . . iLb4+ 1 2 'itf1 tlldxe5 1 3 tllx e5
tll xe5 1 4 'ii'h 5+ tllf7 1 5 iLxh 7 'ii'd4
1 6 iLe3 'ii'xb2 1 7 l:fo1 'Wes 1 8 'ii'd 1
l:l.xh7 1 9 'ii'a4+ il.d7 20 'ii'xb4 l:.c8
21 l:l.xc8+ iLxc8 22 tll g6 'ii'f6 23
'ii'b 1 ?
After this move Black rapidly takes
56
over. The only chance is 23 'ii'f8+!
'it>d7 24 'ii'xf7+ 'ii'xf7 25 tt:'ie5+ 'it>e8 26
ll'ixf7 'it>xf7 27 .txa7 l:th4, when Black
is still better but White has good
drawing chances.
23 . . .tlld6 24 tllf4 tlle4 25 g3 g5
26 tll xd5
Neither 26 ll'ig2 tt:'ixg3+ nor 26 ll'ie2
'ii'f3 27 :g1 :xh2 hit the mark.
26 . . . exd5 27 'ii'b5+ il.d7 28 'ii'xd5
'ii'a 1 + 29 �e2 il.g4+ 30 �d3 'ii'c3+
31 �xe4 l:l.e7+ 32 'ii'e 5 l:.xe5 0-1
Game 26
Emms-Strater
Hastings 1996
1 e4 e6 2 d4 d5 3 tlld 2 tllf6 4 il.d3
c5 5 e5 tll td7 6 c3 b6
This move is an old favourite of Ni­
gel Short's. Before completing his de­
velopment Black seeks to exchange the
light-squared bishops.
7 tlle2
White can also play to retain bish­
ops with 7 f4 i.a6 8 .tb l, e.g. 8 ...tt:'ic6
9 tt:'idf3 cxd4 10 cxd4 :cs 1 1 ll'ie2 and
now:
a) 1 1...i.b4+ should be answered
with 12 .td2, as 12 tt:'ic3? runs into
 3 . . .
12 ...tt:'icxe5! 13 fxe5 .Uxc3.
b) 1 1...f5 12 0-0 i.. e 7 13 .Uf2 0-0 14
h3 with the usual space plus.
7 . . . ila6 8 iLxa6 tll xa6 9 0-0 iLe7
9 ... tt:'ic7, adding further protection
to the e6 and d5 squares, is a serious
alternative. Emms-Conquest, Norton
(rapidplay) 1997, continued 10 ll'if4
'ii'c8 1 1 c4 'ii'b7 12 cxd5 ll'ixdS 13 ll'if3
h6 1 4 'ii'e2 i.. e7 15 tt:'ixds 'ii'xd5 16
dxc5 bxc5 17 .Udl 'i'b7 18 i..e 3, when
I felt that I was a little better, although
the game was eventually drawn.
1 0 tllf4 b5 1 1 'ii'g4 g6
1 2 h4!
This was introduced by the Greek
grandmaster Kotronias against Short.
White wants to probe with h4-h5.
1 2 . . . cxd4
12 ...i.. xh4? allows 13 ll'ixe6! fxe6 14
'ii'xe6+, when the knight on a6 hangs.
1 3 cxd4 h5 1 4 'ii'e 2
Perhaps 14 'ii'h3!? is stronger, pre­
venting 14 ... i..xh4 on account of 15
tt:'ixe6. White c an then think about
ideas involving ll'if3, 'ii' g 3, tt:'ih3 and
i.. g 5, infiltrating the weak dark
squares on the kingside.
1 4 . . . iLxh4 1 5 a4 iLe7
Varying from Kotronias-Short,
tll f6 : Early D e viations fro m t h e M a in L ine
Novi Sad Olympiad 1990, which went
15 ... 'i'b6 16 axb5 tt:'ic7 17 tt:'if3 i.. d 8?!
( 17 ... i.. e 7 is stronger) 18 ll'ih3 ll'ixb5 19
:d1 tt:'ib8 20 i.. g5 ll'ic6 21 'ii'd2 i..xg5
22 'ii'xg5 and White was in the driving
seat.
1 6 axb5 tll c7 1 7 tllb3 'ii'b8 1 8 tlla 5
'if xb5 1 9 'if c2
Naturally White keeps the queens
on. During the game I felt that I
should have a clear advantage, but it is
now apparent that Black has many
defensive resources.
1 9 . . . 'ii'b6 20 tll c6 iLc5!
21 tll x g6! l:lg8!
After 2 1 ...fxg6 22 'ii'xg6+ 'it>f8 23
i..h6+ :xh6 24 'ii'xh6+ 'it>g8 White has
the amazing 25 tt:'id8!! (needless to say,
found by Fritz) as 25 .. .lild8 26 'ii'g5+
picks up the black rook. After the
game move Black can regain his pawn
advantage.
22 tllf4 'ii'xc6 23 b4 tllb5 24 iLe3
l:lg4! 25 bxc5 tllxd4 26 iLxd4 l:lxf4
27 l:lfd 1 l:lf5 28 l:la5!? tllxe5 29 l:le 1
'ii'c 7 30 l:la6 �f8?
Finally going astray in the compli­
cations. After 30....Ug5 White should
bail out with 3 1 'ii'a4+ 'it>f8 32 :xe5
:xe5 33 i..xe5 'ii'xe5 34 l:txa7 lila7 35
57
The Fre n c h Tarra s c h
'ii'xa7. Black is still a pawn up, but the
passed pawn on cS should ensure a
draw.
3 1 'ii'x f5 ! ! exf5 32 iLxe5 'ii'd7 33
l:th6 f6 34 iLxf6 'ifa4 35 .Ue 7 1 -0
Black can only prevent mate with
the ludicrous 35 ...'ii'a l+ 36 i.xal
'it>xe7, after which White is too much
material up.
Game 27
Kolev-Herraiz
Linares Open 1996
1 e4 e6 2 d4 d5 3 tlld 2 tllf6 4 e5
tlltd7 5 il.d3 c5 6 c3 tll c 6 7 tllgf3
Introduced by Vik.tor Korchnoi in
the 1960s, and later advocated by John
Nunn, this leads to a speculative offer
of a pawn.
7 . . . 'ii'b6 8 0-0 cxd4
8 ... g6 is discussed in Game 28.
9 cxd4 tll xd4 1 0 tll xd4 'ii'xd4 1 1 tllf3
'ii'b6 1 2 'ifa4 'ii'b4
White's idea was to swing the queen
to the kingside with, for example,
12 ... i.e7 13 'ii' g4. 12 ...'ii'b 4 prevents
this.
1 3 'ii'c 2
58
1 3 . . . 'ii'c5
The most popular move, although
White has tended to score well against
this. Major alternatives include (in de­
creasing order of importance):
a) 13 . ..ltJcS!? 14 i.d2 'ii'a4 15 b3
'ii'd7 16 i.e2 i.e7 17 i.e3 b6 18 b4
'ii'a4 19 'ii'h2 ltJe4 20 a3 i.a6 21 bs
i.b7 22 .Ufc l l:r.c8 23 ltJd4 0-0 24 f3
.Uxcl + 25 .Uxcl i.xa3 26 'ii'a2 'ii'b4
with great complications in Hellers­
Ludvigsen, Nordic Team Champion­
ship 1985.
b) 13 ...i.e7 14 i.d2 'ii'g4 15 l:tac l 0-0
16 i.xh7+ 'it>h8 17 i.d3 ltJcS 18 h3
'ii'hS 19 i.gS with a distinct advantage
for White in Emms-Tibensky, Douai
Open 1990.
c) 13 ... h6?! 14 i.d2 'ii'b6 15 :ac t
i.e7 16 'ii'a4! is given as clearly better
for White by ECO, as 16 ... 0-0 allows
17 :Xc8 .Uaxc8 18 'ii'xd7.
1 4 'ife2 iLe7 1 5 iLe3 'ifa5 1 6 .Uac 1
0-0 1 7 l:Z.c3 l:.e8
Th alternative 17 .. .fS 1 8 exf6 ltJxf6
19 i.d4 offers White chances of a di­
rect attack. Emms-Giddins, Gausdal
1993, concluded 19 ... i.d7 20 ltJgS h6
2 1 ltJxe6 l:tfe8 22 i.f5 'ii'a6 (or
22 ... i.bS 23 'ii'd2 i.xfl 24 ltJxg7! 'it>xg7
 3 . . .
2 5 .Ug3+) 2 3 'ii'xa6 bxa6 2 4 lbxg7 i.xf5
25 lbxf5 lbe4 26 l:tc6 1-0. After
17 ... .Ue8 Black doesn't weaken the
kingside, but does allow an effective
version of the Greek gift.
1 8 iLxh7+! �xh7 1 9 tll g5+ il.xg5
19 ... 'it>g8 20 'ii' h 5 i.xg5 21 i.xg5
leads to the game, while 19 ...'it>g6
comes to a sticky end after 20 'ii'c 2+ f5
21 exf6+ 'it>xf6 22 i.d4+ 'it>xg5 23 f4+
'it>h4 24 'ii' h7+ 'it>g4 25 l:.g3 mate.
20 'ii'h 5+ �98
20 ... i.h6 21 i.xh6 g6 22 'ii' h 4 .Uh8
White can play 23 i.d2+ 'it>g8 24
'ii'xh8+ 'it>xh8 25 :h3+, but even
stronger is 23 .U.fc l!, which leaves
Black in helpless state, e.g. 23 ...'ii'b6 24
i.g5+ 'it>g8 25 :xc8+ lbf8 26 :Xa8!
lhh4 27 .Uxf8+ 'it>xf8 28 .Uc8+ 'it>g7 29
i.f6+ 'it>h6 30 :hs mate.
2 1 iLxg5
see following diagram
2 1 . . .'ii'x c3
This seems rather extreme, but after
21.. ..Uf8 22 i.f6!! Black is forced to
sacrifice anyway, when 22 ...'ii'xc3
(22...lbxf6 23 exf6 doesn't improve
matters) 23 bxc3 gxf6 24 exf6 lbxf6 25
'ii' g5+ 'it>h7 26 'ii'xf6 is winning for
li:J f6 : E a rly De viations from t h e M a in L ine
White. Note that 21...lbxe5 allows
mate after 22 :h3 lbg6 23 'ii'h7+ 'it>f8
24 'ii'h8+ lbxh8 25 l:.xh8.
22 bxc3 tll xe5 23 f4 1 -0
Black's situation is hopeless, e.g.
23 ... lbc4 24 l:lf3 lbd6 25 l:th3 f6 26
i.xf6. A powerfully played game by
White.
Game 28
Maksimovic-Farago
Copenhagen Open 1989
1 d4 e6 2 e4 d5 3 tlld2 tllf6 4 e5
tllfd7 5 c3 c5 6 il.d3 tllc6 7 tllgf3
'ii'b6
In Emms-S.Brown, Surrey Open
1995, my opponent tried an immedi­
ate 7 ... g6, but here I made use of the
fact that I hadn't castled. The game
continued 8 h4!? i.g7 9 h5 0-0 10 hxg6
fxg6 (10...hxg6!?) 1 1 'ii'e2! and the
semi-open h-file gave White a small
edge.
8 0-0 g6!?
This has become a popular alterna­
tive to grabbing the pawn.
9 dxc5 'ii'c7 1 0 tllb3 tlldxe5
10 ...i.g7 1 1 lbfd4! 'ii'xe5 12 f4 is a
crucial continuation:
59
The F re n c h Ta rra s c h
a) 12 ...'Wbs 13 fs es 14 .tgS! exd4 15
cxd4 lbxd4 16 f6 .txf6 17 i.xf6 lbxf6
1 8 l:txf6 and White was better in
Korneev-Gleizerov, Katowice 1992.
b) 12 ...'ii'f6 13 fS!? gxfS 14 .txfS
lbxd4 15 lbxd4 exfS 16 lbxfS lbxcS 17
'ii'xdS .txfS 18 'ii'xcS 'iVeS 19 'ii'xeS+
i.xeS 20 :xf5 and White was a clear
pawn up in Sutovsky-Yudasin, Israeli
Championship 1994, although he ac­
tually lost the game in the end. It is
difficult to suggest improvements for
Black in these lines.
1 1 iLf4 il.g7 1 2 tllx e5 tllx e5
Natural enough, although I don't
see much wrong with 12 ... .txeS ei­
ther.
1 3 il.b5+?!
After this move Black's central
pawn mass becomes the most impor­
tant factor. 13 :e l is more testing,
when after 13 ...0-0 14 c4! f6 15 l:k l
'ii'f7 16 cxdS exdS 17 lbd4 White had a
small plus in Tkachiev-Komarov,
France 1996.
1 3 . . . il.d7 14 iLxd7+ "ii' xd7 1 5 l:.e 1
tllc4 1 6 l:.b 1 0-0 1 7 tlld2 "ii'b5 1 8
tllxc4 "ii'xc4
The opening battle has been well
and truly won by Black. The rest of
60
the game is a model demonstration of
how to exploit a 2-0 central pawn ma­
jority.
1 9 il.d6 l:Ifc8 20 "ii'b3 "ii'a6 2 1 a4 b6
22 cxb6 axb6 23 iLe5 iLxe5 24
l:.xe5 "ii'xa4 25 "ii'xa4 .Uxa4 26 'itf 1
b5 27 l:Ie3 b4 28 b3 l:.a7 29 cxb4
l:.b8 30 b5 l:l.xb5 31 b4 .Ua4 32 l:.eb3
W.8 33 �e2 �e7 34 �d2 e5 35 �c3
�e6 36 l:t 1 b2 l:Ia7 37 �d3 l:.c7 38
l:.c3 e4+ 39 �d4 l:.xc3 40 �xc3 �e5
41 l:.b1 f5 42 l:.b2 f4 43 l:.b1 d4+ 44
�c4 .Ub8 45 b5 l:l.c8+ 46 �b3 d3 47
b6 �d4 48 �b2 e3 49 b7 l:.b8 50
fxe3+ fxe3 5 1 'iti>c 1 �c3 52 l:.b6
52 . . .d 2+ 0-1
53 @dl e2+ 54 Wxe2 :es+ and the
pawn promotes.
Game 29
Emms-S.Clarke
British Championship 1991
1 e4 e6 2 d4 d5 3 tlld 2 tll f6 4 e5
tll fd7 5 tllgf3 c5 6 c4?!
An interesting possibility, but not
one to cause Black any great deal of
concern. I tried this line out a few
times, but this game was my last expe­
rience with it.
 3 . . . liJ f6 : Ea rly D e vi·ations from the Main Line
:..:....:.
___

: 2 lt:Jc5 31 t5
9
3 1 ...gxf5 32
� :xf5 is strong.
gxf d

33 ...tt:Je4 34 ��4 ltJgS would have

32 e6 txe6 33 6 lt:Jd3? !

been very nasty.

34 :b 1 e5? 35 'Wg 5 :ta 36 °Wh5 :t3

� � ��: �
. r•. ·· i
-

%-, 1 � � �.,,
� J�.
-,
6 , , .lt:Jc6 7 cxd5 exd 5 8 i.d3 cxd4 9
•. .
t.•
�*if'�� �, -
·� � ··1
!ff{���-' "�,.ft
0-0 lt:Jc5 1 0 lilb3 � 4 1 1 lt:Jxc5
9
.txc5 1 2 a 3 i.b6 1£ 3 :e 1 0 -0 1 4

��� �� !i. •
B
JU:,� � ;
i.c2 i.h5!
Planning . .. 1'.g6 . Apart from the
. - :

;�
. doub led' White has no
. is
fact that lt .
real compensat n for the pawn. 37 i.h4!

��
1 5 i.t4 :ea "illd 3 .'11.g6 1 7 "ill� 2 Turning the tabl White threatens
:c8 18 i.xg6 hxg6 1 9 :ad 1 'Wd7 O 38 'ir'h7+ @f8 39 'iV + 'ir'g8 40 1'.e7+.
h3 a5 2 1 g4 a4 22 'Wd3 nae 23 :e2 37 . . .:ta 38 'Wh7+ 'it>t8 39 i.xt6 1 -0
:as 24 :de 1
Game JO
Seu I-Zach
Biel 1997

1 e4 e 6 2 d4 d5 3 '"UC
'""' 3 lt:Jt6 4 e5
tbe4 5 lt:Jxe4 dxe4

24 , , .lt:Jd8'
:
Returning the extra pawn in ex-
change for a · �ood knight versus bad
.
bishop' scenario. White also has to
worry about Black's other d-pawn.
25 lt:Jxd4 i.x 4 26 "illxd4 lile6 27
�
°Wd2 :b 5 28 i.g 3 d4 29 t4 :b3 30 6 .>11.e3
�
�
-_
______

66 11
_
-------
The F re n c h Tarra s c h
Note that although White played 3
tt:Jc3 in this game, T arrasch players
would reach the same position with 3
tt:Jd2. The other main alternative here
is 6 i.c4, intending to ignore rather
than surround the e4-pawn. However,
the pawn does have some nuisance
value. After 6 ...a6 7 a4 b6 8 tt:Jh3 i.b7
9 tt:Jf4 tt:Jc6 10 i.e3 tt:Je7 1 1 0-0 g6 1 2
'ir'e2 tt:Jf5 13 IHdl i.h6! Black retained
counter-chances in Aseev-Lputian,
Russia 1984. Indeed Black was able to
seize the advantage after 14 a5 0-0 15
axb6 cxb6 16 i.xa6?! i.xf4! 17 i.xb7
tt:Jxd4 1 8 'ir'c4 l:hal 19 l:hal i.xe5 20
i.xe4 'ir'h4. After 6 i.e3 Black has to
play very actively, else the e-pawn will
simply disappear for nothing.
6 . . . c5 7 dxc5 lt:Jd7 a 'Wg4 lt:Jxc5 9
.tb5+ lt:Jd7 1 0 lt:Je2 'Wa5+ 1 1 lt:Jc3 a6
1 2 .txd7+ .txd7
1 3 .td4!
13 'ir'xe4 is recommended in ECO,
assigning White a clear plus. However,
13 ... i.c6 1 4 'ir'g4 i.a3! is very embar­
rassing for the first player.
1 3 . . . .tc6 1 4 0-0-0 0-0-0 1 5 °Wf4 ! ?
1 5 @bl threatens tt:Jxe4, but Black
can play an unclear exchange sacrifice:
15 .. Jhd4! 16 .�hd4 'ir'xe5, when the
62
two bishops and kingside pawn pha­
lanx offer reasonable compensation.
1 5 . . . .ta3?
Black just loses valuable time with
this move. Perhaps he should try
15 ...Itd7 16 @b l i.b4 17 lDxe4 :hd8
18 c3 i.d5 19 a3 i.e7, with some pos­
sible play on the light squares around
the white king. As the game goes,
Black gets no compensation for the
pawn.
1 6 'We3 .te7 1 7 'it>b 1 l::.d7 1 a lt:Jxe4
l::.hda 1 9 l::.d3 'it>ba 20 l::.hd 1
White has consolidated successfully
and is now simply a pawn up.
20 . . .'it>aa 21 b3 l::.ca 22 .tc3 l::.xd3
23 l::.xd3 'Wc7 24 lt:Jd6 l::.da 25 .td4
:ta 26 lt:Jc4 .txg2 27 .tb6 'Wc6 2a
f3 .tf1 29 lt:Ja5 'Wb5 30 a4 'Wb4 3 1
l::.d 7 :ca 32 l::.x b7 .tc5 3 3 .txc5
'Wxc5 34 'We4 .td3 35 l::.c7+ 1 -0
Game 31
Balashov-Borngasser
Wisla 1992
1 e4 e6 2 d4 d5 3 tt:Jd2 lt:Jf6 4 e5
lt:Je4 5 .td3 lt:Jxd2 6 .txd2 c5 7 dxc5
lt:Jd7!
Preparing to capture on c5 with the
 3 . . .
knight, when the knight will hassle
the bishop on d3. After 7 ...CfJc6 8 CfJf3
i.xc5 9 0-0 i.d7 1 0 c3 h6 1 1 b4 i.b6
12 a4 a6 13 b5 axb5 14 axb5 .l::!.xal 15
'ir'xal CfJa5 16 'ir'a4! i.c5 17 Ital b6 18
'ir'g4 White had a strong initiative in
Akopian-Naumkin, Vilnius 1988.
8 lt:Jf3 lt:Jxc5 9 0-0 iL.e7
On this or the next move Black
should probably capture on d3. One
recent example of such play was
G.Mohr-Sulava, Croatian Cup 1997,
which continued 9 ...CfJxd3 10 cxd3
i.d7 1 1 i.g5 'iia5 1 2 a3 Itc8 13 Ite l
.l::!.xc l 1 4 'iix cl h6 1 5 b4 'ir'a6 16 i.e3
ii.el 17 'ir'c7 i.d8 18 'iic5 b6 19 'iid6
i.e7 20 'iib 8+ i.d8 2 1 'ir'd6 i.e7 and
the players agreed a draw.
ll:i f6: Ea rly De via tions fro m the Main Line
1 0 l:t.e 1 0-0
1 1 iL.f1 !
No more chances! Now White's
pawn wedge on e5 guarantees a nice
space advantage.
1 1 . . .'irb6 1 2 iL.c3 iL.d7 1 3 iL.d4 �c 7
1 4 l:t.e3 a5 1 5 :tc1 b6 1 6 c3 a4 1 7
g3 iL.c6 1 8 h4 :aea 1 9 iL.h3 ll:le4 20
ll:ld2 f5 21 exf6 iL.xf6 22 iL.xf6 l:t.xf6
23 lt:Jxe4 dxe4 24 �e 1
Black has eliminated the e5-pawn,
but is left saddled with a chronically
weak pawn. As soon as this is rounded
up, Black sees no reason to continue
the game.
24 . . . b5 25 :.d 1 iL.d5 26 a3 l:t.ef8 27
l:t.d2 l:t.f3 28 iL.g2 l:t.xe3 29 �xe3 �e5
30 l:t.d4 1 -0
63
The F re n c h Ta rra s c h

Summary
The 3 ...lbf6 Tarrasch offers plenty of offbeat variations for both sides that are
worth an outing. Of course, due preparation is advisable before venturing out
with a new line.

1 e4 e6 2 d4 d5 3 lt:Jd2 lt:Jf6 4 e5

4...lt:Jfd7
4...lbe4
5 lbxe4 - Game30; 5 .id3 - Game31
5 .td3
5 lbgf3 cS
6 c3 lbc6 7 i.d3 - see Games 27 and 28 (by transposition)
6 c4 - Game29
5 ... c5 6 c3 ttJc6
6 ...b6 - Game26
7 tt:Je2
7 lbgf3 'iib 6 8 0-0 (D)
8 ... cxd4 Game27
-

8 ... g6 - Game28
7 . . . cxd4
7 ...
- 'iib6 8 lbf3 cxd4 9 cxd4 f6 (D)
10 exf6 lbxf6 1 1 0-0 .id6 - Games 16-20
10 lbf4 - Game25
8 cxd4 f6 (DJ
8 ...lbb6 Game24 -

9 exf6
9 lbf4 - Game23
9 ...'Wxf6 1 0 lt:Jf3 h6 1 1 0-0 .td6 1 2 .tc2
12 .ib l Game22
-

1 2 . . . 0-0 Game21
-

8 0-0 9 . f6
. . 8 . .f6
.

64
 CHAPTER FOUR I
3 tiJf6: The Big
. . .

Pawn Front with 5 f4

1 e4 e6 2 d4 d5 3 lt:Jd2 lt:Jf6 4 e5 In this line Black does better to play

lLifd7 5 f4 more quietly and try his luck on the
In many ways 5 f4 introduces queenside. Finally, Game 36 discusses
White's most ambitious plan against the rare 7 . .'ifaS, planning immediate
.

3 ... lbf6. He bolsters the eS-pawn and queenside play with ... b7-b5-b4.
plans to develop his pieces behind an
impressive centre, before slowly Game 32
squashing Black on the kingside. Emms-Kosten
That's the theory, and it happens quite British Championship 1985
a lot in practice too, but of course
Black does have some moves too. The 1 e4 e6 2 d4 d5 3 lt:Jd2 lt:Jf6 4 e5
downside to 5 f4 (compared to 5 .id3) lt:Jfd7 5 f4 c5 6 c3 lt:Jc6 7 lt:Jdf3 cxd4
is that it doesn't contribute towards 8 cxd4 ._b6 9 g3 .tb4+ 10 �2 g5!?
White's development. Consequently,
Black can obtain quicker and more
dangerous counterplay against the d+
pawn. Another facet of this line is that
White is often forced to go on a little
walk with his king, which is not to
everyone's taste. Normally it would
be completely safe behind that huge
phalanx of pawns, but Games 32-34
show that Black does have some in­
genious ways of causing a few prob­
lems. Game 35 sees a relatively new
concept for White, the lavish 8 h4, This move turned a fe w heads when
which puts paid to any ... g7-g5 breaks. it was first unleashed on the touma-

65
The Fre n c h Tarra s c h
ment circuit. White can capture in
two ways, but neither move wins a
pawn.
The more conventional 9 .. .f6 is dis­
cussed in Game 34.
11 �e3
Dealing with the incidental threat
of ... g5-g4. The main alternative 1 1
fxg5 is the subject o f Game 33, while
1 1 h3 should be met by 1 1 ...gxf4. Va­
siukov-M.Gurevich, Moscow 1987,
continued 12 gxf4 f 6 13 i.e3 i.e7!,
when the combined ideas of ...tt:Jf8-g6,
...1'.d7 and . . 0-0-0 ensure that Black
.
has enough counterpla y against the
white centre.
1 1 . . .f6
1 1 ...g4 is possible anyway, and may
be even more reliable than 1 1 ...f6.
Certainly Black seems to be doing
okay after 12 tt:Jd2 f6 or 12 tt:Jh4 i.e7.
1 2 �h3 0-0
This move is often preceded by the
capture on e5. After 12 ... fxe5 13 fxe5
0-0 we have:
a) 14 1'.g4, which transposes to the
mam game.
b) 14 1'.xe6+ 'it>h8 15 @g2 g4! and
bl) 16 .1xg4 tt:Jdxe5 17 dxe5 (17
now:
.1xc8 tt:Jc4!) 17 ... 'ifxe3 18 i.xc8 llaxc8,
when Black is extremely active.
b2) 16 tt:Jg5 tt:Jdxe5 17 'i!fc2 tt:Jg6 18
h 3 .1e7! 19 tt:Jf7+ 'it>g8! 2 0 tt:Jh6+ 'it>g7
with a very unclear position in Eisen­
Chapman, correspondence 1989.
c) 14 'it>e2!? (another king move, but
one that does have its points) 14 ... 1'.e7
15 b3 tt:Jdb8 16 1'.g4 .td7 1 7 tt:Jh3 h6
18 'i'bl, when White has avoided any
catastrophes and therefore stands
slightly better, as in Read-Wikman,
66
correspondence 1993
d) 14 Itel!? @h8 15 i.g4 i.e7 16
'ir'd2 and here Black should probably
play 16 ...i.b4.
1 3 �g4
13 i.xe6+ @h8 14 i.xd5 would be a
brave (some would say foolish) at­
tempt at refuting Black's opening
play. However, it is clear that after
14 ... fxe5 15 dxe5 i.c5 16 'ir'b3 gxf4 17
gxf4 tt:Jb4 Black develops some menac­
ing play for the pawns.
1 3 . . . fxe5 1 4 fxe5
1 4 . . .�c5 ! !
This incredible idea was a sugges­
tion of George Botterill, and was
judged as one of the top theoretical
novelties of the year by the prestigious
Chess Informant. Since Botterill's sug­
gestion had already been published in
the British Chess Magazine, I was quite
aware of it and had even dismissed it
as good for White. Obviously I hadn't
seen far enough.
1 5 �xe6+ �h8 1 6 dxc5 'Wxb2+ 1 7
..td2
see following diagram
1 7 . . . g4!
This was the real shock. I had been

expecting 17 ...tt:JdxeS 18 i.xc8 Itaxc8
19 @g2, when I don't think that Black
has enough for the piece. Instead I was
rudely awoken, and didn't react well
to the new circumstances.
1 8 ..txd5?
There are two other ways for White
to continue the fight:
a) 18 i.xg4 't\Vd4+ 19 @g2 't\Vxg4 20
1'.c3 lDxcS 2 1 e6+ @g8 22 'ifxdS lDxe6
is given by Kosten as slightly better
for Black.
b) 1 8 llb l! looks best. I can't find
anything special for Black after
18 ...'ifd4+ 19 i.e3!, so perhaps
18 ... 'iVxeS 19 i.xd7 't\Vd4+ is stronger.
Then 20 @g2 gxf3+ 2 1 tt:Jxf3 :Xf3 ! 22
@xf3 i.xd7 gives Black a strong at­
tack. A safer path is 20 i.e3 't\Vxdl 2 1
:xdl 1'.xd7 2 2 ItxdS i.e6, with a
roughly equal endgame.
1 8 . . . lt:Jdxe5 1 9 ..txc6
Giving up all the light squares, but
19 @g2 gxf3+ 20 tt:Jxf3 tt:Jxf3 2 1 1'.xf3
Itd8 is just winning for Black. The rest
of the game was not my greatest chess
expenence.
1 9 . . . bxc6 20 .l:1b 1 'Wd4+ 21
tbd3+ 22 'it>e2 gxf3+ 23 tLixf3 'We4
24 'Wxd3
3 . . . ll:i f6 : The B ig Pawn Fro n t with 5 f4
24 . . . ..ta6! 25 'Wxa6 'Wxf3+ 26 'it>d2
l:t.fd8+ 27 �c2 'We4+ 28 'it>b2 l:t.ab8+
29 'ita3 'i'xe3+ 30 l:t.b3 'Wxc5+ 3 1
'itb2 l:t.d2+ 3 2 'itb1 l:t.xb3+ 0-1
After going so far, perhaps I should
have allowed mate after 33 axb3 'iYc2+
34 @al 'ifb2.
Some openings just do not suit cer­
tain players. After this game I decided
that I don't like wandering around
with my king after just ten moves!
Game 33
Slobodjan-De la Villa Garcia
Pamplona 1996
1 e4 e6 2 d4 d5 3 lt:Jd2 ttJf6 4 e5
lt:Jfd7 5 f4 c5 6 c3 lt:Jc6 7 lt:Jdf3 'Wb6
8 g3 cxd4 9 cxd4 ..tb4+ 1 0 'itf2 g5
1 1 fxg5
The safest and hence the most popu­
lar way of dealing with 10 ... gS.
1 1 . . .lt:Jdxe5 1 2 lt:Jxe5 lt:Jxe5
see fol/Ming diagram
1 3 ..te3
Black can also answer 13 @g2 with
..te3 13 ... tt:Jc4, e.g. 14 tt:Jf3 1'.d7 15 b3 tt:Jd6
16 i.f4 tt:Je4 17 Itc1 1'.d6 with equality
in Yakovich-Gleizerov, Russia 1987.
67
The Fre n c h Ta rra s c h

able. For example, Black often goes

7...'iYb6 8 g3 f6 9 1'.h3 cxd4 10 cxd4,
reaching the same position.

1 3 . . . lt:Jc4!
Naturally 13 ... lbc6 is also possible,
but this seems to be the most reliable
way of equalising. After the exchange 1 0 .th3 fxe5 1 1 fxe5 .tb4+ 1 2 'itf1
on c4 Black can lay claim to many 0-0 1 3 'itg2 lt:Jdxe5 ! ?
important light squares. Black strikes before his opponent
1 4 .txc4 dxc4 1 5 lt:Jf3 i.d6 1 6 'Wc2 can consolidate his position.
'Wc6 1 7 :he 1 0-0 1 4 dxe5 lt:Jxe5
There also seems nothing wrong
with the immediate 17 ... bS, planning
1 8 ... 1'.b7.
1 8 .tf4 b5 1 9 .txd6 .tb7! 20 'We4
'Wxe4 2 1 :xe4 .txe4 � - �
After 2 1.. .1'.xe4 22 1'.xf8 @xf8 it is
about equal, although there is still
plenty of fight left in the position.

Game 34
Mohrlok-Schrancz
Correspondence 1989
1 e4 e6 2 d4 d5 3 lt:Jd2 lt:Jf6 4 e5
lt:Jfd7 5 c3 c5 6 f4 lt:Jc6 7 lt:Jdf3 cxd4
8 cxd4 'Wb6 9 g3 f6
Before the discovery of 10 ... gS, this
was Black's most risky and challeng­
ing try against 5 f4. Black is prepared
to sacrifice a piece in order to destroy
White's central pawn cover. Note that
the move order is quite interchange-
1 5 'We2!
This has been shown to be the only
real major test of Black's sacrifice.
Against other lines Black seems to be
holding his own at least.
a) 15 't\Vb3 1'.cS 16 't\Vxb6 1'.xb6 17
1'.f4 lbd3 18 1'.d6 19 IUl eS! 20 1'.xeS
(20 lbxeS?? lbe l+! 21 Itxe l Itf2 is
mate) 20 ... 1'.xh3+ 21 lbxh3 Ite8! 22
Ite 1 Itfe7! and Black regained the piece

68
 3 . . .
in Smagin-Rosjak, Skopje 1987.
b) 15 1'.f4 lbg6 16 Itel? (16 lbe2
looks stronger) 16 ... Itxf4! 17 gxf4
lbxf4+ 1 8 @fl 1'.d7 19 a4 .ms and
White didn't last very long in Yasseen­
Moran, Dubai Olympiad 1986.
1 5 . . .ll:lxf3
15 ...lbc4 16 b3 i.c3 17 Itb l seems to
favour White, e.g. 17 ...1'.f6 18 'ir'c2
lbd6 (or 1 8 ... lbe3+ 19 1'.xe3 'ir'xe3 20
Ite l with a clear edge) 19 1'.a3 Itd8 20
1'.xd6 Itxd6 21 Ite l 1'.d7 22 lbe5!
1'.xe5 23 Itxe5 Itc8 24 'ir'd2 .l::!.dc6 25
Ite2 and White had consolidated his
extra piece in Eisen-Onoda, corre­
spondence 1990.
1 6 ll:lxf3 e5 1 7 .txc8 I:f.axc8 18
ll:lxe5 'We6 1 9 .tf4 lU5 20 :ac 1
20 . . . :ctS
Following a suggestion of Knaak's.
20 ... Ite8 21 'ir'b5! was also good for
White in Ermenkov-Knaak, Bulgaria­
East Germany 1983. Instead, damage
limitation can be achieved with
20 ... Itxcl 2 1 �hcl g5 22 Itc7! (22 'ir'g4
.l::!.xe5 23 Itc8+ 1'.f8! 24 'ir'xe6+ Itxe6 25
i.xg5 @f7 is equal) 22 ... 1'.d6 23 I:.xb7
i.xe5 24 1'.d2!, which Mohrlok as­
sesses as only slightly better for White.
2 1 'We3!
tli f6 : Th e B ig P a wn Fro n t with 5 f4
This is stronger than Knaak's 21 h4,
and puts a big question mark over the
merit of Black's previous move.
2 1 . . . g5 22 ll:lg4 'Wg6 23 .te5 h5 24
.l::t c 7! hxg4 25 :g7+ 'Wxg7 26 .txg7
:t2+
26 ...@xg7 is stronger, but the simple
pawn snatch 27 'ir'xa7 still gives White
a clear plus. The rest is easy.
27 'Wxf2 :xt2+ 28 'itxf2 .tc5+ 29
'ite2 'itxg7 30 :d 1 d4 3 1 .l::t c 1 b6 32
a3 a5 33 'itd3 1 -0
Game 35
Beliavsky-Kindermann
Munich 1991
1 e4 e6 2 d4 d5 3 ll:ld2 ll:lf6 4 e5
tt:Jtd7 5 f4 c5 6 c3 tt:Jc6 7 ll:ldf3 'Wb6
8 h4! ? cxd4 9 cxd4 .tb4+ 1 0 �2 f6
1 1 .te31 .te7 1 2 'Wd21
This is more accurate than 12 Itbl,
e.g. 12 ...aS 1 3 lbe2 0-0 14 lbg3 fxe5 15
fxe5 and now Black can try the typical
sacrifice 15 ...lbdxe5! 16 dxe5 d4 17
1'.d2 lbxe5 .
1 2 . . . 0-0 1 3 :d 1 a5!
Brutal attempts do not work here.
After 13 ... fxe5 14 fxe5 lbdxe5 15 dxe5
d4 White holds on to the e5-pawn
69
The Fre n c h Ta rra s c h

with 16 i.f4! i.b4 17 'iVc l . 13 ... aS! is

an excellent long-term move. Black
sees the queenside as the area of coun­
terplay, and so he starts to claim some
space on that side. It just remains to
block the kingside at the most desir­
able moment.
1 4 'itg3 a4 1 5 .td3 'Wd8 1 6 lt:Je2 f5

45 'Wd7 'Wxd7 46 exd7 l::.b7 47 h6

l::.xd7 48 hxg7 l::. xg7+ 49 l::.g2 l::. a2
50 l::.d2 l::. xd2 51 .txd2 �f7 52
l::. x g7+ .txg7 53 'itg2 �e6 � - �

Game 36
Adorjan-J.Watson
New York 1981
We now enter a heavy manoeuvring
stage of the game, where Black's 1 e4 e6 2 d4 d5 3 ll:ld 2 lt:Jf6 4 e5
chances are not worse. ll:lfd7 5 c3 c5 6 f4 lt:Jc6 7 lt:Jdf3 'Was
1 7 h5 ll:lb6 1 8 'iitf 2 ll:lb4 1 9 .tb 1
ll:lc4 20 'Wc3 .td7 2 1 a3 ll:lc6 22
l::.d g1 b5 23 g4 b4 24 axb4 a3! 25
gxf5 exf5 26 .ta2 .te6 27 ll:lg5 'Wd7
28 lt:Jxe6 'Wxe6 29 .txc4 .txb4 30
'Wc2 dxc4 3 1 bxa3 .txa3 32 .td2
l::.fd8 33 .tc3 .te7 34 °Wd2 'Wd5 35
l::. h 3 l::. a3 36 'Wc2 'We6 37 :1h2 l::.da8
38 I:f. b 1 l::.b3 39 l::.d 1 'Wd5 40 'i'xf5
.l::t a 2 41 'Wh3 ll:lb4 42 'itg 1 lt:Jd3 43
'Wea+ .tta 44 e6 :as?
44 ..Jhe2! would have capped off an
excellent positional performance from Adding pressure to the centre by
the German grandmaster. After 45 pinning the c-pawn. This move was
I:.xe2 lbxf4 46 e7 lbxe2+ 47 @f2 'iVf7+ popular in the seventies, but is hardly
Black is winning, e.g. 48 @xe2 't\Vxe7 + ever seen nowadays. Another quieter
49 @d2 't\Vg5+ 50 @c2 'iYg2+ 5 1 Itd2 move is the blocking 7 .. .f5. Ye Jiang­
't\Ve4+ 52 @dl Itbl mate. 44 ... Ita8 al­ chuan-Short, Lucerne 1989, continued
lows White to scrape half a point. 8 i.d3 cxd4 9 cxd4 Ji.el 10 lt:Je2 lbb6

70
 3 . . .
1 1 h3 0-0 12 g4 (Ye recommends 12 a3 !
aS 13 b3, planning to meet 13 ... a4 with
14 b4) 12 ... aS 13 a4 ltJb4 14 1'.b1 1'.d7,
and Black's queenside counterplay was
well under way.
8 .te3 b5 9 dxc5 b4
1 0 lt:Jd4
This move, a big novelty at the
time, has put the whole line to rest.
Previously 10 a3 had been the main
move, but Black gets sufficient play
with the pawn sacrifice 10 ...b3! 1 1
'ir'xb3 1'.xcS.
1 0 . . .'Wxc5
Or 10 ... 1'.b7 1 1 a3! bxc3 12 b4 'ir'd8
13 ltJgf3, when the c3-pawn will soon
die and Black will be swamped by the
mass of queenside pawns.
1 1 .,a4 bxc3
lU f6 : The Big Pawn Fro n t wit h 5 f4
1 1...1'.b7 is powerfully met by 12
1'.bS, so Black is forced into a ropey
piece sacrifice.
1 2 'Wxc6 cxb2 1 3 l:tb1 'ifa3 1 4 �2
:tb8 1 5 f5 .tb7 1 6 'ifc7 .ta6 1 7
'ifxa 7 'if xa2 1 8 ll:lgf3 .tc4
18 ...'i'xbl loses to 19 i.bS!
1 9 'ifxa2 .txa2 20 .tb5 .ta3 21 fxe6
fxe6 22 lt:Jxe6 .txb 1 23 :txb 1 rJ;;e 7
24 lt:Jed4
Black has managed to regain some
of his lost material, but now the four
White minor pieces run rampant, cre­
ating deadly threats around the black
king.
24 . . .:thcS 25 ll:lc6+ l:txc6 26 .txc6
:ca 27 .tg5+ �e6 28 lt:Jd4+ �xe5
29 .txd7 :tc 1 30 ll:lf3+ '1i;d6 3 1 .tf5
.tc5+ 1 -0
71
The Fre n c h Tarra s c h

Summary
It seems that Black has several ways of creating enough confusion against
White's impressive looking centre. As of yet, no real answer has been found to
10 ... gS, and this is probably the single biggest reason for the decline in popularity
of 5 f4.

1 e4 e6 2 d4 d5 3 lt:Jd2 l2if6 4 e5 lt:Jfd7 5 f4 c5 6 c3 l2ic6 7 lt:Jdf3

7 1i'b6
... (D)
7 ...'ir'aS - Game 36
8 g3
8 h4 - Game 35
8 . . . cxd4 9 cxd4 .tb4+ (D)
9 .. .f6Game 34
-

10 �2 g5 11 .te3 (D)
1 1 fxgS Game 33 -

1 1 . . . f6 Game 32
-

7 "flb6
...
9 ..tb4+
. . . 1 1 ..te3

72
 CHAPTER FIVE I
3 cs 4 exd5 'i'xd5 :
M a i n Line with 10 ttJxd4
. . .
1 e4 e6 2 d4 d5 3 ll:ld2 c5 4 exd5
'Wxd5 5 ll:lgf3 cxd4 6 .tc4 'Wd6 7 0-0
ll:lf6 8 ll:lb3 ll:lc6 9 ll:lbxd4 ll:lxd4 1 0
ll:lxd4
This is one of the most important
positions in the whole of the French
Tarrasch. White has a clear develop­
ment advantage and he will attempt to
utilise this in order to build up a large
initiative, while Black's main trump
card is his pawn structure, including
the 4-3 majority on the kingside. Of­
ten White's attempts at a direct attack
crash disastrously against this solid
defensive wall of pawns. Black also
tries to gain time by developing with
tempo (e.g. ...'if cl, attacking c4, and
... 1'.d6, hitting the h2-pawn) . From
humble beginnings this variation is
now firmly established as a favourite
defence of some of the world's leading
players.
From the starting position of the
main line Black has two main options,
10 . a6 and 1 0 .. 1'.d7. Both moves pre-
. . . 1 e4 e6 2 d4 d5 3 lbd2 c5 4 ll:lgf3
pare a later ...'i'c7 and . . . 1'.d6, but
10 ... a6 (Games 37-42) has generally
been the choice of most of the top
players, as it eliminates any lbbS ideas
for White. Play can then become ex­
tremely sharp, as illustrated by the
number of crazy games in this section.
The theory in certain positions has
even spiralled out of all proportions.
For example, in Games 37 and 38
powerful novelties late on have dra­
matically altered the assessments of
whole variations.
10 .. 1'.d7 (Games 43-45) usually
.
leads to quieter positions where Black
has less chance of being rolled over,
but White often secures the long-term
advantage of the bishop pair. That
said, there is also plenty of scope here
for both sides to liven up the proceed­
mgs.
Game 37
Bowden-Levitt
England 1997
cxd4 5 exd5 .-xd5 6 .tc4 'Wd6 7 0-0
ll:lt6 8 ll:lb3 ll:lc6 9 ll:lbxd4 ll:lxd4 1 0
73
The F re n c h Tarra s c h
l2Jxd4 a6 1 1 :e 1
White's main alternatives here, 1 1
1'.b3 and 1 1 b3, are considered m
Games 4 1 and 42 respectively.
1 1 . . .'Wc7
The slower 1 1...1'.d7 is the subject
of Game 40.
1 2 .tb3 .td6 1 3 l2Jf5 ! ?
This concept is well known here.
White readily sacrifices material for an
attack against the black king. Despite
the fact that it has been tested in nu­
merous grandmaster games, no firm
conclusions have yet been reached.
The quiet 13 h3 causes Black fewer
problems, as we shall see in Game 39.
1 3 . . ..txh2+ 14 'ith1 0-0 1 5 l2Jxg 7!
:dB!
Apparently this defensive resource
was Kasparov's discovery. Earlier this
whole line had been deemed dubious
for Black on account of 15 ...@xg7 16
't\Vd4 (planning 'ifh4 and i.h6+) 16 ...eS
17 'iVh4 ltJg4 18 f3 'ifd8 19 i.gS f6 20
i..c l ! and Black is in trouble. Emms­
W.Lowe, Halifax Open 199 1, contin­
ued 20 ...'ifd4 2 1 fxg4 1'.f4 22 c3 'iVcS
23 1'.xf4 exf4 24 Itadl and White had a
winning advantage.
1 6 °Wf3 'itxg7 1 7 .th6+ 'itg6
74
Not 17... Wxh6 18 'ifxf6+ @hs 19
.l::!.e 3, when Black can safely resign.
1 8 c3 lbh5
Clearing the way for the f-pawn.
We shall look at the other major de­
fence (18 . .. ltJdS) in the next game.
1 9 .tc1
Following Michael Adams's recipe
(see below) . This move, threatening
't\Vg4+, has the advantage of not block­
ing the e-file, which can be vital in
many variations. However, Jonathan
Levitt had prepared an extremely sig­
nificant novelty. The other critical
line is 19 1'.e3, as played by Sergei Ru­
blevsky (Rublevsky has worked with
Kasparov, so on that evidence alone
we must take this move seriously!):
a) Levitt's 19 ...1'.f4 is still possible,
but not quite so effective, as after 20 g4
ltJg3+ 21 fxg3 1'.xe3 22 'ifxe3 White's
pieces are better co-ordinated than in
the main game. In particular, the
white queen patrols the dark squares
around the black king. Nevertheless,
Black still seems to be just about okay.
For example, 22 ...h6 23 1'.c2+ @g7 24
gS hs 25 .l::!.adl b6 26 'iVf4 i.b7+ 27
@h2 'iVxf4 28 gxf4 .tf3 29 Itd3 1'.g4 30
@g3 f 6 31 Itee3 1h-1h was the conclu-
 3 . . .
sion of Forster-Levitt, Hastings 1997.
b) 19 .. .fS 20 g4 lbf6 21 gxfS+ exfS 22
'iYg2+ lbg4 23 f3 i.d7 (23 ... b6!? should
be considered, as developing the
bishop on b7 prevents lie6xc6 ideas)
24 i.d4 't\Vg3 2S fxg4 i.c6 26 lie6+
@gs 27 1Ixc6 bxc6 28 't\Vxh2 't\Vxh2+ 29
@xh2 fxg4 30 lifl .l::!. f8 31 i.f7! (setting
up a mating net) 3 1 ..Jhb8 32 @g3
llxf7 33 1Ixf7 and White won easily in
Rublevsky-S.Ivanov, Elista 1997. No
doubt the last word hasn't been spo­
ken here yet.
1 9 . . ...tf4!
In Adams-McDonald, England
4NCL 1997, Black played 19 .. .fS 20 g4
bS?!, but after 2 1 gxhS+ @g7 22 'iYg2+
@fl 23 't\Vh3 Adams won comfortably.
Of course 20 ... lbf6 is more challeng­
ing, but White can win brilliantly
with 2 1 i.xe6 i.xe6 22 1Ixe6 @f7 23
1'.gS!! @xe6 24 'iYxfS+ @d6 2S 'iYxf6+
@cs 26 b4+ @bs 27 a4+ @c4 28 't\Ve6+
@d3 (or 28 ...lidS 29 't\Ve4+) 29 't\Ve3+
@c4 30 'iYe2+ 1Id3 (30 ... @b3 31 'iYa2+
@xc3 32 %:!.cl+ and 30 ...@dS 3 1 lidl+
@c6 32 't!i'c4+ @b6 33 i.e3+ also wins
for White) 3 1 lla3! and Black must
give up the rook to prevent 'iYa2 mate.
20 g4
c S 4 e x d S 'ik xd S : Main Line with 1 0 Ei:J x d 4
20 . . .ll:lg3+!
This inventive move changes the
entire complexion of the position.
Black gives back the extra piece, but in
return he receives relative king safety
and time to complete development.
Added to this, White's own king also
suffers from too much air. If anything,
I prefer Black.
21 fxg3 ..txc1 22 :axc1 b6 23 ..tc2+
'itg7 24 ..te4 :ba 2s :n 'ike7 26 gs
..tb7 27 :ce1 :d2 2 8 'ikf6+ 'ikxf6 29
gxf6+ 'ith6
The weakness on f6, coupled with
Black's active rook on d2, add up to a
decisive advantage for Black.
30 :t4 ..txe4+ 3 1 :exe4 :bd8 32
:b4 bs 33 'itg1 'itg6 34 :t1 :ads
3S :t2 :xt2 36 'itxf2 aS 37 :g4+
'itxf6 38 'ite3 hS 39 :t4+ :ts 40
:h4 :gs 41 'itf3 :ts+ 42 'ite3 'itgS
43 b4 :es+ 44 'itd3 :ds+ 4S 'ite3
axb4 46 :xb4 fS 47 a3 es 48 c4
bxc4 49 :xc4 e4 SO :ca :d3+ S 1
'itf2 :xa3 S2 :gs+ 'itf6 S 3 :ha
:a2+ S4 'it>g 1 'iteS SS :xhS e3 S6
:ha 'ite4 S7 l:UB :a 1 + SB 'itg2 e2
S9 :ea+ 'itd3 O- 1
A well played game and an impor­
tant theoretical battle.
75
T h e Frenc h Ta rra s c h

22 :xd5
Game 38
Adams-Dreev
Wijk aan Zee 1996

1 e4 e6 2 d4 d 5 3 lt:Jd2 c5 4 exd5
'Wxd5 5 lt:Jgf3 cxd4 6 .tc4 'Wd6 7 0-0
ll:lf6 8 ll:lb3 ll:lc6 9 lt:Jbxd4 lt:Jxd4 1 0
lt:Jxd4 a 6 1 1 :e 1 'Wc7 1 2 .tb3 .td6
1 3 ll:lf5!? .txh2+ 1 4 'ith 1 0-0 1 5
ll:lxg7! :dB! 1 6 °Wf3 'itxg7 1 7 .th6+
'itg6 1 8 c3 ll:ld5 1 9 :ad1 !

White is already winning. Basically

Black cannot cope with the direct
threats of l:tdl and 'iih 3.
22 ... .td7 23 'ifh3 .tf8
No defence works, e.g. 23 .. .f4 24
lle6+ i.xe6 25 't\Vxe6+ @g7 26 l:tg5+
mates, while 23 ... h5 24 't\Vh4 and
23 ...1'.f4 24 l:te7 h5 25 .l::!.exd7 are also
terminal.
24 :e3 'itg7 25 :g3+ 'ith8 26 'Wh4!

A fantastic suggestion from Jon

Speelman which completely changed
the theoretical opinion of 18 ...lbd5.
The point is that 19 ... @xh6 is im­
possible due to 20 lbdS! and now:
a) 20 ... exd5 2 1 'iVf6+ @h5 22 i.dl+
i.g4 23 i.xg4+ @xg4 24 f3+ @g3 25
't\Vg5+ @f2 26 'iVe3+ @g3 27 f4+ @g4
28 't\Vh3+ @xf4 29 'iYxh2+ and 30
't\Vxc7.
b) 20 ... l:txd5 2 1 'iVf6+ @h5 22 l:te3
and Black can resign. 26 . . . .te6 27 .tf4 .te7 28 .txc7 1 -0
Also losing is 19 ... i.e5 20 .tel
(another point of 19 I:.adl - the bishop Game 39
can return to cl without disconnect­ Jansa-Brunner
ing the rooks) 20 .. .f5 2 1 c4 lbb4 22 Bad Worishofen 1989
:.Xd8 'iixd8 23 .l::!.xe5 lbd3 24 't\Vg3+.
1 9 . . .f5 20 .tc 1 .td6 21 .txd5 exd5 1 e4 e6 2 d4 d5 3 ll:ld2 c5 4 ll:lgf3

76
 3 . . .
cxd4 5 exd5 'Wxd5 6 ..tc4 'Wd6 7 0-0
.!Lif6 8 ll:lb3 ll:lc6 9 .l::.e 1 a6 1 0 ll:lbxd4
ll:lxd4 1 1 ll:lxd4 'We 7 1 2 ..tb3 ..td6 1 3
h3 0-0 1 4 ..tg5 b5!
A popular idea. Black readily ac­
cepts a weakened pawn structure on
the kingside, knowing that the bishop
pair and the half-open g-file will pres­
ent him with more than enough com­
pensation. Naturally White has no
need to capture on f6, but then Black
would simply continue ...i.b7, reach­
ing a comfortable position.
1 5 ..txf6 gxf6 1 6 ._h5 ..tb7 1 7 c3
�h8 1 8 'Wh6 .l::.g8! 1 9 'Wxf6+ I:f.g7
Black's dangerous initiative is worth
at least a pawn. Note that White's ear­
lier h2-h3 has presented fresh prob-
c S 4 e x d S 'ik x d S : M a in Line with 1 0 t'iJ x d 4
lems, as 20 g3 can be simply answered
by 20 ...1'.xg3 with a mating attack.
20 f3 J:r.ag8 2 1 g4 'Wc5 22 'itg2 ..tc7
23 J:r.e2 'Wd6 24 .l::.h 1 a5!
Black temporarily concentrates on
the queenside, with the aim of de­
stroying the support for the knight on
d4. Now 25 lDxbS runs into 25 ... 'ir'g3+
26 @f1 1'.d8!, trapping the queen.
25 ..tc2 b4 26 ..td3 a4 27 h4
Understandably White is desperate
for counterplay, but this move seri­
ously weakens the pawn on g4.
27 . . . ..td8 28 'We5 'Wxe5 29 .l::.x e5
l:bg4+ 30 'itf1 .l::.f4
Black's strategy has been a complete
success. The rest of the game is not
significant.
31 .l::.h5 f5 32 'ite2 ..tb6 33 .l::.f 1
..txd4 34 cxd4 l:xd4 35 I:t.h6? I:t.xd3
36 'itxd3 ..ta6+ 37 'ite3 ..txf1 38
l:be6 .l::.g2 39 :ea+ 'itg7 40 :as a3
4 1 bxa3 .l::.x a2 42 :a 7 + 'itg6 43 h5+
'itxh5 0-1
Game 40
Adams-Speelman
England 1991
1 e4 e6 2 d4 d5 3 ll:ld2 c5 4 exd5
77
The Fre n c h Tarra s c h

'Wxd5 5 ll:lgf3 cxd4 6 .tc4 'Wd6 7 0-0

ll:lf6 8 ll:lb3 ll:lc6 9 ll:lbxd4 ll:lxd4 1 0
ll:lxd4 a 6 1 1 :e 1 .td7 1 2 c 3 'Wc7

when the bishop and two pawns

cancel out the rook. After 26 i.c2
'ifb6 27 i.b3 'ir'g6 the chances were
1 3 .tb3 roughly equal in Hiinerkopf-Luther,
13 'ir'e2 is less flexible, but neverthe­ Munich 1992.
less quite playable. The game Emms­ 14 'We2 .td6 1 5 h3 'itb8 ?!
Danielian, Cappelle la Grande 1994, Given that Black soon runs into
continued 13 ...i.d6 14 i.g5 (14 h3 is trouble, perhaps he should consider
obviously also possible) 14 ... 0-0-0! (this beginning kingside operations imme­
strong and natural move was new at diately with 15 ...h6 or 15 ...Ithg8.
the time. As soon as I saw it could be 1 6 a4 h6 1 7 .te3 .l::r.he8 1 8 ll:lf3!
played, I knew my opponent would
oblige. Previously 14 ... 0-0 had been
tried, but note that 14 ...i.xh2+ is too
risky: 15 @h l i.f4 16 i.xf6 gxf6 17
i.xe6! fxe6 18 lDxe6 .ixe6 19 'i!fxe6+
@f8 20 'ir'xf6+ @g8 2 1 .l::!.e7 is winning
for White) 15 tt:Jf3 h6 16 i.h4 1'.c6 17
tt:Je5 i.xe5 18 'ir'xe5 'ifxe5 19 l:1xe5 g5
20 i.g3 tt:Je4! and I was frantically try­
ing to equalise.
1 3 . . 0-0-0
.

A different option is to castle king­

side after 13 ... i.d6 14 h3 0-0. Then a
surprising sequence can arise with 15
i.g5, e.g. 15 ...i.f4 16 i.xf6 gxf6 17
'ifh5 @h8 1 8 i.c2 f5 19 tt:Jxf5! exf5 20
Itel 'if c6 2 1 Itxd7 Itg8! 22 g4! 'ifxd7 23
1'.xfS 1'.h2+! 24 @xh2 'i!fd6+ 25 @gl
.l::!.g 7
With the idea of a4-a5, when White
threatens i.b6. It is now clear that
White's attack will come first.
1 8 . . . .tc6 1 9 a5 ll:ld7 20 ll:ld4 .te4
2 1 .ta4 e5 22 ll:lc2 .tc6 23 .txc6
'i'xc6 24 ll:lb4! .txb4 25 cxb4 ll:lf6
26 :ac 1 'Wa4

78
 3 . . .
The last chance was 26 ...'ir'b5.
27 'Wc4 ll:ld5 28 .tb6 ll:lxb6 29 'Wc7+
�as 30 axb6
Black's back-rank weaknesses are
absolutely fatal.
30 . . .'ii'd 7 3 1 lbe5 nta 32 b5 1 -0
After 32 ...'ifxc7 33 bxc7 Ik8 34 b6
and White can win at leisure.
Game 41
Svidler-Glek
Haifa 1996
1 e4 e6 2 d4 d5 3 ll:ld2 c5 4 ll:lgf3
cxd4 5 exd5 'Wxd5 6 .tc4 'Wd6 7 0-0
ll:lf6 8 ll:lb3 ll:lc6 9 ll:lbxd4 ll:lxd4 1 0
ll:lxd4 a6 1 1 .tb3 .td7
Keeping the pressure on d4.
c s 4 e x d S • x d S : M a in Line with 1 0 liJ x d 4
1 1...'iVcl can transpose to the main
lines after 12 Ite 1. White also has the
independent try 12 'ir'f3, but after
12 ...1'.d6 13 h3 0-0 14 1'.g5 lbd7 15 c3
b5 16 Itadl lbc5! 17 1'.c2 1'.b7 18 'ifh5
1'.e4 Black exchanged the dangerous
bishop and had no problems in
Hellers-Hiibner, Wijk aan Zee 1986.
12 c 3 'Wc7
Also possible is 12 ... 0-0-0 13 Ite l
'ir'c7 14 'ir'e2 i.d6 15 h3, transposing
to the previous game.
1 3 .tg5 h5!?
Black has three alternatives here:
a) 13 ...lbe4 14 i.h4 'ir'f4 15 g3 'ir'h6
16 'ir'e2 lbd6 and now Glek gives 17
'ir'e5, intending 17 .. .f6? 18 lbxe6!! fxe5
19 lbc7 mate.
b) 13 ...1'.d6 14 1'.xf6 gxf6 15 'ifh5
'ir'c5 16 'ir'f3 'ii'e 5 17 g3 0-0-0 18 .l::!.fe l
'ir'g5 1 9 i.c4!, with the idea 1'.f1-g2,
when White had an edge in Kudrin­
Remlinger, Philadelphia 1992.
c) 13 . . 0-0-0 14 1'.xf6 gxf6 15 'ir'h5
.
1'.e8 with a small plus for White.
14 °Wf3 .td6 1 5 h3 .th2+ 1 6 'ith 1
.te5 1 7 .l::.fe 1 0-0-0 1 8 .l::.ad 1 .l::.dg8 ?!
This looks a little peculiar. Given
that Black never really gets going with
his attack, the rook remains largely
redundant here.
1 9 'ii'e3 !
A very powerful move. White's
queen eyes the dark squares on the
queenside and may be able to jump to
a7 at some point.
1 9 . . . .td6 20 ll:lf3 .tc5 21 °Wd 2 ll:lg4
22 .tf4!
The strongest move of the game.
White sacrifices a pawn, but calculates
that the forthcoming attack will prove
decisive.
79
The F re nc h Tarra s c h
2 2 . . .ll:lxf2+ 2 3 'it h 2 'Wd8 2 4 :es
.tb6 25 'Wd6
Threatening 'ir'b8 mate.
25 . . . ll:lxd 1
25 ...i.c7 leads to an amazing posi­
tion after 26 Ik5 1'.c6 27 l:hc6!! bxc6
28 'ifxc6 lbxd1 29 lbd4!! (Har-Zvi).
Despite his material advantage of
two rooks for a bishop, Black is com­
pletely helpless against White's variety
of threats.
26 .l::tc5+ .tc6
Or 26 ... 1'.xc5 27 'ir'b8 mate.
27 :xc6+ bxc6 28 'Wxc6+ 'Wc7
28 ...'ifc7 looks good, until one real­
ises that after 29 lbd4 we have trans­
posed to the note to 25 ...lbxdl.
29 .txc7 .txc7+ 30 'itg1 :da 3 1
80
'Wxa6+ 'it>b8 32 ll:ld4!
The game as a contest is over. The
black king is just too open.
32 . . .:d6 33 ll:lc6+ :xc6 34 'Wxc6
:da 35 .tc2 l:.d6 36 'Wea+ :da 37
'Wb5+ 'itc8 38 'Wa6+ 1 -0
An irresistible attacking display
from Svidler. It seems that Black was
lost after the casual 18 ...Itdg8.
Game 42
G. Timoshenko-Daniel ian
Cappelle la Grande 1994
1 e4 e6 2 d4 d5 3 ll:ld2 c5 4 exd5
'Wxd5 5 ll:lgf3 cxd4 6 .tc4 'Wd6 7 0-0
lt:Jf6 8 ll:lb3 ll:lc6 9 ll:lbxd4 ll:lxd4 1 0
lt:Jxd4 a6 1 1 b 3 'Wc7 1 2 .tb2 .td6 1 3
ll:lf3 b6! ?
I like this move. Previously 13 . . .b5
had been the choice, when after 14
i.d3 1'.b7 15 Itel 0-0 16 lbe5 .l:.ad8 17
'ir'e2 lbd5! 1 8 'ir'g4 f5 19 'ir'h4 lbb4 20
Ite2 lbxd3 2 1 lbxd3 1'.e4 22 Itael Itfe8
Black was equal in Tiviakov-Psakhis,
Rostov na Donu 1993. 13 ... b6 is more
enterprising, as with White's bishops
pointing menacingly towards the
kingside, it seems logical for Black to
castle on the other side of the board.
 3 . . .
13 ...b6 permits this without allowing
White to open up the position with
a2-a4.
1 4 l::te 1 il.b7 1 5 il.d3
15 i.xe6 fxe6 16 .l::!.xe6+ looks entic­
ing, but after the simple 16 ...@d7 I
can't find a decent way to continue
the attack.
1 5 . . . llld 5!
Offering a long-term pawn sacrifice,
since after 16 i.xg7 .l::!. g8 17 i.b2 lt:Jf4
18 g3 lt:Jxd3 19 1i'xd3 0-0-0 Black has
play down the half-open g-file and a
dominating bishop on the long diago­
nal. All in all, this amounts to excel­
lent compensation for the pawn.
1 6 a4 0-0-0 1 7 g3 <it>b8 1 8 il.xg7?
White's position was already noth­
ing to write home about, so I imagine
Timoshenko's philosophy was that he
may as well have a pawn for his trou­
bles. Unfortunately, this runs into
some real danger.
1 8 . . . l::th g8 1 9 iLe5 iLxe5 20 lllxe5 f6
21 lllc4 lllc 3 22 'ii'h 5 'ii'c6
The diagram says it all . White can
hardly hope to survive this position.
23 f3 l:lg5 24 'ii'h 3 l::t xd3 25 cxd3
'ii'xf3 26 llld6 l:lh5 27 'ii'f 1 'ii'h 1 + 28
<it>f2 'ii'f3+ 29 <it>g 1 'ii'h 1 + 30 <it>f2
c S 4 e x d S 'ii x d S : Main Line with 1 0 tlJ x d 4
'ii'x h2+ 31 <it>e3 l::t e5+ 32 llle4 llld5+
0- 1
33 @d4 1i'b2+ 34 @c4 .l::!.xe4+ 35
dxe4 1i'c3 is mate.
A pleasing game from the Arme­
nian; 13 ... b6 looks like a very useful
move.
Game 43
Emms-Bibby
British Championship 1990
1 e4 e6 2 d4 d5 3 llld2 c5 4 lll gf3
cxd4 5 exd5 'ii'x d5 6 iLc4 'ii'd 6 7 0-0
lllf6 8 lllb3 lll c6 9 lllbxd4 lllxd4 1 0
lllxd4 il.d7
This is slightly more solid looking
than 10 ... a6.
1 1 c3
The normal move here. 1 1 b3 is
considered in Game 45.
1 1 . . .'ii'c7 1 2 'ii'e 2 0-0-0
The quieter 12 ... i.d6 is the subject
of the next game.
1 3 a4!
The most direct and the best move.
White simply wants to play lt:Jb5,
which would encourage a favourable
exchange and open the a-file for his
queen's rook.
81
Th e Fre n c h Tarra s c h
1 3 . . .h 5
Black has two other attacking pos­
sibilities here:
a) 13 ...i.d6 14 g3! (after the ex­
change on bS, White will have ample
control of the light squares, so this
move doesn't really weaken the king­
side; and in any case Black seems to be
okay after 14 h3 i.h2+ lS @hl i.f4)
14 ... hs lS tt:Jbs .txbS 16 axbS b6 17
:a4! (the rook usefully patrols the
fourth rank, for both attack and de­
fence) 17 ... @b8? (17 ... h4 is better, al­
though White still maintains an edge
with 18 1i'f3) 18 i.gS! and White was
better in Adams-Djurhuus, Oakham
1992.
b) 13 ...tt:Jg4 14 g3 hs lS .tgs! .l::!. e 8 16
tt:Jbs 'fies 17 .tf4 .txbs 18 .txbs :ds
19 b4 1-0 was the sharp finish to Ra­
chels-Rahman, Los Angeles 199 1.
1 4 h3
The immediate 14 tt:Jbs i.xbS lS
axbS also gives White a plus, e.g.
a) lS ... .tcs 16 .l::!.a4! @b8 (16 ...tt:Jg4
17 g3! - Djurhuus) 17 g3 eS 18 i.e3 h4
19 .l::!.fal hxg3 20 hxg3 .l::!.h3 21 i.xcS
'fixes 22 b4 1i'd6 23 .l::!.xa7 e4 24 1i'e3
was winning for White in S.Pedersen­
Djurhuus, Oslo 1992.
82
b) 1S ...tt:Jg4 16 g3 .tcs 17 @g2 (17
.l::!.a4 is also good, transposing to the
last note) 17 ... @b8 18 .l::!.a4 .l::!.h e8?! (the
more direct 18 ... eS, intending .. .f7-fS
and ... hS-h4, suggests itself as an im­
provement) 19 h3 tt:Jf6 20 i.gS .l::!.d7 2 1
.l::!.fal ltJdS 2 2 i.xdS! .l::!.xdS 2 3 .tf4 eS
24 b4 1i'd7 2S .tgs .tb6 26 c4 .l::!.d3 27
cS iixbS 28 @h2! i.c7 29 .l::!.xa7 @c8 30
.l::!.a8+ .tbs 3 1 .l::!.dl and Black resigned
in Adams-Lautier, Biel 199 1 , as 31...e4
allows 32 .tf4.
1 4 . . . il.d6
Or 14 ....tcs lS b4 i.xd4 16 cxd4
@b8 17 bs .l::!. c8 18 .td3 tt:JdS 19 iif3 fs
20 a5 with a readymade attack in
Kosashvili-Djurhuus, Santiago 1990.
1 5 lll b5 il.xb5 1 6 axb5 iLc5 1 7 'ii'f3
lllg4?
This runs into trouble. A stronger
alternative is 17 . . .'fies. For example,
1 8 i.f4 1i'e4 19 1i'xe4 lDxe4 20 i.e3!?
i.xe3 (20 ... @b8!?) 21 fxe3 lDd2 22
.l::!.xf7 lDxc4 23 .l::!.xa7 tt:Jd6?? (23 ...lDb6
24 .l::!.axb7 .l::!.d7 2S .l::!.bxd7 tt:Jxd7 26
.l::!.xg7 is unclear) 24 .l::!.a8 mate, Emms­
Gunter, London (rapidplay) 1997.
1 8 iLf4 e5 1 9 il.g3 lllxf2 20 iLxf2
This is adequate for an advantage,
but 20 iifS+! would have been even
stronger, e.g. 20 ... @b8 (or 20 ....l::!.d7 2 1
i.xf2 i.xf2+ 2 2 lhf2 1i'xc4 2 3 .l::!.xa7)
2 1 i.xf2 i.xf2+ 22 �f2 1i'xc4 23
iixeS+ 1i'c7 24 1i'xc7+ @xc7 2S .l::!.xf7+
and White is a clear pawn up.
20 . . . iLxf2+ 2 1 'ii'xf2 'ii'xc4 22 'ii'x a7
l::t h6
The material balance is still level,
but Black's shaky king position gave
me confidence. His defence is ex­
tremely tricky, if possible at all. All
the same, I didn't have to work as
 3 . . .
hard as I first thought.
23 b3 'ii'e 6 24 'ii' c5+ '>tb8 25 'ii'a7+
<it>c8 26 :a4 'ii' b6+ 27 'ii'xb6 :xb6
28 l::t x f7 l::t x b5 29 l::t a8 mate
Game 44
Psakhis-Herzog
Vienna 1991
1 e4 e6 2 d4 d5 3 lDd2 c5 4 exd5
'ii'x d5 5 lDgf3 cxd4 6 iLc4 'ii'd6 7 0-0
lDf6 8 lDb3 lDc6 9 lDbxd4 lDxd4 1 0
lDxd4 il.d7 1 1 c3 'ii'c 7 1 2 'ii'e2 il.d6
This is a less ambitious line than
12 ... 0-0-0. Black accepts a slight disad­
vantage consistent with giving up the
bishop pair in an open position.
1 3 lDb5 il.xb5 1 4 il.xb5+ <it>e 7 1 5 g3
c S 4 e x d S 'ik x d S : Main L ine with 1 0 t:fJxd4
1 5 . . . h5?!
This seems very natural, but in fact
it only helps White, as the gS-square is
weakened. Black has two better ideas:
a) 1S ... a6 16 i.a4 Ithd8 17 1i'f3 h6 18
i.e3!? .l::!. a c8 19 i.d4 eS?! {Djurhuus
suggests 19 ...bS and gives 20 i.b3 i.cS
21 i.xcS+ 'fixes 22 a4 :d2 as equal) 20
i.e3 e4 21 1i'e2 �f8 22 :adl iLeS 23
i.b3! and White had increased the tra­
ditional bishop pair advantage in
Rozentalis-Djurhuus, Oslo 1992.
b) 1S ... :hd8 16 :el :ac8 17 1i'f3! a6
1 8 i.fl 1i'c6 19 'i'xc6 .l::!.xc6 20 i.g2,
when White's light-squared bishop
had arrived on its best diagonal in
A.Sokolov-Andersson, Clermont­
Ferrand 1989. Sokolov went on to win
a long endgame.
1 6 h4 lDg4 1 7 <it>g2 a6 1 8 iLa4 :hc8
1 9 il.g5+! <tis
Psakhis gives 19 .. .f6?! 20 .l::!.ae l! fxgS?
2 1 1i'xe6+ @f8 22 i.b3 tt:Jf6 23 hxgS as
winning for White.
20 l::tad 1 b5 21 iLc2 b4 22 iLe4
:ab8 23 cxb4 l::t xb4 24 l::tc 1
Of course White can also simply
grab a pawn with 24 1i'xa6. Psakhis
plays for more.
24 . . . iLc5 25 b3 'ii'd 7 26 iLf3 f6 27
:td 1 :d4 28 iLd2 e5
28 ....l::!. d 8 29 :xcS :xd2 30 .l::!.xd2
1i'xd2 3 1 .l::!.xhS wins.
29 :xc5! l::txc5 30 il.b4 'ii'd 6
The triple pin along the diagonal is
decisive. Neither 30 ...1i'c7 31 .l::!.xd4
exd4 32 i.xg4 hxg4 33 1i'c4 nor
30 ....l::!.xdl 3 1 i.xcS+ :d6 32 1i'xa6 save
Black.
3 1 iLxg4 1 -0
3 1 ...hxg4 32 .l::!.xd4 exd4 33 1i'c4
wms.
83
The Fre n c h Ta rra s c h
Game 45
Kopilov-Kahn
Correspondence 1991
1 e4 e6 2 d4 d5 3 lDd2 c5 4 lDgf3
cxd4 5 exd5 'ii'xd 5 6 iLc4 'ii'd 6 7 0-0
lDf6 8 lDb3 lDc6 9 lDbxd4 lDxd4 1 0
lDxd4 il.d7 1 1 b3
Given that this move leads to im­
mense complications that are not un­
favourable for Black, it is surprising
how often 1 1 b3 has been played, es­
pecially when White has a safe and
good alternative in 1 1 c3. Many of the
critical variations are very similar,
with small nuances such as the posi­
tion of White's rooks as the only dif­
ference.
1 1 . . . 0-0-0 1 2 il.b2 "iic 7 1 3 "iie 2 h5
1 4 lDf3
After 14 h3 tt:'ig4! White has nothing
better than to returning to the main
game with lS tt:'if3, as the greedy lS
hxg4 fails to 1S ... hxg4 16 g3 i.cS 17
Ir.fd l i.c6!, and White can resign.
1 4 . . . lDg4
1 5 h3
The other main try is lS l:.adl. Af­
ter 1S ... i.d6 16 h3 i.c6 17 l:!fe l i.cS
84
18 .l::!.xd8+ 'ifxd8 19 i.xe6+ fxe6 20
'ifxe6+ @b8 (20...i.d7 2 1 1i'c4 1i'b6 22
i.d4 wins for White) 21 hxg4 hxg4 22
i.eS+ @a8 23 'ifxg4 i.xf3 Glek as­
sesses the position as unclear. Looking
a little further, we can discount 24
iixf3 on account of 24....l::!.f8! 2S i.£4
gS 26 hgS lhf3 27 i.xd8 i.xf2+ 28
@ft .l::!.f8 29 i.e7 .l::!.f7 and the bishop is
lost. That leaves 24 gxf3 1i'd2 2S %:!.fl
i.xf2+! 26 .l::!.xf2 'iie l+ 27 @g2 'iih l+
28 @g3 'iig l+ 29 :g2 iie l+ with a
perpetual check.
1 5 . . .iLc6 1 6 l::tfd 1 iLc5 1 7 hxg4
An attempt to improve upon 17
.l::!.xd8+, when 17 .. . iixd8 18 hxg4 hxg4
19 i.xe6+ fxe6 20 1i'xe6+ i.d7 2 1 'iic4
gxf3 22 iixcS+ i.c6 gave Black suffi­
cient compensation in Lanka-Glek,
USSR 1989.
1 7 . . . hxg4 18 iLe 5! l::txd 1 + 1 9 "iixd 1
"iie7 20 lDh2 "iig 5! 21 il.g3
21 .. .l::txh2 ! 22 iLxh2
22 @xh2 'ifh6+ 23 @gl i.xg2! wins
beautifully after 24 @xg2 1i'h3+ 2S
@gl 'ifxg3+ 26 @h 1 1i'h3+ 27 @gl g3.
Now White bails out for a draw.
22 . . . g3 23 il.xg3 'i'xg3 24 "iif 1 "iif4
25 :td 1 il.d6 26 l::txd6 1h -1h
The position is completely level.
 3 . . . c S 4 e x d S 'ik x d S : Main Line with 1 0 t:fJxd4

Summary
I don't believe for a moment that we've heard the last of the 1 1 %:!.e l iic7 12 ..5lb3
..5ld6 13 lt:JfS!? argument, the assessment of which seems to change from day to
day. One thing is clear: both players have to know their stuff well before enter­
ing into the labyrinth of complications. White seems able to keep a nagging edge
against 10 ... ..td7 with the simple 1 1 c3 and 12 'iie 2.

1 e4 e6 2 d4 d5 3 lDd 2 c5 4 exd5 'ii'x d5 5 lDgf3 cxd4 6 iLc4 'ii'd6 7 0-0 lDf6

8 lDb3 lDc6 9 lDbxd4 lDxd4 1 0 lDxd4

1 0 . . . a6
10 ... ..td7 (D)
1 1 c3 iic7 12 iie2
12 ...0-0-0 Game 43
-

12 ... ..td6 Game 44

-

1 1 b3 Game 45 -

1 1 :e 1
1 1 ..5lb3 Game 41-

1 1 b3 Game 42-

1 1 .. .'ii'c7
1 1.. . ..td7 Game 40 -

12 il.b3 iLd6 1 3 lDfS (D)

13 h3 - Game 39
1 3 . . .iLxh2+ 1 4 <it>h 1 0-0 1 5 lDxg7 l::td8 16 'ii'f3 <it>xg7 1 7 iLh6+ <it>g6
1 8 c3 lDh5 (D)
1 8 . tt:JdS Game 38
.. -

19 ..tc1 Game 37
-

10. . . ildl 13 lDf5 1 8 . lDh5

. .

85
 CHAPTER SIX I
3 cs 4 exd5 'ifxd5: Early
Deviations from the Main Line
. . .
1 e4 e6 2 d4 d5 3 lll d 2 c5 4 exd5
'ii'x d5
Given the density of the theoretical
material in the last chapter, it is per­
haps not surprising that some White
players, with less study time on their
hands, have preferred to play one of
the more sedate variations that we
shall study here. In particular, 1 0
1i'xd4 (Games 46-48) is still very
common. In the very early days of this
variation the queen exchange was sup­
posed to give White an easy edge. In
those days, however, the strength of
the queenside pawn majority tended
to be grossly overestimated. It is now
common knowledge that, with a little
care, Black can equalise after 10 1i'xd4,
so nowadays the queen recapture is
basically used by two sets of players:
those who have immense belief in
their ability to extract something from
nothing in the endgame; and those
who are intent on killing the position
for a quick draw. Black players must
respect the first category and look to
punish the second! One more thing
86
about 10 1i'xd4: Don't be fooled by
the cross-section of results in the three
games given here. In real life there is a
far higher percentage of draws.
Moving on to more stimulating ma­
terial, it is worth taking a good look at
Zapata's 9 1i'e2 (Game 49) and Ljubo­
jevic 's 9 %:!.el (Game 50), as there may
well be some hidden ideas for both
sides in these lines. Sergei Smagin's 7
'iie2 (Game 52) is also very interesting,
and can be especially dangerous
against the unprepared player. We
could also see a resurgence of the little
known 5 dxc5, which naturally re­
ceived a bad press after Kasparov's lost
with it against Anand. However, more
recently young Russian players such as
Svidler and Rublevsky have not been
afraid to repeat the line, as we see in
Game 55.
From Black's point of view, the
most important sideline is 6 ...1i'd8, in
conjunction with 7 .. . a6, which has
been used by strong grandmasters
with a varying degree of success (see
Game 54). This line does have the ad-
 3 . . . c S 4 e x d S 'ik x d S : Ea rly De viations from t h e Main L ine
vantage of flexibility, as the king's
knight can decide between e7 and f6.
On the flip side, Black has to be very
careful not to wind up too far behind
in development. Finally, two other
sidelines for Black are also considered:
6 ...'iid6 7 0-0 tt:Jc6 (Game Sl) and
6 ...'iid8 7 0-0 ltJe7 (Game S3) .
Game 46
Wessman-Wiedenkeller
Swedish Championship 1989
1 e4 e6 2 d4 d5 3 llld 2 c5 4 exd5
'ii'x d5 5 lllgf3 cxd4 6 iLc4 'ii'd6 7 0-0
lllf6 8 lll b3 lll c 6 9 lllbxd4 lllxd4 1 0
'ii'xd4
1 0 . . . a6
10 ...'iixd4 and 10 ... .td7 are consid­
ered in Games 47 and 48 respectively.
1 1 iLf4 'ii'xd4 1 2 lll xd4 b5 1 3 iLe2
il.b 7 1 4 lll b 3
The immediate 14 c4 led to speedy
equality in Renner-Levitt, Augsburg
1989, after 14 ... bxc4 lS i.xc4 i.e7 16
.l::!.ac l tt:Jds 17 .tes 0-0 18 4Jb3 a5 19
.l::!.fdl :fcs 20 .tbs a4 2 1 .l::!.xc8+ .l::!.xc8
22 i.xa4 l:.a8 23 .tbs .l::!.xa2. With the
text move White plans to harass the
black bishop with ltJaS.
1 4 . . . il.d5!
14 ....te7 does nothing to halt
White's plan. After lS c4 bxc4 16 tiJaS
.tdS 17 tt:Jxc4 i.cS 18 l:.ac 1 0-0 19 a3
aS 20 ltJxaS .l::!.xaS 2 1 i.c7 i.xa3 22
i.xaS i.xb2 White had some winning
chances in Hubner-Klinger, Biel 1986.
1 5 a4 b4 1 6 a5
Fixing the pawn on a6, but it has to
be said that the c2-pawn is just as
weak. White's next few moves are
hardly the best options, but in any
case Black is already a little better, as
he has a simple plan of attack.
1 6 . . . iLe7 1 7 llld 2 0-0 1 8 lllc4 :ac8
1 9 lll b6? l::txc2 20 iLxa6 l::t x b2 2 1
il. b 5 b 3 2 2 a6 l::t a2 2 3 :xa2 bxa2 24
l::t a 1 iLc5 25 lllc4 lll e4 26 :xa2 g5!
2 7 .te3?
This just loses material. The last
chance was 27 i.g3, although White's
uncoordinated pieces ensure that
Black retains the advantage after
27 ...tt:Jc3 28 :b2 :as.
27 . . . iLxe3 28 fxe3 lllc3 29 e4 iLxc4
0- 1
After 30 i.xc4 lDxa2 3 1 i.xa2 Black
can even round up the a6-pawn with
3 1 . . ..l::!.c 8, intending ... :c6. Some might
say that the result was an apt punish-
87
The Fre n c h T a rra s c h
ment fo r White's insipid opening play,
but see the next game ...
Game 47
Tiviakov-Chernin
Podolsk 1993
1 e4 e6 2 d4 d5 3 llld2 c5 4 exd5
'ii'x d5 5 lll g f3 cxd4 6 iLc4 'ii'd6 7 0-0
lll f 6 8 lll b3 lll c 6 9 lllbxd4 lll xd4 1 0
'ii'xd4 'i'xd4 1 1 lllxd4 il.d7 1 2 iLe2!
Tiviakov shows an excellent under­
standing of this endgame. Without
any provocation, the bishop retreats
to e2, from where it can reach the h 1-
a8 diagonal. This is undoubtedly the
best place for this bishop in this line.
1 2 . . . iLc5 1 3 lll b3 il.b6 1 4 a4 a6 1 5
iLf3 0-0-0 1 6 iLd2 iLc6
Black feels obliged to contest the
long diagonal. This is indeed the cor­
rect decision, but the resulting split
pawns will give White a tiny edge.
1 7 iLxc6 bxc6 1 8 iLc3 :hg8 1 9
.l:.fe1 ! l::t d 5 20 llld2 lllg4 2 1 :e2
:gd8 22 :n il.d4 23 lllc4 .l:.c5 24
iLxd4 l::t xd4 25 b3
Despite the apparent activity of the
black pieces, it is still White who has
the superior structure and the winning
88
chances. In particular, the white
knight has found an impressive out­
post on c4. Black should be able to
hold on with passive defence, but in
the heat of a tournament situation,
this is often easier said than done. In­
stead Chemin makes a practical deci­
sion to sacrifice an exchange for a
pawn, eliminating the knight in the
process. Tiviakov's endgame tech­
nique, however, cannot be questioned.
25 . . .:cxc4 !? 26 bxc4 l::txc4 27 :a 1
lllf6 28 :a3 �c7 29 g3 h6 30 �f 1
a5 3 1 �e 1 llle4 32 l::tee3 g5 33 f3
llld6 34 �d 1 :cs 35 �c1 h5 36
l:tac3 l::tf5 37 l:ted3 lllc8 38 g4! hxg4
39 fxg4 l::t e 5 40 l:td4 lllb6 41 l::th 3 c5
42 l::td 1 :e4 43 l::tf3 l::txg4 44 .l:txf7+
�c6 45 :ta lll xa4 46 :xe6+ �b5 47
l::t d 3! :t4 48 :g3 g4 49 :e 1 c4 50
:eg1 l:tf2 51 :3g2 :t3 52 l::txg4
lll c3 53 l:.h 1 ! a4? 54 h4 a3 55 h5
:t2 56 :gh4!
White still could have blown it. The
careless 56 h6 allows 56 ...tt:'ia2+ 57
@bl 4Jc3+ with a perpetual, as 58 @al
runs into 58 ... .l::!.xc2 and Ir.a2 mate.
Now 56 ...tt:'ia2+ 57 @bl tt:'ic3+ 58 @al
x:f.xc2 can be answered by 59 .l::!.4h2!
The rest is easy, as the h-pawn, sup-
 3 . . . c S 4 e x d S 'ii x d S : Ea rly De v ia t io n s from the Main L ine
ported by the two rooks, marches re­
lentlessly up the board.
56 . . .:f 7 57 h6 l::th7 5a l:tg4 �c5 59
l::tg 7 :ha 60 h7 a2 61 �b2 1 -0
After 6 1 ...l:!b8+ 62 @xc3 l:Ib 1 White
finishes off with 63 .l::!.g5+ @b6 64
.:I.h6+ @a7 65 lag7 +. An impressive
exhibition by Tiviakov, but not all of
us have his endgame powers.
Game 48
Slobodjan-Luther
Lippstadt 1997
1 e4 e6 2 d4 d5 3 lDd2 c5 4 exd5
'ii'x d5 5 lDgf3 cxd4 6 ilc4 'ii'd 6 7 0-0
lDc6 a lDb3 lDf6 9 lDbxd4 lDxd4 1 0
'ii'xd4 ild7
A perfectly playable alternative to
10 . . . a6 and 10 ... iixd4, although as you
would expect, there are many direct
transpositions.
1 1 ile3 'ii'xd4 1 2 ilxd4 :ca 1 3 lDe5
.i.c5 1 4 :ad 1
see following diagram
1 4 . . . ilxd4
14 ...@e7 is also possible. After 15
tt:'ixd7 i.xd4 16 laxd4 Brodsky rec­
ommends the nuance 16 ... .l::!.hd8! 17
.l::!.fdl l::txd7 18 .l::!.xd7+ tt:'ixd7, which he
assesses as equal. If instead 16 ... tt:'ixd7
17 .l::!.fdl l:thd8 1 8 f4 White may be a
shade better.
1 5 :xd4 �e7 1 6 .l:.fd 1 :hda 1 7 f3
17 tt:'ixd7 l::txd7 18 .l::!.xd7+ tt:'ixd7
would transpose to the previous note.
After 17 f3 Black can retain the
bishop, although the position remains
'dishwater dull', as well as totally
equal. The rest of the moves see both
sides displaying ample technique to
obtain a draw.
ilea 1 a :xda l:.xda 1 9 .l:.xda �xda
20 �2 lDd7 21 lDxd7 ilxd7 22 f4
We 7 23 �e3 �d6 24 �4 f6 25 ild3
h6 26 h4 b6 27 g3 ilc6 2a a3 ilea
29 ile4 ilf7 30 ild3 ilea 31 b4 ilf7
32 c3 ilea 33 c4 e5+ 34 fxe5+
fxe5+ 35 '>te3 ilf7 36 ile2 ilg6 37
ild3 ilf7 3a ile2 1h -1h
This is an appropriate juncture to
end our discussion on 10 1i'xd4 lines.
The final position is utterly lifeless.
Game 49
Zapata-Dolmatov
Tilburg 1993
1 e4 e6 2 d4 d 5 3 lDd2 c5 4 lDgf3
89
The F re n c h Ta rra s c h

cxd4 5 exd5 'ii'x d5 6 iLc4 'ii'd 6 7 0-0
lDf6 8 lDb3 lDc6 9 'ii'e 2!?
eventually lost. Instead White can win
with the brilliant 28 .l::!.xb7!! .l::!.xb7 29
tbg6+ @g8 30 1i'e8+ @h7 31 iixh8+
@xg6 32 .td3+ @gs 33 h4+ @xh4 34
'iic 8!, threatening the rook and also
mate with 3S g3+ @gs 36 f4+ @h6 37
'iih 8.

9 . . .iLe7
Also critical is 9 ...a6, after which
White has two options:
a) 10 a4 (m my opinion the inser­
tion of a2-a4 and ...a7-a6 has helped
Black, who now doesn't have to 1 0 il.g5 0-0 1 1 l:.fe1 a6 1 2 :ad1 b5
worry about possibilities of .tbs or 1 3 il.d3 il.b7 14 c3 :tea 1 5 iLb1
tbbs) 10 ... .te7 1 1 g3 (1 1 .tgs es 12 'ii'c7 1 6 lDbxd4 lDxd4 1 7 lDxd4 :ad8
l:tfel .5lg4 13 h3 i.xf3 14 'iixf3 0-0 lS So finally White has recaptured the
i.h4 e4 was clearly better for Black in d4-pawn and a typical middlegame
Siklosi-Brinck Claussen, Copenhagen position has arisen: White's active
1988) 1 1. ..eS 12 t'.DgS 0-0 13 f4 and here pieces against Black's super-solid pawn
Black should play 13 ... i.g4! 14 tbxf7 structure.
i.xe2 lS tbxd6+ i.xc4 16 tbxc4 e4 1 8 'ii'd3 g6 1 9 'ii'h 3 lDh5!?
with an awesome looking centre.
b) 10 .l::!.d l! looks more testing, e.g.
10 ... bs 1 1 .td3 'iic7 12 a4 b4 13 i.c4
.tb7 14 tbbxd4 tbxd4 lS .l::!.xd4 i.cS 16
.l::!. h 4 hs 17 h3 tbds 1 8 .tgs .te7 19
ll.dl as 20 .tbs+ @f8 21 :c4 and
Black's king is awkwardly placed on
f8. Geller-Dolmatov, Moscow 1992,
continued 21...1i'd8 22 i.xe7+ 'iixe7
23 tbes @g8 24 tbc6 1i'f6 2S .l::!.xdS
exdS 26 tbe7+ @f8 27 .l::!.c7 .l::!.b 8.
see following diagram
20 il.h6? !
Here Geller played 28 t'.Dc8? g6!, and In this type of position White often

90
 3 . . . c S 4 e x d S 'ik x d S : Ea rly De via tions fro m the Main Line
only gets 'one bite of the cherry'. A
chance of a sacrifice presents itself
with 20 lt:Jxe6!, when after 20 ... fxe6
(20 ... .l::!.x dl 21 J::.xdl fxe6 22 'ifxe6+
@f8 23 l:td7 wins) 2 1 'ifxe6+ Black has
to be extremely careful, e.g.
a) 2 1...@f8? 22 .l::!.xd8 'ifxd8 23 i.h6+
lt:Jg7 24 i.xg6!! hxg6 25 'ifxg6 ..tf6 26
.l::!.xe8+ 'iixe8 27 'iixf6+ 'iif7 28 i.xg7+
with a winning position.
b) 2 1 ...@g7! 22 l::txd8 i.xd8 23
i.h6+ @xh6 24 iixe8 with an unclear
imbalance of rook and two pawns
against bishop and knight. This may
still be slightly better for White.
20 . . . iLf6 21 'ii'g4 e5 22 lllf 5 :xd 1
23 l::t xd 1 .!:.dB 24 iLe3 lllf4 25 iLxf4
exf4 26 llld4 b4
White's kingside demonstration has
got absolutely nowhere, and Black is
just about to take over operations. In
particular, the pair of bishops will
make White suffer painfully in this
open position.
27 'ii'e 2 bxc3 28 bxc3 'ii'xc3 29 lllb3
l::txd 1 + 30 'ii'xd 1
A pawn up, with a dominating po­
sition, one would normally expect
Dolmatov to convert to the full point
with ease. Strangely, however, the
game drifts towards a draw.
30 . . . f3?!
Looks strong, but this pawn 1s
picked up in the endgame.
3 1 g3 a5 32 'ii'd 2 'ii'x d2 33 lll xd2
il.d5 34 a4 il.g5
After 34 ... i.c6 35 i.c2 i.c3 36 �e4
..5lb4 Black's can still play for the win
with the bishop pair. Perhaps a draw
was all the Russian technician needed.
35 iLe4 iLxd2 36 iLxd5 �g7 37
iLxf3 g5 38 � 1 f6 39 �e2 il.b4 40
iLe4 h6 4 1 �3 iLc3 42 �g4 il.b4 43
�f5 iLe 1 44 f4 gxf4 45 �xf4 il.b4
46 �5 iLc3 4 7 il.d5 il.b4 48 il.b3
iLc3 49 iLd 1 il.b4 50 iLh5 iLc3 51
�e6 il.b4 52 �d5 Yz -Yz
Game 50
Winsnes-Lein
Gausdal 1990
1 e4 e6 2 d4 d5 3 llld 2 c5 4 exd5
'ii'x d5 5 lllgf3 cxd4 6 iLc4 'ii'd 6 7 0-0
lllf6 8 �b3 lllc6 9 :e 1
Ljubojevic's idea, which has never
really caught on. White struggles just
to regain the pawn and this gives
Black the chance to co-ordinate fully
and reach a comfortable middlegame.
91
The Fre n c h Ta rra s c h
9 . . . il.d7 1 0 g 3 iLe7 1 1 iLf4 'ii'b4 1 2
'ii'd 3 0-0 1 3 iLc 7? !
This i s probably too ambitious, al­
though it is quite surprising how
quickly White's position deteriorates.
I prefer 13 a3 'iib 6 14 lt:Jfxd4 lt:Jxd4 15
'iixd4 with a level position.
1 3 . . . iLdB 1 4 a3 'ii' e7 1 5 iLf4 il.b6 1 6
lllfxd4 lllxd4 1 7 lll xd4 'ii'c 5! 1 8 J:tad 1
l:tac8 1 9 il.b3 'ii'h 5 20 'ii'e 2 lllg4 2 1
f3
21 . . . e5! 22 iLxe5 J:tce8 23 f4 <it>h8!
0- 1
White has no defence to .. .f7-f6.
Game 51
M. del Campo-Escobedo Tinajero
Mexico 1991
1 e4 e6 2 d4 d5 3 llld 2 c5 4 lllgf3
cxd4 5 exd5 'ii'x d5 6 iLc4 'ii'd6 7 0-0
lllc6
see following diagram
8 llle 4!
When I first encountered 7 ... lt:Jc6 I
thought that Black's move order was
irrelevant, so I bashed out 8 lt:Jb3, ex­
pecting to transpose to normal lines
after 8 ...lt:Jf6 9 lt:Jbxd4. I then remem-
92
her being slightly surprised and a little
annoyed when my opponent didn't
co-operate. Instead there came 8 ...eS 9
'ife2 ..5le7?! (after 9 ... ..5le6! it is not clear
how White can justify his pawn defi­
cit) 10 .l::!. dl 'iic7 1 1 ..tds ..tg4 12 'iic4
@f8 13 l:Iel ..tf6 14 c3! and White
went on to win in Emms-I.Andersen,
Cappelle la Grande 1992. 8 t'.De4 was
suggested to me by the American GM
Ilya Gurevich, and it does seem a logi­
cal way to exploit the omission of
.. .tt:Jf6.
8 . . . 'ii'd8 9 'ii'e 2 iLe7 1 0 l:td 1 lllf6 1 1
c3!
Recovering the pawn and entering a
favourable IQP position.
1 1 . . .lllxe4 1 2 'ii'xe4 0-0 1 3 cxd4
lllb4 1 4 l:te1 iLf6 1 5 iLf4 llld5 1 6
iLe5 iLxe5 1 7 dxe5 b6 1 8 iLxd5!
'ii'xd 5 19 'ii'x d5 exd5 20 J:tac 1 iLe6
21 llld4
The opening has been a complete
success for White. He has cleverly
converted his advantages and is now
left with a 'good knight versus bad
bishop' situation. Added to this Black
has a weak d-pawn, while White can
advance the kingside pawn majority
without any difficulty. White proba-
 3 . . . c s 4 e x d S 'ik x d S : Ea rly De viations from t h e Main Line

bly already holds a decisive advantage. so it is best to hold on to it as long as

possible, even just for nuisance value.
Other moves include:

2 1 . . :tea 22 l::txc8+ :xc8 23 f4 g6

.

24 <it>f2 �8 25 h3 h5 26 �3 iLf5
27 g4 hxg4+ 28 hxg4 iLe4+ 29 <it>e3
<it>e7 30 f5 a6 3 1 f6+ <it>d7 32 <it>f4
l::te8 33 e6+!
This sacrifice gives the white king a
decisive entry square, where it can
support the f6-pawn. Black's battle
against this monster of a passed pawn
is doomed to failure.
33 . . . fxe6 34 <it>e5 g5 35 f7 .l:.f8 36
<it>f6 .l:.h8 37 tl:lxe6 .l:.h6+ 38 <it>xg5
.tt. g6+ 39 <it>h5 .l:.xe6 40 f8if il.g6+
41 <it>g5 :xe 1 42 'ti'g7+ �e7 43
'ii' xg6 :ea 44 'ii' d3 b5 45 'ti'xd5+
l::td6 46 'ti'xd6+ 1 -0
a) 9 ... ..td7 10 0-0-0 0-0-0 1 1 lDbxd4
ltJxd4 12 lhd4 1i'b6 13 'fies 'fies 14
.l::!.h dl with an edge to White in Palac­
Touzane, Nice 1994.
b) 9 ... iib4+ 10 ..td2 1i'b6 1 1 0-0-0
..td7 12 ..tgs 0-0-0 13 tijfxd4 ltJb4 14
a3 ltJbdS lS i.xdS exdS 16 f3 i.d6 17
g3 and again White was slightly better
in Rozentalis-Glek, Antwerp 1993 .
1 0 0-0-0 b5 1 1 il.d3 iLe7 !
This is safer than 1 1...i.b7. Smagin­
Levitt, Amantea 1993, continued 12
4Jbxd4 lDxd4 13 4Jxd4 0-0-0 ('castling
into it' has never been more apt)

Game 52
Anka-1.Almasi
Hungary 1997

1 e4 e6 2 d4 d5 3 tl:ld2 c5 4 exd5
'ii'x d5 5 tl:lgf3 cxd4 6 iLc4 'ii'd 6 7
'ii'e2 ! ? tl:lf6 8 tl:lb3 tl:lc6 9 il.g5

see following diagram

9 . . . a6
Black has no real route to comfort­
able equality by giving the pawn back, 14 i.xbS!! axbS lS lDxbS 1i'b4 16

93
The Fre n c h Ta rra s c h

�xd8+! @xd8 17 .l::!. d l+ @c8 18 :d4

'iiaS 19 'iic4+ and Black resigned on
account of 19 ... @b8 20 .tf4+ @a8 21
.l::!.d8+! 'iixd8 2 2 'iia4+.
1 2 <it>b1
Naturally White would like to re­
claim the pawn with 12 lt:Jbxd4 lt:Jxd4
13 lt:Jxd4, but then 13 ...'iidS! hits a2
and gS. After 1 4 i.xf6 i.xf6 lS @b l
'fies! 16 .5le4 .l::!.a7 Black's two bishops
gave him an edge in V.Ivanov­
Zakharov, Russia 1994. Note that 17
i.c6+ doesn't help White. After
17 ... <ite7 1 8 1i'e4 x:f.c7 the black king is Game 53
quite safe, while the bishop on c6 is a Yandimerov-R.Nikitin
liability. Now, however, Black can Russ ia 1997
hold onto the extra pawn.
1 2 . . . e5 1 3 h3 iLe6 1 e4 e6 2 d4 d5 3 llld 2 c5 4 exd5
After 13 ... tt:Jds 14 i.xe7 lt:Jdxe7 lS 'ii'x d5 5 lllgf3 cxd4 6 iLc4 'ifdB 7 0-0
l::the l f6 White can break open the lll e 7 8 lllb3 'ii'c7 9 'ii'e2 lDg6 1 0
centre to his advantage with 16 c3! lllbxd4!
14 :he 1 lDd7? 10 lt:Jfxd4, intending f2-f4-fS, was
After this Black's position collapses. shown to be artificial in the game Ad­
14 ....l::!. c 8! is a greater test for White. ams-Dreev, Dortmund 1994, which
The game Smagin-Marjanovic, Yugo­ went 10 ... a6 1 1 f4 i.e7 12 fS?! exfS 13
slavia 1994, continued lS g4 (lS lt:JxeS? lt:JxfS i.xfS 0-0 lS ttJd4 lt:Jd7! 16 i.e3
fails to lS ... iixeS 16 'iixeS lDxeS 17 lt:Jf6 17 <ithl .l::!.ae8 and Adarns's lunge
:xeS lt:Jd7) 1S .. . lt:Jd7 and White didn't had only served to weaken his king.
have enough compensation for the 1 0 . . . a6
pawn. In his notes Smagin suggests lS
i.h4 as the best try fo r White, intend­
ing 1S ...lt:Jd7 16 i.g3, when the pres­
sure persists. This line needs practical
tests.
1 5 iLxe7 <it>xe7 1 6 iLe4 :hc8 1 7
..txc6 'ii'x c6

see following diagram

1 8 lDxe5 lllx e5 1 9 lllxd4! 'ii'x g2 20

'ii'x e5 'ii' g6 21 lllf 5+ 1 -0
2 1...@e8 22 lt:Jd6+ We7 23 lt:Jxc8+
.l::!.xc8 24 1i'd6+ is terminal. Against 10 ... i.e7 White could try 1 1

94
 3 . . . c 5 4 e x d 5 'ik x d 5 : Ea rly D e viations from t h e Main Line
ttJb5 1i'b6 12 i.e3 i.c5 13 i.xc5 'iixc5
14 1i'e3! 1i'xe3 15 fxe3 with an edge.
1 1 il.b3
1 1 i.xe6!? fxe6 12 tLixe6 i.xe6 13
1i'xe6+ is a dangerous piece sacrifice
which worked well for White in For­
ster-Vaganian, Biel 1994. After
13 .. . 1i'e7? 14 1i'c8+ 1i'd8 15 1i'xb7
Black's king didn't survive too long.
13 ... i.e7 would have been more stub­
born.
1 1 . . .il.e7 1 2 :e 1 0-0 1 3 il.g5 il.d6
14 'ii'd 3 h6?
Black's development disadvantage
has reached serious proportions.
However, his position would still be
playable but for this move, which al­
lows a powerful sacrifice.
1 5 lll x e6! ! iLxe6 1 6 l::txe6 fxe6 1 7
'ii'xg6 lll c6 1 8 :d 1 lll e5 1 9 'ii'xe6+
<it>h8 20 :xd6 lllxf3+ 21 gxf3 :ae8
22 'ii'g 6 'ii'a 5 23 iLxh6 1 -0
Game 54
Emms-LB.Hansen
Bronshoj jubilee 1995
1 e4 e6 2 d4 d5 3 llld2 c5 4 lllgf3
cxd4 5 exd5 'ii'xd5 6 iLc4 'ii'd6
In fact, the most common move or-
der to reach the diagram position be­
low is 6 ...1i'd8 7 0-0 a6 8 tiJb3 'iic7,
which avoids 8 tLle4.
7 0-0 a6 8 lll b3
8 tLle4!? may be a way for White to
exploit Black's move order, e.g.
8 ... 1i'c7 9 i.b3 i.e7 10 tLlxd4 tiJf6 1 1
%:!.e l with an edge for White.
8 . . . 'ii'c 7
9 'ii'e2
The major alternative is 9 1i'xd4!?
Practical play suggests that White can
secure an advantage, e.g. 9 ... tiJc6 10
'iih4 i.d6 11 i.d2 ttJge7 (1 1...tiJf6 12
i.h6!) 12 i.d3 i.d7 (12 ...ttJe5 may be
stronger) and now:
a) 13 lladl 0-0-0 14 tiJg5! tiJg6 15
'iih3 tLlge5 16 i.e2 and White was bet­
ter, although he later lost, in Akopian­
Dreev, Linares 1995.
· b) 13 i.c3 f6 14 'iih5+ �f8 15 ttJc5
ttJd5 16 ttJxd7+ 'iixd7 17 i.d2 and on
this occasion White converted his ad­
vantage in Nisipeanu-Priehoda, Buda­
pest 1996.
9 . . .il.d6 1 0 lll bxd4 llle 7
The main nuance in this line. This
knight develops at e7 rather than f6.
1 1 :d 1
1 1 %:!.e l is a critical alternative, e.g.
95
The Fre n c h Ta rra s c h
1 1 ...0-0 ( 1 1 ...tt:Jbc6! looks safer) 12 ltJgS
h6 13 lDxf7!! @xf7 14 lDxe6 i.xe6 15
1i'xe6+ @e8 16 i.xh6 :f6 17 1i'g8+
@d7 18 .l::!.xe7+ i.xe7 19 .l::!.d l+ @c6 20
iids+ @b6 21 i.e3+ i.cs
22 b4 @a7 23 i.xcS+ b6 24 i.d4
.l::!.d6 25 'iieS tt:Jd7 26 1i'xg7 1i'c6 27
i.fl .l::!. c8 28 c3 J::.c7 29 1i'g3 @b7 30
i.e2 tt:Jf 6 31 i.f3 tt:JdS 32 c4 lle7 33
i.xdS .l::!.xdS 34 cxdS 'iia4 35 'iic3 and
Black resigned in Slobodjan-Kaminski,
Halle 1995. A brilliant attacking dis­
play from the soon-to-be World Jun­
ior Champion.
1 1 . . . lDbe6
Black has to be a little careful, e.g.
1 1 ...0-0? 12 i.xe6! fxe6 13 lDxe6 1i'b6
14 lDxf8 @xf8 15 1i'd3 is very strong
for White. Nevertheless 1 1 .. .ltJbc6 is
perfectly satisfactory for Black.
1 2 lDxe6 lDxe6 1 3 il.d3 il.d7
13 ...0-0 once again ' castles into it' :
1 4 i.xh7+ @xh7 15 ltJgS+ gives White
a strong attack. Here I was most con­
cerned over the simplifying move
13 ... ltJeS!, when after 14 lDxeS i.xeS
15 h3 0-0 White has no advantage
whatsoever.
1 4 b3 0-0-0
Now I was ready to answer
96
14 ...ltJeS with the continuation 15
i.b2! lDxf3+ 16 1i'xf3 i.xh2+ 17 @hl
i.eS 18 i.xeS 'fixes 19 iixb7, when
White is better.
1 5 il.b2 e5 1 6 e4 f6 1 7 a3 :he8 1 8
iLe4?
This is far too casual. After 1 8 1i'c2,
clamping down on the light squares, I
still prefer White.
1 8 . . . f5! 1 9 iLxe6 iLxe6 20 e5 iLxe5
21 lDxe5
21 .l::!.xd8+ 1i'xd8 22 tt:'ixeS 'iidS is the
end for White, so a pawn goes down
the drain. The rest of the game proves
to be a highly successful grovel.
2 1 . . ..l:.xd 1 + 22 l::txd 1
22 . . .iLxa3! 23 f4 il.xb2 24 'ii'x b2
'ii'b6+ 25 <it>h 1 l::td8 26 :xd8+ 'ii'xd8
27 h3 'ii'd 5?
27 ... i.e4 keeps Black firmly in con­
trol. After the game move White is
able to create enough counterplay for
a draw.
28 <it>h2 'ii'e4 29 lDxe6 bxe6 30
'ii'xg7 'ii'xf4+ 3 1 <it>h 1 h6 32 'We 7
<it>b8 33 b4 'ii'g 5 34 'ii'd6+ <it>b7 35
'ii'd 7+ <it>b6 36 'ii'd4+ �e7 37 'ii'a7+
<it>d6 38 'ii'e5+ <it>d7 39 'ii'a 7+ <it>e6 40
'ii'x a6 'We 1 + 41 <it>h2 'ii'f4+ 42 <it>h 1
'ii'e4 43 'Wea+ <it>d5 44 'ii'f8 'ii'e 1 + 45
 3 . . . c S 4 e x d S Wk x d S : Ea rly De via tions from t h e Main Line
'>th2 'ii'e 5+ 46 <it>h 1 'We 1 + Yz - 1/:i
Game 55
Rublevsky-Beliavsky
Novosibirsk 1995
1 e4 e6 2 d4 d5 3 llld 2 c5 4 exd5
'ii'x d5 5 dxc5 iLxc5 6 lll gf3 lllf6 7
il.d3 0-0 8 'ii'e 2 lll bd 7!
Played by Anand against Kasparov.
The knight is most useful here, as it
adds extra protection to the kingside
and will not obstruct the black bishop
when it fianchettoes on b7.
9 b3
Kasparov-Anand, Reggio Emilia
199 1 , continued 9 tt:'ie4 b6! 10 tt:'ixcS
'fixe s (10 ...tt:'ixcS is also okay for
Black) 1 1 i.e3 1i'c7 12 i.d4 i.b7 13
0-0-0 tt:'icS! 14 i.eS tt:'ixd3+ lS klxd3
1i'c4 and Black was fine . Indeed, the
game ended in a rare Kasparov defeat
with the white pieces.
9 . . . a5!?
Discouraging White from castling
queenside, although in this game
White will not be deterred. Also feasi-
ble is 9 ...tt:'ig4 10 0-0 tt:'ideS. Now 1 1
tt:'ixeS iixeS! is equal, while Svidler­
Savchenko, Kazan 199S, ended in a
quick draw after 1 1 i.e4 tt:'ixf3+ 12
tt:'ixf3 'fibs 13 h3 tt:'if6 14 i.b2 tt:'ixe4 lS
'ifxe4 f6 16 J::.adl lh-lh. Note that 13
i.b2 would have been met by
13 . ..tt:'ixh2!
1 0 il.b2 b6 1 1 0-0-0 il.b7 1 2 :he1 ? !
This turns out to be too pedestrian.
Beliavsky recommends the direct 12
g4, intending g4-gS, when 12 ... tt:'ixg4
fails simply to 13 i.e4, winning mate­
rial. Now Black gets in first.
1 2 . . .a4 1 3 llle4 'ii'h 5 1 4 lllfg5 'ii'h 6
1 5 <it>b 1 axb3 1 6 axb3 iLa3 1 7
lllxf6+?
The start of a hallucination, based
on a horrendous oversight. 17 i.d4!
was stronger, after which Beliavsky
planned 17 ...tt:'idS!? 1 8 tt:'ixh7 .l::!.fc8 with
good compensation for the pawn.
1 7 . . . lllxf6
1 8 iLxf6? 'ii'xf6 1 9 iLxh7+ <it>h8 20
'ii'h 5
Threatening mate in two, but ...
20 . . . 'i¥b2 mate!
97
The Fre n c h Tarra s c h

Summary
Of White's alternatives in this chapter, perhaps the most dangerous are the early
'iie 2 ideas and the neglected 5 dxcS. It seems that further games are still required
before an accurate assessment can be made on the 6 ...'iid8 and 7... a6 line. '

1 e4 e6 2 d4 d5 3 llld2 c5 4 exd5 'ii'x d5

5 lllgf3
5 dxcS - Game 55
5 . . . cxd4 6 iLc4 'ii'd6
6 ... 'iid8 7 0-0 (D)
7 ...0ie7 Game 53
-

7... a6 8 0ib3 'ifc7 Game 54 (by transposition)

-

7 0-0
7 'iie2 - Game 52
7 ... tllf6
7 ...0ic6 Game 51
-

7 ...a6 Game 54
-

8 lll b3 lllc 6 (D) 9 lll bxd4

9 'iie 2 Game 49
-

9 %:!.el Game 50
-

9 . . .lll x d4 1 0 'ii'xd4 (D) a6

10 ... 'iixd4 Game 47 -

10 ... .td7 Game 48-

1 1 iLf4 Game 46 -

7 0-0 8 lllc6
. . . 10 'ii'xd4

98
 CHAPTER SEVEN
3 cs 4 exd5 exd5:
. . .
Main Line with 6 �b5
1 e4 e6 2 d4 d5 3 lll d 2 c5 4 exd5
exd5 5 lllgf3 lllc6 6 il.b5
All the rage in the seventies and
early eighties, interest has slowly
waned in this line, as Black players
have searched for more dynamic ways
to combat the Tarrasch, including the
...iixd5 variation and ... e6xd5 in con­
junction with an early ...a7-a6 or
... lbf6. The general consensus in the
old 5 ...lbc6 line is that White can play
for a risk-free plus by concentrating
his efforts against the isolated queen's
pawn (IQP) . On the other hand, it
must be said that Black has no prob­
lems with piece activity and this line
has appealed to IQP specialists such as
Vaganian, Dolmatov, Lputian, Uhl­
mann, and of course Korchnoi, all of
whom have scored notable successes.
In Games 56-58 we study the popu­
lar 1 1 i.d3, a slightly paradoxical idea,
given that White's main plan is to pile
up on the dS-pawn. In many positions
White even goes for a kingside attack
with c2-c3, i.c2 and iid3. This is ac­
tually quite a dangerous plan and
I
Black players really need to know
their stuff here. In contrast 1 1 i.g5
(Game 59) heads for a quieter position
more typical of the IQP Tarrasch. The
same can also be said of 10 lbbd4
(Game 60), although this is thought to
be less threatening for Black.
In Games 6 1-62 we see the less
popular 9 ...i.b6, a perfectly natural
looking move. The only problem with
it is that White has the even more
natural idea of 10 %:!.el and 1 1 i.e3,
which seems to secure an advantage.
Finally, 6 ... cxd4 is considered in Game
63.
Game 56
M. Read-Goncalves
Correspondence 1989
1 e4 e6 2 d4 d5 3 llld2 c5 4 exd5
exd5 5 lllgf3 lllc6 6 il.b5 il.d6 7
dxc5
This is a more exact move order
than 7 0-0, which allows Black the idea
of 7 ...cxd4. After 8 lbb3 lbge7 9
lbbxd4, White has been forced into
99
The F re n c h Tarra s c h
the tiJb3-d4 lines, a tempo less fo r each
side (see Game 60).
7 . . . iLxc5
7 ...'iie 7+ crops up from time to
time, but it doesn't really cause White
any problems. For example, 8 iie2
'ifxe2+ 9 @xe2 i.xcS 10 tiJb3 i.b6 1 1
i.e3 i.g4 1 2 h 3 i.hS 1 3 i.xb6 axb6 14
tiJbd4 and Black's wrecked structure
gave White a definite plus in Tal­
Korchnoi, Moscow 1973.
8 0-0 lll g e7 9 lll b 3 il.d6
10 .l:.e1 0-0 1 1 il.d3 h6
Of course Black would like to con­
tinue with 1 1 .. . i.g4, but this falls for
12 i.xh7 +! @xh7 13 tL!gS+, regaining
the piece with a winning position. In­
deed, this is one of the points behind
1 1 i.d3, since after 1 1 ...h6 White has
time to avoid the annoying pin.
1 2 h3 lllf 5
Paving the way for the black queen
to reach f6, where it can support the
advance of the d-pawn and allow a
rook to go to d8.
1 3 c3 1i'f6
The other main alternative here,
13 . .. i.c7, is the subject of Game 58.
14 iLc2 iLe6
Bolstering the dS-pawn, although
100
Black can also play 14 ....l::!.d 8, which
may be stronger (see the next game).
1 5 'ii'd 3 l:tfe8 1 6 .id2!?
White simply wants to double the
rooks on the e-file, when the pressure
on e6 may become surprisingly an­
noying. The direct 16 g4 is also play­
able, when after 16 ...tiJh4 17 tLixh4
ir'xh4 18 iih7+ @f8 19 �g2 'iif6 20
i.e3, Uhlmann assesses White's
chances as slightly better.
1 6 . . . g6 1 7 l:te2
A change of plan with 17 iibS?! is
inadvisable. Ernst-Vaganian, Copen­
hagen 1988, continued 17 ...tiJh4 18
tLixh4 'iixh4 19 'iixb7? i.xh3! and here
White should have bailed out with 20
iixc6, which leads to a draw after
20 ... i.h2+ 2 1 @fl i.xg2+ 22 @xg2
iig4+ 23 @h 1 iif3+ 24 @xh2 iixf2+
25 @h3 iif3+.
1 7 . . .J:tacS
17 ...i.f8 was played in Wolff­
Benjamin, San Francisco 199 1, but
after 1 8 l::tae l tiJd6 19 tiJbd4 tLle4 20
tLixe6 !!xe6 2 1 i.b3 tLlxd2 22 'i'xd2
the isolated d-pawn was doomed. At
first I thought that 17 ... J::.ad8 might be
an improvement for Black. Neverthe­
less, this causes brand-new problems
 3 . . .
with 1 8 .l::!.ael i.f8 19 g4 tt:Jd6 2 0 gS!
(exploiting the fact that a rook is on
d8; 20 ... hxgS 2 1 i.xgS picks up an ex­
change) 20 ...'ifh8 2 1 .l::!.xe6 fxe6 22
'ifxg6+ 'it'g7 23 gxh6 'ifxg6+ 24 i.xg6
and White is clearly better.
1 8 J:tae 1 j_f8 1 9 g4! t2Jd6
19 ...tt:Jh4 20 tt:Jxh4 'ifxh4 21 .l::!.xe6! is
a recurrent theme which runs
throughout this game: 21...fxe6
(2 1.. ..l::!.xe6 22 nxe6 fxe6 23 'ifxg6+
.5lg7 24 'ifxe6+) 22 'ifxg6+ .5lg7 23
.l::!.xe6 .l::!.f8 2 4 'ifh7+ 'i!;if7 2S l:.xh6 is the
end for Black.
20 t2Jc5 t2Je5 2 1 J:txe5 J:txc5
22 j_e3! t2Jc4 23 j_d4 t2Jxe5
Read gives two other tries which
lose quickly:
a) 23 ...tt:Jxb2 24 'ife2 'ife7 2S i.xg6
fxg6 26 i.xcS 'ifxcS 27 .l::!.xe6.
b) 23 ... 'ife7 24 .l::!. xe6 fxe6 2S 'ifxg6+
.5lg7 26 i.xcS.
24 t2Jxe5 j_c8
24 ...'ifh4 once again allows 2S
tt:Jxg6!, e.g. 2S ... .tfS 26 .l::!.xe8! i.xd3 27
.l::!.xf8+ 'i!;ih7 28 .l::!.xf7+ 'i!;ig8 29 .l::!.g7
mate. After 24 ....td7 White can play
2S .l::!.e 3!, as 2S ... .tbs allows 26 tt:Jd7!
'ifd6 27 .l::!.xe8 i.xd3 28 tt:Jxf8 f6 29
i.xcS and White has far too much ma-
c s 4 e x d S e x d S : Main Line with 6 i. b S
terial for the queen.
25 J:te3 J:txe5
Giving back the exchange, but
other moves are no better:
a) 2S ...'ifh4 26 tt:Jxg6 .l::!.xe3 27 tt:Je7+.
b) 2S ...'ifgS 26 tt:Jxg6 'ifxg6 (26 ... fxg6
27 i.xcS and 26 .. ..l::!.xe3 27 lDe7+ i.xe7
28 'ifh7+ 'i!;if8 29 'ifh8+ 'it'g8 30 .tg7+
both win for White) 27 .l::!.xe8 .l::!.c6
(27...'ifxd3 28 .l::!.xf8+!) 28 'ifxg6+ fxg6
29 .5la4 .l::!.c7 (or 29 ....l::!.c4 30 .tbs .l::!.c7
31 .teS) 30 .tes :c4 31 .tbs .l::!.cS 32
i.d6 and White wins (Read) .
26 J:txe5 'iff4 27 J:te8! 1 -0
At first sight this seems like an early
resignation, but in fact Black is com­
pletely lost here. 27 ... .l::!.c6 runs into 28
i.e3 and 29 i.xh6, while 27 ... 'ifcl+ 28
'i!;ig2 .l::!. c 7 29 i.e3 'ife l 30 'ifxdS is also
winning. A model game in which
Read cleverly mixed a kingside attack
with the usual play against the IQP.
Game 57
Asrian-Lputian
Yerevan 1995
1 e4 e6 2 d4 d5 3 t2Jd2 c5 4 exd5
exd5 5 j_b5+ t2Jc6 6 l2Jgf3 j_d6 7
dxc5 j_xc5 8 0-0 tbge 7 9 t2Jb3 j_d6
101
The F re n c h Ta rra s c h
10 J:t e 1 0-0 1 1 i.d3 h6 1 2 h 3 t2Jf5
1 3 c3 'iff6 1 4 i.c2 l:td8 1 5 'ifd3 g6
1 6 'it'd2 ! ?
This very unusual concept of block­
ing the c 1-bishop with the queen has
in fact been witnessed before in this
variation (see Game 58 for the same
idea in a slightly different position) . In
contrast the idea of i.d2, followed by
doubling on the e-file is not so effec­
tive here, e.g. 16 i.d2 aS! 17 a4 b6!
(cutting across White's plan as 18 .l::!.e 2
is hit by 18 ... i.a6) 18 'iie 2 i.a6 19 i.d3
i.xd3 20 'iixd3 tt:Je5 21 tt:Jxe5 (2 1
'iixd5? falls for 2 1 ...ltJd3! 22 'iixd3
i.h2+) 2 1 ...i.xe5 22 .l::!. adl d4 23 cxd4
tt:Jxd4 24 i.c3 tt:Jc6 25 i.xe5 lDxe5 26
'iic 3 .l::!.x dl 27 .l::!.x dl .l::!. d 8 28 .l::!.xd8+
'iixd8 and here the players agreed a
draw in l.Gurevich-Lputian, Philadel­
phia World Open 1994, as after 29
'iixe5 Black regains the piece with
29 ...'iid l+ 30 �h2 'iix b3.
1 6 . . .i.f8 1 7 'iff4
After 1 7 tt:Jh2 'iig7! (prophylaxis
against tt:Jg4) 1 8 ltJg4 h5 19 tt:Je3 (19
i.xf5 i.xf5 20 tt:Jh6+ 'iixh6 21 'iixh6
i.xh6 22 i.xh6 d4 is equal according
to V.lvanov) 19 ...tt:Jfe7 20 'iie2 'iif6
Black had no real problems in
1 02
V.lvanov-Lastin, Russia 1994. He will
follow up with ... i.g7, ... i.e6 and
.. . d5-d4.
1 7 . . .i.g7 1 8 'it'c7??
This move is simply too ambitious.
White should have simply played 18
i.d2, or perhaps the visual 18 'iih 2!?,
intending .tf4, keeping absolute con­
trol over the important h2-b8 diago­
nal.
1 8 . . . t2Jd6!
The simplest ideas are usually the
best! Black just threatens to trap the
queen with ....l::!.d7. To avoid this
White has to give up a vital pawn.
1 9 t2Jc5 i.xh3! 20 gxh3
20 tt:Jxb7 loses to 20 ... .l::!.ac8, as in­
deed it does on move 2 1 .
2 0. . . 'ifxf3 2 1 J:te3 'ifh5
I doubt whether this was the posi­
tion that White had envisaged before
playing 18 'iic 7. White is a pawn
down, his king has no shelter and his
queen has nothing to show for her
'walkabout'. Someone of Lputian's
undoubted class had no difficulty put­
ting this one away.
22 t2Jd7 t2Jf5 23 i.xf5 'ifxf5 24 t2Jc5
d4! 25 cxd4 t2Jxd4 26 t2Je4 t2Jf3+ 27
<it>g2 l2Jh4+ 28 � 1
 3 . . .
Or 2 8 \t> h l .l::!.d l+ 2 9 \t>h2 i.e5+.
28 . . . 'ifb5+ 0-1
It is checkmate in all lines. 29 \t>gl
l:!dl+ 30 \t>h2 .l::!.h l+ 31 \t>xh l 'iif l+ 32
\t>h2 'iig 2 is one possible finish.
Game 58
Smagin-B. La lie
Sochi 1987
1 e4 e6 2 d4 d5 3 t2Jd2 c5 4 l2Jgf3
t2Jc6 5 exd5 exd5 6 .i.b5 .i.d6 7 dxc5
.i.xc5 8 0-0 tbge 7 9 t2Jb3 .i.b6 1 0
J:te 1 0-0 1 1 .i.d3 h 6 1 2 h 3 t2Jf5 1 3
c3 .i.c7
The other main approach. Black
mimics his opponent by piling up on
the b8-h2 diagonal. It is White, how­
ever, who creates the initial threats.
1 4 .i.c2 'it'd6 1 5 'ifd3 g6 1 6 'it'd2!?
Once more we see this queen move.
On first impression the alternative 16
g4 looks decisive, as if the knight re­
treats then White simply gobbles the
h6-pawn. However, Black has the sur­
prising response 16 ... i.b6! (threatening
...'iig3+) 17 \t>g2 i.xf2! 1 8 \t>xf2 'iig3+
19 \t>e2 ltJd6! 20 i.xh6 .l::!. e 8+ 21 \t>dl
'iixh3 with extreme complications.
1 6 . . . h5 1 7 .i.xf5 .i.xf5 1 8 'ifh6 J::tfe8
c S 4 e x d S e x d S : Main L ine with 6 i. b S
18 ... i.e4 19 ltJbd2 f5 was suggested
as an improvement on Black's play
but it now appears that there is a ma­
jor flaw in this move. Here are two
practical examples:
a) 20 g3? l:U6 2 1 lDxe4 fxe4 22 ltJh4
lDe5 23 i.f4 'iie 6 24 i.xe5 i.xe5 25
'iig5 i.c7 26 .l::!.e2 .l::!.af8 27 lDg2 'iif7 28
'iid2 i.b6 29 ftJf4 h4 and Black went
on to win in T olnai-Schmittdiel,
Dortmund 1989.
b) 20 lDxe4 fxe4 21 ltJg5 'iih2+ 22
@fl l:.xf2+ (this was previously
thought to give Black a perpetual
check but White has an amazing re­
source) 23 \t>xf2 'iig3+ 24 @fl! (24
\t>e2 'iid3+ 25 \t>f2 'iig3+ was the sup­
posed draw) 24 ... l:.f8+ 25 i.f4!! .l::!.xf4+
(25 ...'iixf4+ 26 \t>e2 'iif2+ 27 \t>dl and
Black runs out of checks) 26 \t>gl lDe5
27 'iih7+ \t>f8 28 lDe6+ \t>e8 29 lDxc7+
\t>d8 30 'ii'h 8+ \t>d7 3 1 'iie 8+ 1-0
Holzke-Tondivar, Groningen 1997.
1 9 .i.e3 'iff8 20 'ifxf8+ <it>xf8 21
J:tad1 J:tad8
Now White wins a pawn with a
small combination. If instead 21.. .i.e6
22 lDc5 keeps a small advantage.
22 J:txd5 J:txd5 23 .i.h6+ <it>g8 24
J::txe8+ <it>h7 25 .i.d2 .i.b1 ?
103
The Fre n c h Ta rra s c h
After 25...tt:JeS! 2 6 lDxeS :IxeS 27
l:t.xeS i.xeS it would have been very
difficult for White to convert the extra
pawn, as Black would boast the bishop
pair as compensation. After the game
move White can favourably give back
the extra pawn in order to drum up a
direct attack against the black king.
26 c4 l::tf 5 27 j_c3 g5
27 .. .f6 28 tt:Jbd4 lDxd4 29 l:t.e7 + \tg8
30 tt:Jxd4 :I.cs 3 1 lDe6 is also an easy
win. Black just cannot co-ordinate his
pieces for a successful defence.
28 t2Jbd2 j_xa2 29 t2Je4 j_d8 30 .!!Jd6
J:tf4 31 j_d2
31 . . . J:txf3
3 1.. .:f6 32 ltJxgS+ \tg6 33 .l::!.g8+
\th6 34 tt:Je8! is also hopeless.
32 gxf3 j_97 33 t2Jxb7 j_f6 34 j_c3
j_xc3 35 bxc3 j_xc4 36 t2Jd8 j_b5
37 t2Jxc6 j_xc6 38 l:te7 <it>g6 39 J:txa7
j_xf3 40 c4 'itf6 41 r.Ia3 j_b 7 42
l::t e3 h4 43 c5 j_d5 44 <it>h2 'itf5 45
l::t e 8 j_b7 46 J:tg8 1 -0
Game 59
Agnos-Lputian
Linares 1996
1 e4 e6 2 d4 d5 3 t2Jd2 c5 4 exd5
1 04
exd5 5 l2Jgf3 t2Jc6 6 j_b5 j_d6 7
dxc5 j_xc5 8 0-0 l2Jge7 9 t2Jb3 j_d6
1 0 l:te1 0-0 1 1 j_g5
Karpov's positional approach,
which he introduced against Korchnoi
in their 1 974 match. White immedi­
ately activates his dark-squared bishop,
before it can be blunted with ... h7-h6.
In some lines White simply exchanges
on el. The main idea, however, is the
manoeuvre i.h4-g3, swapping the
dark-squared bishops. Once again in
the Tarrasch, this plan, though quite
time-consuming, presents Black with
some difficult problems to solve. Ac­
tive piece play offers the only chance
for him to achieve equality.
1 1 . . _j_g4 1 2 j_h4 J:te8 1 3 j_g3 j_xg3
1 4 hxg3 'ifb6!
A good move. Once the dark­
squared bishops are exchanged, the
queen sits very nicely on this square.
Often Black follows up with ... a7-a5,
planning ... a5-a4, which may loosen
White's control over the d4-square.
1 5 a4 h5!?
15 ...a6 16 i.xc6 i.xf3 17 'ifxf3 bxc6
18 'i'e3 probably keeps an edge for
White, who controls many dark
squares on the queenside.
 3 . . . c S 4 e xdS e xd S : M a in L ine with 6 i. b S

1 6 'ir'd3 j_f5 1 7 'ir'd2 a6 should be considered, especially as the

Possibly this was an intended im­ a5-pawn will not run away.
provement over the earlier move 2 1 . . . t2Jxa5 22 t2Je5 'it>g8 23 'ir'f4 f6
17 ... i.e4, which was played in lvan­ 24 t2Jd7 'ir'd8
chuk-Vaganian, Novgorod 1995. That The greedy 24 ...'ifxb2 runs into a
game continued 1 8 tt:Jfd4 tt:Jxd4 19 brilliant combination: 25 .l::!.xa5 'ifxc3
'ifxd4 tt:Jc6 20 'ifxb6 axb6 2 1 l:.e2 \t>f8 26 nxe4 dxe4 27 tt:Jxf6+ gxf6 28 .l::!.xh5
22 .l::!.d l i.f5 23 .l::!.ed2 .l::!.ed8 24 f3 g6 25 'ifxd4
\t>f2 \t>e7 26 c3 i.e6 27 tt:Jd4 tt:Jxd4 28
.l::!.xd4 .l::!.d c8 29 .l::!.b4 .l::!.c5 30 i.d3 lk6
3 1 i.c2 .l::!. a5 32 i.b3 and White's edge
had persisted. It has to be said that
17 ... a6 doesn't really alter the general
assessment of the position, which is a
little better for White.
1 8 j_f 1 J:tac8 1 9 t2Jbd4 j_94 20 c3
'it>f8 ! ?
This provokes White into some real
action. 20 ...tt:Jxd4 is an alternative,
though after 2 1 tt:Jxd4 tt:Jc6 22 lDxc6
bxc6 23 a5 White fixes the a6-pawn, 29 i.c4+!! .l::!.xc4 30 'i'g4+ \t>f8 3 1
which becomes a weakness. .l::!. h8+ tt:Jg8 (3 1...\t>f7 32 l:!h7+ @f8 33
21 a5!? 'it'g7 mate) 32 'ifxg8+ \t>e7 33 'ifxe8+
\t>d6 34 'ifb8+ \t>c6 35 .l::!. d 8 'i'al+ 36
\t>h2 \t>b6 37 .l::!. d7! and it is unlikely
that the black king will survive very
long. Even after the sober 24 ... 'ifd8 the
attack persists.
25 t2Jxf6+! gxf6 26 J:txe4 dxe4 27
'ir'xf6 t2Jec6 28 'ir'g6+ 'itf8 29 t2Jf5

An inventive pawn sacrifice, which

is hardly typical of this line. Given
that it is difficult to foresee that
White's attack will quickly gather
momentum, one can hardly blame
Lputian for grabbing the pawn. Nev­
ertheless, the super-safe 21...'ifc7

105
The F re n c h Tarra s c h
White is a rook down, but the
threat of b2-b4 followed by i.c4 is
extremely difficult to meet. The queen
and knight once more prove to be an
irresistible attacking force. If now
29 ...'illcl 30 b4 tbe5 3 1 'i!lf6+ \tg8 32
.l::!.xa5 the white rook also enters the
attack and Black cannot set up a de­
fence. The move chosen by Lputian
also fails to stem the flow.
29 . . . J:tc7 30 b4 J:tf7 31 'ifh6+ <it>g8
32 'ifxh5 J:te6 33 bxa5 'iff6 34 j_c4
J:tfe7 35 t2Jxe7+ 1 -0
A powerful attacking display by
Agnos.
Game 60
Emms-Prandstetter
Barcelona 1993
1 e4 e6 2 d4 d 5 3 t2Jd2 c5 4 exd5
exd5 5 l2Jgf3 t2Jc6 6 j_b5 j_d6 7 0-0
tbge7 8 dxc5 j_xc5 9 ll'ib3 j_d6 1 0
t2Jbd4
This position can also arise after
7 ...cxd4 8 tbb3 tbge7 9 tbbxd4 (with
one tempo less for either side). This
variation was given a thorough testing
in the 1974 Karpov-Korchnoi match,
the general verdict being that White's
advantage was just too minute to cause
his opponent serious trouble.
1 0 . . . 0-0 1 1 c3 j_g4 1 2 'ifa4 'it'd7
12 ...i.h5 is a fully playable alterna­
tive. Howell-Psakhis, Bled 1995, con­
tinued 13 i.e2 a6 14 i.e3 'i!lc7 15 h3
tba5! 16 .l::!.a dl l:t.ad8 17 l:tfe l h6 18
tbh4 i.xe2 19 l:txe2 tbc4 (once the
knight reaches this square, Black has
often solved his difficulties) 20 tbdf5
b5 2 1 'i!lc2 (2 1 'i!lxa6? l:ta8 22 'i!lxb5
l:tfb8 23 tbxd6 l:txb5 24 tbxb5 'i!lb7 is
106
better for Black) 21...tbxf5 22 tbxf5
nfe8 23 i.d4 i.f8 24 .l::!.de l 'i!ld7 25 b3
tbd6 and the players agreed a draw.
This is a good example of accurate
play by Black. In the final position
White has absolutely nothing .
1 3 J:te 1
13 i.e3 is probably more precise.
Then after 13 ... a6 14 i.e2 tbxd4 15
'i!lxd4 tbc6 the queen can retreat to d2.
That said, following 16 'i!ld2 .l::!. fe8 17
.l::!.adl .l::!.ad8 18 i.b6 i.c7 19 i.xc7
'i!lxc7 20 l:t.fe 1 h6 White had only an
infinitesimal edge in Karpov-Korch­
noi, Moscow (game 16) 1974.
1 3 . . . a6 14 j_e2 t2Jxd4! 1 5 'it'xd4 t2Jc6
1 6 'ifd 1 J:tfe8
An exchange of minor pieces often
favours the side battling against the
isolated pawn, as the weakness of the
IQP becomes more pronounced as the
endgame approaches. Here, however,
Black is perfectly co-ordinated and
even has a slight development advan­
tage.
1 7 j_93 J:tad8 1 8 'ifc2 h6 1 9 J:tad 1
'it'c7 20 h3 j_h5 2 1 <it>f1 ? !
A ludicrous attempt to centralise
the king for the endgame! Black could
have exploited this immediately by
 3 . . .
2 1 ..Jhe3! 2 2 fxe3 i.g3, when White
has to start thinking about how to
equalise.
21 . . .j_f4 22 'ir'f5 ! ? j_xe3 23 'ir'xh5
i.87 24 j_d3
24 .l::!.xd5 allows 24 ... 'ifb6! 25 ltJd4
'ifxb2, when I prefer Black.
24 . . .'ir'b6?!
This leads to problems. 24 ... d4! 25
cxd4 ltJxd4 26 l:t.xe8+ l:t.xe8 27 ltJxd4
i.xd4 is the simplest way to draw.
25 l:txe8+ l:txe8 26 l:td2 d4 27 cxd4!
Black was hoping for 27 ltJxd4
ltJxd4 28 cxd4 'ifb4. Around this stage
I realised that I actually had some real
winning chances.
27 . . . 'ir'c7 28 J:tc2 'ir'f4 29 'ir'g4! 'ir'xg4
30 hxg4 t2Jb4
The best move. 30 ... ltJxd4 3 1 .l::!.c7!
.i::t b 8 32 i.c4 is clearly better for
White, while after 30 ... i.xd4 White
doesn't play 3 1 .l::!.xc6 bxc6 32 ltJxd4
.l::!.d8, but instead 3 1 i.xa6 ltJb4 32 .l::!.c4
bxa6 33 .l::!.xb4.
31 J:td2?!
As both players were entering into
big time-trouble, the inevitable errors
creep in. 3 1 .l::!. c3 is stronger here, as
31...ctJxa2 can then be met by 32 .l::!.c7.
Now 3 1 ...ctJxa2 should have been
c S 4 e x d S e x d S : Main Line with 6 i. b S
tried.
31 . . . 96?! 32 a3 t2Jxd3 33 l::txd3 J:te4
34 J:tc3 b5?! 35 d5 b4? 36 J:tc7 j_b6
37 J:tc6 j_d4
This drops a piece, but Black's posi­
tion was already hopeless, e.g.
37 ... i.d8 38 .l::!.c8 l:t.e8 39 d6 bxa3 40
bxa3 'itig7 41 d7 .l::!.h8 42 ctJe5.
38 J:tc4 bxa3 39 bxa3 1 -0
A somewhat scrappy game, and a
rather fortuitous victory for White.
Game 61
Jansa-Votava
Prague 1993
1 e4 e6 2 d4 d5 3 t2Jd2 c5 4 l2Jgf3
t2Jc6 5 exd5 exd5 6 .ib5 .id6 7 dxc5
j_xc5 8 0-0 tbge 7 9 t2Jb3 j_b6
A common alternative to the usual
9 ...i.d6.
10 J:te 1 !
The plan involving l:tel and i.e3 is
by far the most popular idea for
White. If truth be told, most Black
players shy away from 9 ... i.b6 .for
precisely this reason. Other ideas pale
by comparison. For example, 10 ltJbd4
0-0 1 1 c3 i.g4 12 i.e2 ctJxd4 13 ltJxd4
i.xe2 14 ctJxe2 ctJc6 15 ltJf4 d4 16 cxd4
i.xd4 gave Black no problems at all in
Renet-Korchnoi, Paris (rapidplay)
1990.
10 ...0-0 11 j_e3 j_g4 1 2 j_xb6 axb6
The sharper 12 ...'ifxb6 will be ex­
amined in the next game. There is
nothing really to this position apart
from Black's dodgy pawn structure,
which gives White a small but obvious
advantage.
1 3 j_e2 'ii'd6 1 4 c3 l:tfd8 1 5 t2Jfd4
j_xe2 1 6 J:txe2 l2Jg6 1 7 t2Jf5 'ir'f6 1 8
107
The Fre nch Ta rra s c h
t2Jbd4 t2Jce5 1 9 g3 t2Jc4 2 0 'ir'c2
l2Jge5 21 b3 t2Ja3?! 22 'ir'd2 t2Jc6 23
J:tae 1 <it>h8 24 'WWd 3
Black has done his best to create
complications, but to no avail. The
knight on a3 is out of the game and
White has control over the all­
important e-file.
24 . . . 96 25 l2Jh6 t2Jxd4 26 cxd4 'ir'g7
27 l2Jg4 h5 28 t2Je5 l:tac8 29 t2Jf3
'ir'f6 30 <it>g2 <it>g7 3 1 J:te7
31 . . J:tc2
This leads to a rather abrupt end.
3 1 .. . .l::!.d6 32 .l::!.xb7 :c2 has been sug­
gested an improvement, but White can
also win here with 33 b4 .l:.dc6 34
.l::!.ee7! .l::!. 6 c3 35 .l::!.xf7+ 'ifxf7 36 l:t.xf7+
\t>xf7 37 tt:Je5+.
32 .l::. 1 e6 ! 1 -0
A nice finish. Both 32 ...'ifxe6 33
nxe6 fxe6 34 tt:Je5 \t>f8 35 'ifxg6 and
32 ...'iff5 33 'ifxf5 gxf5 34 :xb6 lha2
35 tt:Jh4 are hopeless for Black.
Game 62
Dvoirys-Tondivar
Leeuwarden 1994
1 e4 e6 2 d4 d5 3 t2Jd2 c5 4 exd5
exd5 5 l2Jgf3 t2Jc6 6 j_b5 j_d6 7
1 08
dxc5 j_xc5 8 t2Jb3 �6 9 0-0 tbge 7
1 0 :e 1 0-0 1 1 j_93 j_g4 1 2 j_xb6
'ir'xb6
After 12 ...'ifxb6 it is White who
winds up with doubled pawns, but
Black has to sacrifice one of his own
for this privilege.
1 3 j_xc6 t2Jxc6 1 4 'ir'xd5 t2Jb4 1 5
'ir'e4 j_xf3 1 6 gxf3
Black hopes that his opponent's
wrecked kingside pawn structure will
be enough compensation for the
pawn. However, White's pieces be­
come extremely active over the next
few moves, rendering these weak­
nesses insignificant. This line has very
few takers at master level.
1 6 . . . l:tad8 1 7 J:tad 1
This temporary sacrifice of the a­
pawn is probably stronger than 17
.l::!.e2 .l::!.d6!? 18 c3 .l::!.e6 19 'ifc4 .l::!.g6+ 20
\t>hl 'iff6 21 f4 'ifh4 22 tt:Jd2, which
occurred in lvanchuk-Dokhoian, Yer­
evan 1989 . In this position 22 ... 'ifh3!
has been suggested as a way to keep
things going for Black, e.g. 23 f3 tt:Jc6
24 l:t.gl .l::!.d8 and White is certainly not
in complete control.
1 7 . . .J:txd 1 1 8 J:txd 1 t2Jxa2 1 9 J:td7
'ir'f6 20 t2Jd4
 3 . . .
Tondivar was actually repeating one
of his previous games, which had gone
20 l:t.xb7 'ifxb2 21 l:t.xa7 h5 22 \t>g2
ltJcl 23 ltJd4 'ifc3 24 'iff4 'ife l 25 'ife3
.l::!. e 8 26 'ifxel .l::!.xe l 27 %:!.al 1-0 Jansa­
Tondivar, Ostend 1993. No doubt he
had some improvement lined up, al­
though I can't see anything special. In
any case, Dvoirys got his new move in
first.
20 . . .'ir'g5+ 2 1 <it>f1 'ir'c 1 + 22 <it>g2
'ir'xb2
Now that the king is more strongly
placed on g2, White would have an­
swered 22 ...'ifg5+ with 23 'ifg4.
23 �xb7 'ir'c3 24 l::tx a7 lll b4 25 J:ta4!
White has kept the extra pawn, and
now initiates a surprising attack which
is immediately decisive.
25 . . . 'ir'c5 26 lllf5
26 . . .J:tbS
Other moves also lose:
a) 26 ...ltJc6 27 nc4.
b) 26...ftJds 27 'ifxd5 'ifxds 28 lDe7+.
c) 26 . . .fDxc2 allows the prettiest fin-
ish, as 27 lDe7 + \ti h8 28 'ifxh7 +! \t>xh7
29 l:t.h4+ leads to mate.
27 c3!
27 .l::!. a 8 allows the defence 27 ... 'iff8,
so White deflects the queen. Black has
c s 4 e xdS e xd S : Main L ine with 6 i. b S
no choice as the knight has no useful
retreat square.
27 . . .'ir'xc3 28 J:ta8 1 -0
Since 28 ...'ifc7 runs into 29 'if e8+. A
powerful performance from Dvoirys.
The onus is on Black to come up with
a substantial improvement here.
Game 63
T. Ravi-Dolmatov
Calcutta 1996
1 e4 e6 2 d4 d5 3 llld2 c5 4 lllgf3
lllc6 5 exd5 exd5 6 j_b5 cxd4
This is a clever attempt to steer the
game into lDbd4 lines, which would
arise after 7 0-0 .td6 8 lDb3 lDge7 9
lDbxd4. At the moment it is not im­
mediately obvious how White should
best avoid this transposition.
7 'ii'e2+
After 7 lDxd4 .td7 8 'ife2+ 'ife7 9
lD2b3 'ifxe2+ 10 lDxe2 tDf6 1 1 i.e3 a6
12 i.xc6 i.xc6 13 0-0-0 0-0-0 14 .td4
ltJd7 15 ltJg3 .l::!.g8 16 f3 g6 17 fDe2 .td6
the position was equal in Brodsky­
Dolmatov, Novgorod 1995. The d­
pawn is difficult to attack in the ab­
sence of a light-squared bishop.
7 . . .'ir'e7
109
The Fre n c h Ta rra s c h

8 t2Jxd4 change the character of the position.

8 tt:Je5!? leads to some crazy chess 1 6 . . .�e6 1 7 cxd5+ t2Jxd5
after 8 ....td7 9 i.xc6 bxc6 10 0-0
\td8!? (preparing .. .f7-f6) 1 1 tt:Jb3 f6 12
i.g5!? fxg5 13 tt:Ja5 'ife8! The first time
this bizarre position occurred was in
the game Ivanchuk-Dolmatov, Manila
Interzonal 1990, where the players
called it off here and agreed a draw!
Five years later Dolmatov was brave
enough to play on with his extra piece
and was rewarded after 14 .l::!.fel \tc7!?
15 'iff3 i.b4 16 'it'g3 i.d6 17 c4 tt:Jf6!
1 8 tt:Jaxc6?! (18 c5 is the only chance)
1 8 ...tt:Je4! 19 l:t.xe4 dxe4 20 c5 i.xe5 2 1 The minimal advantage of the
lDxe5 \tc8! 22 tt:Jc4 'ife7 23 tt:Jd6+ \td8 bishop against the knight in an open
24 'ifb3 i.c6 25 'ifc4 'iff6 0-1 Svidler­ position means that White can play
Dolmatov, Haifa 1995. White needs on. Indeed White keeps a small edge
something new here, otherwise the all the way to the final position, but
piece sacrifice looks rather dubious. Ravi decided not to test Dolmatov's
8 . . . 'it'xe2+ 9 <it>xe2 j_d7 1 0 t2J2f3 undoubted technique.
j_c5 1 1 j_93 t2Jxd4+ 1 2 t2Jxd4 j_xd4 1 8 l::td2 l::thd8 1 9 <it>f3 t2Jb4 20 a3
1 3 j_xd7 + <it>xd7 1 4 j_xd4 f6 1 5 t2Jc6 21 l::te 1 + �7 22 J:ted 1 J:tac8 23
J:thd 1 tbe 7 1 6 c4 j_c3 l::t xd2 24 J:txd2 <it>e6 25 <it>e4 h5
16 i.c5 .l::!.ac8 17 i.xe7 \txe7 18 c3 26 f4 t2Je7 27 g3 t2Jf5 28 <it>f3 l::tc 7
l:t.c5 is dead level, so White tries to 29 h3 t2Jd6 � - �

1 10
 3 . . . c S 4 e x d S e x d S : Main L ine with 6 i. b S

Summary
In the main line with 9 ... .td6 it seems that White has reasonable chances for an
advantage with either 1 1 .td3 or the older 1 1 .tgS. 9 ....tb6 still has a poor repu­
tation and this line needs some major surgery before it becomes playable. If
Black really has to capture with 12 ... axb6 (as in Game 61) then I suspect we
won't see much of this line in future. The variation with 6 ... cxd4 (Game 63) cer­
tainly merits further investigation, but even here Black players may have to ac­
cept that White can always make an easy draw.

1 e4 e6 2 d4 d5 3 t2Jd2 c5 4 exd5 exd5 5 l2Jgf3 t2Jc6 6 i.b5 j_d6

6 ... cxd4 Game 63
-

7 dxc5
7 0-0 cxd4 8 tt:Jb3 0-0 9 tt:Jbxd4 - Game 60 (by transposition)
7 . . .j_xc5 8 0-0 l2Jge7 9 t2Jb3 j_d6
9 ... .tb6 10 �e l 0-0 1 1 .5le3 .5lg4 12 .5lxb6 (D)
12 ...axb6 Game 61
-

12 ...'ifxb6 Game 62
-

10 �e1
10 tt:Jbd4 - Game 60
1 0 ... 0-0 1 1 j_d3 (D)
1 1 .tgS - Game 59
11 ... h6 1 2 h3 t2Jf5 1 3 c3 'it'f6
13 ... .tc7 Game 58
-

14 j_c2 j_e6 (D)

14....l::!. d8 Game 57
-

15 'it'd3 Game 56
-

12 j_xb6 1 1 j_d3 14 . j_e6

. .

111
 CHAPTER EIGHT I
3 cs 4 exdS exdS:
. . .

Fifth M ove Alternatives

1 e4 e6 2 d4 d5 3 tll d 2 c5 4 exd5 possibilities, but there are also some

exd5 significant independent lines too, as
In the last decade 5 lt'igf3 lt'ic6 has we shall see in Games 71-73.
gradually been overtaken in the popu­
larity stakes by the even more reliable Game 64
5 ...lt'if6. Old references had discarded Tiviakov-Dreev
this natural move (for no particular Wijk aan Zee 1996
reason), but top players such as the
super-solid Russian GMs Evgeny 1 e4 e6 2 d4 d5 3 tlld2 c5 4 exd5
Bareev and Alexei Dreev have now exd5 5 tll gf3 tllf6
shown that Black can obtain a fully
playable game with this move, with
results that reflect their convictions. In
Games 64-67 we look at various at­
tempts for White to squeeze out an
opening edge against 5 ... lt'if6.
5 ... a6 (Games 68-70) is a more risky
and ambitious attempt for Black, and
is ideally suited to those going all or
for the win. White can either capture
on cS immediately, or invite Black to
clamp on the queenside with ... c5-c4.
Note that the position after 5 ... a6 is 6 .ib5+
often reached via the move order 1 e4 6 .lie2 is seen from time to time, but
e6 2 d4 dS 3 lt'id2 a6 4 lt'igf3 cS 5 exdS the general view is that it gives Black
exdS. less problems than the main line. One
5 i.bS+ has various transpositional recent example continued 6 ...lt'ic6 7

1 12
 3 . . .
0-0 ile7 8 dxc5 ilxc5 9 tt:Jb3 il..b6 10
il..g5 0-0 11 c3 l:t.e8 12 il..h4 h 6 13 l:t.el
g5! 14 il..g3 tt:Je4 15 tt:Jfd4 f5! 16 Ji..h5
l:t.f8 and Black had a very active posi­
tion in Onischuk-Kramnik, Tilburg
1997.
6 . . .j_d7 7 j_xd7+ t2Jbxd7 8 0-0 j_97
9 dxc5 t2Jxc5 1 0 J:te 1
The start of a logical plan which
was introduced by the Swiss GM Lu­
cas Brunner. White's idea is quite
clear: He plans tt:Jfl and il..e3-d4, fol­
lowed by tt:Je3 and possibly 'ifd3 and
l:t.adl. The white pieces would then be
very harmoniously placed and the d5-
pawn would also come under pres­
sure.
10 tt:Jd4 and 10 tt:Jb3 are considered
in Games 66 and 67 respectively.
1 0 . . . 0-0 1 1 t2Jf1 l::te 8!
The stem game Brunner-Atalik,
Kecskemet 199 1, varied with
1 1 ...tt:Jce4?! 12 il..e 3 .l::!.c 8 13 il..d4 il..d6
14 lDe3 il..b8 15 'ifd3! l1e8 16 .l::!. adl
'ifd6, and now instead of 17 g3?!
White should have just played 17
il..xf6! tt:Jxf6 18 c4 with a big advan­
tage, as the d-pawn drops. It was
hardly surprising that the idea begin­
ning with 10 .l::!.e 1 became popular for
c s 4 e x d S e xd S : Fifth Mo ve A lt e rn a ti v e s
White after this game, although
Black's play was far from inspiring.
1 2 j_93 t2Jfe4
Apart from the challenging 12 ... a4,
which is the subject of the next game,
Black has also tried:
a) 12 ... b5 13 il..d4 tt:Je6 14 il..e5 tt:Jg4
15 tt:Je3 lDxe3 16 .l::!.xe3 'ifd7 17 'i'd2
.i::!.ad8 18 l'!dl tt:Jc5 19 il..c3 Ji..f8 20
.l:txe8 'ifxe8 21 il.. d 4 tt:Je4 22 'ifaS and
Black's pawns were looking rather
vulnerable in Yurtaev-Ulibin, Russia
1997.
b) 12 ... a6 (too slow) 13 c3 l:tc8 14
'ifc2! 'ifc7 15 .l::!.a dl b5 16 il..d4 and
once more White has reached an ideal
set-up, as in Kasparov-Short, London
(rapidplay) 1993.
1 3 c3 'ifd6 14 l2Jg3 j_f8 1 5 'ifc2
'ifa6 1 6 J:ted 1 J:tad8 1 7 l2Jg5! t2Jxg5
After 17 ...'ifg6 18 lD5xe4 dxe4 19
.l::!.xd8 l:t.xd8 20 ilxc5 ilxc5 White
doesn't play 21 lDxe4? because of
21...l:t.e8 22 l:t.el f5 winning a piece,
but 21 'ifxe4!, which leaves him a clear
pawn up.
1 8 j_xg5 J:td7 1 9 j_93 b6 20 'iff5
g6?!
This concedes a slight weakness on
f 6, which White is able to utilise.
20 ...'ifb7 would have kept White's
plus to a minimum.
21 'it'g4 t2Je6 22 j_d4! 'it'b5 23 l2Jh5
j_97 24 j_f6 d4?
24 ... il..f8 would have been a better
defence, although after 25 ile5 ile7 26
.l::!. d2 the weaknesses around the black
king still give White a clear advantage.
After 24...d4? White reaches a very
pleasant endgame.
25 j_xe7 'ifxh5 26 'ifxh5 gxh5 27
j_f6 d3 28 J:td2 t2Jf4 29 <it>f1 l:.e2 30
1 13
Th e Fre n c h Ta rra s c h

J:tad 1

For the first 30 moves White's play Game 65

has been quite exemplary. It might l.Gurevich-Bareev
seem that Black is quite active here, Hastings 1992
but appearances can be a little decep­
tive. The d3-pawn is in fact a real de­ 1 e4 e6 2 d4 d5 3 t2Jd2 c5 4 exd5
fect in this position, and should be exd5 5 l2Jgf3 t2Jf6 6 i.b5+ j_d7 7
picked off soon. If you add this to j_xd7+ t2Jbxd7 8 0-0 j_97 9 dxc5
Black's other weak pawns on the t2Jxc5 1 0 J:te 1 0-0 1 1 t2Jf 1 J:te8 1 2
kingside, White has virtually a win­ j_93 a5!?
ning position. Tiviakov's technique,
however, couldn't cope with Dreev's
gritty resistance and his bamboozling
knight moves. In fact, in the end he
was quite relieved to make a draw!
30 . . . t2Je6 3 1 g3 J:txd2 32 J:txd2 .l::.d5
33 <it>g2 h6 34 c4! ? J:td6 35 <it>f3 <it>h7
36 j_e5?
36 \t>e3 tt:Jcs 37 iLd4 l:.e6+ 38 \t>f3
lDe4 39 .l::!.xd3 gets rid of that annoying
d-pawn.
36 . . .t2Jg5+! 37 <it>f4 l:td7 38 h4 l2Jh3+
39 'iiff 3? tbg 1+ ! 40 <it>e4 J:te 7! 41 J:td 1 A very deep move indeed. Black is
l2Jh3 42 �f5 t2Jxf2 43 J:td2 l2Jg4 44 not merely content to centralise his
j_d4? t2Je3+ 45 'iiff4 t2Jxc4 46 J:txd3 pieces and await events. With 12 ... aS
J:td7 ! 47 b3 t2Ja3 48 J:td2 he aims to gain space on the queenside
There is still time for White to blow and possibly also weaken White's
it completely. 48 \t>e4?? lDc2, intend­ pawn structure.
ing .. .f7-fS+, wins for Black. 1 3 'ife2
48 . . . t2Jb1 49 l:td3 Yz -Yz After 13 iLd4 tt:Je6 14 ii.es ii.cs! 1S

1 14
 3 . . .
.l::!.e2 tl'ie4 16 tl'ie3 tl'i6g5 White has to
play with care just to equalise, e.g. 17
tl'ixgS (17 'ifxds?! i.xe3 1 8 'ifxd8
i.xf2+ 19 l:t.xf2 .l::!.axd8 20 l:t.e2 tl'ixf3+
2 1 gxf3 l:t.xeS 22 fxe4 fS wins a pawn)
17 ... 'ifxgS 1 8 i.d4 tl'ixf2 19 .l::!.xf2 i.xd4
20 'ifxd4 'ifxe3 2 1 'ifxdS .l::!.e7 22 l:t.dl
h6 and the players shook hands on a
draw in Rozentalis-M.Gurevich, Bel­
fort 1997.
1 3 . . .'it'd7 14 l2Jd4 a4 1 5 a3
White feels obliged to prevent the
possible ... a4-a3, but now Black ob­
tains a bind on the light squares in the
centre and on the queenside.
1 5 . . .t2Jce4 1 6 'it'b5 'ifc 7 ! 1 7 l:tad 1
tbd6 1 8 'it'd3 t2Jc4 1 9 j_c 1 j_c5 20
t2Jf5 t2Jg4 2 1 l2J5e3 j_xe3 22 fxe3
t2Jf6
Without making any obvious blun­
ders, White has managed to reach
quite a hideous position. Both c4 and
e4 are excellent outposts for the black
pieces, while White has been saddled
with weak pawns on e3 and b2. Given
the way things have gone so far, it is
quite a miracle that Gurevich survived
this position.
23 'it'd4 t2Jd6 24 'iff4 J:te4 25 'it'g3
J:tc4 26 c3 'it'b6 27 'iff3 t2Jde4 28
c S 4 e x d S e x d S : Fift h Mo ve A lternatives
J:td4 'ifc6 29 J:ted 1 J:te8 30 J:t 1 d3
l2Jg5 31 'it'd1 l2Jge4 32 J:txc4 dxc4 33
l::td4 b5 34 j_d2 t2Jc5 35 j_e 1 t2Jfe4
36 'iff3 'it'e6 37 l2Jg3 g6 38 t2Jxe4
t2Jxe4 39 j_h4 h6 40 h3 <it>g7 41 <it>h 1
'it'c6 42 j_d8 l::te6 43 <it>g 1 'it'e8 44
.l::.d 5 l2Jg5 45 'iff4
Black has lost control and no longer
possesses any real advantage. Now
45 ... .l::!.xe3 loses to 46 i.f6+ \t>h7 47
.l::!.d 8, so Black bails out.
45 . . .t2Je4 46 j_c7 Yz - Yz
Game 66
Karpov-Bareev
Linares 1994
1 e4 e6 2 d4 d5 3 t2Jd2 c5 4 exd5
exd5 5 t2Jgf3 t2Jf6 6 j_b5+ j_d7 7
j_xd7+ t2Jbxd7 8 0-0 j_97 9 dxc5
t2Jxc5 10 t2Jd4 'it'd7!
The queen often finds itself parked
on a light square in this line. Here it
patrols the sensitive fS-square, which
the knight on d4 was eyeing.
1 1 t2J2f3 0-0 1 2 j_f4
The main alternative is 12 tl'ieS 'ifc8!
and now:
a) 13 'iff3 .l::!. e8 14 tl'ifS tl'ice4 15 c3
and here not 15 ... i.d8?! 16 tl'ig4 .l::!.e6
1 15
The Fre n c h Tarra s c h
17 h3 with an edge t o White in Van
der Wiel-Korchnoi, Wijk aan Zee
1992, but either 1 5 . . . i.c5 or 15 ... i.f8,
both of which seem perfectly satisfac­
tory for Black.
b) 13 CL'id3 CL'ice4 14 i.f4 a6 15 l:!el
i.d6 16 c3 .l::!. e 8 17 i.xd6 CL'ixd6 18
.l::!.xe8+ 'ifxe8 19 'ifc2 'ifd7 20 %:!.e l l:!e8
with a dull and equal position in Ar­
nason-Dolmatov, Moscow 1990.
1 2 .. .J:tfe8 13 J:te 1
1 3 . . .j_fB
An early game in this line went
13 ... CL'ice4 14 CL'ie5 'ifd8 15 CL'id3! lk8 16
c3 i.f8 17 'ifb3 and the knights were
very nicely placed on d3 and d4, en­
suring White an edge in Adams-Short,
English Championship 199 1 .
1 4 t2Je5 'ir'a4!?
Naturally Black can also consider
14...'ifc8 and 14 ...'ifd8.
1 5 c3 'ir'a6 1 6 'ir'e2 'ir'xe2 1 7 J:txe2
j_d6
Allowing a neat trick, although it is
not clear that this changes the assess­
ment that much. White has a minus­
cule plus, which shouldn't really be
enough for victory.
1 8 t2Jd7! j_xf4
1 8.. Jhe2? 19 CL'ixf6+ gxf6 20 i.xd6
1 16
saddles Black with further worries on
the kingside, after which White would
have real winning chances.
1 9 J:txe8+ J:txe8 20 t2Jxc5 j_c7! 2 1
t2Jd3
Karpov suggests 21 @fl i.b6 22
lt:Jcb3 as a way of keeping a pull, but
once again it is nothing too serious for
Black.
21 . . . j_b6 22 t2Jb3 'ii/fa 23 l::td 1 a5 24
�1 J:tc8 25 t2Jd2 a4 26 a3 g5 27
tbt3 g4 28 l2Jh4 d4 29 cxd4 j_xd4
30 t2Jf5 j_b6 3 1 t2Jb4 t2Je4 32 f3 gxf3
33 gxf3 t2Jc5 34 h4 l::td8
This error is a prelude to a 'blunder
of the year' contender. After 34 ... i.d8!
35 tt:Jds lt:Je6 Black is very close to the
draw.
35 J:td5
35 . . .j_a7?? 36 J:txd8 mate
A shocking finish and a reminder
that it happens to the best of us! After
such a disaster it only seems fair to
include an instructive Bareev victory.
Game 67
l.Gurevich-Bareev
Biel 1993
1 e4 e6 2 d4 d5 3 t2Jd2 c5 4 exd5
 3 . . .
exd5 5 l2Jgf3 l2Jf6 6 .i.b5+ .i.d7 7
.i.xd7 + t2Jbxd7 8 0-0 .i.e 7 9 dxc5
t2Jxc5 1 0 t2Jb3 t2Jce4
It is not a good idea for Black to al­
low an exchange of these knights:
a) 10 ...tt:Jxb3?! 1 1 axb3 and the semi­
open a-file can only help White.
b) 10 ... 0-0 1 1 lDxc5 i.xc5 12 i.g5
.l::!. c 8 13 c3 l:tc6 14 tt:Je5 .l::!.e6 15 tt:Jg4
i.e7 16 tt:Je3 with a comfortable plus
for White in Tal-Benko, Skopje 1972.
1 1 .i.f4
Another plan is 1 1 tt:Jfd4 'it'd7 12
'it'f3, planning to plonk a piece on the
f5-square. White then has two options
after 12 . . 0-0:
. A brilliant move which exploits the
a) 13 'it'f5 g6! (Black doesn't fear a
queen exchange) 14 'it'xd7 tt:Jxd7 and
the d-pawn is quite safe.
b) 13 tt:Jf5 i.d8! 14 i.e3 g6 15 tt:Jg3
ne8! 16 c3 a5 17 tt:Jd4 .l::!.a6 18 l:.adl
tt:Jg4 19 tt:Jc2 .l::!.f6 and Black had a
strong initiative in Kotronias-Psakhis,
Chalkidiki 1992.
1 1 . . .0-0 1 2 'it'd3 lies 13 .i.e5!? 'it'd7
14 a4 a6 1 5 l:lfd 1 l:tac8 1 6 c3 .i.d8!
Once again we see this clever move.
From d8 the bishop may shoot to c7
or b6, where it applies pressure on the
central squares. There now follows a
c5 4 exd5 e x d 5 : Fifth Move A l ternatives
long bout of manoeuvring in which
nothing much happens, which is quite
typical for a position like this.
1 7 h3 h6 1 8 a5 l:lc6 1 9 .i.xf6 t2Jxf6
20 t2Jfd4 J::tc 8 2 1 t2Jc2 .i.c 7 22 t2Je3
l:tcd8 23 t2Jd4 l:le4! 24 t2Jf3 'ifc6 25
b4 l::tde8 26 � 1
26 tt:Jxd5? tt:Jxd5 27 'it'xd5 l:t.e l+
wins material. However, White is
tempted to take the bait a few moves
later.
26 . . .l:ldS 27 l:lab 1 .i.f4 28 t2Jf5 'it'c7
29 t2J5d4 l::t c8 30 l::t b 3 l:lee8! 31 t2Je2
.i.e5 32 l::tc 1 t2Je4! 33 'it'xd5?
33 . . ..i.g3 ! !
temporary disharmony in the white
ranks.
34 c4
Other moves are also bad for
White:
a) 34 tt:Jxg3 .l::!.cd8! 35 'it'f5 'it'c4+.
b) 34 fxg3 .l::!.cd8 35 'it'f5 'it'c4!
c) 34 .l::!.b2 is probably the best of a
bad bunch, although 34 ... i.xf2 is still
grim for White.
34 . . . l:tcd8 35 'iff5 'it'xc4! 36 l:le3
36 .lhc4 .l::!.d l+ 37 lDe l tt:Jd2+ 38
�gl l:txel is mate.
36 . . ..i.bS 37 g3 'it'xb4 38 <it>g2 t2Jd6
117
The Fre n c h Ta rra s c h
3 9 'it'd3 J:txe3 4 0 'it'xe3 l:te8 41 'it'd3
'it'xa5
The flurry of tactics is over and
Black is completely winning. The rest
of the game is of little note.
42 t2Jf4 'iff5 43 'it'b3 t2Jb5 44 tbd5 h5
45 h4 j_d6 46 t2Jg5 g6 47 'it'c4 <it>g7
48 f 4 <it>h6 4 9 t2Jc3 J:tc8 50 'it'b3 <it>g7
5 1 t2Jxb5 J:txc 1 52 t2Jd4 'iff6 53
'it'xb7 j_97 54 'it'd5 l::t c8 55 f5 l::td8
56 t2Jde6+ fxe6 57 t2Jxe6+ <it>h6 58
t2Jxd8 j_xd8 59 fxg6 <it>xg6 60 'it'd3+
<it>g7 61 'it'd1 'ifc6+ 62 <it>h2 j_c7 63
'ifd2 'ifc5 64 <it>h3 'it'e5 65 'it'd7+ <it>f6
0-1
Game 68
Gel ler-Dreev
New York 1990
1 e4 e6 2 d4 d5 3 t2Jd2 a6 4 l2Jgf3
c5 5 exd5
Of course White is not obliged to
capture on d5 here. The alternative 5
dxcS is considered in Chapter 9, Game
77.
5 . . . exd5
This position also frequently arises
from the move order 3 ... cS 4 exds
exdS 5 t'Dgf3 a6. However, with this
1 18
sequence Black can also play the
slightly strange looking immediate
5 ... c4. The drawback to this idea is
that after 6 b3 he is forced to exchange
pawns, as 6 ... bS walks into 7 a4. After
6 ... cxb3 7 axb3 ..5lb4 8 t'DeS t'Df6 9 ..5ld3
0-0 10 0-0 ..5lc3 1 1 l:t.a4 ..5ld7 12 t'Dxd7
t'Dbxd7 13 t'Db l! ..taS 14 ..td2 .tel 15
t'Dc3 White has unravelled successfully
and the bishop pair ensures an edge, as
in Svidler-Korchnoi, Groningen 1996.
6 j_e2
This move invites Black to claim
space on the queenside with ... c5-c4
and ... b7-b5. Naturally Black could
also continue in the normal fashion
with ...t'Dc6 and ... c5xd4, but this
would leave the move ... a7-a6 rather
out of place.
The alternative 6 dxcS is considered
in Games 69 and 70.
6 . . . c4 7 0-0 j_d6 8 b3 b5 9 a4! ?
The most critical response to
Black's idea, attacking the pawn chain
before Black has a chance to consoli­
date his development. White has to be
willing to sacrifice material, as other­
wise Black may just wind up with a
crippling bind on the queenside. Two
other moves are also worthy of con-
 3 . . .
sideration:
a) 9 %:!.e l (vacating the fl-square for
the knight before starting queenside
operations) 9 ...CfJe7 10 a4 .5lb7! (10 ... c3
loses its appeal here: 1 1 CfJfl b4 12 CfJg3
0-0 13 .5ld3 gave White the advantage
on the kingside in Serper-Legky,
USSR 1986) 1 1 CiJe5 0-0 12 bxc4 dxc4
13 .tf3 CfJd5 with equality in Spasov­
Dreev, Tunja 1989.
b) 9 c3 (physically preventing ... c4-
c3!) 9 ... CfJe7 10 a4 .5lb7 1 1 .5la3 .5lxa3
12 l:t.xa3 CfJd7 13 axb5 axb5 14 l:.xa8
.5lxa8 1 5 bxc4 bxc4 16 'ifal 0-0 with
equal chances in lvanchuk-P.Nikolic,
Moscow Olympiad 1994.
9 . . ,j_b7
Here 9 ... c3 ! is the most testing re­
sponse. White can hardly contemplate
locking in his entire queenside army
with 10 CfJbl , so he must go 'all in'
with 10 axb5 cxd2 1 1 .5lxd2 and now:
a) 1 1 .. ..5lb7 12 bxa6 CfJxa6 13 .5lxa6
.l::!.xa6 14 'ife2+ 'ife7 15 'ifxa6!! .5lxa6 16
.l::!.xa6 'ifc7 17 %:!.e l+ \t>d7 18 c4 dxc4 19
bxc4 f6 20 c5 .5lxc5 21 dxc5 CfJe7 22
CiJd4 1-0 was the brilliant game Geller­
Kekki, CSKA-Matynkyla 1986. In the
final position White's many threats
include 23 l:t.a7 'ifxa7 24 .l::!.xe7+ \t>xe7
25 CfJc6+.
b) 1 1...CfJf6! 12 c4 dxc4! (12 ... 0-0? 13
c 5 .5lc7 14 b6 .5lxb6 15 cxb6 'ifxb6 16
.t aS was good for White in Dolgener­
Liepold, Germany 1991) 13 bxc4 0-0
and although White obviously has
some compensation for the piece, it is
not clear whether it is really enough.
Kr.Georgiev-P.Nikolic, Elenite 1993,
supported this claim after 14 c5?!
(weakening the d5-square) 14 ... .5le7 15
l1e 1 .te6 16 'ifb l .td5 17 .taS 'ifc8 18
cS 4 e x d S e x d S : Fifth Mo ve A l tern a tives
bxa6 .5ld8 19 .5lxd8 .l::!.xd8 2 0 a7 CfJc6 21
'ifb6 CfJd7 and, on realising that the a­
pawn was lost, White abruptly re­
signed.
1 0 bxc4! bxc4
1 1 j_xc4!
An inspired sacrifice which is not
only good, but also quite necessary. If
Black were able to complete his devel­
opment then his clamp on the queen­
side would start to become a serious
concern for White.
1 1 . . . dxc4 1 2 t2Jxc4 i.e 7
12 ... CfJe7 runs into 13 CfJxd6+ 'ifxd6
14 .5la3!, after which Black is in grave
trouble, e.g. 14 ...'ifc7 15 l:t.el CfJbc6 16
ds.
13 J:te 1 'it'c7 1 4 : b 1
Threatening 15 l:t.xb7 'ifxb7 16
CfJd6+. The white rooks really play
their part in this game.
1 4 . . . 'it'xc4 1 5 :xb7 lbc6 1 6 t2Jd2!
'it'xd4 1 7 j_b2! 'it'xa4 1 8 :e4 'ifa2 1 9
j_xg7 0-0-0 20 l::t b3
see following diagram
Black's position is beyond repair.
Now 20...CiJf6 loses after 21 .l::!.c4! l:t.xd2
22 .l::!.xc6+ \t>d7 23 'ifxd2+ \t>xc6 24
'ifc3+! \t>d7 25 .5lxh8.
1 19
The Fre n c h Ta rra s c h
2 0 . . . j_f6 2 1 'it'g4+ � c 7 22 'iff4+
<it>c8 23 j_xf6 t2Jxf6 24 'ifxf6 'it'xc2
25 'iff5+! 1 -0
25 ...'itt c l 26 l:t.e7+ or 25 ...l:t.d7 26
.l::!.e 8+ both win the queen. A powerful
performance by Geller, but we await
more tests with 9 ... c3 !
Game 69
Donchev-Eingom
Debrecen 1992
1 e4 e6 2 d4 d 5 3 t2Jd2 a6 4 l2Jgf3
c5 5 exd5 exd5 6 dxc5 j_xc5 7 t2Jb3
j_a7
Here Black utilises the al-square
that was vacated by ... a7-a6. The re­
treat to this haven is obviously the
most natural move, but there is also
nothing wrong with 7 ... il..e 7, a favour­
ite of Dolmatov. An example of
Black's chances is shown in the game
Einarsson-Dolmatov, Reykjavik 1988,
which went 8 il..d3 tt:Jf6 9 0-0 0-0 10
liite l tt:Jc6 11 c3 il..g4 12 il.. gS (12 h3
Jl.. h s 13 il..gS 'ifd6 14 .l::!.e3 looks more
testing) u ... h6 13 il.. h4 'ifb6 14 ilc2
tt:Je4! 15 'ifxdS?! ilxf3 16 gxf3 il..xh4 17
'ifxe4 ilxf2+ 1 8 'itifl g6 and White's
king didn't last many more moves.
120
8 j_d3
8 il..gS is the subject of the next
game.
8 . . . 'it'e7+!
This annoying queen check is the
reason why 8 il..d3 is not seen very
much. White has to choose between
losing a move with 9 ile2, or an un­
welcome exchange of queens.
9 'ife2 tt:Jc6 1 0 c3
An early indication that Black has
no problems here was the game Mata­
novic-Korchnoi, Wijk aan Zee 1968.
After 10 0-0 il..g4 11 h3 il..h S 12 il..f4
'ifxe2 13 il..xe2 tt:Jf6 14 c3 0-0 15 fHel
l:t.fe8 16 g4 il..g6 17 il..fl .l::!.e4 18 .l::!.xe4
ilxe4 19 ilg2 hs 20 gs tt:Jd7 2 1 .l::!. dl
tt:Jf8 22 il..e 3 il..xe3 23 fxe3 tt:Je6 24 h4
l:t.e8 Black had taken over the initiative
and went on to win.
1 O . . ,j_g4 1 1 0-0 'it'xe2 1 2 j_xe2 t2Jf6
This endgame holds absolutely no
worries for Black, whose active pieces
more than offset the weakness of the
dS-pawn.
1 3 h3 j_h5 1 4 g4 j_g6 1 5 t2Jfd4 h5
1 6 <it>g2 hxg4 1 7 hxg4 t2Je5 18 f3
0-0-0 1 9 J:tf2 l:tde8 20 j_f4 t2Jc4 2 1
j_d 1 t2Jd7 2 2 j_c2 j_xc2 2 3 J:txc2
23 . . .j_bS!
 3 . . .
From a very quiet-looking position,
Black suddenly goes for a mating at­
tack! 24 i.xb8 lbe3+ wins material, so
the white king ventures into the un­
known.
24 <it>g3 ..ixf4+ 25 <it>xf4 l:.h3! 26 l:.f2
llld e5 27 1:%.d 1 lll g6+ 28 <it>g5 l:.h6 29
l:te2 0-1
Donchev didn't wait around to see
29 .. .f6+ 30 'it>f5 lbd6 mate.
Game 70
Van der Wiel-Seirawan
Biel 1985
1 e4 e6 2 d4 d5 3 ltJd2 c5 4 exd5
exd5 5 lll gf3 a6 6 dxc5 ..ixc5 7 lllb 3
..ia7 8 ..ig5 lll e 7 9 °ir'd2
Planning both to castle queenside
and exchange Black's most active mi­
nor piece with i.e3.
9 . . .lDbc6 1 0 ..ie3 ..ixe3 1 1 'ir'xe3 0-0
1 2 0-0-0 ..if5
The exchange of dark-squared bish­
ops gives White a tiny edge in this po­
sition, but with careful play, as in this
game, Black can often neutralise this
completely. On the other hand,
Black's winning chances are very slim.
1 3 lDfd4 ltJxd4 1 4 ltJxd4 ..ig6 1 5
c5 4 e x d5 e x d 5 : Fifth Mo ve A lternatives
..id3 l::. e8 1 6 l:.he 1 'ir'd7 1 7 ..ixg6
ltJxg6 1 8 'ir'g5 l:.xe1 1 9 l:.xe 1 l:.e8 20
l:txe8+ 'ir'xe8 21 'ir'e3 'ir'xe3+ 22 fxe3
The major pieces have been un­
ceremoniously hoovered off and the
only conceivable result is a draw.
22 . . . lll e 5 23 lllf5 'itf8 24 a4 b6 25
<it>d2 g6 26 lll d 6 ltJd7 27 lll c 8 <it>e8
28 <it>c3 <it>d8 29 llld6 <it>e 7 30 lllc8+
<it>d8 31 llld6 � - �
Game 71
Rublevsky-Bareev
Rubinstein Memorial 1997
1 e4 e6 2 d4 d5 3 ltJd2 c5 4 exd5
exd5 5 ..ib5+ ..id7
The alternative 5 ... lbc6 is seen in
Game 73.
6 'ir'e2+ ..ie7
The inferior 6 ...'ii' e 7 is dealt with in
the next game.
7 dxc5
Otherwise the queen check would
have no point. White must now make
his opponent work hard to regain the
pawn.
7 . . .lDf6 8 lllb3 0-0 9 tllf3 l:te8 1 0
..ie3 a6 1 1 ..id3
1 1 . . . ..ia4
121
The Fre n c h Ta rra s c h
A very direct move. Black simply
wishes to eliminate the knight on b3
and regain the c5-pawn. If Black pre­
fers to play in sacrificial mode then
1 1 . ..lbg4 is worth considering. In the
game Van Wely-Ehlvest, Biel 1993,
White quickly gained a substantial
advantage after 12 0-0 a5 13 a4 i.f6? 14
c3 lba6 15 i.b5 lbc7 16 lbbd4 lbxb5
17 axb5 'ii' c7 18 b4 i.xd4 19 b6 'ii'b 8
20 cxd4 axb4 21 'ii'd2. However,
13 ... lba6! is a stronger move. After 14
i.xa6 l::!.xa6 Black's two bishops give
him definite compensation for the
pawn.
1 2 lDfd4 lDbd 7 1 3 0-0-0
The main move here, although the
quieter 13 0-0 must not be underesti­
mated. Here Black should probably
continue with 13 ...lbxc5, as 13 ...i.xb3
was met by the zwischenzug 14 c6! in
the game Kosashvili-Sjodahl, Arnhem
1989, when after 14 ...bxc6 15 axb3
i.c5 16 'ii'd l 'ii'b 6 17 l::!. a4 a5 18 h3 lbf8
19 'ii'f3 Black's weak queenside pawns
gave White an edge.
1 3 . . . ..ixb3 1 4 lDxb3
Naturally 14 axb3 is also possible,
although this may give Black more
chances as the b3-pawn can be a target
for the lunge ... a6-a5-a4. Van Wely­
Brenninkmeijer, Holland 1993, con­
tinued 14 ... i.xc5 15 'ii'f3 lbe5 16 'ii'h3
'ii'b 6 17 lbf5 i.xe3+ 18 lbxe3 a5 19
lbxd5 lbxd3+ 20 'ii'xd3 lbxdS 21 'ii'xd5
a4, when White is a pawn up but faces
some tricky moments on the queen­
side.
1 4 . . . a5
Or 14 ...lbxc5 15 lbd4 (15 'ii'f3 is
probably stronger) 15 ...lba4!? 16 c3
.Uc8 17 i.c2 lbb6 1 8 'ii'd3 g6 19 lbb3?!
122
lbc4 20 i.d4? i.a3!! 21 bxa3 'ii'd6 22
lbc5 lbe4 with favourable complica­
tions for Black, as in Yudasin-Psakhis,
Beersheva 1993.
1 5 a4 ltJxc5 1 6 ..ib5 lDxb3+ 1 7 cxb3
'ir'c7+ 1 8 <it>b1
White now has doubled pawns on
the queenside, and the d5-pawn is also
passed, as well as isolated. On the
other hand, the white king is well pro­
tected now, and he can hope to use his
two bishops in an open position.
1 8 . . Jled8 1 9 ..id4 ltJe4 20 f3 lDc5
21 ..ie5 ..id6
2 1...'ii'b 6 allows 22 i.xg7, so Black
has to enter an inferior endgame,
where the bishop is much stronger
than the knight. Rublevsky, however,
makes a few slight inaccuracies and
this allows Bareev to hold the draw.
22 l:.xd5 ..ixe5 23 'ir'xe5 'ir'xe5 24
llxe5 lDxb3 25 <it>a2
25 l::!.e7 looks like a better winning
try.
25 . . .ltJd4 26 llc 1 llab8 27 l:.c7 ltJe6
28 l:.d7 l:.dc8 29 llf5 f6 30 l:.d6 ltJc5
31 g4 b6 32 h4 <it>f7 33 g5 lld8 34
llc6 l:.dc8 35 gxf6 l:.xc6 36 ..ixc6
gxf6 37 llh5 l:.h8 � - �
Game 72
Moya-Roldan
Correspondence 1984
1 e4 e6 2 d4 d5 3 lDd2 c5 4 exd5
exd5 5 ..ib5+ ..id7 6 'ir'e2+ 'ir'e7?!
The endgame reached after this
move is known to be comfortably
better for White.
7 ..ixd7+ ltJxd7 8 dxc5 ltJxc5 9 lllb3
'ir'xe2+ 1 0 ltJxe2 lDxb3 1 1 axb3 ..ic5
1 2 ..id2
 3 . . .
As well as the customary weakness
of the dS-pawn, Black also has to
worry about White's pressure on the
semi-open a-file. All in all this makes
the endgame unpleasant to defend.
1 2 . . . ltJe7 1 3 ..ic3 0-0 1 4 ..id4 ..ixd4
1 5 ltJxd4 a6 1 6 <it>d2 l:.fe8 1 7 l::the 1
<it>f8 1 8 l:.e5 l:.ad8 1 9 l:.ae 1 ltJc6 20
l::txe8+ l:.xe8 2 1 l:.xe8+ <it>xe8 22
ltJxc6 bxc6 23 b4
An instructive position, where
Black's weaknesses tell in the king and
pawn endgame. Black has fewer
'passing' moves than White and con­
sequently cannot prevent a decisive
infiltration of the white king.
23 . . . <it>d7 24 <it>d3 <it>d6 25 <it>d4 f6 26
f3 h6 27 f4 g6 28 g4 h5 29 gxh5
gxh5 30 h4 'Oti>d7 3 1 <it>c5 <it>c7 32 f5
<it>d7 33 <it>b6 1 -0
Game 73
Brynell-Schmidt
Naestved 1988
1 e4 e6 2 d4 d5 3 ltJd2 c5 4 exd5
exd5 5 ..ib5+ ltJc6
White can now obviously play 6
lbgf3, transposing to lines in Chapter
7, while avoiding 5 lbgf3 a6 and 5
c 5 4 e x d 5 e x d 5 : Fifth Mo ve A l tern a t i ves
lbgf3 lbf6. In this game White keeps it
independent with the queen check.
6 'ir'e2+ ..ie7
6 ...'ii'e7 once again leads to an infe­
rior endgame. For example, 7 dxcS
'ii'xe2+ 8 lbxe2 i.xcS 9 lbb3 i.b6 10
i.d2 lbge7 11 i.b4 led to another
grim, turgid defence for Black in
Euwe-Botvinnik, World Champion­
ship 1948.
7 dxc5 lDf6 8 lDb3 0-0 9 ..ie3 l::te8
1 0 lDf3 a6 1 1 ..id3 d4! ?
An interesting attempt to liven the
position up. Funnily enough, it seems
to lead to a forced draw!
1 2 lDfxd4 ltJxd4 1 3 ltJxd4 ..ixc5 1 4
c 3 ltJg4 1 5 0-0 'ir'h4 1 6 h 3 ltJxe3 1 7
fxe3
1 7 . . ...ixh3 !
Striking before White can consoli­
date. Luckily for White he can steer
the game to a drawn ending.
1 8 gxh3 ..ixd4 1 9 cxd4 'ir'g3+ 20
�h1 l:.xe3 21 ..ixh7+! <it>h8
Or 21. ..'itxh7 22 'ii' h S+ and 23
'ii'xf7+.
22 'ir'h5 'ir'xh3+ 23 'ir'xh3 l:.xh3+ 24
<it>g2 l:.xh7 25 l:.xf7 l:.d8 26 l:.xb7
l:.h4 27 l:.d1 l:.dxd4 28 l::txd4 l:.xd4
29 'itf3 <it>h7 � - �
1 23
The Fre n c h Tarra s c h

Summary
Despite a variety of attempts against 5 tbgf3 lbf6, Black's move has stood firm
and the onus is clearly on White to produce something new. The variation 5 ... a6
6 .ie2 c4!? is especially good for Black players determined to go for the full
point, as many of White's most promising lines involve sacrificing material. 5
.ibS+ continues to be less popular than 5 lbgf3, as it seems that Black has suffi­
cient resources after either 5 ... .id7 or 5 . . lbc6 6 'ii'e2+ .ie7.
.

1 e4 e6 2 d4 d5 3 lDd2 c5 4 exd5 exd5 5 lDgf3

5 .ibS+(D)
5 ....id7 6 'ii'e2+
6 ... .ie7 Game 71
-

6 ...'ii'e7 Game 72
-

5 ...lbc6
6 lbgf3 - Chapter 7
6 'ii'e2 Game 73
-

5 . . .lDf6
5 ... a6
6 .ie2 Game 68
-

6 dxcS .ixcS 7 lbb3 .ia7 (D)

8 .id3 Game 69
-

8 .igS Game 70
-

6 i.. b 5+ i.. d 7 7 i..x d7+ lDbxd7 8 0-0 i..e7 9 dxc5 ltJxc5 (D) 1 0 lle1
10 lbd4 - Game 66
10 lbb3 - Game 67
10 ... 0-0 1 1 lDf1 l::te8 1 2 i..e3 ltJfe4
12 ... aS Game 65
-

1 3 c3 -Game 64

S i.. bS+ 7. . . i..a 7 9 lDxc5

124
 CHAPTER NINE I
3 . . . cs 4 ltJgf3

1 e4 e6 2 d4 d5 3 ltJd2 c5 4 lDgf3 previous chapter with 4 ... a6 5 exdS

4 lbgf3 is a sly transpositional move exdS, as we see in Game 77. After
which White players often employ in 4 ...lbc6 White normally plays 5 exdS,
order to direct the game into their but in Game 78 we discuss the little
opening terrain, rather than their op­ played 5 .tbs.
ponent's. For example, after 4 ...cxd4
White need not play 5 exdS 'ii'xdS, but Game 74
can choose 5 lbxd4, which gives an Adams-Levitt
entirely different type of position (see Dublin 1993
Games 74 and 75). White also has
other options against Black three main 1 e4 e6 2 d4 d5 3 ltJd2 c5 4 lDgf3
alternatives, 4 ...lbf6, 4 ... a6 and 4 ...lbc6, cxd4
and Black needs to be prepared for all Inviting the ...'ii'xdS lines seen m
of these eventualities. The good news Chapters 5 and 6 via 5 exdS 'ii'xdS.
for Black is that none of these lines are 5 ltJxd4 lDf6
particularly terrifying. In general
White players straying from the main
path are merely looking to reach a
playable position rather than striving
hard for a theoretical edge. Game 76
sees 4 ...lbf6, which together with the
similar line 4 ...cxd4 5 lbxd4 lbf6
(Game 74), is probably Black's most
reliable move. 4 lbgf3 is a good idea
against ...a7-a6 adherents, as White
certainly has more options than sim­
ply transposing to Games 68-70 in the

1 25
The Fre n c h Ta rra sc h
We will look at the alternative
move 5 ... lbc6 in Game 75.
6 exd5
6 es looks quite tempting, but after
6 ...lbfd7 Black's position is extremely
resilient. Svidler-Dreev, Rostov 1993,
continued 7 lD2f3 lbc6 8 lbxc6 bxc6 9
i.d3 i.a6! (this exchange of White's
best bishop ensures Black a comfort­
able game) 10 0-0 i.xd3 1 1 'ii'xd3 i.e7
12 c4 0-0 13 'ii'c2 aS 14 l::!.d l a4 15 i.f4
lbb6 and Black was fully equal.
6 . . .'ir'xd5 ?!
Black loses time after this move.
6 ... lbxdS is stronger, when after 7
lb2f3 i.e7 the game transposed to
Emms-Bronstein (see note to Black's
sixth move in Game 76) .
7 lDb5! 'ir'd8 8 ltJc4 ltJd5
An exchange of queens brings no re­
lief to the black camp, as 8 ... 'ii'xdl+ 9
'it>xdl leaves the squares c7 and d6 de­
void of sufficient protection.
9 lDe3 ltJc7
Levitt suggests 9 ... a6 10 lbxdS exdS
1 1 lbd4, when White has an edge, but
at least Black can reach a reasonable
position.
1 0 1Ld2 lDxb5 1 1 1Lxb5+ 1Ld7 1 2
1Lxd7+ 'ir'xd7 1 3 'ir'g4 ltJc6 1 4 0-0-0
126
Black is already in danger. White's
development advantage is very serious,
and Black has further problems devel­
oping the dark-squared bishop with­
out losing the g7-pawn.
1 4 . . . 'ir'd4 1 5 'ir'e2 'ir'a4 1 6 'Oti>b1 1Le7
1 7 1Lc3 0-0 1 8 l:.d7 tll b4
This move loses, but it doesn't
really deserve a question mark as it is
probably Black's best practical chance
and requires some quite brilliant play
to refute it.
1 9 'ir'g4 g6 20 'ir'd4! 'ir'xa2+ 2 1 <it>c 1
f6 22 1Lxb4 1Lxb4
In his book Chess in the Fast Lane
Adams admits that he had intended 23
'ii'xb4 l::!.ad8 24 l::!. xd8 .Uxd8 25 'ii'a3 and
White wins. Only here did he see that
Black can play 24 ...'ii'a l+! 25 'it>d2
:xd8+ 26 'it>e2 'ii'x hl 27 'ii'e7 l::!.f8 28
lbg4 'ii'c l and Black is slightly better.
Half an hour in the 'think tank' pro­
duced the answer.
23 c3! ! 'ir'a 1 + 24 <it>c2 'ir'xh1 25
'ir'xb4 :I.ads
There is absolutely no way out.
25 ....Uae8 allows 26 'ii'h4 l::!.f7 27 .Udl!
and the queen is neatly trapped on the
bl-square.
26 l:.xd8 l:.xd8 27 'ir'e 7 1 -0
 3 . . . c5 4 lb g f3
------�
-------
ttJ';'!ii
�:--:=� '
(9 ...lDf6 !?) 10 i.b4+ 1 1 <tie2 'ii'a5
12 i.e3 i.e7 1 3 'iixd2+ 14 i.�d2,
as m . Van der Wi·el-Ehlvest, Hanmge
1990.
8 i..d3 'iic7
9 � "' 2 .>ild6
Black can als develop the kmght
.
first, whi�h has th advantage of tak­
ing the sung :;_ut o; White's e4-e5. For
examp le, 9 . . .'1.Je7
lilf lO 0-0 t2Jg6 1 1 ltJf3
i.d6 12 :el 4 13 .hf4 �� xf4 14 c4

0-0 15 g3 i.h6 16 e d5 cxd5 was equaI

m Zakharov-Petrosian,
. � USSR Cham­
After 27 ...�f8 28 ltJ 4 White wins, . nship 1976.
pio
t e "ucial difference ;eing that the�e
h 1 0 lllf3 dxe4?.1
.
This move, releasmg the tension
•

is now no ... 11i'c! defence. A fantastic

performance_ b y Adam� but Black
. .
should certam ly replace 6 ...'ii'xd5 with
6 ...ltJxd5.
premature1y, is not to be recom-
mended. In parucular, Black now has
. his . .
real problems ushenng king mto
. .
safety. A muc h better alternauve is
.
10 ... ltJe7 1 1 0-0 0-0 12 c4, when Whlte
.
is only slightly better.
1 1 'iixe4 lDf6 1 2 ""'4 l!bB
.
Black wo':1ld like to play 12 ... 0-D•
. ·

but 13 i.g5! is s1mply very annoymg.

1 3 0-0 ltJd5 1 4 lle1 . e 7 1 5 'ilg4 g6
. .
_ i.. is
Once agam cast!mg out of the
.
uesuon: 15 0-0 16 i.h6 i.f6 17 ·we
··· .
�
41 ·

;nd h7 is droppmg.
1 6 c4 f5?

7 lllxc6 bxc6
7. . i.xc6 can lea_d to an endgame
. Blac k' s spht pawns on the
Where
. .
queenside · v white a minute advan­ .
tage, e.g. � �xc6+ bxc6 9 c4 dxe4 Thmgs were already becoming

127
The Fre n c h Ta rra s c h
rather serious for Black, but this move
allows White a pretty combination,
which accelerates the end.
1 7 1Lxf5! lDf6
17 ...gxf5 18 'iifh 5+ 'it>d8 19 cxd5
cxd5 20 lbg5 and 17 ...exf5 18 'ii'd4 :f8
19 cxd5 cxd5 20 i.g5 are also very un­
desirable. Black seeks salvation in the
endgame, but his airy king still causes
too many problems.
1 8 'ir'f4 'ir'xf4 1 9 1Lxf4 llxb2 20
1Lxe6 1Lc5 21 1Lxd7+ <it>xd7 22 ltJe5+
<it>c8 23 ltJd3 1 -0
After 23 ...l::!.c2 24 lbxc5 l:.xc4 25
i.e3 White is a clear piece up.
Game 76
Popovic-P .Nikolic
Yugoslavia 1991
1 e4 e6 2 d4 d5 3 ltJd2 c5 4 lDgf3
lDf6 5 exd5
Here 5 e5 lbfd7 6 c3 lbc6 7 i.d3
transposes to lines considered in
Chapter 3, Games 27 and 28. After 5
exd5 Black can obviously play
5 ...exd5, converting to Games 64-67,
but here Black chooses another en­
tirely playable move.
5 . . . ltJxd5 6 lDb3
128
6...ltJd7
6 ... cxd4 is probably more accurate.
Emms-Bronstein, Maidstone 19 94,
continued 7 lbbxd4 i.e7 8 g3!? (8 i.d3
is also possible) 8 ... 0-0 9 i.g2 b5! 10 0-0
(10 lbxb5 'ii'aS+ 1 1 lbd2 'ii'xb5 12 c4
'ii'a6 13 cxd5 exd5 14 i.xd5 lbc6 gives
Black excellent play for the pawn)
10 ... a6 1 1 'ii'e2 'iifb6 12 c3 lbd7 13 i.g5
i.xg5 14 lbxg5 lb7f6 and Black had
equalised.
7 1Lg5! 1Le7 8 1Lxe7 'ir'xe7 9 1'.b5
cxd4 1 0 'ir'xd4 0-0 1 1 0-0-0 a6 1 2
1Lc4 lD5f6 1 3 1:%.he 1 b5
14 1Ld5!
14 i.d3 i.b7 would not have pre­
sented Black with any problems, but
now White can steer the position into
an endgame, where his control of the
d-file will be the most important fac­
tor. Note that 14 ... :b8 15 i.c6 l::!.b6
fails to 16 lbe5!
14 . . . ltJxd5 1 5 'ir'xd5 lla7 1 6 'ir'd6
'ir'xd6 1 7 llxd6 llc7 18 ltJe5 lDb8
This looks rather ugly, but
18 ...lbxe5 19 :xe5, with the idea of
llc5, also favours White.
1 9 ltJa5 f6 20 ltJd3 e5 21 lDb4 <it>f7
22 l:.ed 1 1Le6 23 c3 1Lc8 24 g3 h5
25 h4 lle8 26 lDd5 lla7 27 lDb4 lic 7
28 : 1 d2 :ea 29 :da :ea 30 :ada
:ea 3 1 :da :ea 32 :xe8 <it>xe8 33
:da :a7 34 :ba lDd7 35 l:.ea+ �7
3a :ca :as 37 lDd5 e4?
Black has been forced to defend a
grim endgame for quite a while, and
finally cracks under the pressure.
37 ...:b8 was the only way to hang on.
38 :c7
38 . . . <it>ea
This loses a piece, but 38 ... 'it>e8 fails
to 39 :xd7, while after 3 8 ...'it>f8 39
lbc6 Black is completely tied up.
39 :xd7 1 -0
White wins the rook on a8 after
39 ... 'it>xd7 40 lb b6+ or 39 ... .i.xd7 40
lbc7+.
Game 77
Wolff·Yermolinsky
USA Championship 1993
1 e4 ea 2 d4 d5 3 ltJd2 aa 4 lDgf3
c5
This position can also be reached
via the move order 3 ... c5 4 lbgf3 a6. If
now 5 exd5 exdS we have reached the
standard position of the ... a7-a6 line
(see Chapter 8, Games 68-70) .
5 dxc5 1Lxc5 a 1Ld3 ltJe 7 7 0-0
lDbca 8 c3 0-0
9 exd5
The other way to play this position
is to maintain the pawn on e4. Rozen­
talis-P.Nikolic, Moscow 1994, contin­
ued 9 'ii'e2 lbg6 10 lbb3 dxe4 (perhaps
10 ....i.a7!?, maintaining the tension, is
more appropriate) 1 1 i.xe4 i.d6 12
i.e3 'ii'c7 (12 ...lbf4? loses 13 i.xf4
i.xf4 14 i.xh7 + 'it>xh7 15 'ii'e4+) 13
:adl :d8 14 'ii'c4! i.d7 15 i.xg6 hxg6
16 lbg5! and White had drummed up a
menacing initiative.
9 . . . exd5
9 ...lbxdS is also very reasonable, af­
ter which White should probably con­
tinue with 10 lbe4. After 9 ...exd5 we
have actually reached a position simi­
lar to that discussed in Chapter 7 (see
Games 56-58). The difference here is
that Black has committed himself to
...a7-a6, which is only a semi-useful
move.
1 0 lDb3 1'.da 1 1 :e 1 ha 1 2 h3 1Lf5
After 12 ...lbf5, which is normal
procedure in Games 56-58, Black's
...a7-a6 proves to be a serious loss of
time. White can simply play 13 i.c2
i.e6 14 'ii'd3, when the threat of g2-g4
is hard to deal with.
129
The Fre n c h Ta rra s c h
1 3 ..ie3 :ea
1 4 lDbd4
14 i.cS!? 'ii'd7 15 i.xfS lbxfS 16
'ii'd3 is a sensible suggestion for White,
when the pressure on d5 ensures a
small edge.
1 4 . . . ..ie4 1 5 'ir'c2 ltJg6 1 6 ..ixe4
dxe4 1 7 ltJd2 ltJxd4 1 8 ..ixd4 f5!
The nature of the position has sud­
denly changed. Black now has a po­
tentially dangerous pawn roller on the
kingside. Wolff now decided to grab a
hot pawn on b7, but Alex 'the Yermi­
nator' Yermolinsky utilised his initia­
tive to the full.
1 9 'ir'b3+ �h7 20 'ir'xb7
20 . . ..i.e5! 21 J:.ad 1 .i.xd4 22 cxd4
'ir'g5!
130
Now it is clear that White is in real
trouble. 23 g3 lbf4 24 'it>h2 lbd3 is
probably the best of a bad bunch, al­
though Black's attack would still be
senous.
23 lDf1 ? lDh4 24 lDg3 f4 25 l:.xe4
fxg3 26 fxg3 l:.ab8 27 'ir'c6 lDf3+ 0-1
28 gxf3 'ii'xg3+ 29 'it>fl 'ii'xf3+ wins
a rook.
Game 78
Kuijf-Uhlmann
Amsterdam 1990
1 e4 e6 2 d4 d5 3 ltJd2 c5 4 lDgf3
ltJc6 5 ..ib5
An interesting bid to avoid the
normal lines that arise after 5 exdS
exdS (see Chapter 7).
5 . . .dxe4
This is the theoretical choice, al­
though 5 ... a6 also doesn't look bad,
e.g. 6 i.xc6+ bxc6 7 0-0 lbf6 8 es lbd7
9 c4 i.e7 10 dxcS lbxcS 1 1 lbd4 'ii'b6
12 'ii'g4 0-0 13 lb2f3 fS with an unclear
position in Sutovsky-Psakhis, Haifa
1996.
Note that 5 ... cxd4 6 lbxd4 simply
transposes to Game 75.
6 lDxe4 ..id7

7 ..ie3
This doesn't really cause Black any
problems at all. Black has to be more
careful after the aggressive 7 i.gS!? A
typical Tal miniature concluded
7 ...'ii'aS+ 8 lbc3 cxd4 9 lbxd4 i.b4 10
0-0! i.xc3 11 bxc3 'ii'xc3 12 lbfS!! exfS
13 l':.e l+ i.e6 14 'ii'd6 a6 15 i.d2!
'iifxc2 16 i.b4! axbS 17 'ii'f8+ 'it>d7 18
:edl+! 'it>c7 19 'ii'xa8 1-0 Tal­
Uhlmann, Moscow 1971. Instead of
8 ...cxd4, Black should play 8 ... a6 9
i.xc6 i.xc6, when his bishop pair off­
sets White's slight development lead.
7 . . .'ir'a5+ 8 ltJc3 cxd4 9 ltJxd4 ..ib4
1 0 0-0 ..ixc3 1 1 bxc3 ltJge7 1 2
ltJxc6 ..ixc6 1 3 ..ixc6+ !Lixc6 1 4 c4
0-0 1 5 °ir'b1 'ir'c7 1 6 c5 llfd8
Black has no weakness, whereas
White's doubled c-pawns are always a
cause for concern. Optically things do
not look too bad for White, who can
use the d6-outpost. However, some
3 . . . c s 4 li:J g f3
extremely calm play by Uhlmann
shows that Black holds all the posi­
tional trumps.
1 7 °ir'b5 lld7 1 8 llad 1 llad8 1 9 lld6
a6
19 ...l':.xd6 20 cxd6 l':.xd6? fails to 21
i.f4, as 21...eS 22 i.xeS lbxeS 2 3 'ii'e8
is mate.
20 °ir'b2
20 . . . ltJe7!
This regrouping manoeuvre spells
the beginning of the end for White.
Black wins a pawn after both 2 1 .Ufdl
l':.xd6 22 cxd6 lhd6 23 .Uxd6 'ii'xd6
and 2 1 i.d4 lbfS 22 l:.xd7 lhd7 23 c3
eS, while White's chosen move also
sheds a pawn.
2 1 ..ig5 llxd6 22 cxd6 llxd6 23 ..if4
'ir'b6! 24 'ir'e5 lld7 25 ..ie3 'ir'c7 26
°ir'b2 lDf5 27 ..ib6 'ir'c6 28 g3 h5 29
h4 lld2 30 'ir'b4 ltJxg3 0-1
A fine positional game from the
French Defence stalwart.
1 31
The F re n c h Ta rra s c h

Summary
As long as Black knows his stuff and doesn't get confused by any move order
tricks, he should have nothing to fear from 4 tbgf3 . Black should avoid 4 ...cxd4 5
lbxd4 lbf6 6 exd5 'ii'xd5?!, but alhhe other lines discussed in this chapter look
quite playable.

1 e4 e6 2 d4 d5 3 ltJd 2 c5 4 lll gf3 cxd4

4 ...lbf6
5 e5 lbfd7 6 c3 lbc6 7 i.d3 - see Games 27-28
5 exd5 (D)
5 ...exd5 Games 64-67
-

5 ...lbxd5 Game 76
-

4 ... a6
5 exd5 exd5 - Games 68-70
5 dxc5 Game 77
-

4 ...lbc6
5 exd5 exdS - Chapter 7
5 i.b5 (D)
5 ...cxd4 6 lbxd4 Game 75 (by transposition)
-

5 ...dxe4 Game 78
-

5 lllxd4 (D)
5 exd5 'iifxdS - Chapters 5-6
5 ...lllf6
5 ... lbc6 Game 75
-

6 exd5 - Game 74

S exdS S i..bS 5 lllxd4

132
 CHAPTER TEN I
Third Move Alternatives
for B lack
1 e4 e6 2 d4 d5 3 ltJd2
Although 3 ... cS and 3 ...lbf6 con­
tinue to be the overwhelmingly popu­
lar choices at all levels, Black does
have some playable alternatives, which
we shall look at this final chapter. One
option for Black is 3 ...lbc6 (Games 79-
8 1), an unconventional move which to
be honest has never really caught on.
Black obstructs the c-pawn, so much
of his typical queenside counterplay is
eliminated, or at least delayed. Black
often has to accept a passive, blocked
position (see Game 8 1), but there are
some positional ideas here that may
appeal to certain players.
3 ... i.e7 (Game 82) is also seen infre­
quently, but it does seem more reliable
than 3 ...lbc6. It is favoured by the
Armenian GM Smbat Lputian, who
has employed it with some success. In
particular, White has to be quite
mindful not to stumble into an unde­
sirable transposition. In Game 83 we
take a look at 3 ... a6 (which usually
transposes to the 3 ... cS variations that
we have already discussed after 4 lbgf3
cS) and some other unusual moves for
Black.
The final two games are a bit of a
bonus, as the Rubinstein Defence does
not really belong to the Tarrasch
proper. Even so, White players would
do well to flick through the summary
of main lines. The introduction of
some new ideas for Black, such as the
aptly named 'Fort Knox Defence'
(Game 84), has given the Rubinstein a
new lease of life. The classical 4." .. lbd7
is seen in Game 85.
Game 79
Xie Jun-Brunner
Shanghai 1995
1 e4 e6 2 d4 d5 3 lDd2 lDc6 4 c3
This tricky line generally leads to
very open positions. The more com·
mon 4 lbgf3 is considered in Games 80
and 8 1 .
4 . . . e5!
Black does best to strike immedi­
ately in the centre here. This out­
wardly extravagant second move of
1 33
The F re n c h Tarra s c h
the e-pawn is warranted by White's
non-developing 4 c3.
5 exd5 'ir'xd5 6 lll gf3
6 exd4
...
The natural looking 6 ...i.g4? is ac­
tually a blunder here. After 7 i.c4!
i.xf3 8 'ii' b 3! Black is in difficulty, e.g.
8 ... 'ii'd7 9 lbxf3 lba5 10 i.xf7+! 'ii'xf7
1 1 'ii'b 5+, regaining the knight with a
pawn to boot.
7 ..ic4 °ir'h 5!
With 7 ...'ii'h5 Black is ready to sac­
rifice a pawn. In my opinion this
move is stronger than 7. . . 'ii'f5, when
after 8 lbxd4 lbxd4 9 cxd4 i.d7 10 0-0
0-0-0 1 1 'ii'b 3! White has annoying
pressure on the fl-pawn. Emms­
Crouch, Isle of Man 1993, continued
1 1 . ..lbh6 12 lbf3 i.d6 13 Itel Itdf8 14
i.xh6 gxh6 15 lbe5 i.e8 16 l':.e3 l':.hg8
17 .Uf3 'ii' h5 18 i.d5 with a clear ad­
vantage to White.
8 cxd4 ..ie6 9 ..ixe6 fxe6 1 0 'ir'b3
0-0-0 1 1 0-0 lllf6
This looks like the most natural
move, but also playable is 1 1...lbh6,
intending to put pressure on d4 with
...lbf5. After 12 lbe4 lbf5 13 'ii'xe6+
'it>b8 14 i.g5 lbfxd4 15 lbxd4 11xd4 16
i.e3 'ii' d5 1 7 lbg5 'ii'xe6 1 8 lbxe6 l':.d7
134
19 lbxf8 .Uxf8 an equal position was
reached in Lane-Landenbergue, Zug
1989.
1 2 'ir'xe6+
After 12 lbe4 Black can shield his
weak e6-pawn with 12 ... .UdS. An ex­
ample of Black's chances in this line is
illustrated by this following horror
story for White: 13 .Ue l lbxe4 14 lhe4
i.d6 15 h3 Itf8 16 l':.xe6 l':.df5 17 g4??
'ii'xh3! 18 gxf5 lbxd4 19 lbg5 'ii'h2+ 20
'it>fl 'ii'h l mate 0-1 Lawson-Landen­
bergue, Haifa 1989 .
1 2 . . .<it>b8 1 3 llle4 lllxe4 14 'ir'xe4
:ea 1 5 'ir'd3 ..id6 1 6 ..id2 g5!
After this lunge Black manages to
create sufficient counterplay to com­
pensate for the sacrificed pawn. The
greedy capture on g5 would of course
lead to a disaster for White: 17 i.xg5?
l':.ef8 18 h4 h6 19 i.d2 Ithg8 20 'it>hl
.Uxg2 21 'it>xg2 'ii'g4+ 22 'it>h l Itxf3 and
White can safely resign.
1 7 g3 g4 1 8 lllh4 llhf8 1 9 ..ic3 lld8
20 llae 1 ..ie7 2 1 lllg 2!
Preparing to offload an exchange,
which releases much of the pressure
on the white kingside. Black still has
to play carefully to maintain the bal­
ance.

2 1 . . . ltJe5 22 llxe5 'ir'xe5 23 'ir'xh7
'ir'd6 24 ltJe3 ..if6 25 :!.d 1 llg8 26
'ir'f5 llg5 27 'ir'e4 lldg8 28 lle 1 a6
29 ltJc4 'ir'd7 30 'ir'e6 'ir'g 7 31 ltJe5
..ixe5 32 dxe5 llg6 33 'ir'c4 lle8 34
lle4 'ir'd 7 35 <it>g2 °ir'd 1
The g4-pawn gives Black just
enough counterplay to ensure the
draw.
36 'ir'e2 'ir'd5 37 'ir'c4 'ir'd 1 38 'ir'e2
'ir'd 5 39 'ir'c4 1h -1h
Game 80
Krishilovsky-Doroshkievich
Novgorod 1997
1 e4 e6 2 d4 d5 3 ltJd2 ltJc6 4 lDgf3
lDf6 5 e5 ltJd7 6 ..ib5
Third M o ve A l terna tives for B l a c k
By putting pressure o n the a4-e8 di­
agonal, White discourages his oppo­
nent from playing the desirable .. .f7-f6
advance.
Of course the natural 6 .id3 is also
important here. After the critical 6 .. .f6
White can simply capture, but the real
fun starts after 7 lDgS!? Now 7 ... fxgS?
actually loses after 8 'ii'hS+ g6 9
.ixg6+ hxg6 10 'ii'xg6+ �el 1 1 lt:Je4!!
.ih6 12 .ixgS+ .ixgS 13 'ii'g7+ '1te8 14
'ii'xh8+ lt:Jf8 15 'ii'hS+. Fortunately for
Black he can play 7 . .. ttJdxeS! 8 dxeS
fxgS 9 'ii'hS+ g6 10 .ixg6+ 'it>d7 with a
very messy position. Note that 1 1 lt:Jf3
is a blunder here due to 1 1 ...hxg6,
when 12 'ii'xh8 fails to 12 ... .ib4+.
The other main alternative, 6 lt:Jb3,
is the subject of Game 81.
6 . . . a6
Another option for Black is the
strange looking 6 ... aS!?, which pre­
pares ...lt:Ja7, hitting the bishop on bS
and preparing .. . c7-c5. Nevertheless,
this all sounds rather time-consuming,
just to play something which could
have been played at move three! After
the simple 7 0-0 lt:Ja7 8 .id3 White
must be better.
7 ..ixc6 bxc6 8 lDb3! c5
The game A. Rodriguez-Pecorelli
Garcia, Cuba 1997, went instead 8 ... aS
9 .id2! a4 10 lt:JaS lt:Jb8 1 1 c4 a3 12 b4
dxc4 13 0-0 .ia6 14 :e1 .ibS 15 'iifc l
'ii'dS 16 'ii'xa3 lt:Jd7 17 'ii'c3 lt:Jb6 18 a3
h6 19 'ii'c2 .ia4 20 'ii'c l .ie7 2 1 .ie3
.ibS 22 tt:Jd2 0-0 23 tt:Jb l! fS 24 exf6
:xf6 25 lt:Jc3 'iifhS 26 .if4! and Black
was in a massive bind. After 8 ... cS
Black plans to exchange off his dou­
bled c-pawn.
9 ..ig5 ..ie7 1 0 ltJa5!
1 35
The F re n c h Ta rra s c h
The point to White's previous play.
Now 10 ... i.xg5 1 1 tbc6 traps the
queen, so Black is forced into contor­
tions.
1 0 . . . ltJba
The other try is 10 ...tbxe5 1 1 dxe5
i.xg5 12 tbc6 'ii'd7 13 tbxg5 and now:
a) 13 ...'iifxc6 14 'ii' h5 'iifd7 (14 ... g6 15
'ii'h6 i.d7 16 'ii'g7 0-0-0 17 tbxf7 is
very good for White) 15 tbxh7, when
Black has to deal with the menacing
threat of tbf6+.
b) 13 ... h6 14 tbxf7 'iifxf7 15 0-0 0-0
16 'ii'd2 i.d7 17 tba5 l':.ab8 18 c4 'iiff4
19 'iife 2 and Black's weak pawns give
White the advantage, as in Yudasin­
Drasko, Tbilisi 1987.
1 1 ..ixe7 'ir'xe7 1 2 c3 0-0 1 3 0-0
cxd4
White also holds the advantage after
13 ... c4 14 b4 cxb3 15 'iifx b3!, as 15 .. . c5
runs into 16 'ii'a3 !
1 4 cxd4
A very pleasant position for White.
Black is forced to play . .. c7-c5 before it
is prevented forever, after which
White is ahead in development, has
nice outposts for his knights, and can
utilise the c-file with his rooks.
Meanwhile, Black has problems find-
136
ing a useful square for the bishop on
c8.
14 . . ..c5 1 5 dxc5 'ir'xc5 1 6 °ir'd2 ltJc6
1 7 ltJxc6 'ir'xc6 1 a llac 1 'ir'b6 1 9 llc3
..id7 20 l:.fc1 lltba 21 b3 h6 22 lDd4
a5 23 a4?!
Around here White starts to drift
slowly but surely, until the players
agree a draw in a virtually level posi­
tion. 23 a3 a4 24 b4!, securing a pro­
tected passed pawn, would have been
preferable.
23 . . . 'ir'b4 24 'ir'f4 l::tca 25 llxca+
..ixca 26 h3 ..id7 27 llc7 .tea 2a
'ir'e3 llba 29 f4?! 'ir'a3 30 'ir'c3 'ir'a2
3 1 :ca 1h -1h
Game 81
Spassky-Drasko
Sarajevo 1986
1 e4 e6 2 d4 d5 3 ltJd2 ltJc6 4 lDgf3
lDf6 5 e5 ltJd7 6 lDb3
This is the theoretical recommenda­
tion for an advantage against 3 ...tbc6.
White lets the c 1-bishop out and pre­
pares for action against Black's ... c7-c5.
6 . . . a5 7 a4 b6 a ..if4 ..ie7 9 h4 ..ib7
10 c3 'ir'ca 1 1 ..ib5 ltJda
We are now treated to a bout of
 Third Move A lt e rn a t ives fo r Black

heavy duty manoeuvring from both After 4. . . tbf6 5 eS tbfd7 6 ild3 cS 7 c3

sides. The assessment of the position tbc6 White has slipped into a lbgf3
doesn't change. White has a nice, line against 3 ... tbf6, which is not eve­
healthy space advantage. ryone's cup of tea.
1 2 i.. g 5 i..f8 1 3 h5 h6 1 4 i.. h4 c6 1 5
i..e 2 i.. a 6 1 6 0-0 i..x e2 1 7 'ir'xe2
'ir'a6 1 8 'ir'd 1 c5 1 9 i..xd8 llxd8 20
lDe 1 i.. e 7 21 f4 f5 22 lDc2 l:.g8 23
ltJe3 lDf8

4 . . . c5 5 dxc5 ltJd7!
With the bishop already having
moved to e7, this is the most favour­
able way of recapturing on cS. We
now reach a position similar to those
24 <it>h2? that arise after 3 ... cS 4 exdS exdS 5
24 'it>f2 is more accurate. The white tbgf3 tbf6 (Games 64-67), the only
king may find itself surprisingly ex­ difference being that the light-squared
posed on the h-file. bishops remain on the board. In any
24 . . . g5 25 hxg6 h5! case, Black's position looks perfectly
Now 26 'ii'xhS tbxg6 27 'it>gl 'it>f7 al­ playable.
lows Black to take over operations on 6 exd5 exd5 7 lDb3 ltJxc5 8 lDf3 lDf6
the kingside, so Spassky bailed out 9 0-0 0-0 1 0 i..e3 ltJce4!
with a timely draw offer. 10 . ..tbxd3 1 1 'iifxd3 nets a bishop
26 lDc1 % -% for a knight, but helps White to co­
ordinate his remaining pieces. As is
Game 82 normal in these IQP positions, it is
Adams·Lputian generally useful for Black to keep the
Manila Interzonal 1990 pieces on.
1 1 i..e 2 i..d6 1 2 1:%.e 1 1:%.eS 1 3 i..d4
1 e4 e6 2 d4 d5 3 ltJd2 i..e 7 lDh5?!
Here Black starts to go somewhat
see following diagram
astray, weakening his position with­
4 i..d 3 out any provocation. The simple
One of the ideas behind the 3 ... ii.el 13 .. . ilfS would have been perfectly
line is seen after the normal 4 lbgf3. acceptable.

137
The F re n c h Tarra s c h
1 4 g 3 f 5 1 5 1Lf 1 1Le6?! 16 1Le5!
Adams is at his best in these type of
positions. Here he offers an exchange
of dark-squared bishops, which would
emphasise Black's newly formed
weaknesses.
1 6 . . .lDhf6 1 7 °il'd4 1Le7 1 8 h3 llf8
1 9 1Lg2 "il'e8 20 °il'd3 ltJd7 2 1 1Lf4
ltJdc5 22 ltJxc5 1Lxc5 23 lle2 °il'a4?
24 ltJg5! "il'c6 25 ltJxe6 "il'xe6 26
�h2 °il'f7
Visually Black's position looks
okay, but in fact it is about to fall
apart at the seams. Note that 26 ...i.xf2
27 l::!.xf2 lbxf2 28 i.xdS wins for
White.
27 f3 lDd6
28 llae 1
138
28 i.xd6 i.xd6 29 f4 is also very
strong, but the game move is even
more telling.
28 . . .ltJc4 29 c3 llad8 30 b4 1'.b6 3 1
lle7 °il'f6?
3 1 ...'ii'hS 32 .Uxb7 leaves White in a
position of total dominance. After
3 1 ...'ii'f6 the end is accelerated, as
Black's queen goes west.
32 :!. 1 e6 1 -0
Game 83
011-Topalov
Zaragoza 1992
1 e4 e6 2 d4 d5 3 lDd2 a6
Here is a quick look at two rare
moves:
a) 3 .. .fS has been played from time
to time, but it looks rather ugly and
gives White a nice juicy outpost on eS.
The game Karpov-Enevoldsen, Skopje
Olympiad 1972, which went 4 exfS
exfS 5 lbdf3 lbf6 6 i.gS i.e7 7 i.d3
lbe4 8 i.xe7 'ii'xe7 9 lbe2 'ii'b4+ 1 0 c3
'ii'xb2 11 0-0 0-0 12 c4 dxc4 13 i.xc4+
'it>h8 14 :b1 'iifa3 15 lbeS with a clear
advantage to White, has put most
people off this line.
b) 3 ...lbd7 4 lbgf3 lbe7!? is an un­
usual idea of Petrosian. Now after 5 c3
cS 6 eS lbc6 we have suddenly reached
a main line 3 ...lbf6, with White's
king's knight committed to f3. Instead
White can consider keeping the ten­
sion with 5 g3 cS 6 i.g2 cxd4 7 lbxd4,
or 5 i.d3 cS 6 c3, both of which look
slightly better for the first player.
After 3 ... a6 most White players con­
tinue with 4 lbgf3 cS, transposing to
lines considered in Chapters 8 and 9.
Here we look at another possibility.

4 e5 1Ld 7!
A cute idea. Black's notorious prob­
lem child will pop out at bS, where it
can exchange itself for one of White's
minor pieces. This move seems to se­
cure a comfortable game for Black.
5 lDe2 c5 6 lDf3 1'.b5 7 c3 ltJc6 a
1Le3 cxd4 9 ltJexd4 1Lxf1 1 0 <it>xf1
ltJge 7 1 1 g3 ltJxd4 1 2 cxd4 ltJc6 1 3
<it>g2 iLe 7 1 4 :!.e 1 'ir'b6 1 5 lle2 :ca
1 6 lie 1 0-0 1 7 h4 h6
White has the normal space advan­
tage, but with the exchange of two sets
of minor pieces, this becomes rather
insignificant. Now the game heads for
a predictable draw as all the rooks are
hoovered off.
1 a h5 'ir'b5 1 9 llec2 ltJa7 20 lDh2
Third M o ve A lt e rn a tives for B l a c k
llxc2 2 1 llxc2 llca 2 2 llxca+ ltJxca
23 b3 ltJa7 24 ltJg4 ltJc6 25 'ir'd2
°ir'b4 26 'ir'c 1 'ir'a3 27 'ir'c2 'ir'b4 2a
'i'c 1 'ii'a3 29 'ii'c2 °ir'b4 1h -1h
4...i.d7 looks like a complete an­
swer to 4 eS.
Game 84
Shirov-Hi.ibner
Munich 1993
1 e4 e6 2 d4 d5 3 lDc3 dxe4 4 ltJxe4
1Ld7
The classical 4 ...tbd7 is considered
in the next game.
5 lDf3 1Lc6 6 1Ld3 lDd 7
This is the 'Fort Knox Defence', so­
called because Black achieves a super
solid structure, albeit with a rather
passive position. This line has found
some distinguished advocates, includ­
ing FIDE World Champion Anatoly
Karpov.
7 0-0 lDgf6 a ltJg3
8 tbxf6+ 'ii'xf6 is fine for Black, who
has some annoying pressure on f3.
However, Black must beware after the
enterprising 8 tbegS, e.g.
a) 8 ... h6 9 tbxe6! fxe6 10 i.g6+ We7
1 1 :el with a very strong attack.
139
Th e F re n c h Ta rra s c h
b) 8 ...i.d6! is more accurate.
Korchnoi-Dreev, Brno 1992, contin­
ued 9 l':.e 1 h6 10 lbh3 i.xf3 1 1 'ii'xf3 c6
12 lbf4, and now instead 12 ...0-0 13
lbhS, which gave White a promising
attack, Black should have played
12 ...'ii' c7 or 12 ... 'ii'aS, followed by
... 0-0-0, with about equal chances.
8 . . . ..ie7 9 b3 0-0 1 0 ..ib2 ..ixf3 1 1
'ir'xf3 c6
This is all very typical of the Fort
Knox. White has the bishop pair and
extra space, but it is still difficult to
break down Black's position.
1 2 c4 :I.ea 1 3 llfe 1 lDfB 1 4 h4! ? 'ir'c7
1 5 llac1 ltJg6!
Good defence. Against 15 ... l':.ad8
Shirov had planned 16 dS! cxdS 17
cxdS 'ii'd 7 1 8 d6! and now:
a) 1 8 ... i.xd6 19 i.xf6 gxf6 20 lbhS
i.eS 2 1 :xeS wins for White.
b) 18 ...'iifxd6 19 i.bS lb8d7 20 l':.edl
'iifb 6 2 1 i.xf6 'iifxbS 22 i.xe7 llxe7 23
l':.c7! and the pinned knight on d7
causes Black all sorts of problems.
1 6 h5 lDh4 1 7 'ii'e3 ..id6! 1 8 ltJe4! ?
An enterprising idea. Initially White
gives up the exchange, but later he
wins Black's knight on hS, which runs
out of squares. However, the position
140
still remains quite unclear, as Black
picks up a couple of pawns while the
white king rushes to safety.
1 8 . . ...if4 1 9 lDxf6+ gxf6 20 'ir'e4 f5
2 1 'ir'e2 ..ixc 1 22 :!.xc1 f6! 23 g3
'Oti>f7! 24 gxh4 'ir'f4 25 lle 1 llg8+! 26
'itf 1 llae8 27 'ir'e3 'ir'xh4 28 <it>e2!
'ir'xh5+ 29 <it>d2 llg2! 30 :!.e2 'ir'g5 31
f4 :!.xe2+
Shirov gives 3 1...'ii'g3 !? 32 'ii'xg3
:xg3 33 l':.h2 'it>g6 as unclear.
32 ..ixe2 'ir'g2 33 'ir'f3 'ir'xf3 34 ..ixf3
The two bishops give White a small
edge in this tricky endgame, but Hilb­
ner now starts to make some serious
inaccuracies, which make Shirov's task
much easier.
34 . . .<it>g6? ! 35 <it>e3 lld8 ?! 36 ..ic3
<it>f7 37 a4 'Oti>e7 38 a5 <it>d7?! 39 d5!
exd5 40 cxd5 cxd5 4 1 ..ixf6 :!.f8 42
..ih4! <it>c6 43 <it>d4 <it>b5 44 ..ixd5
<it>xa5 45 ..ixb7 llb8 46 ..id5 h5 47
..ie6 <it>a6 48 ..ic4+ <it>b6 49 <it>e5 a5
50 <it>xf5 <it>c5 51 <it>e5 1 -0
Game 85
Aseev-Nevostrujev
St Petersburg 1994
1 e4 e6 2 d4 d5 3 ltJc3 dxe4 4 ltJxe4

tt:Jd7 5 tt:Jt3 tt:Jgf6 6 tt:Jxf6+ tt:Jxf6
7 i.b5+
A small refinement, which doesn't
make much difference, apart from
adding an extra move to each player's
scoresheet. The normal move order is
7 .td3 cs.
7 . . . c6 8 i.d3 c5 9 dxc5 i.xc5 10 0-0
10 'ife2!?, retaining the option of
castling either side, is an ambitious
alternative to 10 0-0. Here are a couple
of examples:
a) 10 ... 0-0 1 1 .5lg5 'i!faS+ 12 c3 .5le7
13 lbes h6 14 .th4 l:rd8 15 0-0 'ifcl 16
l:radl b6 17 l:rfe l .5lb7? 18 lbxf7! and
White was clearly better in the game
Karpov-Speelman, Reykjavik 199 1, as
18 ...@xf7 fails to 19 'ifxe6+ @f8 20
.5lxf6 .5lxf6 2 1 .5lc4.
b) 10 ...'ifc7 1 1 .5ld2 .5ld7 (1 1 ...0-0
looks quite sensible and may well be
stronger than 1 1 .. ..5ld7) 12 0-0 .5ld6 13
l:tfe l 0-0-0?! (Nunn gives 13 ....5lc6 14
lbe5 .5lxe5 15 'ifxeS 'ifxeS 16 l:rxeS as
just slightly better for White) 14 l:radl
.5lc6 15 h3 h6?! 16 b4! and White's
offensive on the queenside will be
first, as in Nunn-Korchnoi, Amster­
dam 1990.
Th ird M o ve A lt e rn a tives for B l a c k
10 . . . 'ii'c 7
Also possible is 10 ...0-0 1 1 'ife2 (1 1
.5lg5!?) 1 1 . ..h6 12 b3 b6 13 .tb2 .tb7
14 l:radl 'ife7 15 c4 l:rfd8 16 lbeS l:rd6
and Black had equalised in I. Rogers­
Vaganian, Manila Interzonal 1990.
1 1 i.g5 b6 1 2 'ii'e 2 0-0! 1 3 tt:Je5
13 .5lxf6 gxf6 14 'ife4 looks tempt­
ing, but the rook is poisoned! 14 .. .fS
15 'ifxa8 .5lb7 16 'ifxa7 bS traps the
queen.
1 3 . . . tt:Jd7?
A blunder which immediately de­
cides the game. Black should have
played 13 ... .5lb7, not fearing 14 .5lxf6
(14 l:radl is less committal) 14 ...gxf6 15
'i!fg4+ @h8 16 'ifh4 fS, when White
should probably take the perpetual,
otherwise he is likely to stand worse.
14 i.f4 i.b7
Or 14 ... lbxeS 15 .5lxe5 'ife7 16 b4!
.5lxb4 17 'i!fe4 and White wins.
1 5 i.xh7+! 'it>xh7 1 6 tt:Jg6!
Black has only one way to deal with
the two threats .
1 6 . . . i.xf2+ 1 7 l:!.xf2 'ii'c5 1 8 i.e3
'ii'a 5 1 9 b4! 1 -0
After 19 b4 'i!fdS 20 c4 Black cannot
cover hS any longer.
14 1
The Fre n c h Ta rra sc h

Summary
Out of all of Black's third move alternatives, 3 ... .5le7 looks to be the most under­
rated and it is surprising that it is not seen more often. After this move Black
often reaches uncharted positions similar to those that arise after 3 ...lbf6 and
3 ... cS. On the other hand, 3 ... lbc6 should cause well prepared White players no
problems whatsoever.

1 e4 e6 2 d4 d5 3 tt:Jd2

3 ...tt:Jc6
3 ... .5le7 - Game 82
3 ... a6 (D)
4 lbgf3 cS
5 exdS exdS Games 68-70
-

5 dxcS Game 77
-

4 eS - Game 83
3 ... dxe4 4 lbxe4 (D)
4 ....5ld7 Game 84
-

4 ...lbd7 - Game 85
4 tt:Jgf3
4 c3 - Game 79
4 . tt:Jf6 5 e5 tt:Jd7 (D) 6
. . tt:Jb3
6 .tbs Game 80
-

4 ... a5 Game 81
-

3 a6
. . .
4 ttJxe4 5 ltJd7
. . .

142
 INDEX OF GAMES I

Adams-Dreev, Wijk aan Zee 1996 ................................................................................. 76

Adams-Levitt, Dublin 1 993 ......................................................................................... 125
Adams-Lputian, Manila Interzonal 1 990 .............. ..................................................... 137
Adams-Speelman, England 1 991 ................................ . ..................... ............................. 77
Adorjan-Watson.J, New York 1981 ......... .................. ............................. ...................... 70
Agnos-Lputian, Linares 1 996 ................................ ...................................................... 104
Alvarado-Matamoros, Telde 1993 ............. . . . . ............................................................... 29
Anka-Almasi.I, Hungary 1997....................................................................................... 93
Antonov-Balinov, Djuni 1 987 ............................................. . . . ...................................... 20
Aseev-Nevostrujev, St Petersburg 1 994 ........................................ .............................. 140
Asrian-Lputian, Yerevan 1 995 .................................................................................... 101
Balashov-Borngasser, Wis/a 1992 .................. ............................................................... 62
Beliavsky-Kindermann, Munich 1991 .......................................................................... 69
Bowden-Levitt, England 1997 ........................ ................ ............. . . . . . . .. . ...... . . . . . . . ............ 73
Brynell-Schmidt, Naestved 1988 ............ ..................................................................... 123
del Campo.M-Escobedo Tinajero, Mexico 1991 ........................................................ 92
Donchev-Eingom, Debrecen 1992 . . ............................................................................ 120
Dvoirys-Tondivar, Leeuwarden 1 994 ......................................................................... 108
Dvoirys-Ulibin, Cheliabinsk 1990. . ............................................................................... 23
Emms-Bibby, British Championship 1990 ..... ............................................................... 81
Emms-Clarke.S, British Championship 1991 ................................ ............................... 60
Emms-Hansen.LB, Bronshoj Jubilee 1 995 ................................................................... 95
Emms-Kosten, British Championship 1985 .................................................................. 65
Emms-Menoni, Montecatini Terrn e 1 996 ..................................................................... 34
Emms-Prandstetter, Barcelona 1993 .......................................................................... 106
Emms-Strater, Hastings 1996.......................... ............................................................... 56
Geller-Dreev, New York 1 990 ..................................................................................... 1 1 8
Georgiev.Kr-Dolmatov, Sofia 1985 ................................ ............................................. 40
Gurevich.1-Bareev, Biel 1 993 ...................................................................................... 1 1 6
Gurevich.1-Bareev, Hastings 1992 .............................................................................. 1 14
Henao-Sequera, MeriM 1992 . . ............ .................... ......................... .............................. 2 7
Jansa-Brunner, Bad Worishofen 1989 ........................................................................... 76
Jansa-Votava, Prague 1 993 107
.................. ......................................... ................................

Karpov-Bareev, Linares 1 994....................................................................................... 1 1 5

143
The Fre n c h Tarra s c h

Kolev-Herraiz, Linares Open 1 996 . . . ............................................................................ 58

Komarov-Barsov, Reims 1994 ....................................................................................... 21
Kopilov-Kahn, Correspondence 1 991 ............................................................................ 84
Komeev-Knaak, Bad Worishofen Open 1992............................................................... 50
Kramnik-Ulibin, USSR Championship 1991 ............................................................... 18
Krishilovsky-Doroshkievich, Novgorod 1997................... ....................................... 135
Kuijf-Uhlmann, A msterdam 1 990 .............. . ............................................................... 130
Lputian-Agzamov, USSR Championship 1 985 ........................................................... 54
Majer-Ruzicka, Correspondence 1 993 ...... ..................................................................... 43
Maksimovic-Farago, Copenhagen Open 1989 ............................................................. 59
Marjanovic-Ti mman, Sarajevo 1984 ............................................................................ 38
Menke-Wilde, Correspondence 1957 ............................................................................. 55
Mohrlok-Schrancz, Correspondence 1989 .................................................................... 68
Moya-Roldan, Correspondence 1984 . . . . . ........ ...................... ........................................ 122
Muratov-Bata, Correspondence 1967 . . . . ........... .............. . . . . . . . . . . . ........................ . . . . . . . . . . . . 46
Nijboer-Farago, Dieren Open 1988 ........................... ................................................... 4 1
011-Topalov, Zaragoza 1992 . . . . . . . . ................................................................................ 138
Pirrot-Hertneck, German Bundesliga 1 990 ........................... ........... ........................... 53
Popovic-Nikolic.P, Yugoslavia 1991 .......................................................................... 128
Psakhis-Herzog, Vienna 1991 ....................................................................................... 83
Ravi.T-Dolmatov, Calcutta 1996................................................................................ 109
Read.M-Goncalves, Correspondence 1989 ................................................................... 99
Rozentalis-Hergott, Montreal 199 5 .............................................................................. 3 1
Rublevsky-Bareev, Rubinstein Memorial 1997 ......................................................... 1 2 1
Rublevsky-Beliavsky, Novosibirsk 1995 . . .................... . . .............................................. 97
Rublevsky-Gleizerov, USSR 1991 ................................................................................ 16
Seul-Zach, Biel 1 997 ........................................................................................................61
Shirov-Hiibner, Munich 1 993...................................................................................... 139
Slobodjan-De la Villa Garcia, Pamplona 1996........................................................... 67
Slobodjan-Luther, Lippstadt 1997................................................................................. 89
Smagin-Dolmatov, USSR Championship 1986 ........................................................... 37
Smagin-Lalic.B, Sochi 1987 .......................................................................................... 1 03
Sowray-Martin.B, London 1 996 ................................................................................... 45
Spassky-Drasko, Sarajevo 1986.................................................................................... 136
Stigkin-Basin, St Petersburg 1 992 .................................................................................. 28
Svidler-Glek, Haifa 1 996 ........................................................... ..................................... 79
Svidler-Ulibin, Russian Championship 1 994 ................................................................ 25
Timoshenko.G-Danielian, Cappelle la Grande 1 994 ................................................. 80
Tiviakov-Chemin, Podolsk 1 993 .............. .................................................................... 88
Tiviakov-Dreev, Wijk aan Zee 1 996 ........................................................................... 1 1 2
Todorovic-Dimitrov, Belgrade 1 995 ..... ....................................................................... 5 2
Van der Wiel-Seirawan, Biel 1985 .............................................................................. 1 2 1
Wells-Fries Nielsen, Copenhagen Open 1 995 .............................................................. 4 7
Wessman-Wiedenkeller, Swedish Championship 1989............................................... 87
Winsnes-Lein, Gausdal 1 990 .......................................................................................... 91
Wolff-Yermolinsky, USA Championship 1 993 . ........................................................ 129
Xie Jun-Bronner, Shanghai 1 995 ................................................................................ 133
Yandimerov-Nikitin.R, Russia 1997............. . .............................................................. 94
Yemelin-Kostenko, Szeged 1994 .................................................................................... 36
Yudasin-Gulko, Biel 1 993 ............................................................................................ 127
Zapata-Dolmatov, Ti/burg 1 993 ................................................................................... 89

144