1 d4 d5 2 c4 c6

To my parents.

The Slav

CHESS PRESS OPENING GUIDES

Other titles in this series include:
1 90 1 25 9 05 6

Caro-Kann Advance

Byron Jacobs

1 90 1 25 9 06 4

Closed Sicilian

Daniel King

Dutch Leningrad

Neil McDonald

1 90 1 25 9 0 3

1 90 1 25 9 1 0 2

French Advance

Tony Kosten

1 90 1 25 9 02 1

Scandinavian

John Emms

1 90 1 25 9 OS 0

Semi-Slav

Matthew Sadler

1 90 1 25 9 0 1 3
1 90 1 25 9 09 9

Sicilian T aimanov

James Plaskett

Trompowsky

Joe Gallagher

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

Chess Press Opening Guides

The Slav

Matthew Sadler

ir
[1Illj

The Chess Press, Brighton

First published 1 997 by The Chess Press, an imprint of First Rank Publishing,
23 Ditchling Rise, Brighton, East Sussex, BN1 4QL, in association with
Everyman Books plc
Reprinted with corrections 1 999
Copyright 1 997 Matthew Sadler
Distributed by Everyman Chess, Gloucester Mansions,
1 40a Shaftesbury Avenue, London WC2H SHD.

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

retrieval system or transmitted in any form or by any means, electronic,
electrostatic, magnetic tape, mechanical, photocopying, recording or
otherwise, without prior permission in writing from the publishers .
A CIP c atalogue record f o r t h i s b o o k i s available from the British Library
ISBN 1 90 1 259 00 5

Cover design by Ray Shell Dcsign

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

CONTENTS
1 d4 d5 2 c4 c6

Bibliography

Introduction

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

xc4 b4 8 0-0) : Black plays to prevent e4

13

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

xc4 b4 8 0-0) : Black allows e3-e4

27

The New Main Line (3 ttJf3 ttJf6 4 ttJc3 dxc4 5 a4 fS 6 ttJeS) :

Black fights for control of e4

39

The New Main Line (3 ttJf3 ttJf6 4 ttJc3 dxc4 5 a4 fs 6 ttJeS) :

Black counterattacks

53

The Smyslov Variation (3 ttJf3 ttJf6 4 ttJc3 dxc4 5 a4 ttJa6)

62

The Bronstein Variation (3 ttJf3 ttJf6 4 ttJc3 dxc4 5 a4 g4)

72

The 4 . . . a6 Slav (3 ttJf3 ttJf6 4 ttJc3 a6) : White plays 5 e3

87

The 4 . . . a6 Slav (3 ttJf3 ttJf6 4 ttJc3 a6) :

Aggressive options for White

98

The Exchange Variation (3 cxds cxdS)

1 08

10

Move-Orders and Transpositions

1 18

11

Odds and Ends

130

2
3
4

Index of Complete Games

142

BIBllOGRAPsH Y

Books

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

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

Periodicals

In/ormator
ChessBase Magazine
New In Chess Yearbook
British Chess Magazine
Chess Monthly

INfRODUCrlON

1 d4 d 5 2 c4 c6

The skill of preparing an opening is

frequently misunderstood: many play
ers (including some strong grand
masters) believe that to play an open
ing well, it is necessary to analyse a
great many variations; that no prepa
ration is complete without at least one
queen sacrifice and that to stop before
move 20 is akin to criminal negli
gence. I know from experience that
the sheer volume of opening theory
can be overwhelming, and this is es
pecially true for the non-professional
player who has little time to keep up
with the latest fashions.
And yet, having been a professional
player now for six years, I know that
I have won more games from 'nor-

mal' openings than from any 30-move

piece of analysis (and it's not because I
haven't done any!) . The brilliant 'I
had this position after move 80 on my
board at home' games that we see in
magazines are the exceptions: beauti
ful, treasured by every chessplayer,
but very, very rare. Chess is a sport
and most games are a struggle, and we
win games because we fight harder
than our opponents, or because we
understand the position better.
In my opinion, opening prepara
tion can be successfully reduced to
three simple steps:
1. Knowing the main aim of our
opemng.
2. Knowing the value of move
orders.
3 . Understanding typical positions.
Therefore, let's apply these ideas to
the Slav.
Opening Aims

With 2 c4, White challenges the black

centre. The natural 2 . . . e6, allowing
Black to develop his kingside pieces,
has the drawback of blocking the
light-squared bishop inside the pawn
chain. 2 . . . c6 aims to hold the centre,
to develop the light-squared bishop
9

Th e S l a v

outside the pawn chain, and then to

play . . . e7-e6 and conclude the black
development. However, the course of
chess ideas, like love, never runs
smoothly! Black must be careful
when he develops his light-squared
bishop: after 1 d4 ds 2 c4 c6 3 CDf3
CDf6 4 CDc3

that Black can only play a quick

. . . fs if he can successfully defend b7
with his queen. Thus, 1 d4 ds 2 c4 c6
3 CDf3 CDf6 4 e3 fs

is fine for Black since S cxdS cxdS 6

'iYb3 can easily be met by 6 . . . 'iYc7;
however 1 d4 ds 2 c4 c6 3 CDc3 CDf6 4
e3 fS?!
Black would like to play 4 . . . fS , but
he will have great difficulty defending
b7 after S cxds cxds 6 'iYb3 !

Now 6 . . . 'iYb6 loses a pawn to 7

CDxds 'iYxb3 8 CDxf6+ exf6 9 axb3 and
6 . . . b6 weakens the queenside light
squares too much: 7 e4! dxe4 8 CDeS e6
(to stop 'iYxf7+ mate) 9 bs+ CDfd7 10
g4 g6 1 1 h4! , intending h4-hS, trap
ping the bishop. The general rule is
10

S cxds cxds 6 'iYb3 is not good,

since 6 . . . 'iYc7 loses a pawn to 7 CDxdS .
So how can Black carry out his main
idea? Black either has to stop White
from playing 'iYb3 , or he has to find a
good way to defend b7. This is a typi
cal opening dilemma: whether to pre
vent an opponent's threat directly, or
whether to arrange the pieces in such
a way that the threat is nullified.

In t r o du c tio n

The main line of the Slav runs

dS 2 c4 c6 3 tt'lf3 tt'lf6 4 tt'lc3 dxc4.

d4

First, Black wins a pawn and

threatens . . . b7-b5, making this gain
permanent. Second, the b3-square is
cunningly taken away from the
queen, which means that White can
not attack b7, and hence that . . . f5
becomes possible. While White recap
tures the c4-pawn, Black will develop
the light-squared bishop to f5 or g4
and will be looking to complete his
kingside development: 5 a4 (surroun
ding the c-pawn by preventing . . . b7bS) 5 . . . f5 6 tt'le5 (intending tt'lxc4) or
6 e3 (intending xc4) 6 . . . e6 are the
main lines. So far I have been very
enthusiastic about Black's strategy,
but now I have to reveal the downside
of his play. This sort of schizophrenia
is necessary when you play both sides
of the Slav, as I do!
4 . . . dxc4 relinquishes control of e4,
which makes it easier for White to
cramp Black with two central pawns
on d4 and e4. But White must be care
ful that his pawns do not become
weaknesses as Black first immobilises,
then attacks them. The bottom line is
that the player who has the better

understanding of the line will get the

best results with either colour.
The second idea is to play 1 d4 d5 2
c4 c6 3 tt'lf3 tt'lf6 4 tt'lc3 a6.

The move 4 . . . a6 was first played in

Britain by Grandmaster Jon Levitt,
but it is Grandmaster Julian Hodgson
who has upheld this variation at the
highest level, and introduced the most
significant ideas.
The original idea of 1 d4 d5 2 c4 c6
3 tt'lf3 tt'lf6 4 tt'lc3 a6 was to meet iVb3
with . . . b7-b5, moving the pawn to a
protected square. However, ways
were found to exploit the drawbacks
of . . . b7-b5: the dark-square weak
nesses on c5, b6, as and the slightly
exposed black queenside. Then, in a
brilliant piece of unstereotyped think
ing, Hodgson realised that . . . a7-a6
could allow the rook to defend b7
from a7. No one had dared to play
such a strange move before, but Julian
did, and this has made some previ
ously dodgy lines completely viable.
However,
although
avoiding
. . . d5xc4 helps to prevent e2-e4, Black's
position is less dynamic than in the
. . . d5xc4 lines, as it is much harder to
break against White's centre with
11

Th e Sla v

. . . c6-cS and . . . e7-eS. My own prefer

ence as Black is for the 4 . . . dxc4 lines,
as they are richer in content and offer
a wider range of possibilities to suit
many different styles.
Move-Orders

Move-orders are a much underrated

part of opening preparation. Oppo
nents don't always play fair! Imagine
the scene: you sit down to play, con
fident that you know your opening at
least as well as your opponent, and
what happens? He plays the opening
in some unusual move-order, and you
emerge a bit dazed to find yourself
playing a different line to the one you
wanted! And unless you work out
your move-orders thoroughly, this
will continue to happen, time and
time again.

12

S o how can this happen in the Slav?

If you want to play the 4 . . . a6 Slav,
then there is nothing that White can
do to muddy the water, which is one
of the attractions of this line. 4 . . . dxc4
lines, however, require some care.
First, White can try to sidestep them
by playing an early e2-e3, protecting
c4, e.g. 3 !bc3 !bf6 4 e3 . Black has
many reasonable moves here, but
none of them fit in with the idea we
want to play. The other way for
White to play is c4xdS, leading to the
Exchange variation; 4 . . . a6 Slav fans
should study these positions particu
larly carefully, since this sort of posi
tion is very typical of this line, and
there are many transpositions.
Understanding Typical Positions

Well, for this part, read on . . .

CHAPTER ONE
The Old Main Line:
Black plays to prevent e3-e4

1 d4 d5 2 c4 c6 3 LiJf3 LiJf6 4 LiJc3

dxc4 5 a4 .i.f5 6 e3 e6 7 .i.xc4 .i.b4

8 0-0

This traditional system of devel

opment for White is especially popu
lar at club level. White quietly recap
tures the pawn and puts his king to
safety before he starts his plan of e3e4, to obtain a 'perfect' pawn centre.
This chapter examines Black's at
tempts to prevent White from achiev
ing this goal.
First, we need to ask ourselves a
few questions in order to establish our
approach:
Question 1: How will White try
and achieve e3-e4?
Answer: White has two major ap
proaches:
a) 'i'e2. This is the most dangerous
idea, which we shall examine first.
b) ctJh4, to remove the bishop on
f5, which is helping Black to prevent
e3-e4.
Question 2: How can Black fight
against 'i'e2 and e3-e4?
A nswer: Black has three pieces at
tacking the e4-square: the bishops on
b4 and f5 and the knight on f6. When
White plays 'i'e2 he is supporting the
e3-e4 push with only two pieces: the

queen and the knight on c3. How

ever, he will achieve the e3-e4 advance
with tempo because the e-pawn at
tacks the bishop on f5. If Black wants
to, he can simply pre-empt this by re
treating the bishop to g6, so that e3-e4
no longer attacks the bishop. Now if
White plays e3-e4 regardless, Black
can win a pawn by playing . . . xc3
and . . . ctJxe4.
Question 3: What move-order
should I play this in?
Answer: My own favourite has
been to play 8 . . 0-0 9 'i'e2 g6;
8 . . ctJbd7 9 'i'e2 g6 is sharper since
White can offer a dangerous pawn
sacrifice.
.

Gamel
Richardson-Sadler

Islington Open 1995

This was a crucial game for me: I was

leading the Islington Open by only
half a point and only a win would
guarantee first place. However, even
more importantly, only a win would
be good enough to pip Keith Arkell
for the Leigh Grand Prix!
1 d4 d5 2 c4 c6 3 LiJf3 LiJf6 4 LiJc3
dxc4 5 a4 .i.f5 6 e3 e6 7 .i.xc4 .i.b4
13

Th e Sla v

S 0-0 0-0 9 'iVe2 jL g 6 1 0 LDe5

10 e4? ! xc3 ! 1 1 bxc3 tbxe4 wins a

safe pawn. Consequently, White re
verts to 'Plan B ' : he exchanges his
knight on f3 for my bishop on g6 and
removes an attacker of e4.

1 0 . . . LDbd7 1 1 LDxg6 hxg6 1 2 .lci.d 1

12 e4 tbb6! wins a pawn as 13 e5 (or

13 Mdl xc3 14 bxc3 tbxc4 15 'i'xc4
tbxe4) fails to 13 . . . 'i'xd4 14 xe6 fxe6
15 exf6 'i'xf6 with a clear extra pawn.
1 2 ...'iVe7!?

An interesting move, though I

imagine that it is not the most accu
rate - for Kramnik's 12 . . . 'i'a5, see the
next game.
1 3 e4

ing t o prevent e3-e4? Well, smce

White's aim was to push a pawn
within his own territory, it was al
ways unlikely that we could prevent
it for ever. However, by resisting for
as long as possible, we have forced
White to make a concession, namely
that he has had to play tbe5xg6 before
being able to play e3-e4. Although
White gains the two bishops with this
manoeuvre, he exchanges off the piece
that would be most affected by the e3e4 advance; on g6, the bishop has little
scope if White can maintain his pawn
on e4. Moreover, the departure of the
knight from f3 means that White loses
some control over the central dark
squares, d4 and e5 . This last point is
seen to great effect in this game.
1 3 . . . e5 1 4 d 5 nacS !

I spent a great deal of time at this

stage and realised that I had to force
White to release the tension in the
centre and play d5xc6. The explana
tion for this has a lot to do with the
central dark squares: without the d
pawn, I can transfer a knight to e6
(via c5) and exploit the outpost that
my pawn on e5 creates on d4. By
placing my rook on c8, I was hoping
to get my opponent worried about
possible threats on the c-file, in order
to tempt him into playing d5xc6.
1 5 jL g 5 :!:;lfdS 1 6 dxc6?

Wait a minute! Wasn't Black play14

Here it is! After this mistake, White

has to be very careful to avoid a dis
advantage. The correct plan is ex
tremely ingenious. Black has two ave
nues of pressure: he has possible
threats along the c-file and he can de
velop some pressure against e4 by
means of . . . tbc5. How can White deal

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

with both these threats? With the

manoeuvre 16 l':rd3 ! 4.Jc5 17 l':re3 ! ! On
e3, the rook covers e4 and defends the
knight on c3 along the rank, thus pro
tecting White's queenside against c-file
play. White is slightly better after 16
l':rd3, but the game is still very compli
cated.

game, I was very impressed with

White's attitude: realising that his po
sition had worsened considerably,
White regrouped and concentrated
totally on defence.
1 9 iLxe6 xe6 20 .!:Id 3 gxd 3 2 1
xd 3 b3 2 2 .t:!.b 1

l:!:b8 2 3 iL d 2

iLa 5 !

1 6 . . bxc6 !

My opponent had underestimated

this recapture. Although it weakens
Black's queens ide pawn structure,
Black protects the central light
squares, over which he previously had
little control due to the exchange of
his light-squared bishop. This move is
so strong, because White's queenside
is so weak: the pawn on a4 gives
Black a comfortable slot on b4 for the
queen, from which it can attack the
a4- and b2-pawns.
1 7 .!:Id 3 ? CLlc5 1 8 .!:Ih 3 ? CLle6 !

White's 17th and 1 8th moves were

excessively optimistic as he had no
chance of an attack along the h-file.
Meanwhile, Black threatens . . . 4.Jd4.
With his control of the d-file, and
White's weakened queens ide as a clear
target, I believe that Black can already
be thinking about victory. During the

Protecting the d8-square, so that the

rook can use either of the open files
on the board.
24 f3 .!:Id8 25 e2 CLle8 !

Black's knight is the least active of

his piecs. The text prepares to bring
it to d4 via c7 and e6. When this hap
pens, all of Black's troops will be on
their optimal squares.
26 iLe 1 CLlc7 27 l:!:c 1 CLle6 28 c2
b6+ 29 f2 CLld4

Intending 3o . . . ihc3 31 xc3 4.Je2+

winning the exchange, as 32 '>jVxe2 is
impossible since the queen is pinned
to the king.
30 h 1 b3 31

b 1 l:!:b8 3 2 h3

c4 3 3 f4 exf4 34 xf4 ge8 3 5

d 1

iLxc3 3 6 iLxc3 CLle2 3 7 f3

gxe4

The first weak pawn falls.

38 a5 a6 39 d3 CLlxc3 40 xc4
gxc4 41 bxc3 .!:!.c 5 !
15

Th e Sla v

what both sides should be aiming for

in this variation.
... .
.

Game 2

Karpov-Kramnik

Monte Carlo (blindfold) 1995

1 d4 d 5 2 c4 c6 3 CZJf3 CZJf6 4 CZJc3
dxc4 5 a4 jLf5 6 e3 e6 7 jLxc4 3l.b4
8 0-0 0-0 9 e2 jLg6 1 0 CZJe 5 CZJbd7
1 1 CZJxg 6 hxg6 1 2 .lld 1 a 5 !

I think that the rook ending is

winning now. 4 1 . . .Mxc3 42 Md8+ h7
43 Ma8 would have regained the a
pawn, but now after 42 Md8+ h7 43
Ma8 , Black can simply play 43 . . . Mxa5.
42 .l:i.d8+ 'it> h 7 43 gc8 g 5 44 'it>g 1
'it>g6 45 'it>f2 f5 46 c7 f6 47
.l:i.xg 7 .l:i.xa 5 48 e3 .l:!.e5+ 49 'it>d4
f4 50 ga7 ge6 !

5 1 b7

5 1 Mxa6 c5+! wins a rook.

5 1 . . .'it> g 3 5 2 .l:i.b2 .l:i.e 5 5 3 llb6 c5+
54 'it>c4 f 5 5 5 gxa6 'it>xg 2 5 6 .llg 6 f4
57 h4 f3 58 .llx g 5+ .l:i.xg 5 59 hxg5
f2 60 xc5 f 1 61

c4 f5+ 6 2

'it>d6 g 6 + 0 - 1

This was a very important game for

me, and an instructive example of
16

This move i s more active, and

probably more logical, than 12 ... 'iYe7.
While Black is not threatening to win
a pawn immediately with 13 . . . xc3
due to 14 bxc3 'iYxc3 15 d2 'iYc2 16
d3 ! 'iYb2 17 Mdb 1 , winning the
queen, it does prevent 13 e4, as with
White's centre slightly weakened,
Black can get away with taking the
pawn: 13 . . . xc3 14 bxc3 'iYxc3 1 5
d2 'iYxd4 (15 . . . 'iYc2!?) 16 b4 'iYe5
17 xf8 Mxf8 when with two pawns
for the exchange, Black stands very
well. Note that 13 ctJa2 allows
13 . . . 'iYxa4 14 ctJxb4 'iYxa 1 15 ctJa2
(hoping to trap the queen) 15 . . . 'iYb 1 ! ,
escaping t o f5!
1 3 jLd2

Protecting c3, and intending e3-e4,

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

but . .

3 3 a2 h2+ 34 \t>f 1 xh3+ 3 5

1 3 . . . e5 1 4 d 51Iad8

W e 2 tLJ e 5 36 iL e 2 xg4+ 3 7 W d 2

14 . . . cxds? 15 ctJxds ctJxds 16 xds

'i'xds 17 xb4 wins for White.

!!ed8+ 38 We 1 xd 1 + 3 9 iLxd 1 g 1

40 iLd2 tLJd3+ 4 1 <;t>e2 ldb2+ 0 - 1

1 5 dxe6 bxe6 1 6 iLe1 e4 !

Securing an outpost on d3 for the

knight.

S o White's plan of 9 'iVe2 and 10

ctJes seems harmless. Let us take a
look at the more direct 9 ctJh4.
Game 3
Yusu pov-Kram n ik

Riga 1995
1 d4 d 5 2 e4 e6 3 tLJf3 tLJf6 4 tLJe3
dxe4 5 a4 iLf5 6 e3 e6 7 iLxe4 iL b4
8 0-0 0-0 9 tLJh4!

The most testing idea: White elimi

nates the bishop on fs without wast
ing time on 'iVe2.
1 7 iLb3 e5 1 8 1Le2 iLd6 1 9 g 3
tLlc5 2 0 \t> g 2 f 5 2 1 lId2 1L e 5 ! 22
gad 1 J::i: b 8 !

The white queenside is looking

very weak.
23 iL b 1

iLxe3 24 bxe3 tLJxa4 25

ga2 tLJb2 26 U d 2 tLJe4 2 7 1::i: d 1 tLJe 5

28 h3 tLJf3 29 lIa4 tLJ g 5 30 g4 e5
3 1 e2 lIfe8 3 2 !!xa7 tLJf3

9 . . . tLJbd7 ! ?

A typical stratagem: Black's dou

bled f-pawns will take over the
bishop's task of preventing e3-e4.
Question 4: What is wrong with
9 . g6?
Answer: White can play 10 ctJxg6
hxg6 1 1 'iVc2!
Question 5: Why is it important to
put the queen on c2 and not e2?
Answer: First, the queen neutralises
.

Threatening . . .'iVh2+. Now Black is

just winning.

17

Th e Sla v

Black's most active plan of . . . 'iYa5,

threatening . . . jl,xc3 . Second, with the
queen on c2, Black must be careful
that when he plays . . . e6-e5 he does
not allow 'iYxg6! ( . . . e6-e5 has opened
up the a2-g8 diagonal and the f7-pawn
is now pinned to the king, so Black
cannot recapture on g6) . Of course,
we are dealing with subtle nuances
rather than big differences, but it is
important to understand them none
theless.

Game 4
Ivanchuk-Bareev

Dortmund 1 995
1 tLlf3 d 5 2 d4 tLlf6 3 c4 c6 4 tLlc3
dxc4 5 a4 jLf5 6 e3 e6 7 jLxc4 jL b4
S 0-0 tLlbd7

1 0 tLlxf5 exf5 1 1 c2 g 6 1 2 f3 b6

Preventing e3-e4 by attacking the

d4-pawn, which has been weakened
by the absence of the white knight
from 3 .
1 3 h 1 aeS 1 4 f2 c5 1 5 h4
.ld:cS ! ? 1 6 jLa2 fdS 1 7 jLd2 tLlfS 1 S
a 5 a6 1 9 fd 1 c4

Shutting out White's light-squared

bishop.
20 jLe 1 .l:!.eS 2 1 e4 jLxc3 22 jLxc3
fxe4 23 d5 tLlSd7 24 d4 Y2 - Y2

The draw was agreed in a very

murky position.
It is now time to consider the other
move-order: 8 . . . tLlbd7, intending to
meet 9 'iYe2 with 9 . . . jl,g6 as above.
18

Question G: What does Black gain

from delaying castling?
Answer: 8 . . . tLlbd7 is directed against
the plan of an early tLlh4, which we
saw in Yusupov-Kramnik. After 9
tLlh4, Black will reply 9 . . . jl,g6, as 10
tLlxg6 hxg6 is extremely dubious for
White. Since Black has not castled, his
rook is well placed on the semi-open
h-file, pointing towards White's king!
Black will play . . . 'iYc7 (attacking h2) ,
castle queenside and then double
rooks on the h-file, which is not what
White was hoping for when he sensi
bly (he thought!) took the bishop
pair!
So what does White do after
8 . . . tLlbd7 9 tLlh4 jl,g6? Give up? Cry?
Well, if he's a genius like Ivanchuk,
he chooses a third option: he gets
sneaky.
9 tLlh4 jLg6 1 0 jLe 2 ! ?

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

White wants to take on g6 only

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

1 2 . J:lc8 ! ?
.

10...0-0

Fans of tactics can investigate

10 . . . ]Lxc3 1 1 l2lxg6 (not 1 1 bxc3 l2lds
12 l2lxg6 l2lxc3! 13 iVc2 l2lxe2+ 14
'iYxe2 hxg6, winning a pawn)
1L.]Lxb2 12 l2lxh8 ]Lxa1 13 ]La3 or
13 iVc2 (unclear - Ivanchuk) and
when they have, 1'd be grateful if they
could tell me what is going on! How
ever, more positional players can be
happy with Bareev's move. Although
White's queen will go straight to c2,
the bishop is more passive on e2 than
on c4: after . . . e6-eS, Black no longer
has to fear iVxg6 (in fact he'd be quite
pleased to see it!) as the white bishop
is not on the a2-g8 diagonal; and this
also means that White cannot reply so
easily with d4-ds after . . . e6-eS or . . . c6c5.
11 ct:Jxg6 hxg6 1 2 'iVc2

Since the white queen i s o n the c2,

Black tries to inconvenience it by
opening the c-file. The immediate
12 . . . cS would be met by 13 l2la2 ! , net
ting Black's other bishop since
13 . . . ]LaS loses a pawn to 14 dxcS, in
tending b2-b4. Bareev therefore plays
the rook to the c-file in order to facili
tate . . . c6-cS . The . . . c6-cS break is
played less often than . . . e6-eS in the
Slav, but it is a typical idea that is well
worth remembering.
1 3 e4 ! ?

Since White does not want the c-file

to be opened, he prepares to meet
13 ... cS with 14 ds . If only he still had
his bishop on c4! This move
19

Th e Sla v

introduces a sharp pawn sacrifice that

is probably not quite good enough, so
13 Md1 was suggested by Ivanchuk as
an alternative, when he claims a slight
advantage for White. 1 3 . . . cS 14 ds
exdS 15 ttJxds ttJxds 16 Mxds VlJie7 1 7
b3 ! i s indeed rather better for White.
Black's main problems are the weak
ness of his light squares and his bishop
on b4, which is shut off from the rest
of Black's pieces by the pawn on cS .
The immediate 1 3 . . . VlJie7 is stronger:
14 e4 (also interesting is 14 b3!?,
intending 14 ... cS [ 14 ... eS is more
sensible] 15 dS iLxc3 16 d6! and VlJixc3
with advantage) 14 . . . cS 1 5 dS exdS 1 6
ttJxds ttJxds 1 7 exdS (17 Mxds loses
the e-pawn to 17 . . . ttJf6)

ous play due to his threat of . . . Me2.

1 3 . . . c5 1 4 d 5 exd 5 1 5 exd 5 ! ? l:l:e8 ? !

The start of a series of slight inaccu

racies that Ivanchuk exploits brutally.
Is . . . iLxc3 16 bxc3 ttJxds 17 Md1
ttJ7f6! 1 8 iLf3 (18 c4 ttJb4! unpins)
1 8 . . . VlJie7! is Ivanchuk's recommenda
tion, as 19 iLxdS ttJxdS 20 MxdS al
lows mate after 20 . . . VlJiel .
1 6 IId 1 c4

1 7 d6!

A very strong move: the d6-pawn

exposes the light-squared weaknesses
in the black position by opening the
hl-a8 diagonal and freeing dS for the
knight on c3 .
1 7 . . . l:l:e6 1 8 i.f4 "i'b6 1 9 l2lb5 i.c5
20 i.g3!

1 7 . . . c4! , freeing the bishop and pre

venting White from establishing a
light-squared blockade of the queen
side with b2-b3 and iLc4. After 1 8
iLxc4, Black can regain the pawn with
18 . . .tZJb6 19 b3 ttJxc4 (19 . . . ttJxdS 20
iLb2! [20 iLxds Mxc2; 20 Mxds VlJiel+!J
leaves White more active due to his
two raking bishops) 20 bxc4 VlJih4!,
when 2 1 f4 (the only move to save the
c-pawn) 2 1 . . .Mfe8 (22 . . . Mel+ is now a
threat) 22 g3 VlJig4 gives Black danger20

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

Calmly protecting f2.

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

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

lLlxg 3 2 6 a 5 1 -0

26 hxg3 would also have won.

White is just the exchange up with a
winning pOSltlOn.
Game 5
T opalov-Gelfand

Belgrade 1995

1 d4 d5 2 c4 c6 3 tLJf3 tLJf6 4 tLJc3

dxc4 5 a4 .liLf 5 6 e3 e6 7 .liLxc4 .liLb4

Black's passive play has gIven

White a huge space advantage.

8 0-0 tLJbd7 9 tLJh4 .liLg6 1 0 h 3 ! ?

1 9 . . . tLJd6 20 tLJxg 6 hxg6 2 1 .liLf4

Another waiting move. However,

Black's bishop has a square!

tLJe8 22 \\Wf3 .liLd 6 23 .liLe3 tLJdf6 24

10 . . . .liLh 5 1 1 g4 tLJd 5 !

26.liLc2? !

Black makes a discovered attack by

the queen on the knight on h4. Once
the knight retreats, Black will again
have g6 for his bishop.

26 dS! eS 27 liJe4 (Topalov) would

have led to a clear edge for White.

12 tLJg 2 .liLg6 1 3 tLJa2 .liLe7 ? !

exd 5 3 2 tLJxd 5 tLJxd 5 3 3 .liLb3 <;t>fS

Topalov rightly suggests 1 3 . . . d6

f3 hS ! as an improvement, striking
immediately against the exposed king
side pawns.
14

g5 tLJh 5 25 !:tfd 1 tLJc7

26 . . .f5! 27 h4 b 5 2S axb6 axb6 29

l:haS \\WxaS 30 \\WxaS J::!. x aS 3 1 d 5
34 .liLxd 5 l:!.a4 35 .liLxb6 l:!.xh4 36
.liLc6 .liLe7 3 7 J::!. a 1 tLJf4 3S .l:!.aS+ <;t>f7
39 .liLeS+ e6 40 J::!. a 6 d 5 41 .liLf7+
<;t>e4 42 il..e 3 Yz - Yz

1 4 \\We2 tLJ 5 b 6 1 5 .liLb3 c 5 1 6 tLJc3

0-0 1 7 a 5 cxd4 1 8 exd4 tLJcs 1 9

lLlf4

Instead of 9 liJh4, White has a more

testing plan: 9 ik'e2 and 10 e4!
21

Th e Sla v

Game 6
Ivanchuk-Lautier

Linares 1994
1 d4 d 5 2 e4 e6 3 ctJf3 ctJf6 4 ctJe3
dxe4 5 a4 iif5 6 e3 e6 7 iixe4 iib4
8 0-0 ctJ b d 7 9 'iWe2 iig 6 1 0 e4 ! ?

This pawn sacrifice is the problem

with this move-order.
1 0 . . . iixe3

The more restrained 10 . . . 0-0 is con

sidered in the next chapter.
1 1 bxe3 ctJxe4 1 2 iia3 !

c6! and 13 . . . QJexc5 with 14 .ib5 ! 0-0

15 .ixd7 QJxd7 16 .ixf8 .
Question 9: How then can Black get
his king to safety?
Answer: Black can castle queenside
instead of kingside.
Question 1 0: You don't sound very
impressed!
Answer: Black's position is horrible!
White has a simple and extremely ef
fective plan: a4-a5-a6, softening up the
black queenside, and then, after mov
ing the bishop on c4, c3-c4-c5 finish
ing off the job. In reply, Black must
seek to play . . . e6-e5 and activate his
kings ide pawns.
1 2 . . .. e7 1 3 afe 1

From el, the rook protects the c3pawn and supports the c3-c4-c5 push.
1 3 . . 0-0-0
.

Question 6: What i s going on?

Answer: In return for the sacrificed

centre pawn, White has gained the

two bishops and prevented Black
from castling kingside.
Question 7: Can't Black just grab
another pawn with 12 .. .ct:Jxc3?
Answer: NO!! 13 'i'b2 (hitting the
knight and the pawn on b7) 13 . . . QJxa4
14 'i'b3 ! , threatening both 'i'xa4 and
.ixe6, is horrible for Black.
Question 8: Can't Black just block
the a3-f8 diagonal with 12 . . . c5 and
then castle kingside?
Answer: This is logical, but 13 dxc5
is awkward, meeting 13 . . . 0-0 with 14
22

13 . . . c5 14 QJe5 ! QJxe5 (14 . . . 0-0 15

QJxg6 wins a piece) 15 dxe5 'i'xe5 (or
else White plays f2-3 and h2-h4, trap
ping the knight) 16 3 wins a piece,
while 13 . . . QJd6 (blocking the a3-f8
diagonal) 14 .ixe6! 0-0 (14 . . . fxe6 15
'i'xe6+ is crushing) 15 .ib3 favours
White due to his bishop pair.
1 4 a5

1 4 . . . ctJd6 ! ?

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

The young Rumanian player Gab

riel Schwartzman tried 14 .. J;fhe8
against Razuvaev in Dortmund 1993,
but after 1 5 a6! b6 16 ttJh4 ttJd6 17
jLb3 es 1 8 ttJxg6 hxg6 19 iVg4!
(preventing Black from activating his
kingside pawns with .. .f7-fS by attack
ing g6) 19 . . -,,?b 8 20 jLxd6! iVxd6 2 1
jLxf7 e7 2 2 iVxg6, White stood
clearly better. 19 . . . cS cutting out the
bishop on a3 , was suggested as an im
provement, but after 20 dxcS bxcs 2 1
ab 1 ! (preventing 2 1 . . .b8 due t o 22
jLxf7+) 2 1 . . .e7 (defending f7) 22
cd1! (eyeing the knight on d6) the
only positive course of action open to
Black is to wring the neck of the man
who made this suggestion!
1 5 1l.b3 1l.h5

1 7 . . . b6 1 8 c4 !

Threatening c4-cs.
1 8 . . . c 5 1 9 a 4 e 5 20 dxc5 b x c 5 2 1
e3 ge6 22 tLlg 5 .l::rf 6 23 b 2 h 6 24
tLle4 tLlxe4 25 xe4 ge6

I would not recommend this posi

tion to anyone. The game finished:

An idea of Bareev's I believe, trying

to inconvenience White by the pin on
the knight. Note that 1 s . . . he8 is met
by 16 ttJh4!

gxa6 3 3 xg7 h5 34 f4 f 5 35 e5+

1 6 h3

'it>b7 36 gc7+ 'it>b6 3 7 'it>h2 'it>a5 38

16 'iVe3 unpinning, and eyeing the

a7-pawn, is also interesting.

4 1 .l::rx a7 'it>xc4 42 1::!. a 8 c6 43 .l:!.a3

16 . . J!he8

26

ge 1

g6

27

a8+

b8

28

1l.xd7+ gxd 7 29 xb8+ 'it>xb8 30

gxe5 gdd6 3 1 .l:!.xc5 .l:!.xa6 3 2 gxa6

'it>g3 e8 39 'it>h4 'it>b4 40 g 3 gg6

e8 44 ge3 d 5 45 f6 d 7 46
c3 e6 47 b4 d7 48 xh 5
gg8 49 c3 d6 50 'it>h6 e6 5 1
h7 ga8 5 2 b4+ d7 5 3 g7
l::!. a4 54 'it>f6 d 5 5 5 c3 e4 56
g4 fxg4 57 hxg4 c2 58 .l:!.e7+ d8
59 e5 ga6+ 60 l::!. e 6 gxe6+ 6 1
xe6 e8 6 2 d6 d 1 6 3 g 5 1l.c2
64 f5 1 -0

Game l
Ehlvest-Schwartzma n

New York

Open 1996

17 a6!

Softening up the protection around

the black king.

1 tLlf3 d5 2 d4 tLlf6 3 c4 c6 4 tLlc3

dxc4 5 a4 f5 6 e3 e6 7 xc4 b4
23

Th e S l a v

8 0-0 ttJ b d 7 9 'We2 g 6 1 0 e4 xe3

1 1 bxe3 ttJxe4 1 2 a3 'We7 1 3 J:l:fe 1
0-0-0 1 4 a 5 b8 1 5 e7 ! ?

prepares to activate the light-squared

bishop on the long diagonal.
44 . . . 'i'xf3 loses simply to 45 'i'd6+.
44 . . . a 5 45 iLg2 e6 46 'Wd8+ a6
47 'We7 ttJd7 48 f4 'We7 49 e6+
ttJb6 50 'Wxg 6 a4 51 'We8 a3 5 2 'i1\Vf8
a 2 53 'i1\Va3+ ttJa4 54 'i1\Vxa 2 'i1\Vd 7 55
d4 1 -0

If this isn't enough to convince you

of the danger in accepting the pawn
sacrifice, then try this!
Game 8
Hubner-Beliavsky

An interesting manoeuvre, transfer

ring the bishop to the annoying h2-b8
diagonal. The bishop has already ful
filled its task on a3 by preventing the
black king from castling kingside.

Munich 1994
1 d4 d 5 2 e4 e6 3 ttJe3 ttJf6 4 ttJf3
dxe4 5 a4 f5 6 e3 e6 7 xe4 b4
8 0-0 ttJbd7 9 'i1\Ve2 g6 1 0 e4 xe3

1 5 . . . .i:!.de8 1 6 h4 a8 1 7 'Wb2 f5

11

1 8 ttJd2 f4 1 9 f3 ttJxd2 20 'Wxd 2

J::!. f e 1 ! ?

J:l:hf8 2 1 f 1 e 5 22 d x e 5 J:l:xe 5 2 3

A very aggressive alternative to the

old 13 Mfc l . White sacrifices yet an
other pawn, reasoning that this will
merely open more lines for his pieces.

a6

Ehlvest criticises this move, prefer

ring White after 23 Ma4 Md5 24 'i'a2
ctJc5 25 Md4. This may well be more
accurate, but the essential point is that
White will always have very good
chances because Black's king is weak
and White's bishops are strong.
23 . . . b 6 24 :!la4 .i:!.d 5 2 5 :!ld4 :!lxd4
26 'Wxd4 ttJ e 5 2 7 .i:!.d 1 :!le8 28 e4
iLe2 29 J::!. e 1 llxe 1 + 30 1i.xe 1 1i.f5
3 1 1i.d 2 'Wd 7 32 'li'xf4 b5 3 3 f 1
ttJxa6 34 e3 b 7 3 5 f2 g6 3 6
h4 e 5 3 7 'W e 3 b6 3 8 h 2 'We8 3 9
'Wg 5 h 6 40 'Wd 2 'W e 6 4 1 g3 f6
42 d7 e4 43 f2+ ttJe5 44 g 3 !

After a time-scramble and a little

confusion, White re-establishes con
trol with this evil little move, which
24

bxe3 ttJxe4 1 2 a3 'i1\Ve7

13

1 3 . . . ttJxe3

Best and the most critical. 13 . . . 0-0-0

was extremely unpleasant for Black in
the game Beliavsky-Akopian, No
vosibirsk 1993: 14 'i'b2 Mhe8 15 as e5
16 Mab 1 c5 17 f1 f6 and now 1 8
ctJh4! ctJxc3 19 'i'xc3 exd4 2 0 'i'b3
JiLxb 1 21 Mxb 1 gives White an over
whelming initiative, as Beliavsky
pointed out. Clearly in such lines, the
king's rook is much more actively
placed on e 1 than on c 1 (as in lines we
have seen previously). This is also true
of 13 . . . c5, as in Hubner-Hertneck,
Munich 1994, when 14 d5! e5 15 d3!
ctJef6 1 6 ctJxe5 0-0-0 17 ctJxd7 Mxd7 1 8

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

c4 (Hubner) is the (unpleasant) best

that Black can hope for.
14 'i'b2 ctJe4 1 5 a 5 ! ?

now becomes a little random, due to

mutual time pressure, but White pulls
through in the end.

T o break up the black queenside

with a5-a6. 15 I[ac 1 is also interesting.
15 . . . ctJdf6

15 .. .liJd6 (intending . . . 0-0) is met by

16 .!xe6! (16 Vib4 c5 ! [not 16 . . . lZ'lxc4??
1 7 'ife7 mate] 17 dxc5 lZ'lxc4 1 8 c6

4Jxa3 ! 19 cxd7+ Vixd7 20 Vixa3 Vie7

'ifa4+ Vi d7 22 Vi a3 [preventing
kingside castling], which leads to a
draw by repetition after 22 . . . Vie7)
22 . . 0-0 17 ltxd7 Vixd7 1 8 lZ'le5 Vic7
19 a6! , breaking up the queenside with
an advantage.
Instead 1 5 ... a6 (preventing a5-a6) is
best, when Hubner suggests 16 I[e3
4Jd6 17 .!xe6 0-0-0 18 ltxd6 Vixd6 19
.!c4 and I[b3 with a dangerous attack.

f 4 Il:.d 5 3 3 .l:i.e4 g 5 3 4 a4 e 5 3 5

16 tLJe5 a6 1 7 .l:i.ac 1 .l:i.d 8 1 8 jLxa6 ! !

fxe 5 fxe 5 3 6 a 7 g 8 3 7 JJ.e7

21

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

. .

22 . . . ctJd6 23 c7 .l:i.d 7 24 c3 0-0

25 ctJe5 Il:.dd8 26 c6? xc6 27
.l:i.xc6 ctJb5 28 jLxf8 xf8 29 gxa6
ctJxd4 30 f3 f6 ? ! 31 ctJxg6+ hxg6 3 2

N ow White is winning again.

'i'xa5 2 1 ctJxg4 b5 2 2 c2 ? !

37 . . . ctJc6 38 JJ.e6 gc5 39 g6 f7

This i s White's first inaccuracy in

this fascinating game! Hubner notes
that the simple 22 Vixb5 axb5 23 I[c7
I\,a8 24 ltb4 I[a4 25 I[c8+ 'it'd7 26
I\,xh8 I[xb4 27 lZ'le5 'it'c7 28 I[c1 + 'it'b7
29 f3 lZ'lf6 30 lZ'lxg6 would have been
clearly better for White. The game

40 xg 5 g 6 41 h4 f6 42 h2 ctJe7
43 a4 c6 44 .l:i.g 3 ctJf5 45 gf3
g7 46 l:te4 e6 47 h5 ctJe7 48
hxg6 ctJxg6 49 h3 ctJe7 50 g3
ctJg6 51 f5 e7 5 2 g4 l:ta7 5 3
g 5 ctJh8 5 4 g 4 ctJ f 7 + 5 5 <;t> h 5 +
f 8 56 g6 e 8 5 7 h 4 1 -0
25

Th e S l a v

S u m m a ry

The alert reader will have noticed my profound mistrust of the line 8 . . . 'bbd7
9 'i'e2 g6 10 e4 xc3 1 1 bxc3 'bxe4. I honestly cannot understand the at
traction of these lines for Black. Therefore, if Black wishes to try to prevent
e3-e4, then Kramnik's 8 . . . 0-0 9 'i'e2 g6 is the line for you; Yusupov's plan of
a quick 'bh4 is the most testing response.
1 d4 d 5 2 c4 c6 3 ttJt3 CUt6 4 ttJc3 dxc4 5 a4 .ltf5 6 e3 e6 7 .ltxc4 .ltb4
8 0-0
8 . . . 0-0

8 . . . 'bbd7 (D)
9 'bh4 g6
10 e2 - game 4
1 0 h3 - game 5
9 'i'e2 g6 (9 . . . g4 - see next chapter) 10 e4 xc3 (10 . . . g6 see next chapter) 1 1 bxc3 'bxe4 12 a3 'i'c7
13 .sfe 1 - game 8
1 3 .sfc1 0-0-0 14 as (D)
14 . . . 'bd6 game 6
14 . . . c,t>b8 - game 7
-

9 e2

9 'bh4 game 3
-

9 . .lt g 6
. .

9 . . . 'bbd7 - see next chapter

1 0 ttJe5 ttJ b d 7 1 1 ttJxg6 hxg6 1 2 .l:td 1 (D)

12 . . . 'i'e7 - game 1
12 . . . 'i'aS game 2
-

B... ttJ b d7

26

14 a5

12 'iJ.d1

CHAPTER TWO
The Old Main line:
Black allows e3-e4

1 d4 d 5 2 c4 c6 3 l2lf3 l2lf6 4 l2lc3

dxc4 5 a4 i.f5 6 e3 e6 7 i.xc4 i.b4

8 0-0

This chapter deals with lines arising

from 8 . . .cubd7 9 'Wie2 0-0 10 e4 iLg6
(or 9 . . . iLg6 10 e4 0-0, turning down
the dangerous pawn sacrifice on the
way) and 9 . . . iLg4.
Question 1 : When White plays e3e4, he gains a commanding central
presence with pawns on e4 and d4.
Why is Black playing this position?
Isn't he just worse?
Answer: Central pawns are strong if
they are dynamic and able to advance
and chase away the opposing pieces.
Otherwise, they can present easy tar
gets for the enemy pieces. In this case,
White cannot advance d4-dS, and e4eS leaves a hole on dS for the black
pieces. Moreover, Black is threatening
to win a pawn with . . . iLxc3 and
tiJxe4, now that his king is safely cas
tled. Therefore, while the d4-e4 centre
gives White a definite space advantage,
Black has plenty of threats against the
white centre, which is the basis of his
counterplay.
Question 2: What is the difference
between playing 9 . . . iLg6 first or
9 . . . 0-0 10 e4 iLg6?

Answer: Good question! Generally,

black players play 9 . . iLg6 to pretend
that they are willing to take on the
pawn sacrifice after 10 e4. Even if
they don't intend to take the pawn,
the idea is to make White waste a lit
tle time on the clock thinking about
his variations! You never know - a
few minutes might be handy later!
.

Game 9
G ofshtein-Sadler

Ischia 1996
1 d4 d 5 2 c4 c6 3 l2lf3 l2lf6 4 tLlc3
dxc4 5 a4 i.f5 6 e3 e6 7 i.xc4 i.b4
8 0-0 tLlbd7 9 e2 0-0 1 0 e4 i.g6
1 1 i.d3

White had to deal with the threat

27

Th e S l a v

of . . . iLxc3 and . . . tbxe4. It is a general

rule that the longer you can delay
committing your centre, the better,
since the later you reveal your hand,
the less time your opponent has to
adjust to it.
11

. . .

it h 5

the knight on f3 to the queen and

thus threatens . . . e6-eS. The alternative
1 1 . . .h6 is considered in Game 12.
1 2 itf4

Trying to avoid e4-eS for a while

longer, White brings another piece to
bear on eS. Strangely enough, this is
probably not the best move. The di
rect 12 eS is considered in the next
game.
12

. . .

'Vj'e7 ! ?

Black threatens 1 3 . . . iLxf3 1 4 'iVxf3

eS! , equalising comfortably. I think
that this is a novelty: 12 . . . Me8 had
been played before.
1 3 e5

Absolutely necessary.
13

. . .

tZ:l d 5 1 4 tZ:lxd5 cxd 5

Question 3 : What is the pomt of

this move?
Answer: The first place to look for

counterplay, is with your pawn

breaks. Pawn breaks have two func
tlons:
a) They attack the opponent's
pawn structure and force him to react
to you.
b) They are a 'breakout': they gain
space and therefore give more room
for your pieces to become active.
Black has two pawn breaks in this
position: . . . e6-eS and . . . c6-cS . Usually
he prefers to aim for . . . e6-eS, since this
stops White from playing e4-eS him
self, inconveniencing the black
knight. For example, 1 1 . . .cS 12 eS!
tbdS 1 3 tbxds exdS (13 ... iLxd3 14
'iVxd3 exdS 1 5 tbgS ! is unpleasant) 14
iLxg6 hxg6 15 tbgS, with threats of
eS-and 'iVg4-h4 and 'iVh7 mate, is nasty
for Black. The move in the game pins
28

White has a space advantage due to

his pawns on d4 and eS. Normally,
with his bishops, knight and queen
pointing towards the black king,
White could consider launching a
kingside attack, but here Black's
bishop on hS interferes with this plan:
it can exchange itself for the knight
on f3 or return to g6 to block any
white attack on the b 1-h7 diagonal.
Meanwhile Black will challenge for
the c-file, exchanging the bishop on

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

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

c2 as an entry square for the black
major pieces. Black will also transfer
his knight to c6 via b8 from where it
not only attacks d4, but can invade
the white queenside by as-b3 or via
b4. So what on earth can White do?
Stay calm! White does not want to
exchange pieces on the c-file since this
would help Black to free his cramped
position, so he has two plans. First
(my own favourite) , he can concen
trate on the kingside where White
holds most of the trumps: a space ad
vantage and a large concentration of
minor pieces. I would try to push my
kingside pawns: 1 5 h3 !!fc8 16 g4 ltg6
17 h4, intending h4-hs. This plan
demonstrates the drawback to ltf4,
however: White would like to throw
the f-pawn forward as well, but the
bishop gets in the way. The chances
after 15 h3 are, I believe, about equal.
White's choice is interesting, but there
is always a danger in choosing plans
based mainly on tactical points: if
there is just a little hole in your calcu
lations, then you often find that you
have just wasted time and must retreat
in disarray. On the other hand, such
plans are often the most unexpected
and the most difficult for the oppo
nent to deal with!

1 7 . . . Jixd 3 1 8 xd 3 .l:!.e4 !

Black is now planning . . . b7-bs and

19 b3 is met by 19 . . . Mc7 ! , when 20
'ik'bS does not win a piece as Black has
20 . . . ltc3 , when he stands well. Realis
ing that his activity on the queenside
has come to naught, White goes back
to 'Plan A' and expands on the king
side, but he is several tempi down on
what he could have had earlier.
1 9 h4 h 6 20 h5 ctJ e 5 !

1 5 e3 1:He8 1 6 a 5 ! ? Jig6 1 7 .l:!.a4 ! ?

At first I thought about playing

17 . . Mab8, intending 18 . . . bS to drive
away the white rook. Then to my
horror I noticed 18 ltbS ! White is
threatening to take the knight on d7
and then take my bishop on b4, so
18.. .ctJf8 is natural, but then 19 'ik'b3 !
and my bishop i s trapped!
.

Oops! My opponent had missed

that one. However, after a big think,
he came up with an active defence.
2 1 dxc5 nxf4 22 a6 ! bxa6 23 c6 a 5
2 4 .l:!.e 1

g e 8 2 5 b 3 ge7 26 .l:!. e 2

ne4? !

A rather casual move. 26 . . . Mfs 27

29

Th e S l a v

Mal Mxh5 28 g4 Mh3 29 'It>g2 h4!!

(this lovely move was pointed out to
me by Julian Hodgson after the game)
was the way to play.
27 '!::!'a 1 jL c 5 ? ! 28 .!::!.x a5 xc6 ? ?

35 'It>gl Mc1+ leads to immediate

mate.
So 12 f4 doesn't seem all that
promising for White. What about 12
e5 instead?
Game 10
Gelfand-Lautie r

.t-urich 1994

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

dxc4 5 a4 jLf5 6 e3 e6 7 iLxc4 jL b4
8 0-0 lLlbd7 9 e2 jLg6 1 0 e4 0-0
1 1 jLd3 iLh5 1 2 e 5

Probably the best move.

29 .!::!. b 5 ? ?

My God! I had missed that White

could simply win a piece with 29
b5 , forking rook and bishop. I had
thought that I could play 29 . . . xf2+
with a discovered attack on the rook
on c2, but White just plays 30 Mxf2!
Luckily White shared the same blind
spot! After 29 Mb5, White is just lost.
29 . . . c7 30 a2 iLb6

Trapping the rook.

3 1 <;t>h2 .l::!. c3 3 2 d2 c6 33 gxb6

1 2 . . . lLld 5 1 3 lLlxd5

axb6 34 lLld4 .!::!. h 4+ 0- 1

The alternative, 13 liJe4,

with in the next game.

1S

dealt

1 3 . . . cxd 5

13 . . . exd5 has been suggested, but

since White already has a space advan
tage on the kingside, I am sceptical
about conceding a pawn majority as
well in that area.
1 4 e3 h 6 ? !

A debatable decision. I would pre

fer 14 . . . e7, followed by a rook to
the c-file and . . . liJb8-c6.
1 5 lLle 1 !
30

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

pletely dominates the black knight.

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

A very instructive plan: White in

tends to play f2-f4 and f4-f5, which is
particularly dangerous once Black has
weakened his kingside with . . . h7-h6.
Moreover, the knight will eventually
come to d3 with tempo, hitting the
bishop on b4.
15 .. .f5 1 6 exf6 'i'xf6 1 7 3i.b5 ttJb8
1 8 tLld 3 ! a6

Fixing the g 7-pawn.

34 . . . 'i'd8 3 5 'i'g3 'i'd7 36 3i.d6 'i'f7
37 'i'e5 'i'f5 38 'i'e2 b4 39 g4 'i'f6
40 3i.e 5 'i'g 5 41 g2 b3 42 3i.d 6 !

Preventing the knight from becom

mg active.

18 ... 1td6 19 tLJe5 is not pleasant for

Black.
1 9 tLlxb4 axb5 20 a5 ttJa6 2 1 ttJ d 3 !
l:tfe8 22 3i.d 2 g e 2 23 gae 1 g a e 8 24
l:txe2 gxe2 25 ge 1 gxe 1 + 26 3i.xe 1
iLg6 2 7 ttJe5 'i'f5 2 8 h 3 'i'e2 29
lLlxg 6 !

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

45 3i.e 5 'i'h4 46 3i.g 3 'i'f6 47 'i'd2
'i'e7 48 'i'e3 'i'f6 49 3i.d 6 h 7 50
'i'x b3 'i'xd4 51 'i'e2+ \tlh8 5 2 'i'e8+
h7 53 'i'e2+ h8 54 'i'e8+ h7
5 5 'i'xe 6 !

Protecting g4.
Excellent judgement. In the result
ing position, White's bishop com-

55 . . . 'i'xb2

56

3i.e 5 'i'c2

57

'i'e7

'i'e4+ 58 g3 'i'd3+ 1 -0
31

Th e S l a v

After 59 f3 , Black cannot stop

mate. His knight has not moved since
move 20!
Black should be fine after the ex
change of knights on dS , providing he
avoids weakening his kingside. Let us
take a look at 1 3 tbe4.

Game 1 1
Xu Jun-Akopian

Moscow Olyrt;lpiad 1994

1 d4 d 5 2 c4 c6 3 CLlf3 CLlf6 4 CLlc3

queen to the kingside. Yes, this is the

hacker's option!
1 3 . . . e7

13 . . . cS is unpleasantly met by 14
ltgS! 'i'aS 15 ltbS! Once Black moves
the knight on d7, he will lose the cs
pawn, and he cannot protect it with a
rook due to the bishop on gS. If he
protects the knight with l S . . . 'i'c7,
then 16 Mac1 is unpleasant. The text
prevents ltgS and prepares . . . c6-cS .
1 4 CLlg3

The direct approach. 14 as has also

been tried.

dxc4 5 a4 f5 6 e 3 e 6 7 xc4 b4
8 0-0 CLlbd7 9 e2 g6 1 0 e4 0-0
1 1 d 3 h5 1 2 e 5 CLld 5 1 3 CLle4 ! ?

Question 4: Why does White play

14 as?
Many white players d o not enjoy
the positions that we have seen in the
first two games of this chapter. With
the centre closed, and the prospect of
exchanges on the c-file, they feel un
easy about their winning prospects; so
recently the plan with 13 tbe4 has
come to prominence. Of course,
Black keeps his knight outpost on ds
and his pawn-break against the centre
with . . . c6-cS. However, White's space
advantage remains and he retains e4 to
transfer first his knight, then his
32

Answer: a4-aS gains queenside

space, preventing Black from using
the as or b6 squares for his pieces.
Question 5: So what's the verdict? Is
it a good move?
Answer: Absolutely not! These
aims are completely irrelevant. 14 . . . cS
is logical, striking at the d4-pawn.
Neither 1 5 dxcs tbxeS nor 15 tbxcs
tbxcs (lS . . . ltxcS!?) 16 dxcS MCS (of
course not 16 . . . ltxcs 17 ltxh7+ xh7
1 S iVc2+ gS 19 tbgS ltg6 20 iVxcs,
winning a pawn) , followed by . . . ltxcs
or . . . MxcS, promise White anything.

Th e O ld M a in L in e : B l a c k a //o ws e 3 - e 4

14 . . JLg 6
.

1 5 JLxg6 hxg6

15 ... fxg6 used to be played almost

automatically, primarily for defensive
reasons: black players were worried
that if they recaptured with the h
pawn, White would play his knight
on g3 to gS via e4 and his queen to h4,
when Black would have no defence to
'Ih7 mate. By taking with the f-pawn
Black retains the option of ... h7-h6 to
keep a knight out of gS, and of course,
he gains the semi-open f-file for coun
terplay. And then people realised that
White's attack was hardly automatic
after 1 5 . . . hxg6, so this move gradually
became the main line!
1 6 tLle4 c 5

1 7 tZJc3

This is a perfectly reasonable plan,

challenging Black's knight on ds once
Black has weakened its support by
playing . . . c6-cS, but why did the Chi
nese player avoid 17 CLlegS? 17 . . . cxd4
18 'iVe4 's'e8 19 'iVh4 CLlf8 seems to be
a good reason. The knight on f8 de
fends against 'iVh7 and if by some
miracle White manages to threaten to
get a rook on h3, then Black can hit
the 'panic button' and chase the
knight away with .. .7-f6. White
could, however, try and open up the
black kings ide with h2-h4-hS , possibly
after 16 'iVe4 cS 17 h4 cxd4 18 hS !?
And now it's up to you, the reader!
We'll have to wait for practical tests
before a conclusion can be reached.
1 7 . . . b6

17 . . . CLl7b6 18 as CLlxc3 19 bxc3 CLlds

20 c4 CLlb4 also seems reasonable.
1 8 tZJxd 5 exd 5 1 9 dxc5 tZJxc5 20
JLe3 e6 21 a5 a6 22 .l:!.ac 1 .l:!.ac8
23 .l:!.fd 1 tZJb3 24 .l:!.xc8 .l:!.xc8

Black has an isolated d-pawn, but

White's queenside is weak. The posi
tion is about equal.
25 JLb6 JLd8 ? ! 26 JLxd 8 .l:!.xd8 27
e 1 ? !

33

Th e S l a v

J::i b 6 J::i d 5 3 6 b4 d 3 3 7 c!;e 1 c!;d7 38

l1.. .'iYaS has a similar idea. After 12

f4! lIfe8 (12 . . . xc3 13 bxc3 'iYxc3
loses to 14 d2 'iYc2 15 d3 'iYb2 16
lIfb l) 13 h3 lIac8 14 12Ja2!, White had
a slight advantage in Beliavsky-Short,
Linares 1995, as 14 . . . f8 (14 . . . 'iYxa4
15 ttJc3! 'iYb3 16 c4 wins the queen)
15 b4! gains queenside space with
tempo: 15 . . . 'iVxa4 16 lIfb l and 12Ja2-c3
traps the queen, while 15 . . . xb4 16
12Jxb4 'iYxb4 17 lIfb l 'iYaS 18 lIxb7 is
unpleasant for Black.

d2 c!;c7 3 9 J::i d 6 J::i x d6 40 exd6+

1 2 f4 J::i c 8 1 3 l:Ifd 1 J::i e 8 1 4 h3 a6

xd6 41 xd 3 d5 42 tLJ d 2 tLJc6

1 5 l:Iac 1 !

43 c!;c3 tLJe5 44 tLJb3 tLJc4 45 f4 f6

Black's slightly incautious 25th

move allowed White the chance to
activate his queen by the lovely 27
'iYe4!, intending 12Jg5 and 'iYh4, as
27 . . . clxe4 allows 2 8 lIxd8+ h7 29
ttJg5+ winning the queen (analysis by
Xu Jun) . The rest is hard-fought, but
it was always going to be a draw.
27 . . :Viii e 7 28 'Viii c 3 tLJ c 5 29 'iVb4 f8
30 f 1

tLJe6 3 1 'Viiix e7+ c!;xe7 3 2

J::i d 3 d4 3 3 J::i b 3 J::i d 7 34 g 3 tLJd8 3 5

1 d4 d 5 2 c4 c6 3 tLJf3 tLJf6 4 ct:Jc3

After 15 ttJa2 f8 16 b4 Back has

16 . . . aS! (an excellent idea, breaking
White's grip on the c5-square) 17 bxaS
'iYxaS 18 d2 'iYc7 19 e5 xd3 20
'iYxd3 12JdS 21 ttJc3 ttJxc3 22 xc3
ttJb6 23 as ttJdS 24 el c5 and Black
had no problems in LSokolov-Oll,
Moscow Olympiad 1994.

dxc4 5 a4 f5 6 e 3 e6 7 xc4 Si.b4

1 5 . . . b6

8 0-0 tLJ b d 7 9 'Viii e 2 Si.g6 1 0 e4 0-0

No better is 15 . . .'iYaS 16 ttJd2! b5

(16 . . . xc3 17 bxc3 'iYxa4 18 lIal traps
the queen) 17 axb5 axb5 18 12Jb3 'iYb6
19 e5 12Jd5 20 12Jxd5 exdS 21 xg6
fxg6 and Black was probably happy
he couldn't see his position in Lautier
Gelfand, Monaco (blindfold) 1999.

lh - lh

Game 12
Sadler-M iles

British Championship 1 998

1 1 d 3 h6

1 6 b 1 !

I really like this development

scheme. White consolidates his queen
side and mobilises all his pieces, ready
for any of Black's breaks.
1 6 . . . h 7 ? !

A slightly risky idea. Black devel

ops quietly and waits for an opportu
nity to break with either . . . c6-c5 or
... e6-e5. The problem is that, as in this
game, Black can really get sat on!
34

Black should really have taken the

plunge with 16 . . . c5, although 17 12Ja2!
wins the bishop pair with a nice ad
vantage for White.
1 7 tLJe5?

This is rather careless though!

Th e O l d M a in L in e : B l a c k a ll o w s e 3 - e 4

17 . . :i'd8 ?

17 ... ctJxeS 1 8 dxeS "iVc7! 19 "iVc4

f8 is okay for Black.

tremely solid and has no pawn weaknesses.

1 8 ClJc4!

N ow

Black is suffering.

18 . . . ClJb6 1 9 ttJ a 2 ! .ll. f 8 20 b 3 !

Maintaining the knight on c4, as

Black will now have to improve
White's structure to get rid of it.
20 . J::i. a 8 2 1 h 1 ttJc8 22 a5 ttJd7
.

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

itJf6 26 f3 ttJa7 27 .ll. g 3 ttJd7 28

d5! exd 5 2 9 exd 5 i.xb 1 30 J::i. x b 1

cxd 5 3 1 ttJcb6 ttJxb6 3 2 ttJxb6 gad8

33 ClJxd 5 'i'f8 34 ttJc7 ge7 3 5 J::i. x d8

1 5 . . . a6 1 6 ..\te2 J::i. f d8 1 7 ga4 ..\te7

xd8 36 ttJ d 5 J:i.e 1 + 3 7 J::i. x e 1 i.xe 1

1 8 f4 CLle8 1 9 .ll. e 3 gac8 20 .ll. f3

38 b4 Wiie 8 ? ? 3 9 ttJc7 ! c6 40 'iVe3

ttJd6 21 i.e2 g 6 22 J::i. a a 1 1h - Y2

ixb4 41 Wiix a7 .ixa5 42 b8+ h7

43 ClJe8 b 5 44 ttJd6 f6 45 Wii a 7 i.b4

46 ClJf5 i.f8 47 f7 'i'c8 48 CLld4
ia3 49 ttJe6 'i'g8 50 'iVd7 h8 5 1
ic7 .ib4 5 2 f4 .ic3 5 3 f5 b4 54
id6 1 -0

Here Kasparov shows a more criti

cal approach for White.
Game 14
Kasparov-Bareev

Occasionally 9 . . . .,tg4 is seen instead

of 9 .. .,tg6, as in the next two games.
.

Novgorod 1994

1 d4 d5 2 c4 c6 3 ttJf3 ttJf6 4 <1Jc3

dxc4 5 a4 i.f5 6 e 3 e6 7 i.xc4 i. b4

Game 13

8 0-0 ttJbd7 9 'iWe2 i.g4 1 0 h 3 !

Khalifman-Kir Georgiev
.

Elenite 1994
1 d4 d5 2 ttJf3 ttJf6 3 c4 dxc4 4
itJc3 c6 5 a4 Jl. f5 6 e3 e6 7 i.xc4
ib4 8 0-0 ttJ b d 7 9 e2 i.g4 ! ?

A rather unusual
bore White to tears
into a dull ending.
10 J::i. d 1 'i'a 5 !

11

move that aims to

by exchanging off
e4 'i'h5

1 2 h3

ixf3 1 3 'i'xf3 'i'xf3 1 4 gxf3 0-0 1 5

a5

White has the two bishops and a

space advantage, but Black is ex-

The most aggressive idea: now if

1 0 . . . .,thS, White can play for e3-e4
35

Th e S l a v

without fear of the . . . iVaS-hS manoeuvre.

1 0 . . . SLXf3 1 1 jVxf3 0-0 1 2 .i:::l. d 1 jVa5
1 3 e4 e5 1 4 d 5

This bears a distinct similarity to

Richardson-Sadler from Chapter 1 .
White's queen i s better placed o n f3
than e2, though of course Black's
queen is more actively placed as well.
1 4 . . . ttJb6

1 5 b3 xc3

1 6 bxc3

cxd 5 1 7 exd5 .i:::l. a c8 1 8 c4! !

Brilliant. 1 8 . . . lLJxc4 19 gS iVa6 20

d6! lLJxd6 21 xf6 gxf6 22 iVxf6 lLJc4
23 iVxa6 bxa6 24 Md7 is clearly better
for White (Kasparov) .

Which is the better move?

A nswer: I can offer no definitive
conclusion: in theory, 8 . . . 0-0 and
8 ... lLJbd7 are equally good. However,
practical chess is not only about find
ing the very best move: when making
a final decision whether to play 8 ... 0-0
or 8 ... lLJbd7, it is also necessary to
take the strength of your opponent
mto account. The following game
shows why.
Game 15
D . Strauss-Lakdawala

USA

1992

1 d4 d5 2 c4 c6 3 ttJf3 ttJf6 4 ttJc3

dxc4 5 a4 f5 6 e3 e6 7 xc4 b4
8 0-0 ttJbd7 9 'li'b3

1 8 . . . .i:::l. f e8 1 9 d 2 jVa6 20 d 6 ! ttJbd7

21 e3 .i:::l. c 6 2 2 a 5 ! .i:::l.x d6 23 a4
.i:::l. x d H 24 .i:::l. x d 1 .i:::l. d 8 25 c 5 !

Black has been tied up in quite bril

liant fashion.
25 . . . h 6 26 jVf5 g 6 27 c2 ! xa5
28 .i:::l. d 6 <;t>g7 29 jVd 1 c7 30 h4
h8 31 h 5 g 5 32 'if3 <;t>g7 33 'if5
b6 34 SLxd 7 ! ttJxd7 3 5 xg 5 ! 1 -0

Kasparov gives 3 S . . . hxgS 36 iVxgS

'.tf8 37 h6 bxcS 38 h7 winning. A
really magical game.
Question 6: In your own games you
have played both 8 . . . 0-0 and 8 . . . lLJbd7.
36

This move forces a draw by repeti

tion if White desires after 9 . . . a5 (best)
10 lLJa2 (chasing the bishop away)
10 . . . e7 1 1 iVxb7 Mb8 12 iVa6 (12
iVxc6 Mb6!) 12 ... Ma8 1 3 iVxc6 Mc8 14
iVbS Mb8, as the queen cannot escape
from the rook's attack. So if your op
ponent is much weaker than you, or
you desperately need a win, you must
play 8 . . . 0-0, as 9 iVb3 iVe7! (protecting
b7 and facilitating ... c6-cS) is nothing

Th e O l d M a in L in e : Bla c k a ll o w s e 3 - e 4

for White. Alert readers will observe

1 2 ....txa5

that I chose 8 . . .0-0 against Richardson

in just such a must-win situation. The

tiJe4 1 5 tiJxe4! .txe4 1 6 "lif7 ! !

1 3 "lixe6+ '>t>dS

1 4 e5

attempt to avoid the repetition in this

game is brutally dealt with.

9 :tlVb6 1 0 e4
..

brilliant

discovery

of

David

Gliksman. The queen moves to allow

the e-pawn to advance.
1 6 . . . MfS

10 . . . .tg6

10 .. .ctJxe4 1 1 CLlxe4 xe4 1 2 xe6!

is clearly better for White.
11 xe 6 ! fxe6 1 2 a 5 !

The situation is already desperate.

1 6 ... xf3 1 7 e6! (D.Strauss) 1 7 ... CLlf6
1 8 e7+ c8 19 e6+ c7 ( 1 9 . . . CLld7
20 e8 20 f4 + wins, while instead
1 6 . . . g6 17 xg7 Me8 1 8 dS ! ! (again
D.Strauss) ,

opening

more

central

lines, is crushing as 1 8 . . . cxdS 19 g5 +

c8 20 Mfc 1 + is appalling for Black.
17

'llIVx g7

Si.d5

1S

e6!

.txe6

19

"lig5+!

Now the point of 1 2 as is revealed:

this check picks up the loose bishop
on as !
1 9 .. . '>t>cS 20 Mxa5 gS 2 1 "lih5 .tg4
22 Mg 5 ! .txh5 23 MXgS+ tiJfS 24
MxfS+ '>t>d7 25 tiJe5+ '>t>c7 26 MxaS

Diverting the bishop to an inferior

square.

'i'xd4 27 Me 1 .te2 2S MeS c5 29

e7+ '>t>cS 30 tiJc6 1 -0

37

Th e S l a v

S u m mary

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

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

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

8 0-0
8 . . 0-0
.

8 . . . LtJbd7 (D)
9 "iVe2 g4
10 dl - game 13
10 h3 - game 14
9 "iVb3 - game 1 5
9 e2 CL'l b d 7 1 0 e 4 iL g 6 1 1 iL d 3 iLh 5 (D)

11. . .h6 - game 12

1 2 e5

12 f4 - game 9
1 2 . . . CL'l d 5 (D)

13 LtJxds - game 1 0
1 3 LtJe4 - game 1 1

38

. . .

CL'lbd 7

11

. . .

iLh 5

12

. . .

CL'ld5

CHA PTER THREE

The New M ain Line:
Black fig hts for contro l of e4

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

dxc4 5 a4 iLf5 6 iZJe5

The sequence 1 d4 d5 2 c4 c6 3 ttJf3

ttJf6 4 ttJc3 dxc4 5 a4 f5 6 ttJe5 is the
latest fashion at the highest level. 6
ttJe5 truly is a 'modern' move: White
does not worry about putting his king
into safety; he tries immediately to
achieve e2-e4 by moving the knight to
e5 and playing f2-3 . In principle
Black should not allow White to
achieve e2-e4, since here it is ideally
defended by a pawn on f3 , leaving the
white pieces free for active operations.
In this chapter, Black continues the
fight for e4 with an interesting piece
sacrifice:
6 e6 7 f3 iL b4
. . .

Pinning the knight on c3 and thus

fighting for e4.
8 e4

Question 1 : What is the material

balance?
A nswer: Black has four pawns for a

piece, though it is touch and go

whether he can hang on to the c4pawn. Even three pawns, however, is
good material compensation for the
temporarily rather inactive bishop on
f1 .

... and after this there is no going

back!

Game 16
Lal i c-Sadler

8 .lil.xe4 9 fxe4 iZJxe4 1 0 iLd2

10 'i'f3 leads to a well-known
. . .

draw
by repetition after 10 . . . 'i'xd4 1 1
'fxf7+ d8 1 2 g5+! ttJxg5 1 3 'i'xg7
.ixc3+ 14 bxc3 'i'xc3+ 15 e2 'i'c2+
16 el 'i'c3+, etc.
10

. . .

xd4 1 1 iZJxe4 xe4+ 1 2 e2

iLxd 2 + 1 3 xd 2 d 5 +

Hastings

1995/96

1 c4 c6 2 d4 d 5 3 iZJf3 iZJf6 4 iZJc3

dxc4 5 a4 iLf5 6 iZJe5 e6 7 f3 iLb4
8 e4 iLxe4 9 fxe4 iZJxe4 1 0 iLd2
xd4

11

iZJxe4

xe4+

12

e2

iLxd2+ 1 3 xd2 d5+ 1 4 c3

39

Th e Sla v

After 14 c2 lLJa6, which is consid

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

1 4 . . . 0-0 ! 1 5 e3

This looks incomprehensible - what

is wrong with 15 lLJxc4? The problem
is 15 . . . b5! 16 lLJe5 and now 16 . . . b4+!
The king's exposed placement gives
Black a vital extra tempo for the at
tack. 17 xb4 is met by 17 . . . lLJa6+! 1 8
a3 ( 1 8 'Yxa6 l'hb8+! 19 c3 'Yxe5+
winning) 18 . . . Mab8 19 'i'e3 'i'd6+ 20
a2 lLJb4+ 21 b 1 'i'd1+ 22 'i'c1
'i'd4! (threatening . . . 'Yxe5 and
. . . 'i'e4+) 23 'i'e 1 lLJc2!
see

follo wing diagram

Thanks to this powerful knight

thrust, Black is now winning by
force. Fasten your seat belts, a rather
long variation lies ahead!
40

The attempt to turn down the gift

with 17 c2 is met by 17 . . . b3+ 1 8
c3 lLJa6! 19 'i'e3 ( 1 9 'Yxa6 'Yxe5+ 20
xb3 Mab8+ wins for Black)
19 . . . Mab8 (intending . . . lLJb4) 19 xa6
'i'a5+ 20 d3 'Yxa6+ 2 1 e4 with a
crazy position, where anything could
happen (especially to the white king!) .
15 'i'e3 first of all aims to take con
trol of some dark squares; secondly,
White frees the bishop to develop and
threatens xc4.
1 5 . . . b 5 ! 1 6 .i1L.e2

Th e N e w M a in L in e : Bla c k fig h ts fo r c o n t r o l o f e 4

Instead 1 6 axb5 cxb5 1 7 e2 tiJd7!

tLlxd7 'li'xd7 19 f3 l:I.ac8 20 l:I.xa7
b4+! 2 1 c2 (Hubner points out that
2 1 xb4 loses to 2 1 . . .l:I.b8+ 22 a3
'i'd6+ 23 a2 l:I.b3 24 'li'f2 l:I.fb8)
2 1 . . .'li'b5 gave White nothing in the
stem game Gelfand-Hubner, Munich
1992. Even if White wins both of the
black queenside pawns for his b
pawn, and manages to swap off the
queens and both sets of rooks, the
resulting ending is likely to be drawn
since White has the wrong-coloured
rook's pawn for his bishop . Hence, all
Black needs to do is to aim for a posi
tion like
18

and he will draw Slllce the best

White can achieve is

which is stalemate! The game con

tinued 22 l:I.d1 b3+ 23 b1 g6 24 g4
(taking f5 from the black queen)
24 . . . 'li'b8 25 l:I.a4, when 25 . . . 'li'b5 26
l:I.a7 'li'b8 would have led to a draw by
repetition according to Hubner, while
25 . . . 'li'xh2!? 26 l:I.c 1 c3 27 bxc3 l:I.fd8
led to great complications.
1 6 . . . CLld7 1 7 CLlxd 7 'iVxd 7 1 8 'iVc5

A new idea. Piket had earlier

played 18 f3 against Kramnik, but
Black's strategy is similar in both
cases.
1 8 . a6! 1 9 Uhd 1 Yz - Yz
. .

Exciting stuff, this grandmaster

chess! 18 . . . a6! was in fact a strong new
idea and showed (I think) good under
standing of the position. Black has
41

Th e Sla v

four pawns for the piece, which is

ample. However, his pieces are pas
sive and he has to find a way to acti
vate them. White is strong on the
light squares (he has a light-squared
bishop) , but weak on the dark
squares, so I have to put my queen in
contact with some dark squares. The
c7-square is the obvious spot since
from there the queen eyes as, eS , f4
and the h2-pawn. However, I obvi
ously couldn't play 1 8 . . :iWc7, as 19
axbs would win for White. Therefore
1 8 . . . a6! seemed logical, and after a lit
tle calculation I saw that it was indeed
the best move. For example, after
19 . . . 'i'c7, 20 Md6? Mad8 ! 21 Mxc6 'i'f4!
is extremely worrying for White.

The more active alternative,

16 . . .'!iJcS, is considered in Game 22.

1 7 1le2

17 . . . 'i'xg2 wins a pawn, but after 1 8

Mhg1 'i'xh2 19 Mxg7 h e has problems
defending his second rank.
1 7 . . . a8 1 8 g4!

The next game is intended as a cau

tionary tale for black players, and I
hope that my opponent will forgive
me for using it in this way. Cynics
may point out that I am demonstrat
ing one of my rare wins from a cata
strophic British Championship!

Game 1 7
Sadler-Ferguson

British Championship 1996

1 d4 d 5 2 c4 c6 3 lZlf3 lZlf6 4 lZlc3
dxc4 5 a4 llf5 6 lZle5 e6 7 f3 ll b4
8 e4 llxe4 9 fxe4 lZlxe4 1 0 ll d 2
xd4

11

lZlxe4

xe4+

12

e2

llxd 2+ 1 3 xd 2 d5+ 1 4 c2 lZla6

Black cannot hold on to the c4pawn with 14 . . . bS as 15 Md1 'i'cs 1 6

axbS axbs 1 7 'i'f3 ! , attacking f7 and
the rook on a8 , is just one way of ex
ploiting Black's mistake.
1 5 lZlxc4 0-0-0 1 6 e3 b8
42

Black has played rather slowly, put

ting his king to safety in the corner,
so White begins to take control. This
nice move takes away the annoying
check on fS from the black queen.
1 8 . . . d 7

A novelty. Kramnik-Kir.Georgiev,
Moscow Olympiad 1994, had contin
ued 18 . . .f6 19 Mhd1 'i'g2 20 Mxd8
Mxd8 2 1 'i'xe6 with a clear advantage
for White.

Th e N e w M a in L in e : Bla c k fig h ts fo r c o n t r o l o f e 4

1 9 Bad 1 tLJ b4+ 2 0 b 1 tLJ d 5

Black has got his knight to a central

outpost on dS , but it is hard to suggest
another active thing for him to do.
2 1 a3 c7 22 a5 b8 23 a 6 !

Softening up the black queenside.

23 . . . b6 24 f3 .l::!. h e8 2 5 .l::!. h e 1 f6 26
b3 a8 2 7 h4! b5 28 tLJe3 a5

A desperate attempt to break out,

but one that is easily refuted.

Question 2: What went wrong?

Why did Black lose without seem

ingly being able to put up any sort of
fight?
Answer: Black knew very little
about this line and was extremely un
fortunate that this is one of those
variations where knowledge is essen
tial: the big decisions for Black are
strategical - there is almost no chance
to calculate your way out of trouble.
Black must understand where and
when to exchange queens; where to
aim to put his knight; and how to
arrange his pawns. Without this
knowledge, Black has little chance of
success.
Question 3: How does one acquire
this knowledge?
Answer: One must study games in
this line and draw conclusions from
them.

29 tZlxd 5 exd 5 30 nxe8 nxe8 3 1

Game 18

!:!'xd 5 !

K ra m n i k -Lautier

Linares 1994
1 tLJf3 d 5 2 d4 tLJf6 3 c4 c6 4 tLJc3
dxc4 5 a4 f5 6 tLJe5 e6 7 f3 b4
8 e4 xe4 9 fxe4 tLJxe4 1 0 .i1Ld2
xd4

11

tLJxe4

xe4+

12

e2

xd2+ 1 3 xd2 d5+ 1 4 c2 tLJa6

1 5 tLJxc4 0-0-0

It>c2 c7 3 5 n d 7 + b6 36 xb5

Black has castled queenside, and not

kingside. Why? First, castling queen
side brings a rook immediately to the
open d-file; second, if White ex
changes queens, the black king is well
placed to protect the queenside pawns
from a potential attack by the white
knight.

1 -0

1 6 e5 f6 1 7 xd 5

31

. . .

b8

3 1 . . .exdS is met by 32 'iYxds + and

mates.
32 nd6 xa6 33 xc6 Be 1 + 34

43

Th e Sla v

The more critical 17 'ie3

cussed in Games 20 and 2 1 .

lS

dis

1 7 . . . exd 5

26 . . . g 6 27 h e 1 e5 28 a5 tZ:la2 ! 29
xe5 bxe5 30 a 1

tZ:lb4 3 1

tZ:le2

tZ:le6 3 2 .lixa6 !;la8 3 3 .li b 5 !;lxa5 34

Here Black has exchanged queens

on his own terms: he has forced
White to take on ds . We can conclude
that the exchange of queens is only
acceptable to Black if it improves the
black pawn structure. Also, Black
should recapture on ds with the c
pawn: after . . . e6xdS, Black has just a
4-2 majority on the queenside; after
. . . c6xdS, Black has a pawn chain of
five against just two white kingside
pawns on g2 and h2. Black is more
likely to be able to create passed
pawns and a pawn chain that will re
strict the white pieces with the latter
rather than the former.

!;lxa5 tZ:lxa5 3 5 tZ:l e 1 tZ:lb7 3 6 tZ:lf3

We7 3 7 tZ:lg5 tZ:ld6 3 8 tZ:lxh7 tZ:lf7 !

Just in time!
39 g4 tZ:lxh6 40 gxf5 gxf5 4 1 We3
e5

1 8 tZ:la3 tZ:l b4+ 1 9 W d 2 Wd7 20 !;le 1

!;le8 2 1 !;le3 b6 22 .li b 5+ Wd6 23
.lie2 a6

Preventing lLlbs+.

Black's pawns are now very dan

gerous and White must play accu
rately.

24 h4 !;lhf8 2 5 h 5 f5 26 h6

42 .lie6 Wd6 43 .lib7 tZ:lg4+ 44 Wf3

A typical attacking idea for White.

Although White runs the risk of los
ing this pawn, as it is now cut off
from the rest of its troops, if White
can get a knight to gS or a bishop to
g8 . . .

e4+ 45 Wf4 tZ:le5 46 .lia6 e4 47 tZ:lg5

44

tiJd3+ 48 We3 f4+ 49 Wd4 e3 50

tiJf3 e2 51 .lib7 tiJe5 52 tZ:le 1 tZ:ld3
53 tiJf3 tiJe5 54 tiJe 1 f3 5 5 We3 We5
56 Wf2 Wd4 5 7 tiJxf3+ tiJxf3 58
Wxe2 tZ:le5 59 Wd2 tiJd3 60 .lixd5

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

Iiix d5 6 1 b 3 Y:l - Y:l

So what if White does not swap

queens?

squares, which makes it easy for

White to blockade them with his
bishop.

Game 19
Karpov-Hjartarson

Tilburg

1988

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

dxc4 5 a4 .ltf5 6 t2Je5 e6 7 f3 itb4

8 e4 iLxe4 9 fxe4 t2Jxe4 1 0 itd2

xd4

11

t2Jxe4

xe4+

12

e2

ltxd2+ 1 3 'kt>xd2 d 5+ 1 4 'kt>c2 t2Ja6

1 5 tLlxc4 0-0-0 1 6 e 5 f6 1 7 e3 ! ?

At the time, this was a new idea.

White is claiming that he has forced
his opponent to weaken the pawn
structure around his king.

23 . . . xb3+

24

x b3

t2Jxb3

25

'kt> x b 3 .l:! d 4 26 h4! .!:!.hd8 2 7 itc4

'kt>c7 28 h5 .!:!.g4 29 h 6 !

1 7 . c5 ! ?
.

17 . . . b8 is considered in the next

two games.
1 8 'kt>b3 t2Jb4

It's that plan again! Now the black

kingside pawns are softened up and
White gradually assumes complete
control.
Black's plan seems very logical: he
is aiming to put a knight on d4.

l:!.d6 32 .l:!h6 e4 33 .l:!hxf6 h 5 34

19 gc 1 ! t2Jc6 20 'kt>a3 ! t2Jd4 2 1 t2J a 5 !

e5 2 2 c3 ! b6 23 t2J b 3

.l:!.xg7 3 7 .l:!.f4 .!:!.g3+ 38 'kt>c2 .!:!.g2+ 3 9

This game is still the model for

dealing with . . . c6-c5 and . . . lZJb4.
White weakens Black's light squares
by forcing all his pawns to dark-

29 . . . l:!.xg2 3 0 hxg7 .l:!.xg7 31

.l:!.cf 1

l:i:6f4 .!:!.d4 3 5 gf7+ .!:!. d 7 36 .l:!.xg7

'kt> c 3 .!:!. g 3 + 40 'kt>d2 It g 4 4 1

l:!.f7+

'kt>d6 42 'kt>e3 a6 1 -0

A really impressive game from

Karpov.
45

Th e Sla v

21 b3 ctJd5 22 f3 f4 23 1;;!; f 1

Game 20

d4 24 f2 xf2 25 llxf2 c8 !

S h i rov- Ba reev

Biel 1991
1 d4 d 5 2 c4 ctJf6 3 ctJc3 c6 4 ctJf3
dxc4 5 a4 lLf5 6 ctJe5 e 6 7 f3 lL b4
8 e4 lLxe4 9 fxe4 ctJxe4 1 0 lLd2
xd4

11

ctJxe4

xe4+

12

e2

lLxd2+ 1 3 xd2 d5+ 1 4 c2 ctJa6

1 5 ctJxc4 0-0-0 1 6 e5 f6 1 7 e3
b8 1 8 1Le2 xg 2 ! ?

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

contest the second rank.
26 lLg4 f5 27 lL h 5 lld7 28 1;;!; f g2
1;;!; h d8 29 ctJe5 llxg 7 30 llxg 7 ctJf4
31

llxh7 lld 5 !

32 ctJg6 ctJxg6 3 3

lL x g 6 lld 7 !

Seeking t o reduce White's winning

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

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

Centralising the knight with tempi.

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

This is probably fine in general for

Black, but not very inspiring. I prefer
46

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

the more active plan that Black

adopted in the next game.

Came 21
K ra m n i k -Ivanchuk

Linares 1994
1 CLlf3 d 5 2 d4 l2lf6 3 c4 c6 4 l2lc3
dxc4 5 a4 iLf5 6 l2le5 e6 7 f3 iLb4

8 e4 iLxe4 9 fxe4 l2lxe4 1 0 iLd2

xd4

11

l2lxe4

xe4+

12

e2

xd 2+ 1 3 xd2 d5+ 1 4 c2 l2la6

1 5 4:Jxc4 0-0-0 1 6 e5 f6 1 7 e3

29 l2le3 c5 30 l2lc4 b6 3 1 b 1

Wb8! 1 8 iLe2 e 5 !

Y:, - Y:,

Neither side has an obvious way to

contmue.

This i s the best set-up for Black:

pawns on f6 and eS, keeping the
queenside pawns where they are,
while transferring pieces to the two
outposts d4 and dS . Sometimes, Black
will transfer the knight to d4 via cS
and e6.

It is time to draw a few conclusions:

1. Black should castle queenside.
2. If the queens are to be ex
changed, Black wants them to be ex
changed on dS, when he will improve
his pawn structure with . . . c6xds .
3 . Black's ideal outpost for his
knight is on d4 and not ds .
4. Black's best pawn structure in
the middle game is to place his pawns
on a7, b7, c6, eS, f6, g7 and h7, since
this creates two central outposts, d4
and ds .
But this is not the end of the story.

Came 22
K ramni k-Shirov

1 9 93 e6 20 b3 .nd 5 21 b2 Rhd8
22 :gad 1 l2l b4 23 c3 e7 24 l2le3

Dortmund 1996

Sxd 1 25 Rxd 1 .Q.xd 1 26 iLxd 1 9 6 2 7

d2 c7 28 l2lc2 a 5 !

1 l2lf3 d5 2 d4 c6 3 c4 l2lf6 4 l2lc3

An important move, securing the

knight on b4. If White could success
fully play a4-aS, then the knight could
become vulnerable and Black would
have to be careful.

dxc4 5 a4 iLf5 6 l2le5 e6 7 f3 iL b4

8 e4 iLxe4 9 fxe4 l2lxe4 1 0 iLd2
xd4

11

tLlxe4

xe4+

12

e2

iLxd2+ 1 3 xd2 d5+ 1 4 c2 l2la6

1 5 l2lxc4 0-0-0 1 6 e3 !
47

Th e S l a v

a2 is j ust winning for White: his

king is fact perfectly safe. 20 . . . .:gS ! ?
and 2 0 . . :i'g2 (threatening . . :i'g6+) are
both interesting, but the onus is
clearly on Black to find a reasonable
continuatIOn.
20 xd4 xe2+ 21 ct:Jd2 gd8 22
c5 I:Ixd 2+ 23 b3 I:Ixb2+ 24 a3

With hindsight, very obvious! This

is the very latest idea in this line:
White avoids giving Black the extra
tempo .. .f7-f6, forcing Black to look
for another defensive formation.
1 6 . . . eiJc 5

1 7 e2 xg2 1 8 ghg 1

xh2 1 9 I:Ixg7 !

This move was originally thought

Amazingly White's king is quite

to be impossible due to 19 . . . .:cI3, but

20 ':h 1 ! is very strong for White after

safe, and now it is j ust a matter of the

20 . . . .:xe3 2 1 ':xh2 or 20 . . :iUxh 1 2 1

material telling in the end.

ilxcl3 .

24 . . . I:Id2 25 I:Ig3 '>iVe4 26 .i::i. b 3 b6 27

f8+ b7 28 xf7 + a6 29 f3

1 9 . . . I:Id4? ! ?

xf3 30 gxf3 a5 3 1 .i::i. f4 gd3+ 32

b2 .i::i. h 3 33 ge 1

I:Ih2+ 34 b3

.i::i. h 3+ 3 5 c2 I:Ih2+ 36 d3 I:Ih6 37

I:Ife4 c 5 3 8 I:Ixe6

The black pawns are insufficiently

advanced to cause White any real
problems.
3 8 . . . I:Ih3+ 39 g 1 e 3 I:Ih 1 40 I:I3e4
I:Ih3+ 4 1 c4 g g 3 42 gh6 a6 43
I:Ixh7

gg5 44 gb7

gh5

45

ge6

I:Ih4+ 46 d5 g b4 47 I:Ic6 gd4+ 48

e6 gb4 49 .i::i. b 8 1 -0

An amazing attempt that j ust falls

short. When
position,

started to analyse this

w anted to play 1 9 . . . .:hg8,

White will win easily by attacking

the black pawns from the rear with
his king.

aiming to meet 20 ':xf7 with 20 . . . .:g2,

but 2 1 'iYxcs ':xe2+ 22 b3
48

':cI3+

23

Postscript: Since the first edition of

Th e N e w M a in L in e : Bla c k fig h t s fo r c o n t r o l o f e 4

this book went t o press, there have

been several important games in this
variation, which have focused mainly
on the previously relatively neglected
idea of Black castling kingside
(15 . . . 0-0) instead of queenside. This is
based on the fine idea of
1 5 . . 0-0 1 6 "Vje5 l:!.ab8 !
.

This wonderful idea has two

points: first, Black supports the . . . b7bS advance to open up the queenside
against the exposed white king; and
second, Black protects the pawn on
b7 against the typical White manoeu
vre 'iWxdS followed by ctJaS or ctJd6.
So far, White has even been struggling
in this position!
17 .Le2

17 Md1 ctJb4 + 18 el ctJa2 + 19

c2 ctJb4 + was a draw by repetition
in P.Cramling-Hector, Malmo 1998.
1 7 . ctJb4+ ! 1 8 Wc3 b 5 1 9 l:!.hd 1 f6 !
. .

20 'iVg 3

e4

21

ctJe3 ctJ d 5+

22

exercises less influence from f3 than

eS, but on the other hand, White
threatens the immediate Md1 , chasing
the queen from dS.
1 6 . . . c5

1 8 Wb1

Kramnik's improvement over

19 . . . 'iWgS 20 'iWg4! 'iWcs 2 1 Ma3 Md4 22
'iWhS, when Black had wasted rather a
lot of time in Beliavsky-Shirov, Bel
grade 1997.
20 l.1a3 l.1fd8 21 g4 g 5

Black is beautifully mobilised.

22 l:!.b3 a5 23 h4 "Vjg6+ 24 Wa 1
ctJc2+ 25 Wa2 ctJ b4+ 26 Wa 1 ctJc2+

% -%

So the idea of castling kingside

may, after all, be Black's best course
of action.
To finish this section, here are two
games featuring slightly offbeat at
tempts by White.

tLlxd 5 cxd 5

Yes, it really is as bad as it looks! I

don't know how, in the game Kram
nik-Van Wely, Tilburg 1998, White
managed to hang on!

1 7 Si.e2 ctJ b4+

l:!.ad8 1 9 l:!.c 1 l:!.d4 !

Game 23
Adianto - K ramnik

London (Intel Grand Prix) 1994

23 .Ld3 l:!.fc8+ 24 Wd2 b4+ 25

1 d4 d5 2 c4 c6 3 ctJf3 ctJf6 4 ctJc3

'iit> e 2 bxa4 26 'it>f 1 l:!.b6 2 7 e3 b3

dxc4 5 a4 .Lf5 6 ctJe5 e6 7 f3 .Lb4

28 l:!.d2 g 6 2 9 "Vjd4 b4 3 0 Si.e2

8 Si.g5

'i'xd4 % - %

Fighting for e4 by pinning the

black knight, so that 9 e4 is now a
threat. Instead 8 ctJxc4 0-0 9 gS h6
10 h4 was met by 10 . . . cS! 1 1 dxcS
'iWxd1+ 12 xd1 (12 Mxd1 c2! 1 3
Mel xa4! 14 xf6 gxf6 1 5 Mal b3
16 ctJb6 ctJc6 17 ctJxa8 Mxa8 gave
Black excellent compensation for the
exchange in Beliavsky- Bareev, USSR
1986) 12 . . . Md8+ 13 el ctJc6! 14 e4
h7 1 5 f2 ctJd7! in Akopian-Oll,

A few months after that game,

these two players, Kramnik and Van
Wely, continued their debate of this
variation at Wijk aan Zee 1999 - but
this time with colours reversed! Van
Wely varied from the standard 16
i"eS with
1 6 f3

The latest idea. The white queen

49

Th e Sla v

New York Open 1994. 1 5 .. .'Jd7! intends . . . tLJxc5, highlighting the weakness on b3, and . . .f7-f5 activating the
light-squared bishop on h7. After 16
d3 xc5! 17 xc5 tLJxc5 18 c2 f5!
1 9 exf5 tLJd4! Black stood clearly bet
ter.

22 lLle3 0-0 23 !:!'a 1 !:!.b3 24 !:!.xa4

SLxe3+ 25 \tg3 z:!.xb2 26 SLf 1 f 5 !

White is in big danger, but some

how he just hangs on.
27 !:!.xe6 f4+ 28 \th3 e8 29 !:!.xe8
!:!.xe8 30 c4+ \th7 3 1 d 5 e5 32
!:!.c4 !:!.h5 3 3 g3 e5 34 !:!.c7 !:!.b6 35
8 . . . h6 9 SLh4 c5 1 0 dxc5 a 5 ! 1 1
l'i'd4 lLlc6

1 2 lLlxc6

bxc6

1 3 e4

!:!.f7 \tg6 36 !:!.xa7 !:!.b8 3 7 !:!.d7 h5

38 e7 l:ib2 39 f8 fxg 3 40 !:!.xg7+

itxc 5 !

\tf6 41

1 3 . . . g6 14 f2 i s less good for

Black.

lh - lh

1 4 xc4 g6

1 5 a6 xa6

16

SLxa6 !:!. b 8 !

!:!.f7+ \tg6 42 !:!.g7+ \tf6

Game 24
I .Sokolov-Bareev

Leon 1 995
1 d4 d 5 2 c4 c6 3 lLlc3 lLlf6 4 lLlf3
dxc4 5 a4 SLf5 6 tLle5 e6 7 g 3

White already has trouble with his

queenside.
1 7 lLld 1 lLl d 7 ! 1 8 !:!.C 1 b4+ 1 9 \tf2
lLlc5 20 e2 lLlxa4 2 1 !:!.xc6 c5+
50

Th e N e w M a in L in e : Bla c k fig h ts fo r c o n t ro l o f e 4

White fights for the e4-square in a

different way: by putting the bishop
on g2. This also helps to dissuade
Black from playing the pawn break
... c6-c5, as b7 will be hanging.
7 . b4 8 -li g 2 e4!
. .

Forcing White to block the long

diagonal, which will allow Black to
play ... c6-c5 without fear of xb7.

ctJb5 xe5 14 f4 c5 1 5 ctJc7+ e7

16 ctJxa8 ctJa6! wins for Black)
1 3 . . . ctJc6 is also good.
1 0 . . . c5 1 1 4:Ja2 a5 1 2 d x c 5

Or 12 ctJxc4 cxd4 13 ctJxa5 xa5 14

xd4 ctJc6 1 5 c4 b6+ 16 h 1
ctJa5!, intending . . . ctJb3 .
1 2 . . . d 5 1 3 xd 5 exd 5 1 4 QJxg6
hxg6 1 5 l::!. b 1 4:Jbd7 1 6 e3 l::!. c 8 1 7
b4 cxb3 1 8 4:J c 1 b2 1 9 4:J b 3 c3 20
.!::!. f d 1 4:J e 5 21 d4 -lixd4+ 2 2 '!::!' x d4
b6 23 l::!. x b2

9 f3 -lig6 1 0 0-0

10 e4 c5 11 e3 cxd4 12 xd4
'iVxd4 13 xd4 ctJc6 14 ctJxc6 bxc6 1 5
0-0-0 0-0-0 i s the theoretical recom
mendation, but 16 f1 is more pleas
ant for White due to his superior
structure and Black's inactive bishop
on g6. 13 . . . ctJfd7!?, instead of
13 ... ctJc6, was my first idea in order to
reactivate the bishop on g6 with .. .f7f6 and . . . f7. However, 14 ctJxc4 f6
15 0-0-0 ctJc6 16 f2 (intending
... ttJc6) 16 . . . e7 17 ctJa2! wins the
bishop pair, giving White a small ad
vantage, as 17 . . . c5 loses to 1 8
lhd7+! xd7 1 9 xc5, winning two
pieces for a rook. In fact, 1 1 .. .c7! is
stronger: 12 ctJxc4 cxd4 (attacking the
knight on c4) 1 3 xd4 ctJc6 is fine for
Black and 1 2 0-0 cxd4 13 xd4 (13

Better was 23 cxb6 axb6 24 's'xb2

's'c4 with an equal position according
to Ivan Sokolov.
2 3 . . . 4:Jc4 24 l::!. b 1 bxc5 25 QJxc5 0-0
26 4:Jd3 4:Ja3 27 l::!. b 7 4:Jc2 28 l::!. f4
l::!. f e8 29 h 3 l::!. c 3

29 . . . ,S,b8! (I.Sokolov) gave chances

for an edge for Black.
30 f2 g5 3 1 l::!. f 5 g4 3 2 xg4 4:Je3
3 3 l::!. x f6 4:Jxg4+ 34 fxg4 gxf6 3 5
l::!. x a7 l::!. a 3 36 4:Jf4 l::!. a 2 3 7 a 5 l::!. e 4
38 l::!. a 8+ h 7

39

l::!. d 8 l::!. x a5 40

4:Jxd 5 l::!. x d 5 41 l::!. x d 5 l::!. x g4 42 f3

.!::!. a 4 43 g4 g6 44 h4 '!::!' a 1 45 h 5+
g7 46 l::!. f 5 l::!. g 1 47 e3 l::!. f 1 + 48
e4 l::!. g 1 49 l::!. f4 .!::!. g 3 'h - 'h

A tough endgame.

51

Th e Sla v

S u m m a ry

The sidelines do not seem to cause Black any problems, but undoubtedly the
most crucial line at the moment is Kramnik's 14 Wc2 ctJa6 15 ctJxc4 0-0-0 16
'iVe3 . In general, such positions are easier to play for White than for Black.
1 d4 d5 2 c4 c 6 3 lLlf3 lLlf6 4 lLlc3 dxc4 5 a4 f5 6 lLle5
6 . . . e6 7 f3 (0)

7 g3 game 24
-

7 . . . b4 8 e4

8 g5 - game 23
8 . . . xe4 9 fxe4 lLlxe4 1 0 d2 xd4 1 1 lLlxe4 xe4 1 2 e2 xd2+
1 3 'It>xd 2 d5+ 1 4 'It>c2

14 Wc3 game 1 6
-

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

16 'iVe3
16 . . . Wb8 game 1 7
1 6 . . . ctJc5 game 22
-

1 6 . . . f6 1 7 e3

17 'iVxd5 - game 18
1 7 . . . 'It>b8

17 . . . c5 game 1 9
-

1 8 e2 (0)

1 8 . . . 'iVxg2 - game 20
18 . . . e5 - game 21

7 f3

52

1 6 e5

18 e2

CHA PTER FOUR

The New Main Line:
Black co unterattacks

1 d4 d 5 2 c4 c6 3 ltJf3 ltJf6 4 ltJc3

dxc4 5 a4 i.f5 6 ltJe5

In this chapter we consider lines in

which Black does not doggedly fight
on for e4, but switches his attention
to the d-pawn, arguing that 6 ctJe5 has
weakened White's control of d4. Re
cent attention has concentrated on
6 . e6 7 f3 c5! ? 8 e4 cxd4, which is a
specialty of both Boris Gelfand and
Alexei Shirov. In the following games
you will see the wildly different ways
in which they handle this line!

man-Salov, 199 1 , but had not been

tried since, as, in his notes, Khalifman
had pointed out a continuation that
seemed to equalise for Black.

. .

Game 25
Piket-Gelfa n d

Wijk aa n Zee 1996

. . .

.Lg6 1 0 i.b5+ ltJfd 7

10 . . . ctJbd7?? simply loses to 1 1 g5 .

1 d4 d 5 2 c4 c6 3 ltJf3 ltJf6 4 ltJc3

1 1 'ixd4 a 6 !

dxc4 5 a4 i.f5 6 ltJe5 e6

In the 1920s and 1930s, 6 . . . ctJbd7 7

ctJxc4 -&rc7 8 g3 e5 (attacking d4) was
popular, but after 9 dxe5 ctJxe5 10
14 gd8 11 -&rc 1 d6 12 ctJxd6+
'ixd6 13 g2, White stands better.
He has the two bishops and Black
cannot activate his queenside pawn
maJonty.
7 f3 c5 8 e4 cxd4 9 .Lxc4 ! ?

A n unusual move that was success

ful in its first appearance in Khalif53

Th e Sla v

And this was it. If now 12 iLe2,

then 1 2 . . . CDxeS 13 'ilxeS CDc6 is very
good for Black, so White is forced
into exchanges.
1 2 .liLxd 7 + tLlxd 7 1 3 xd 7 + xd 7
1 4 tLlxd 7 xd 7

1 5 e 2 .liL b 4 1 6

gd 1 + ctJe7 1 7 .liLf4 g h c 8 1 8 gac 1

%-%
White obviously has more crucial
possibilities. Who better to test the
black position than Garry Kasparov?

Game 26
Kasparov-Shi rov

Dos Hermanas 1 996

1 d4 d 5 2 c4 c6 3 tLlf3 tLlf6 4 tLlc3
dxc4 5 a4 .liLf5 6 tLle5 e6 7 f3 c 5 8
e4 cxd4 9 exf5

e7 13 CDxhS, 13 . . . cxb2 is discovered

check, so Black wins. White can try
10 f2!?, since after 10 . . . dxc3 1 1
'ilxds+ xdS 1 2 CDxf7+ e7 13
CDxhS, 13 . . . cxb2 is no longer discov
ered check, but Black can exploit the
other exposed piece in White's posi
tion: the knight on eS. He can play
10 . . . 'ilc7!, threatening both 1 1 . . .'ilxe5
and 1 1 . . .dxc3 , as White can no longer
exchange queens with 'ilxdS+. It
seems that White can stop both these
threats with 1 1 'ilxd4, but Black has
the last laugh after 1 1 . . .iLcs, picking
up the queen. Sadly Black is not com
pletely winning after 1O . . . 'ilc7, as
White can play 1 1 CDa2, attacking the
bishop on b4, but after 1 1 . . .'ilxeS 12
CDxb4 'ilcS ! Black has powerful com
pensation for the piece: two pawns
and the exposed white king. I think
that Black is better here. Garry played
the morc natural. . .
1 0 .liLxc4 d 6 ! ?

An amazing move, adding t o the

confusion by attacking another piece.
1 1 .liLb5+ tLlc6 1 2 tLlc4 c5?

The main line and the only real test

of Black's play.
9

. . .

.liL b4 ! ?

Black could not take the knight on

c3 , regaining his piece, because after
9 . . . dxc3 10 'ilxdS+ xdS 1 1 CDxf7+
White wins a rook. However, if we
imagine that it is Black's move after
9 . . . iLb4, then 10 . . . dxc3 is possible be
cause after 1 1 'ilxdS+ xdS 12 CDxf7+
54

A serious and, in such a sharp posi

tion, fatal mistake. Black could simply
have retreated with 12 . . . 'ild7 when,

Th e Ne w M a in L in e : Bla c k c o u n t e ra t t a c k s

due to the threat o f . . . d4xc3, White

probably has nothing better than to
repeat moves with 13 ctJe5 d6.
1 3 .1Ld2!

Black cannot regain the piece now,

as 13 . . . dxc3 14 bxc3 a5 loses a piece
to 15 ctJxa5 . Obviously Shirov did not
miss this move; but I believe that he
overlooked something extremely
cunning later on.
13 0-0 1 4 ctJa2 ..txd2+ 1 5 xd 2
CiJ e 7
. . .

Black's is threatening ... a7-a6, win

nin g the bishop, while he can also try
... LZJxf5, intending . . . ctJe3 . White
seems to have problems but . . .
1 6 b4 !

Forcing the exchange of queens as

16.. .'i'xf5 allows 17 xe7.
16

. . .

xb4+ 1 7 ctJxb4 a6 1 8 ctJ b 6 !

White does not lose the bishop af

ter all! The rest is easy for Kasparov.
1 8 axb5 1 9 ctJxa8 xa8 20 fxe6
bxa4 21 exf7+ xf7 22 d2 e6
23 J:i.hc 1 d6 24 b3 b 5 25 bxa4
bxa4 26 .8.c4 ctJf5 27 ctJc2 ctJd7 28
gcxa4 .8.xa4 29 .l:1.xa4 ctJb6 30 ctJxd4
1 -0
. . .

fine game by Kasparov, but not

one that refutes Black's idea. Let us

take another look.

Game 27
Gelfand-Shirov

Dortmund 1996
1 d4 d 5 2 c4 c6 3 ctJc3 ctJf6 4 ctJf3
dxc4 5 a4 ..tf5 6 ctJe5 e6 7 f3 c5 8
e4 cxd4 9 exf5 ..tb4 1 0 ..txc4 d6

Shirov could not resist trying this

idea a second time, but this time he is
convincingly mauled. I do wonder
why Black has been avoiding
10 . . . dxc3 1 1 xds+ xds 12 0-0 (12
ctJxf7+ e7 1 3 ctJxhs cxb2+) 12 . . . cxb2
13 xb2 e7 14 fxe6 fxe6.

White's two bishops, the weak (but

extra) pawn on e6 and the slightly
exposed king on e7 obviously offer
compensation for the pawn, but I
don't see an advantage for White.
Black's bishop on b4 prevents his op
ponent from playing Me 1 and ganging
up on e6, so 15 ctJd3 is tempting:
15 . . . d6 is met by 16 Mfe 1 while
1 5 . . . a5 16 a3+ is also sub-optimal,
as Jon Speelman would say! However,
15 . . . McS ! , attacking the bishop on c4,
is the best defence: 16 ctJxb4 Mxc4 is
55

Th e Sla v

good for Black and 1 6 iLxe6 xe6 17

tDxb4 leaves an equal position. Fi
nally, 16 iLb3 is met by 16 . . . tDc6, pro
tecting the bishop, when 17 tDxb4
tDxb4 18 Mfe 1 Mc6! (the point of
1S . . . Mc8) 19 iLa3 as ! , intending . . . f7,
is fine for Black. I feel that the onus is
on White to demonstrate more than
just sufficient play for the pawn.

Suddenly it is Black's king that is in

danger!
1 8 . . . xh 2 1 9 5lxc6 .l::!. a c8 20 xf6
xf4+ 21 c2 g6 22 .l:!.df 1 h 2+ 23
.l::!. g 2 h3 24 fxg6 fxg 6

1 1 5l b 5+ tLlc6 1 2 5lf4 !

25 .i:!xg6+ hxg6 26 xg 6+ h8 27
.l::!. h 1 !

Winning the queen. The game is

over.
27 . . . .i:!f2+ 28 b3 .i::!. b 8+ 29 a3 1 -0

This prevents the capture 1 2 . . . dxc3

due to 13 tDxc6 cxb2+ 14 tDxb4+! ,
when, thanks t o the great strength of
the discovered check, White wins the
whole house!
1 2 . . . 0-0 1 3 tLlxc6 xf4 14 xd4 !

So it seems as if, despite these two

reverses, Shirov's 9 . . . iLb4 may well be
just about playable. Let us now take a
look at the more restrained continua
tion 9 . . . ctJc6.

5lxc3+ 1 5 bxc3 g 5 1 6 f4 xg 2 1 7

Came 28

O-O-O ! bxc6 1 8 .i:! hg 1 !

l I Iescas-Gelfand

Dos Hermanas 1 996

1 d4 d 5 2 c4 c6 3 tLlf3 tLlf6 4 tLlc3
dxc4 5 a4 5lf5 6 tLle5 e6 7 f3 c 5 8
e4 cxd4 9 exf5 tLlc6
see

follo wing diagram

The older and more solid move.

1 0 tLlxc6 bxc6
5lxc4

56

11

fxe6 fxe6

12

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

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

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

Game 29
Van der Sterren-Petu rsson

San Bernardino Open

1 992

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

dxc4 5 a4 lLf5 6 ttJe5 e6 7 f3 c5 S
e4 cxd4 9 exf5 ttJc6 1 0 ttJxc6 bxc6
11

1 2 e2 i s considered i n Game 30,

while more adventurous players may
like to consider the alternative 12 ltJa2
d3, keeping the piece and hoping to
unravel later with g2-g3, g2 and 0-0,
although Black's counterplay is very
dangerous!
12

. . .

fxe6 fxe6

1 2 lLxc4 dxc3

13

bxc3

A different move-order that should

be met by 13 . . . xd1+ 14 <;t>xd1 <;t>d7,
with the same ideas as in Illescas
Gelfand above.
1 3 . . . '>il\Va 5 ? 1 4 '>il\Ve2 ! !

dxc3 1 3 '>il\VxdS+ <;i>xdS 1 4 bxc3

This rook sacrifice is a magnificent

concept!
White is a bit better in this ending,
since he has a slightly better pawn
structure (fewer pawn islands) and the
two bishops, but Black's pieces are
active.
14 .':tJd 5 1 5 <;i>d2 lL d 6 1 6 <;i>c2 <;i>d 7 !
. .

The king protects both weak

pawns and helps to cover the only
open file on the board: the b-file.
1 7 lLd2 ..hfS 1 S gab 1

gabS

19

1 4 . . . xc3+ 1 5 <;i>f 1 xa 1 1 6 '>il\Vxe6+

<;i>dS

16 . . . e7 loses to 17 xc6+ <;t>f8 1 8

xa8 ctJe8 19 <;t>e2 xa4 20 d5 ctJd6
2 1 d3 with a crushing attack
(Petursson) .
1 7 <;i>e2 ! !

This quiet move, allowing the rook

to j oin in the attack, justifies White's
brilliant idea.
57

Th e Sla v

1 7 . . : xa4 1 8 .l:!.d 1 + xd 1 + 1 9 xd 1
e 5 20 f7 ge8 2 1 xg 7 ctJd 7 22
iLf7 .l:!.f8 23 iLe6 ctJf6 24 b7 .l:!.e8
25 xa8+ e 7 26 iLf4+ 1 -0

Came 30

This ending is somewhat more

awkward for Black since his king is a
little more open. Nonetheless, it is
surprising how quickly his position
goes downhill.
1 8 . . . .l:!.hd 8 ? !

Topalov-G elfand

Dos Hermanas 1 996

Black must try and activate his

knight: ls . .ebds! is stronger, when 19
xg7 .l:IhgS ! regains the g2-pawn. Af
ter 1 9 .l:Iac 1 , 19 . . . .l:IheS is best, prevent
ing 20 xds cxds 2 1 .l:Ic7 + due to
2 1 . . .d6+, a discovered check that
wins the rook!
.

1 d4 d 5 2 e4 e6 3 ctJf3 ctJf6 4 ctJe3

dxe4 5 a4 iLf5 6 ctJe5 e6 7 f3 e 5 8
e4 exd4 9 exf5 ctJe6 1 0 ctJxe6 bxe6
1 1 fxe6 fxe6 1 2 e2 ! ?

An interesting novelty that aims for

a slightly different endgame.

1 9 :l:!hd 1 :l:!ab8 20 g3

This places the g-pawn on a pro

tected square and thus prevents the
black knight from moving. Black
now has serious problems which he is
unable to overcome.
20 . . . h 5 2 1 gd4 a5 22 a3+ e8
23 :l:!ad 1

.l:!.xd4 24 .l:!.xd4 .l:!.b 1

25

d3 ge 1 + 26 f2 f7 27 e5 :l:!a 1
28 e4+ e8 29 d3 f7 30 xa 7
.l:!.a2+ 3 1 f 1 :l:!a 1 + 32 g2 ga2+
33 h3

N ow White is just winning.

33 . . . ga3 34 f4 e 1 35 c5 .l:!.e3 36
1 2 . . . dxe3 1 3 xe6+ e7 1 4 iLxe4

e4+ e8 3 7 b4 Be 1 3 8 xe 1

xe6+

:l:!xe 1 39 a5 Ba 1 40 a6 e7 4 1 gd2

15

xe6

exb2

16

iL b4+ 1 7 e2 e7 1 8 e4
58

iLxb2

1 -0

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

Both 6 . . . e 6 7 f3 b4 and 6 . . . e 6 7 f3
are popular counterattacking sys
tems at all levels of play. However,
Black can also play more slowly, aim
ing to break out from a cramped posi
tIOn.
c5

Game 3 1
Kramni k-Short

Novgorod 1 994
tiJf3 d 5 2 d4 tlJf6 3 c4 dxc4 4
tZJc3 c6 5 a4 .iLf5 6 tlJe5 tlJbd7 7
tilxc4 tlJb6 8 tlJe5 a 5

Preventing a4-as-a6, breaking up

the black queenside. 8 . . . e6 is consid
ered in the next game, while 8 . . .'Jbd7,
still seeking the exchange of knights,
was crushed by 9 iVb3 ! ctJxe5 10 dxes
CLlg4 1 1 iVxb7 ctJxes 12 f4 ctJg6 13 e4
_d7 14 fs ctJes 15 f4 f6 16 xes
[xeS 17 .s,dl in Kasparov-Timman,
Riga 1995 .

The standard plan of development:

the queen supports e3-e4 from e2,
leaving the rook to cover d4 from d l .
1 4 . . . . tlJfd 7 1 5 tlJd3 ! ?

1 5 ctJxd7 ctJxd7 16 e4 is also slightly

better for White.
1 5 . . . e7 1 6 e4 e 5 1 7 d 5 gfd 8 1 8
.iLe 3? !

A mistake, allowing Black to

weaken the white queenside. 18 d2
would have kept an edge according to
Kramnik.
1 8 . . . .iL x c 3 1 9 b x c 3 cxd 5 20 exd 5
tlJc4 ! ?

9 g3!?

A novelty. 9 f3 ctJfd7 10 ctJxd7

CLlxd7 1 1 e4 g6 12 e3 e6 13 c4
b4 14 0-0 is normal, with a slight
advantage for White.
9 . . . e6 1 0 .iLg2 .iLb4 1 1 0-0 0-0 1 2
e3 h6 1 3 e 2 ! .il.. h 7 1 4 J::i: d 1 !

I think that Short may have missed

White's next, but this is actually a
good move!
21

.iLxh6 gxh6 22 g4+ g 5 23

59

Th e Sla v

xc4 ac8 24 b 5 l::t x c3

1 2 xf2 ik'xd4+ and . . . ik'xeS.

Black has good counterplay in this

murky position.

9 . . . a 5 1 0 e4

31 ttJf4 1Le4 32 5d2 c5 3 3 xc 5

10 g4 ttJfdS! 1 1 h4 (1 1 gxfS ik'h4+ 12

d2 ik'f4+ wins back the piece) l 1 . . .f6
1 2 gxfS fxeS (Ruzele) is extremely un
clear.

ttJ x c 5 34 1Lxe4 .l:!.xe4 3 5 ttJ d 5 <Jif8

1 0 . . . 1Lg6 1 1 1Le3 1Lb4 1 2 1Le2 O-O ? !

3 6 ttJxb6 ttJxa4 37 ttJxa4 .6!xa4 38

1 2 ... ttJfd7 is better since 13 ttJxg6

hxg6 gives Black play on the h-file,
while 13 ttJxd7 ttJxd7 transposes to
S . . . aS 9 f3 .

2 5 ttJe 1 b 6 26 d 6 e4 2 7 d 5 f6 28
J:!ad 1 e3 29 fxe3 xe3 3 0 ttJd3 c3

J:! d 5

White has a tiny edge in the end

game but he is unable to make any
thing of it.

1 3 0-0 ttJfd 7

4 1 J:! a 5 a3 42 g4 'it>g7 43 <Ji g 3 <Jig6

Now White can take the two bish

ops.

44 h 3 <Jig7 4 5 'it>h4 'it>g6 46 .l:!.a8

1 4 ttJxg 6 !

<Ji h 7 % - %

1Lg 1 fd8 1 7 b3

38 . . . l:i.d7 39 <Jig2 .l::!. a 2+ 40 'it> h 3 a4

hxg6

1 5 <Jih 1

e7

16

Game 32
Ruzele-Thorsteins

Lyon (European Club

Cup) 1994

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

dxc4 5 a4 1Lf5 6 ttJe5 ttJ b d 7 7 ttJxc4
ttJb6 8 ttJe5 e6 ! ?

White has effortlessly obtained a

wonderful version of the old main
line with 6 e3.
1 7 . . .c5

18

ttJ a 2 !

cxd4

19

ttJxb4

xb4 20 xb4 axb4 2 1 a5 d 3 22

1Lxd3 ttJe5 23 1Lb5 ttJbc4 24 f4 ttJd6
25 1Le2 ttJc6 26 1Lb6 .l:!.e8 27 e5
ttJc8 28 1Lf3 ttJxb6 29 axb6 ttJe7 30
9 f3

1Lxb7 ab8 3 1 a7 ttJc8 3 2 1Lxc8

A sensible reply. 9 as ttJbds

(threatening . . . ttJb4 or . . . .\tb4) 10 a6!?
is consistent but very risky, while 9 g3
is met by 9 . . . .\tb4 10 .\tg2 ttJe4 when
the natural 1 1 .\td2 loses to 1 1 . . .ttJxf2!

exc8 3 3 c7 nxc7 34 bxc7 .l:!.c8

60

35 .l:!.c 1 <Jif8 36 <Jig 1 <Jie7 3 7 <Jif2

<Jid7 38 <Jie3 xc7 39 l::t x c7+ <Jixc7
40 <Jid4 <Jib6 41 <Jic4 1 -0

Model strategy by White.

Th e Ne w M a in L in e : Bla c k c o u n t e ra t t a c k s

S u m m a ry

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

Shirov's handling of the line with 8 e4 cxd4 9 exfS b4. 9 . . .':tJc6 is for the
calmer players amongst you who don't mind taking on a slightly worse end
ing. Probably it is important to choose the right opponent: 9 . . .':tJc6 will be
ideal against an impatient attacking player, while 9 . . b4 would unsettle a
more positionally inclined player.
If you prefer the more solid 6 . . . ctJbd7 7 ctJxc4 ctJb6, and don't mind the
slightly cramped positions that arise from this line, then Thorsteins's 8 . . . e6
looks like a good move-order, since it avoids Kramnik's 9 g3 .
.

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

6

. . .

e6

6 . . . ctJbd7 7 ctJxc4 ctJb6 8 ctJeS (D)

8 . . . aS - game 3 1
8 ... e6 game 32
-

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

9 xc4 game 25
-

. . .

ctJc6

9 . . . b4 10 xc4 'iVd6 1 1 bs + ctJc6

12 ctJc4 game 26
12 f4 game 27
-

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

12 'iVe2 game 30
-

12

. . .

dxc3

13 'iVe2 game 28
13 bxc3 game 29
-

8 ctJe5

9 exf5

l '

. .

. fxe 6

61

CHA PTER FIVE

The S m yslov Variation

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

dxc4 5 a4 ttJa6

The move 5 . . . ctJa6 was originally an

idea of Emanuel Lasker; and it was
later taken up and played with success
by another World Champion Vassily
Smyslov. Recently Ivanchuk and
Short have also used it to good effect.
Question 1: 5 . . ctJa6 looks a little
strange. What does it do?
Answer: 5 a4 (preventing . . . b7-b5)
has weakened the b4-square. In the
main lines with 5 . . . itf5 6 e3, Black
puts a bishop on b4 and develops his
queen's knight to d7; here, Black
leaves his bishop on e7 to keep b4 free
for the knight. Black's light-squared
bishop will now go to g4 to put pres
sure on the d4-square.
Question 2: So what difference does
this make?
Answer: In the 6 e3 lines, Black's
bishops on f5 and b4 combine to pre
vent White from easily achieving e3e4. In this line, Black exerts virtually
no pressure on e4, and very little on
d4, which means that White pretty
much has the centre to himself.
Question 3: Well that doesn't sound
very promising for Black, does it?
What am I supposed to do as Black?
.

62

Answer: Smyslov's style as Black is

perfectly reflected in this system: he is
prepared to accept a slight space dis
advantage and will just place his pieces
on good squares where they coordi
nate well with each other. Since
Black's position is very solid, the op
ponent will not be able to launch a
sudden attack and Smyslov will pa
tiently unravel, gaining space little by
little until he frees himself.

Game 33
Ivanchuk -Smyslov

Tallinn (rapidplay) 1 996

1 d4 d 5 2 ttJf3 ttJf6 3 c4 c6 4 ttJc3
dxc4 5 a4 ttJa6 6 e4

Very straightforward play.

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

. . .

i!.. g 4 7 i!.. x c4 i!.. x f3

7 . . . e6 is dealt with in the next game.

A different approach which leads to

a complicated middlegame.

8 gxf3 e6 9 i!.. x a6 bxa6

8 i!.. e 3 tiJb4 9 a 5

Black has no problems here. First,

he has exchanged two sets of minor
pieces, and such exchanges always
help the player with less space, since it
means that there are fewer pieces in a
confined area; second, Black's doubled
a-pawns give Black the b-file on which
to activate his major pieces and attack
the vulnerable white queenside. Black
would be much less active if his a6pawn were on b7! Certainly Ivanchuk
is happy to exchange queens and es
cape with a draw.

A typical idea from White, aiming

to prevent either . . :iVaS, activating the
black queen, or . . . a7-aS, cementing the
knight on b4. Black has to be a little
careful that this knight, protected
only by the bishop on e7, does not get
cut off from the rest of his army.
9

. . .

i!.. e 7

9 . . . xf3 , forcing 10 gxf3 (10 'iVxf3

ctJc2+) , was still possible but Black
prefers natural development.
1 0 i!.. e 2 0-0 1 1 0-0 b 5 !

1 0 \We2 a 5 1 1 \Wc4 Mc8 1 2 Mg 1 g6

1 3 r;t>f 1 i!.. g 7 1 4 c5 b6 1 5 xb6
axb6 1 6 i!.. e 3 0-0 1 7 MC 1 tiJ d 7 Y:z - Y:z

If Black delays exchanging on f3 , a

more complex situation arises, as we
shall see in the next game.

Game 34
N ovikov-Greta rsson

Berlin Open 1995

1 d4 d5 2 c4 c6 3 tiJf3 tiJf6 4 tiJc3
dxc4 5 a4 tiJa6 6 e4 i!.. g 4 7 i!.. x c4 e6

This far from obvious move is the

black plan for survival in these middlegames.
63

Th e Sla v

Question 4: What is the pomt of

this move?
A nswer: The main idea is that Black

weaker and hence it is easier for Black

to achieve the . . . c6-c5 break.
1 6 ttJe 1 1. x e 2 1 7 ttJxe2 J:t a c 8 1 8

gains just a little more space for his

pieces - remember what I said about
Smyslov patiently improving his posi
tion, taking extra territory little by
little. The other point is that Black
would like to strike at the white cen
tre with . . . c6-c5, but first he needs a
reasonable square for his queen: he
can't put it on the c- or d-files, since
after . . . c5xd4 these files will be opened
and the queen will be in the firing line
of white rooks on d1 and c 1 . 1 1 . . .b5
frees b7 for the black queen, where it
is absolutely safe. A nice bonus is that
after . . . c6-c5, the black queen will join
with the knight on f6 in attacking the
e4-pawn.

z::!. c 3 c 5 ! 1 9 dxc5 ttJa6 ! !

1 2 b3 c7 1 3 Rfc 1 b7 1 4 g 5

d 5 25 ttJc3 d4 26 f3 e 5

R f d 8 1 5 .i1Lxf6 gxf6

Black's control of the dark squares

gives him good chances.

A very neat idea. 19 . . . 1lLxc5 would

have lost to 20 's'xc5 's'xc5 2 1 iVxb4.
20 ttJd3 ttJxc5 2 1 ttJxc5 axc5 22
axc5 xc5 23 c3 xe4 24 xf6

27 g3 b 4 2 8 Rd 1 xd 1 + 2 9 ttJxd 1
d4 30 d 3 g7 3 1 b3 h6 32 a6

%-%
This game is a model illustration of
Black's middlegame strategy in this
vanatlOn.
c5 33 f 1 f6

N ow we move on to look at 6 e3.

Game 35
Benz-Gretarsson

A forced recapture as 15 . . . 1lLxf6

loses a piece to 16 iVxb4. The weaken
ing to Black's kingside is not too seri
ous, however, since White has no
pieces in that area. Moreover, without
his dark-squared bishop, White loses a
lot of control over the central dark
squares, which means that d4 is
64

Oberwart Open 1996

1 d4 d5 2 ttJf3 ttJf6 3 c4 c6 4 ttJc3
dxc4 5 a4 ttJa6 6 e3

The most solid option and proba

bly the best move. White does not
give Black the chance to double his f
pawns with . . . 1lLxf3 .

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

A very sharp reply, but I would be

intrigued to discover what Helgi had
in mind against the ECO recommen
dation of 14 iLf4! (taking c7 away
from the queen) 14 . . . 'iYaS 15 ctJd2!,
intending ctJb3 to harass the queen
some more. 1s . . . cxd4 16 ctJb3 'iYb6 1 7
a s i s not nice for Black and the reck
less lS . . . ctJc2 loses a piece to 16 ctJb3
'iYb4 17 ctJa2 'iYxa4 18 ctJc3 ! (18 'iYxc2
iLxe4 causes some problems) 18 . . . 'iYb4
19 Ra4 'iYb6 20 'iYxc2.
6 . . . 1Lg4 7 1Lxc4 e6 8 h 3 1L h 5 9 0-0

1 4 . . . exd 5 ! 1 5 e 5 d4 ! ? 1 6 exf6 1Lxf6

iLlb4 1 0 e2 !

1 7 1Lf4

A typical manoeuvre in queen's

pawn openings: the queen moves to
e2, supporting the e4 push, while the
rook is played to d 1 , supporting the
d4-pawn and discouraging . . . c6-cS due
to the opposition of the rook to the
black queen on d8 .
1 0 . . . 1Le7 1 1 J:::!. d 1 0-0 1 2 g4 1L g 6 1 3
e4 c5 ! ?

Question 5: What is going on?

A nswer: As compensation for the

An unusually active move at this

of the Smyslov variation, but
Helgi Gretarsson, a fanatic of this
variation, has an interesting idea in
mind. The more restrained 13 . . .':tJd7 is
considered in the next two games.

piece, Black has two pawns, a strong

centre and a tempo on the queen with
. . . Re8 . Unfortunately, I don't think
that this is quite enough; and this is
almost entirely due to the bad placing
of the bishop on f6, which takes away
a brilliant square for the black queen.
Perhaps Black could try 1 S . . . Re8 ! ? to
meet 16 exf6 (16 iLbs ctJc6) with ei
ther 16 . . . iLd6 or 16 . . . iLf8 , intending
to recapture on f6 with the queen,
though I would be the first to admit
that it all looks a bit speculative!

14 d 5 ! ?

1 7 . . . J:::!. e 8 1 8 f 1 a6

stage

65

Th e Sla v

To prevent LLlb5 .

8 h 3 iL h 5 9 0-0 ctJb4 1 0 e2 iLe7

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

1 9 .a.d2? !

White starts to go wrong around

here and drifts very quickly into a lost
position. 19 LLla2! LLld5 (19 . . . LLlc6 20
e l ! beginning to exchange pieces) 20
ilg3 , intending e l , would have
given White the better chances.

Black anticipates the threat of llJeS

and f2-f4, intending f4-f5 to trap the
bishop on g6.
1 4 ctJe5 ! ?

axb5 22 iL x b 5 '>&e4 23 iLxe8 '>&xf4

A very double-edged decision.

White allows his central pawns to be
doubled, but also frees the f-pawn to
advance.

24 .a.e 1

1 4 . . . ctJxe 5 1 5 dxe5 a5 1 6 f4 .l::!. a d8

1 9 . . . '>&d 7

20 g 2?

'>&c6

21

ctJb5

h6 25 iL b 5 iLe4 26 .a.xe4

xe4 2 7 iLc4 f4 28 b3 .)de8 29

1 7 iLe3 h 6 !

g 3 '>&xg 3+ 3 0 fxg 3 ctJc6 31 iLd5

ctJa5 3 2 b4 .a.d8 3 3 iLe4 cxb4 34
ctJe 1 b3 3 5 iLd3 iL g 5 36 tIb2 tIc8
3 7 f 1 g 6 3 8 e2 iLc 1 39 tIb 1 b2
40 ctJf3 ctJb3 0 - 1

The main line for Black i s consid

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

Game 36
K ra m n i k-Ivanchu k

Monte Carlo (blindfold) 1 996

Making an escape square for the

bishop.
1 8 g2 iL h 7 1 9 .a.xd8 .a.xd8 20 .a.d 1
g5!

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

dxc4 5 a4 ctJa6 6 e3 iLg4 7 iLxc4 e6
66

2 0 . . . a 6 i s considered i n the next

game.

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

2 1 :i::!. x d8+ xd8 22 d 2 gxf4 23

ixf4 iL b 6 24 iL b 5 ! !

Black must give perpetual check

due to the threat of mate on hS.
24 xh6 allows 24 . . . iVxeS, so
White sacrifices a bishop to keep the
queen boxed in on as .
24

. . .

cxb5

25 iLxh6 c 5

26 d7

Game 37
Kramni k-Short

Moscow {Intel Grand Prix} 1996

ig6 2 7 c8+ h7 28 iLg5

With his threat of f6 and iVhs

mate, White just seems to be winning,
but now it is Black's turn to sacrifice a
piece, this time to free his queen.

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

dxc4 5 a4 tLla6 6 e3 iLg4 7 iLxc4 e6
8 h 3 iLh5 9 0-0 tLl b4 1 0 e2 e7
1 1 :i::!. d 1 0-0 1 2 g4 iLg6 1 3 e4 tLld7
1 4 tLle5

28

. . .

tLl d 5 ! ! 29 xc5

Not 29 exdS when 29 . . iVb4!

launches a powerful counterattack.
.

29

. . .

tLlxc3 30 xc3 xa4 3 1 g3

xe4 3 2 iLf6 b4 3 3 c8 e 1 + 34
f4 f2+ 35 g 5 d2+ 36 h4
h6+ 37 g3 e3+ Yz - Yz

Kramnik obviously believes in this

continuation for White, but it seems a
little hasty to me. Since Black is
threatening little in the centre, a sen
sible move like 14 f4, taking c7
away from the black queen and
67

Th e Sla v

intending perhaps h3-h4-hS, makes

more appeal to me.

'iVxe 5 30 'iVb6 ctJd4 3 1 .1Ld 1 .1L g 8 !

1 4 . . . ctJxe5 1 5 dxe5 a5 1 6 f4 h 6
1 7 g2 J::r a d8 1 8 .1Le3 .1L h 7 1 9 J::r x d8
J::r x d8 20 t1.d 1 a 6 !

The new idea, activating Black's

queenside majority.

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

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

2 1 .1L b 3 h8 !

So that the bishop on h7 can reacti

vate itself by means of . . . iLgS , . . .f7-f6
and . . . iLf7! This position would not
be to everyone's taste, but Short wins
a mce game.
22 J::r d 2 b 5

23 axb5 cxb5 24 f5

ctJc6!

Yes it's free! White will not be able

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

Stressing the new weakness on eS.

25 J::r x d8+ x d 8 26 .i.f4 .i. c 5 2 7
.i.e3 'b6 28 .1Lxc5 xc 5 29 'iVf2
68

Kramnik's approach with CDeS is

rather impatient. A quieter method is
demonstrated in the following game,
the last of a match between France's
13-year-old star Etienne Bacrot and
ex-World Champion Vassily Smyslov,

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

whose name this variation bears. The

match score was a rather crushing 5- 1 ,
and one player was made t o look
vastly inferior in the endgame. But
not the player one might have ex
pected!

Game 38

f4-fs or h3-h4-hs with a clear advan

tage. Black must play . . . ctJd7 to pre
vent ctJeS, either before or after . . . b7bS, with a typicaI S . . . ctJa6 position.
1 2 . . J c 8 ? ! 1 3 .:iLb3 c 5 ?

This is excessively active from

Black at this early stage.
1 4 CL:l b 5 !

Bacrot-S myslov

Albert (sixth match game) 1 996

1 d4 d 5 2 c4 c6 3 CL:lc3 CL:lf6 4 CL:lf3
dxc4 5 a4 CL:la6 6 e3 .:iLg4 7 .:iLxc4 e6

8 0-0 .:iLe7 9 e2 CL:lb4 1 0 J::i. d 1 0-0

1 1 h3 .:iLh5 1 2 a 5 ! ?

This fine move threatens ctJxa7 and

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

We have already seen White's idea

in Novikov-Gretarsson. Our first
thought should be therefore to im
plement the plan of . . . b7-bS, followed
by the transfer of the queen to b7:
12 bs 13 b3 "Wic7 14 e4 (14 g4 g6
15 ctJeS! ? , intending a quick h3-h4-hS
trapping the bishop) 14 . . . "Wib7. It is
obvious that White is better prepared
for his opponent's plan than in the
above game. The rook covers d4 from
d 1 , while the queen on e2 both pro
tects e4 and attacks bS, making . . . c6-cS
more difficult to achieve. After 15 g4
itg6 16 ttJeS White intends either f2. . .

The ending is very nasty for Black,

since his knights have no outposts.
23 ga3 CL:le4 24 .:iLe 1 CL:ldf6 25 J::i. c 1
&i.d8 2 6 f 3 CL:ld6 2 7 .:tId 1 CL:lfe8 28
69

Th e Sla v

a4 gd 5

29 gxd 5 exd 5

30 gd3

37

f2

CiJxc7

38

gb3

CiJe6

39

CiJc7 31 b4 CiJ d b 5 3 2 c 5 f5 3 3

.a.xb7+ f8 40 .a.b8+ e7 41 .a.b7+

b6 gd6 34 e4 !

.a.c7 42 .a.xc7+ CiJxc7 43 c6 d4 44

b4 1 -0

This allows White to make a mas

sive stride forwards with his e-pawn as
34 . . . dxe4 loses a piece to 35 Mxd6.
34 . . . f7 3 5 e5 .a.c6 36 xc7 .a.c 1 +

70

A fine vlCtOry for the young

Frenchman, crowning an amazing
match result.

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

S u m m a ry

Theoretically, S . . . ctJa6 is doing well for Black and if a system has been played
by Smyslov, Ivanchuk and Short then it must have some merit! If you don't
mind playing slightly cramped positions, then it could be the system for you.
6 e3 is the most critical test.
1 d4 d 5 2 c4 c6 3 ctJf3 ctJf6 4 ctJc3 dxc4 5 a4 ctJa6
6 e4 (D)

6 e3 iLg4 7 iLxc4 e6 8 h3 iLh5 9 0-0 ctJb4 10 'Wie2 iLe7 1 1 Md1 0-0

12 g4 iLg6 13 e4 (D)
1 3 . . . cS game 35
13 . . . ctJd7 14 ctJeS ctJxe5 15 dxeS 'Wia5 16 f4 Mad8 17 iLe3 h6
18 g2 iLh7 19 Mxd8 Mxd8 20 Md1
20 . . . g5 game 36
20 . . . a6 game 37
12 as game 38
-

. . .

ilLg4 7 ilLxc4 (D)

7 iLxf3 - game 33
7 . . . e6 game 34
. . .

6 e4

1 3 e4

7 ilLxc4

71

CHA PTER SIX

The Bronstein Variation

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

dxc4 5 a4 jLg4

Like Smyslov's 5 .. .'Ja6, 5 . . . iLg4 is a

variation that has been around for a
long time without ever gaining wide
spread popularity.
Question 1 : What is the point of
5 . . . iLg4?
A nswer: 5 . . . iLg4 looks to delay e3e4 by putting pressure on the d4pawn. White should avoid 6 e4 as af
ter 6 . . . e6 7 iLxc4 iLb4 (threatening
. . . ctJxe4) , he has problems holding his
centre. Therefore 6 ctJe5, gaining a
tempo on the bishop, is almost always
played. Now after 6 . . . iLh5 White still
cannot play 7 e4, as this would allow
7 . iLxd l !
Question 2 : What plans does White
have?
A nswer: Since White can take the
c4-pawn at his leisure, there is no need
for him to hurry with ctJxc4. His two
most dangerous plans both aim to
exploit the slightly precarious posi
tion of the bishop on h5:
a) 7 f3 , which blocks the h5-dl di
agonal and threatens to achieve e2-e4,
while supporting g2-g4 iLg6, h2-h4,
intending to trap the bishop.
b) 7 h3, intending to gain space on

the kingside with g2-g4 and develop

the bishop to g2. This variation is so
complicated, however, that a whole
book would be needed to explain its
ramifications! I will do my best, but
I'm afraid you'll only get a brief
taster!
For the less savage, there is also the
quiet 7 g3 , planning a fianchetto.

Game 39
Kram nik-Damljanovic

Moscow Olympiad 1994

1 d4 d 5 2 tLlf3 tLlf6 3 c4 c6 4 tLlc3
dxc4 5 a4 jLg4 6 tLle5 jLh5 7 f3
tLlfd7

. .

72

Question 3: This is a strange

looking move. What does it do?

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

A nswer: Black must challenge the

knight on eS or his bishop will be

caught by g2-g4 and h2-h4. Now after
8 g4, 8 . . . g6 9 h4 4JxeS ! 10 dxe5
'i'xdl+ 1 1 xdl hs is fine for Black.
However, another point is to give
Black an unexpected opportunity to
continue his pressure against the d4pawn.

knight's natural square, while his

bishop is biting granite on hs. 9 4J e4
seeks to exploit this temporary confu
sion by threatening a devastating
check on d6, and also shields the
knight on c4 along the fourth rank; so
10 dxeS is now a threat as . . . iVh4+ will
no longer achieve anything.
9 . . . SL b4+

10

SLd2 Wife7

11

SLxb4

8 ttJxc4 e 5 !

Wifxb4+ 1 2 Wifd2 Wifxd2+

And this i s it! The knight's move

from f6 has freed the h4-d8 diagonal
for the queen, giving Black the oppor
tunity to exploit the slight weakening
on the e l-h4 diagonal created by 7 3 .

Forced, as 12 . . .iVxc4 loses to 1 3

4J d6+.

9 ttJe4

1 3 xd 2 exd4 1 4 iLled6+ e 7

The alternative, the enterpnslllg

14 . . . d8, is considered in the next
game.
1 5 iLlf5+ f6 1 6 iLlxd4 J:;!d8

The old move and a very sensible

one. White cannot play 9 dxeS as
9 ... iVh4+! wins the knight on c4,
while 9 4JxeS 4Jxe5 10 dxeS 4J d7 1 1
f4 JLb4 is a very risky pawn grab.
Black has a substantial lead in devel
opment and will follow up with
. . . 'i'e7 and . . .f7-f6 or . . . g7-gS, opening
further lines. 9 g3 is considered in
Games 4 1 and 42.
Question 4: What does 9 4J e4 do?
Answer: Black's pieces are a little
strange at the moment: his king's
knight is on d7, which is the queen's

We have reached an ending, but the

tactical complications continue for
some while yet.
1 7 c3 iLlc5 1 8 e4 ! B.xd4 1 9 xd4
iLlb3+ 20 c3 iLlxa 1

21 .iLe2 e 7

22 J::i. x a 1 iLl d 7 23 b4 !

Question 5: Isn't this just a boring,

equal ending?
A nswer: Unfortunately for Black,
no. White does enjoy a definite edge
here, and it all boils down to that
wayward bishop on hs. First, Black is
going to have to spend a tempo with
73

Th e Sla v

. . . f7-f6 to bring it back into play; and

second, if it was still on cS , Black
wouldn't have such an annoying
weakness on b7! The white knight is
excellently placed on c4, as it can at
tack b7 via d6 (with the help of a rook
on dl) or as .
Question 6: You mean Black is
lost?!
A nswer: No, not at all. He only has
one real weakness, so he should be
able to defend, but it isn't really that
much fun.

stopping the knight from activating

via eS . Black obviously felt very un
comfortable round here, since he
starts just moving his knight around
for no reason.

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

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

23 . . . f6 2 4 gd 1 iLl b 6

Kramnik suggests that 24 ... Mbs

more solid.

IS

2 5 iLl a 5 !

Exchanging knights would greatly

simplify Black's defensive task. Now
White forces unpleasant weaknesses
in the black queenside.
25 . . . iLlxa4+ 26 b3 iLl b 6 27 iLlxb7

Threatening a breakthrough with

e4-eS.

f7+ 28 c3

3 7 . . . iLla8? 38 3L a 4 J:!.c 7 39 lIc5 iLlb6

The trade of the a4-pawn for the

b7-pawn has been profitable for
White, as now he has two targets: a7
and c6.

1 -0

And Black lost on time in this

hopeless position: he is just going to
lose his c-pawn.

28 . . . gb8 29 a6 e8 30 J:!.a 1 iLld7

31 f4 !

Gaining space on the kingside and

74

Let us now take a look at 14 . . . d8

instead of 14 . . . e7.

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

Game 40
Schandorff-Hellsten

Copenhagen 1996
1 d4 d 5 2 c4 c6 3 ctJf3 ctJf6 4 ctJc3
dxc4 5 a4 g4 6 ctJe5 h 5 7 f3
IiJfd 7 8 ctJxc4 e 5 9 ctJe4 b4+ 1 0
d2 e7 1 1 x b4 x b4+ 1 2 d 2
xd2+ 1 3 '.t>xd2 exd4

1 4 ctJed6+

d8 ! ?
1 9 '.t>c3 he8 ! 20 ctJac4 .!;l;ab8 2 1
d 3 ? ctJxd3 2 2 '.t>xd3 ctJc5+ 2 3 '.t>e2
f5! 24 '.t>f 1 f4 2 5 ctJd 1 ctJb3 0 - 1

An amazingly quick defeat for

White!

Game 41
Epishin-Pomes

Manresa 1995
1 d4 d 5 2 c4 c6 3 ctJf3 ctJf6 4 ctJc3

An interesting idea: in the ending

above, Black would have loved to
have his king on c7 protecting the
weak b-pawn! One drawback is that
White can take on b7 with check, but
at least the position is unbalanced,
unlike the safe edge which White eas
ily obtained in the game Kramnik
Damljanovic.

dxc4 5 a4 g4 6 ctJe5 h 5 7 f3
ctJfd 7 8 ctJxc4 e 5 9 g 3

1 5 ctJxb7+

15 g4!? JlLg6 16 f4 f6 17 f5 JlLeS 1S

tLlxb7+ c/c7 19 tbba5 does not really
improve the white position, but in
stead 15 h4!? (threatening g2-g4 and
h4-h5) 1 5 . . .f6 16 g4 JlLeS 17 tbf5!?, at
tacking g7 and d4, is an interesting
attempt.

This has been the most popular

choice recently. White prevents
. . . h4+ and threatens d4xe5 .

1 5 . . '.t> c 7

9 . . . b4

16

ctJba5

dxe3+ 1 8 ctJxe3 ctJ b4

ctJa6

17

e4

For 9 . . .f6 see Game 43.

75

Th e Sla v

1 0 dxe5 0-0 1 1 ilL h 3 ! ?

White now threatens e5-e6.

1 1 .. .'e7 1 2 f4

is ttJh6+ gxh6 19 xe6 f7 20 f5 h5

21 0-0 would have led to a murky poslttOn.
1 7 ilL e 3

17 'iVe4!, intending ttJh6+, was bet

ter.
1 7 . . . LiJa6 1 8 0-0 ilLf7 1 9 LiJe4 It>h8
20 gad 1 c7 21

b3 ilLf8 22 b2

LiJec5 23 LiJxf6 ilLxb3 24 !:!.xd8 !:!.xd8

25 LiJg4 ilLd5 26 LiJd4 LiJxa4? ! 27
a 1 b 5 ?

12 f4 is considered in the next

game.
1 2 . . . .l:!.d8 1 3 c2 f6

Absolutely necessary in order for

Black to free himself.
1 4 e6

14 exf6 ttJxf6 gives Black some

counterplay for the pawn.
1 4 . . . LiJc5 1 5 LiJe3

Or 15 f5 Md4 16 ttJe3 ttJba6 17 0-0

MadS with counterplay (Epishin) .
1 5 . . . LiJxe6 1 6 LiJf5

Black is determined to ignore his

kingside defences. The punishment is
swift.
28 f5 ilL b4 29 LiJe6 ! ilLxe6 3 0 fxe6
ilLc3 3 1 b 1 e7 32 LiJ h 6 !

The pawn capture 32 ... gxh6 loses t o

33 Mf7.
32 . . . .l:!.f8 33 LiJf7+ It>g8 34 ilLf5 ':!:!'xf7
35 ilLxh7+ It>h8 36 ':!:!'xf7 e8 37
f5 1 -0

Game 42
Parker-Hellsten

Copenhagen 1996
1 d4 d 5 2 c4 c6 3 LiJf3 LiJf6 4 LiJc3
1 6 . . . d 7 ? !

dxc4 5 a4 ilLg4 6 LiJe5 ilLh5 7 f3

A mistake according t o Epishin.

The alternative 1 6 . . :eS 1 7 'iVe4 ttJa6

LiJfd 7 8 LiJxc4 e 5 9 g 3 ilL b4 1 0 dxe5

76

0-0 1 1 ilLh3 e7 1 2 ilLf4

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

1 7 . . . Ile8

17 .. .lbd7 fails to 18 g5 ! 'MUxe5

(18 .. .f6 19 exf6 gxf6 20 'MUxg6+) 19
xd8 'MUxg3+ 20 f1 Mxd8 21 'be4,
when Black has insufficient compen
sation for the exchange.
18

0-0 c5+

19

<;t>g2 xc3

20

xc3 xc3 21 bxc3 ctJd7

A very interesting idea. 12 . . . g5, to

chase the bishop away, is simply met
by 13 d2 (13 xd7 'bxd7 14 xg5
'iVxg5 15 'MUxd7 b5 16 f4 is very messy)
13 . . . 'bxe5 14 'bxe5 'MUxe5 15 'be4
Jtxd2+ 16 'MUxd2 g4 17 g2 with a
better position for White.
1 2 . . Rd8 1 3 c2 g6 1 4 f5
.

14 e4 is very risky but not easy to

refute. For example, 14 . . . 'MUc5 15 'bd6
fLlxe5 16 'bxb7 (forking queen and
rook) 16 . . . 'bxf3+ 17 f1 'MUc4+ 1 8
'iVe2 seems fine for White, while
14 . . . b5 15 axb5 cxb5 16 'be3 'bxe5 17
0-0 is also difficult to judge.

Black just manages to hold the en

suing endgame, but the whole line
seems extremely uncomfortable for
him.

1 4 . . . ctJb6 1 5 xg6 hxg6 1 6 ctJxb6

ctJ d 7 30 Rbxc4 ctJxe5 3 1 xe5 Ilxe5

axb6 1 7 h4

22 gfb 1 ga6 23 a 5 gea8 24 gd 1

ctJf8 25 gab 1

b5 26 c4 bxc4 27

gxb7 I!xa 5 28 Il b4 R d 5 29 R c 1
3 2 e4 ga2+ 3 3 <;t>h3 f5 34 gxc6
fxe4 3 5 fxe4 Ilxe4 36 1::!. x g6 <;t>h7 37
gg 5 P.e7 38 Id.h 5+ '/z - '/z

Game 43
Dautov-Ni kolic

Ter Apel 1994

1 d4 d 5 2 c4 c6 3 ctJf3 ctJf6 4 ctJc3
dxc4 5 a4 g4 6 ctJe5 h 5 7 f3
ctJfd 7 8 ctJxc4 e 5 9 g 3 f6

Preventing . . . g7-g5 .

A very solid continuation, just pro

tecting the e5-pawn.
77

Th e Sla v

<;t>e5 J::l: e 8+ 41 f5 ge2 42 b4 e8

43 gh4 J::l: f 2+ 44 e6 ge2+ 45 d6
a5 46 bxa5 <;t>f7 47 gh7+ <;t>xf6 48
gxb7 ne4 49 xc6 gxa4 50 gb5
1 -0

This is all very sensible. However,

White has another rather crazy idea.
Game 44
Shirov- N i kolic

Wijk aa n Zee 1993

1 0 dxe5 ctJxe5 1 1 \'i{{x d8+ <;t>xd8 1 2
ctJxe5 fxe5 1 3 iL g 5+ <;t> c 7 1 4 0-0-0
ctJd7 1 5 iL h 3 iLe8 1 6 iLe3 !

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

dxc4 5 a4 iLg4 6 ctJe5 iL h 5 7 f3
ctJfd 7 8 ctJxc4 e 5 9 e4

And this is it! This very natural

move was completely ignored until
recently.
9 . . . \'i{{ h 4+ 1 0 g 3

The wacky 10 e2 is considered in

the next game.
1 0 . . : 'IIf 6 !

This very nice move prepares f3-f4,

breaking the position open.

Having softened up the kingside

with . . . "iVh4+, the queen retreats to f6,
where it helps attack f3 with the
bishop on h5 , and d4 with the pawn
on e5.
1 1 dxe 5 ! xf3 1 2 ctJd6+ ! <;t> d 8

1 6 . . . ctJc5 1 7 f4 iLd7 1 8 iLxd7 ctJxd7

1 9 f5

19 ctJe4 Me8 20 f5 was even more

accurate. The ending is basically very
pleasant for White.
1 9 . . . iLe7 20 g4 h 6 21 ctJe4 ctJf6 22
ctJxf6 iLxf6 2 3 h4 iLe 7 24 iL f2 gad8
25 iLg3 iLf6 26 <;t>c2 .!cl.xd 1 27 J::l: x d 1
gg8 28 g 5 h x g 5 29 hxg 5 iLxg 5 30
iLxe5+ <;t>c8 31 gg 1 ge8 3 2 iLxg7
l:;l;.xe2+ 3 3 d3 l:;l;.d2+ 34 <;t>c3 iLe3
3 5 gg3 iLf4 3 6 gf3 gg2 3 7 gxf4
gxg7 38 f6 g g 8 39 <;t>d4 <;t>d7 40
78

Amazingly, 12 ... xd6 loses to 1 3

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

'iVxd6 ixh l 1 4 ,g5 (threatening

'iVe7+ mate) 14 . . .f6 15 exf6 gxf6 16
'iVe6+ WdS (16 . . . WfS 1 7 ilh6+ mate)
17 ilxf6+ lLlxf6 i s ixf6+ Wc7 19
'iVe5+! followed by ixh5 or ixhs
with a crushing position, as pointed
out by Ivan Sokolov.

a5 25 h4 b6 26 c2? !

1 3 xf3 xf3 1 4 ctJxf7+ e8 1 5

e6 ! ?

The first new move o f the game! In

I.Sokolov-Lautier, Belgrade 199 1 ,
White had played 1 5 CLlxhS but
15 . . . CLlxe5 ! (Shirov) 16 ilf4 CLlSd7 is
nice for Black, as after 17 fIg l ,
1 7 . . . ilc5 ! makes sure the rook does
not escape.
1 5 . . . ctJ c 5 1 6 c 4 x h 1

1 7 ctJxh8

JLxe4 ! 1 8 g 5

Either 26 Wd2 or 26 g5 would have

kept a slight advantage for White ac
cording to Shirov.
26 . . . ctJca6 27 ctJe4 a7 28 ctJg 5+
d5 29 ctJhf7 Ide7 3 0 gh8 ctJd7 3 1
xh 7 ctJe5 3 2 ctJxe5 M,xh7 3 3 ctJxh 7
xe5 34 ctJf8 I!>f4 3 5 ctJxg6+ xg4
36 ctJe5+ xh4 37 ctJxc6 g5 Y2 - Y2

Game 45
Nesterov-Imanaliev

Bishkek Zonal 1 993

1 d4 d 5 2 c4 c 6 3 ctJf3 ctJf6 4 ctJc3
dxc4 5 a4 g4 6 ctJe5 h 5 7 f3
ctJfd 7 8 ctJxc4 e 5 9 e4 h4+ 1 0
e 2 ! ?

i s b4 seems to win a piece, but

ls . . . ild3 ! (Shirov) 19 e7 il.xc4
(19 . . . ilxe7 20 ilf7+ followed by bxc5
keeps the fun going) 20 exfSi + xfS
21 bxc5 WgS favours Black.
1 8 . . . f5

1 9 0-0-0 e7

20 xe7

ri;xe7 21 f 1 g 6

A slight error according t o Shirov.

21 . . . re6 22 ilxe6 Wxe6 23 fIfS as is
suggested instead, but this also seems
quite nice for White.
22 g4 xe6 23 xe6 xe6 24 Idf8
79

Th e Sla v

A magnificent idea, the tactical just

ification of which lies in my all-time
favourite opening trap!
1 0 . . . exd4

Tempting but not the best. Black

should react more calmly with
10 . . . jLb4, intending . . . 0-0.

knight on c4. The move is extremely

aggressive: White will expand on the
kingside with g2-g4 and h2-h4 and try
to win Black's light-squared bishop,
while rapid queenside castling is also
on the agenda.

Game 46

1 1 xd4 3L c 5 1 2 lZJd6+! f8

I . Sokolov- Hellsten

12 . . . c,t>e7 loses to 13 ctJfS+ while

12 . . . c,t>ds is met by 13 ctJxb7+.

Malmo 1 995

1 3 xg 7+ ! !
1 d4 d5 2 c4 c6 3 lZJc3 lZJf6 4 lZJf3
dxc4 5 a4 3Lg4 6 lZJe5 3L h 5 7 f3
lZJfd 7 8 lZJxc4 e5 9 3Le3 3Lg6

Question 9: This looks odd as well!

A nswer: This is a typical idea in this

line. Black realises that the bishop is

doing nothing on hS, where it merely
bites against the pawn on f3 . There
fore, he moves it to a more active di
agonal, delaying the decision of which
piece to put on b4: the bishop on f8
or the knight on bS, via a6.
1 3 . . . xg7 1 4 1ZJf5+

1 0 h4!

Regaining the sacrificed queen and

winning a pawn, with a good position
to boot!
1 4 . . . f6 1 5 lZJxh4 lZJa6 1 6 3Lh6 lZJe5
1 7 g4 lZJxg4 1 8 fxg4 3Lxg4+ 1 9 d2
ad8+ 20 c 2 lZJ b4+ 21 b3 3Le6+
2 2 3Lc4 3Lxc4+ 23 xc4 .l:!.d4+ 24
xc 5 lZJc2 2 5 3Lg7+ 1 -0

Finally, White can try the develop

ing move 9 jLe3 .
Question 8: This looks very strange,
doesn't it?
A nswer: With 9 jLe3 , White de
fends the d4-pawn and prepares to
meet 9 . . :h4+ with the simple 10
jLf2 . However, 1 0 dxeS is still not a
threat due to 10 . . :iVh4+, winning the
80

This not only aims to harass the

bishop on g6 with hS , but also threat
ens d4xeS by removing Black's re
source of . . . iVh4+.
1 0 . . . 3Le7 1 1 h 5 ! 3Lf5

1 1 . . .h4+ is met simply by 12 jLf2

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

xf2+ 1 3 'it'xf2 i,f5 14 liJd6+, win

nmg a pIece.
1 2 dxe5 0-0 1 3 h 6 ! lLla6 1 4 hxg7
l:!.e8 1 5 g4 XL g 6 1 6 f4 !

Game 47

K rasen kov-Sapis

Polish

Championship 1995

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

dxc4 5 a4 XLg4 6 lLle5 XLh5 7 f3!
lLlfd 7 8 lLlxc4 e 5 9 XLe3 XLb4

A more natural developing move

than 9 . . . i,g6.
1 0 g4 XLg6 1 1 dxe 5 !

This is possible now, since

l 1 . . .h4+ no longer hits the knight
on c4!
1 1 . . . 0 -0 ! ?

Quite amazing! White is only mov

ing pawns, but Black just seems to be
helpless!
1 6 . . . lLlb4

17

Uc 1

lLld5

18

XLg2

Ibxe3 1 9 lLlxe3 XL h4+ 20 f 1 XLg3

21 f5 b6 2 2 fxg6 hxg6 23 d2
xe 5 24 lLled 5 1 -0

24 . . . cxd5 loses to either 25 Mh8+

\t>xg7 26 h6+ 'it'f6 27 liJxd5+, fork
ing king and queen, or instead 25
tt'lxd5 d4 26 Mh8+ 'it'xg7 27 h6+
mate! A game of astonishing ferocity,
even by Ivan Sokolov's remarkable
standards!

This i s a new idea. The alternative

1 1 . . . e7+ was played in the original
game Granda Zuniga-Nikolic, Biel
Interzonal 1993, and now 12 liJd6+
i,xd6 13 xd6 would have given
White a safe edge due to his two bish
ops and the weakness of Black's dark
squares. Instead in the game, White
went for the crazy complications of
12 f4 h4+ 13 i,f2 xg4 14 b3
liJa6 15 liJd6+ i,xd6 16 xb7 liJb4 17
xa8+ b8 which is just unfathom
able.
1 2 h4 h6 1 3 h 5 XL h 7 1 4 g 5 ! ? hxg 5
1 5 h6 g 6

81

Th e Sla v

1 9 axb5 cxb5 20 'J}f/e4 !

bxc4 2 1

'J}f/xaS ttJxe5 22 'J}f/xa 7

Question 1 0: Why is White playing

so aggressively? He was a pawn up
with more space; why did he sacrifice
his g-pawn?
A nswer: White has played some
very strange moves in the opening.
First, he put his bishop on e3 in front
of the e-pawn, blocking in his light
squared bishop . Then, instead of de
veloping his pieces, he advanced his
kingside pawns in order to chase
Black's light-squared bishop. White
has won a pawn, but his pawn ad
vances have left many weak squares in
his position. For example, if Black
could get a rook to dS and then play
. . . ctJc5, aiming for the weak b3-square,
then White's position would become
critical. White has raised the stakes
with his risky opening play - he must
continue in the same aggressive man
ner or Black will develop and exploit
White's weaknesses.
1 6 Rg 1 'J}f/e7 1 7 Rxg 5 :i:!.dS 1 S 'J}f/c2
b5?

A tactical miscalculation. Chekhov

recommends lS . . . 'i'e6! 19 'i'e4 ctJa6!
(intending . . . ctJdc5) 20 'i'h4 b5!, when
I think that Black has good counter
chances.
82

White is winning now due to his

large material advantage.
22 . . . ttJbd7 23 '>jj'd 4 f6 24 Rg3 ctJc5
25 .l'::!. a S ! l:baS 26 '>jj' d 5+ ttJ e 6 27
'J}f/xaS+ ttJfS 2S Wf2 iLd6 29 ttJe4
iLb4 30 iLd4 WhS 3 1 f4 ttJed7 3 2
Re3 '>jj'f 7 33 ttJ g 5 '>jj' g S 3 4 ReS 1 -0

And now things get even more

complicated. Let us take a look at 7
h3 .

Game 48
Gelfand - N i kolic

Manila Interzonal 1990

1 d4 d5 2 ttJf3 ttJf6 3 c4 c6 4 ttJc3
dxc4 5 a4 iLg4 6 ttJe5 iL h 5 7 h3
ttJa6 S g4 iLg6 9 iLg2

I find it hard to recommend this

line to players of either colour, unless
they have six months in which to ana
lyse the mind-boggling complications!
9 e3 is discussed in the next game.
9 . . . ttJb4 1 0 0-0 iLc2 1 1 'J}f/d2 iLb3

Question 7: Why is Black doing

this?
A nswer: If Black were to play nor-

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

mally, then White's space advantage

would guarantee him a substantial
advantage. Black therefore keeps the
pawn and challenges White to make
something of his lead in development.

iVxg5, which looks like good com

pensatIon to me.

1 8 e4 d8 1 9 f4 e7 20 e3 0-0
21 e2 c 5 22 dxc5 xc 5 23 d 2
1 2 a5

Black does not really have enough

for the exchange.

Threatening a5-a6 to break up the

black queenside. 12 tL'le4, attacking the
knight on b4, is the other major con
tinuation, aiming for the attractive
trap 12 . . . tL'lxe4 13 iVxb4 tL'ld6 14
'i'xb7! ! , when 14 . . . tL'lxb7 loses to 15
coxc6+.
1 2 . . . a6

Stopping a5-a6, but in fact this may

not be so dangerous: 12 . . . e6 13 a6 iVc7
14 axb7 iVxb7 15 g5 tL'lfd5 16 e4 tL'lb6
17 d5 looks impressive, but 17 . . . :l':rd8
18 tL'lxc6 tL'lxc6 19 dxc6 iVxc6 20 e5
4Jd5 21 :l':rxa7 c5, with . . . 0-0 to fol
low, was nice for Black in the game
Moreno-Rogers, Manila Olympiad
1992.

2 3 . . . CLl d 5 2 4 h 1 d 7 25 l::i. a c 1 a4
26 f3 c3 27 bxc3 xa 5 28 f5 c4
29 fxe6 ! liJe7 30 xh7+ xh 7 3 1
e4+ liJg6 3 2 xc4 fxe6 3 3 g4
l::i. x f 1 + 34 .ti.xf 1 iVb5 35 c4 c6+ 36

1 3 liJa4 e6 1 4 g 5 liJ d 7 1 5 liJxd 7

g 1 g8 37 f3 d 7 3 8 l::i. f 2 a 5 3 9

xd 7 1 6 tLlb6 d8 1 7 tLlxa8

e4 tLle7 4 0 xa5 tLlf5 4 1 d 2 1 -0

see

follo wing diagram

1 7 . . . xa 8 ?

Ftacnik suggests 1 7 ... tL'lc2 1 8 :l':rb 1

'i'xa8 19 e3 b4 20 iVe2 iVd8 2 1 d2

If White wishes to duck the critical

lines that we saw in the previous
game, he can play 9 e3 instead of 9
g2.
83

Th e Sla v

Came 49

iLlc2!? 15 xc2 xd4+ 16 g2 xc4

and 17 f5 (Rogers) .

Klarenbeek-Rogers

Dutch Team Championship 1 996

1 d4 e6 2 e4 d 5 3 tZlf3 tZlf6 4 tZle3
dxe4 5 a4 ]Lg4 6 tZle5 ]Lh5 7 h 3
tZl a 6 8 g4 ]L g 6 9 e 3

A quieter attempt, making sure that

White regains the pawn.
see

follo wing diagram

9 . . . tZlb4 1 0 ]Lxe4 e6

1O . . . 'Llc2+ loses to 11 xc2 xc2

12 xf7 mate, but 1o . . . 'Lld7 1 1 iLlxg6
hxg6 12 f3, intending f1-g2, is
more normal. White has a small ad
vantage here due to his slight space
advantage and bishop pair.

14

e2

tZlxb4? !

xd 7
]Lxb4

15

tZla2

e4

17

]Ld2

0-0

16
18

]Lxb4? !

White is just playing for a draw,

but he is doing this badly. The text
makes the a-pawn very weak.
1 8 . . . axb4 1 9 b3 Ra 5 ! 20 Rad 1 b5
21 axb5 exb5 22 ]Ld3 ]L b 7 !

Black is not going to exchange this

bishop, while the as-h 1 diagonal is so
tempting!
23 \t>h2 d 5 24 %:!.b 1 Re8 2 5 Rfd 1
g5!

An unexpected and really strong

move.
26 :!:!.f 1
1 1 0-0 a5 1 2 f4 tZld7 1 3 tZlxd 7

Rogers recommends instead 13 e4

h4 14 g2 with a slight advantage
for White, but with the threat of 15
iLlf3 , intending f4-f5, this looks horri
ble for Black.
1 3 . . . ]L e 2 !

We have seen this before. Black

makes sure that the bishop does not
get shut in behind the e4-pawn.
13 . . . xd7 is strongly met by 14 e4!
84

26 fxg5 loses to 26 . . . d6 (Rogers),

as 27 g1 g3+ is terminal.
26 . . . Re3 27 Rb2 gxf4 28 Rxf4 Ra1
29 \t>g3 Rg 1 + 30 \t>h4 h 6 0 - 1

And White lost o n time i n this

hopeless position.

Came 50
Leitao-Beliavsky

Erevan Olympiad 1 996

1 d4 d 5 2 e4 e6 3 tZlf3 tZlf6 4 ctJc3

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

dxe4 5 a4 g4 6 tLle5 h 5 7 g 3

The quiet option. White ignores

the bishop on h5 and just develops
normally.
7 . . . e6 8 g 2 b4 9 tLlxe4 tLl d 5

Attacking c3 .
1 0 'i'b3 0-0 1 1 d2

11 0-0 fails to 1 1 . . .itxc3 12 bxc3

Jtxe2, winning a pawn.
1 1 . . a 5 1 2 e4 tLl b 6 !
.

An important and typical manoeu

vre. Black exchanges a pair of knights,
relieving his slightly cramped position
while attacking d4.

'i'd6 22 h2 .!:!.ae8 23 .l:!.ee 1 g6 24

1 3 tLlxb6

'!:!'e2 .l:!.e7 25 tLlb5 'i'e5 26 f3 g d 7

This rather helps Black. Beliavsky

suggests 1 3 ite3 instead.

2 7 g2 h8 28 tLle3 .l:!. f d 8 29 gd 1
f6 30 ged 2 f7 3 1 .!:!.d3 h 5 32 'i'b5

1 3 . . . 'i'x b6 1 4 e3 e 5 ! 1 5 d5 tLl d 7

e4 3 3 'i'xe4 h4 34 tLle2 h 7 3 5 gd4

1 6 0-0 tLle5 1 7 h 3 tLlf3+ 1 8 h 1

ge7 36 'i'd3 '!:!'de8 3 7 gxh4 '!:!'e2 3 8

itJd4

d6 '!:!' x b 2 39 d 7 .l:!.d8 40 h 1 h 5 4 1

Clearly Black has now taken over

the initiative. The bishop on h5 has
suddenly become a major player, sup
porting the incursion of the black
knight into the vulnerable kingside
light squares.

1 9 'i'e4 tLle2 20 '!:!'ae 1 tLlxe3 2 1 fxe3

xh 5 'i'xh 5 42 e5+ f5 43 tLlg 1 e 5

4 4 .!:!. g 4 .l:!. x d 7 45 'i'xd 7 fxg4 4 6
'i' d 3 + h6 47 'i'e3 g 3 4 8 'i'xb2
'i'xd 1 49 g2 'i'xa4 50 'i'xb 7 'i'e2+
51 xg3 xe3 0 - 1

A fine game b y Beliavsky.

85

Th e Sla v

S u m m a ry

I cannot really recommend the 5 . . . itg4 line for Black, not because it is a par
ticularly bad line, but simply because unless you have loads of time for de
tailed analysis, you won't be able to feel comfortable playing it. There are
many theoretical problems to solve: 7 f3 ttJfd7 8 ttJxc4 e5 9 ttJe4 gives White a
safe endgame edge, while 9 g3 is also dangerous. Even the crazy 9 e4 and 9
ite3 pose difficult problems! By contrast 7 h3 gives Black too many counter
chances, while 7 g3 is a little tame.
1 d4 d 5 2 c4 c6 3 ctJf3 ctJf6 4 ctJc3 dxc4 5 a4 g4 6 ctJe5 h 5 7 f3

7 h3 ttJa6 8 g4 g6 (D)
9 itg2 - game 48
9 e3 game 49
7 g3 - game 50
-

7 . . . ctJfd 7 8 ctJxc4 e 5 (D) 9 ctJe4

9 g3
9 . . . b4 10 dxe5 0-0 1 1 ith3 Wlic7
12 f4 game 41
12 itf4 - game 42
9 .. .f6 game 43
9 e4 Wlih4 +
10 g3 - game 44
10 e2 game 45
9 ite3
9 . . . itg6 game 46
9 job4 - game 47
-

..

9 . . . b4+ 1 0 d2 Wii e 7 1 1 xb4 Wiix b4+ 1 2 Wii d 2 Wiix d2 1 3 xd2 exd4

1 4 ctJed6+ (D)

14 . . . e7 - game 39
14 . . . d8 game 40
-

8
86

. . .

g6

. . .

e5

14 ctJed6+

CHA PTER SEVEN

The 4

a6 Slav :

White plays 5 e3

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

a6

In the next two chapters we shall

look at the move-order 1 d4 d5 2 c4 c6
3 ctJf3 ctJf6 4 ctJc3 a6.
Question 1 : What is the point of
4 .. a6?
Answer: 4 . . . a6 allows the bishop to
develop outside the pawn chain to g4
or f5, since after . . . iH5 , 'i"b3 attacking
b7 can be met by . . . b7-b5, advancing
the b7-pawn to a safe square, or
... J.:!a7!, an ugly looking but brilliant
thought of Julian Hodgson's.
Question 2: Sounds great. Any
drawbacks?
A nswer: You had to ask. Black is
placing a lot of pawns on light
squares , so he can often suffer from
weak dark squares.
.

Game 5 1
O i l -Anand

Biel Interzonal 1993

1 d4 d 5 2 c4 c6 3 lL:lf3 ttJf6 4 lL:lc3
a6 5 e3

White protects the c4-pawn and

prepares to develop his kingside.
White's numerous alternatives here
are discussed in the next chapter.

5 . . . b5

Black would like to develop his

bishop outside the pawn chain by
playing 5 . . . g4, for example, but 6
'i"b3 is strong, as 6 . . . b5 7 cxd5 cxd5 8
a4! breaks up the black queenside.
Black must be able to meet a2-a4 with
. . .b5-b4; after . . . b5xa4, his a-pawn be
comes very weak. In fact Black could
play an interesting tactical idea here:
after 8 . . . xf3 9 gxf3 he can try
9 . . . b4!? , so that after 10 'i"xb4, 1 0 . . . e5!
attacks the white queen. 1 1 'i"b3 exd4
1 2 exd4 ctJc6 13 e3 b4 gives com
pensation for the pawn due to
White's weakened pawn structure,
but 1 1 'i"b7! gains a tempo by attack
ing the rook on a8 : 1 1 . . .ctJbd7 1 2 dxe5
ctJxe5 13 xa6, when Black has
87

Th e Sla v

enormous problems on the light

squares. I'm sure that Julian Hodgson
would suggest 6 . . . Ma7(!) , but after 7
ctJe5 (threatening 8 ctJxg4 ctJxg4 9
cxd5, winning a pawn) 7 . . . e6 (as usual
in the Slav Black does not mind
swapping off his light-squared bishop
for White's knight) 8 f3 ! 11Lh5 9 g4
11Lg6 10 h4! , White's threat of h4-h5,
trapping the bishop, forces Black to
play the disastrous 1 O ... h6, when 1 1
ctJxg6 fxg6 1 2 c2 is just winning for
White.
The inclusion of 5 . . . b5 6 b3 takes
the b3-square away from the white
queen, allowing Black to develop his
light-squared bishop in greater com
fort.

For 8 gxf3, see the next game.

8

. . .

e6

This was a novelty at the time of

the game, as black players had been
experimenting with the violent 8 ... eS,
to exploit the absence of the queen
from the queenside and the slight
weakness of the knight on c3 (it is no
longer protected by a pawn on b2) . I
prefer Anand's simple move, which
carries the same threats but without
the risk.
9 :iL.d2 :iL.b4 1 0 d 1

6 b3

The exchange 6 cxd5 is considered

in Game 55.
6

. . .

:iL.g4 7 h 3

A very natural reaction, putting the

question to the bishop. 7 11Le2 is dealt
with in Games 53 and 54.
7

. . .

:iL.xf3 !

7 . . . 11Lh5 8 g4 11Lg6 9 ctJe5 , intending

h2-h4, is rather awkward, as we have
seen.
8 xf3
88

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

c2, but this move i s a little meek. A
few years ago, I played 10 11Ld3 aS
1 1 Me l ! ? 11Lxc3 12 MXc3 xa2 13 d1
against Jon Levitt, sacrificing a pawn
in order to gain the advantage of two
bishops against two knights. In fact I
won a nice game after 13 . . . 0-0 14 0-0
a3 15 a1 ! xa1 16 Mxa1 Ma7!? 17
cxb5 cxb5 (hoping for 1 8 11Lxb5 ctJe4,
swapping off one of my bishops) 18
Mc2 ! , allowing my dark-squared
bishop to activate itself via b4. With
control of the c-file and Black's pas
sive pieces, I quickly gained a decisive
advantage. Black should have played
. .. ctJe4 at some point before Mc2 in

Th e 4

a 6 Sla v : Wh i t e p la y s 5 e 3

order to force the exchange of one of

White's bishops, but White has rea
sonable compensation for the pawn.
1 0 . . . 0-0 1 1 e2 bxc4 ! 1 2 bxc4 c 5 !
This i s a typical freeing manoeuvre
for Black. The immediate 1 1 . . .c5
would of course have lost a pawn to
12 cxb5, so Black first exchanges on
c4 and then breaks in the centre.
1 3 dxc 5 ?

This is a serious mistake, and after

Anand's superb play it almost looks
like the losing move. White had to
play for equality with 13 cxd5 cxd4 14
exd4 1i.xc3 15 1i.xc3 tLlxd5 .
1 3 . . d4! 1 4 exd4 xd4
.

17

a4 ad8

18

e1

tDd4!

19

xb4?

Retreating with 19 1i.d1 was the

only (but rather miserable) way to
avoid material loss.
1 9 . . . tDxe2+ 20 tDxe2

Or 20 h l :d3 !
20 . . :xa 1 2 1 tDc3 c 1 22 a5 f4
23 xa6 l:l:a8 24 d6 xc4 0 - 1

White had had enough. This game

is a really impressive demolition job
by Anand.
Let us now see what happens if
Black recaptures on f3 with the pawn
instead of the queen.
White has a worse pawn structure
and real tactical problems, as Black's
pieces quickly become amazingly ac
tive.

Game 52
Van der Sterren-Shi rov

Biel Interzonal 1993

1 5 c2

15 iVc 1 (to avoid Black gammg a

tempo on the queen after . . . tLlc6-d4)
15 . . J::t d S ! prevents White from cas
tling.
1 5 . tDc6
. .

15 .. J::r d S is well met by 16 :d1 !

1 6 0-0 e 5 !

Freeing the
knight.

d4-square

for the

1 d4 d 5 2 c4 c6 3 tDf3 tDf6 4 tDc3

a6 5 e3 b 5 6 b3 g4 7 h 3 xf3 8
gxf3 ! ?

The recapture with the queen

leaves White vulnerable to a quick
. . . 1i.b4. This recapture aims to suffo
cate Black by playing f3-f4
(preventing Black's . . . e7-e5 break) and
c4-c5 (preventing the . . . c6-c5 break) .
89

Th e Sla v

5 liJf3 b 5 6 b3 g4 7 e2

8 . . . liJ b d 7 !

Threatening . . . c7-eS .
9 f4 bxc4 1 0 bxc4 dxc4!

By taking the c-pawn, Black pre

vents c4-cS . Now White cannot stop
Black from playing . . . c6-cS himself,
and the game soon fizzles out.
1 1 xc4 e6

Peter Wells and Glenn Flear are

well known for their deep knowledge
of Slav systems, so this game is espe
cially interesting.
7 . . . e6 8 0-0 ..IiLd6? !

An inaccuracy that has unpleasant

consequences. When White plays
LiJeS, attacking c6, Black needs to be
able to exchange it as quickly as pos
sible for one of his own knights.
Therefore Black should either play
8 . . LiJbd7 or 8 . . e7 (to meet 9 4Je5
with 9 . . . xe2 10 llix e2 and then
1O . . . LiJfd7!) . In the game, White gets a
grip on the dark squares and wins in
model fashion.
.

9 h 3 h5 1 0 liJe5 xe2 1 1 liJxe2!

0-0 1 2 liJf4 Wlic7 1 3 liJfd 3 ! liJbd7 1 4
b2 Wlib7

1 5 '!:!c 1

ac8

1 6 Re2

1 2 d2 ..IiL b4 1 3 0-0 0-0 1 4 liJe4 a5

.!:!fd8 1 7 Wlif3 f8 1 8 J:!fc 1 !

1 5 ..IiLxb4 axb4 1 6 liJxf6+ liJxf6 1 7

White's pieces are coordinating

beautifully.

Wlid3 liJ d 5

1 8 fc l .!:!a5

1 9 J:!ab 1

d 6 Y:z - Y:z

see

follo wing diagram

Game 53

1 8 . . . bxc4 1 9 bxc4 liJxe5 20 liJxe5

Wells-Flear

d6 21 liJd3 liJe4 22 e2 Wlib8 23

Oakham+1994

liJc5 xc5 24 dxc5 Wlib7 2 5 cxd 5

xd5 26 ..IiLxg7 xg7 27 Wlig4+ LtJg5

1 d4 d5 2 c4 c6 3 liJc3 liJf6 4 e3 a6
90

28 e4 J::!. e 5 29 f4 J::!. x e4 30 Wlixg 5+

Th e 4

a 6 Sla v : Wh i t e p la y s 5 e 3

Wf8 3 1 g d 2 e7 32 h6+ g8 3 3

White back.

l:!: c 3

1 7 . . . b7 1 8 e2 ga 7 ! !

h8

34

gg3

gg8

35

gd7

xe5+ 3 6 h 2 1 -0

Game 54
K rasenkov-Epishin

Bmo 1 994

I f White now defends the eS-pawn

with 19 f4, then 19 . . .tDb S ! 20 l:\,c3 b4!
21 l:\,d3 l:\,cs (Epishin) gains the c-file
for Black.
1 9 z:!. e 1 LZlxe 5 ! 20 LZlxe6 xe6 2 1
LZlxf8 xe2 2 2 gxe2 LZle4! ! 2 3 bxe4

1 d4 LZlf6 2 e4 e6 3 LZlf3 d 5 4 LZle3

dxe4

a6 5 e3 b 5 6 b3 JiLg4 7 JiLe2 e6 8 h3
Jih5 9 0-0 LZl bd 7 ! 1 0 LZle5 JiLxe2 1 1
CLJxe2 LZlxe5 1 2 dxe5 LZld7

13 exd 5 exd 5 14 JiLb2 JiLe7 1 5 LZld4

The white knight is trapped. Black

regains his piece and his queenside
pawns prove to be far too much for
his opponent to cope with.

b6 1 6 l:i.e 1 0-0 1 7 ge6

24 JiLd4 ge7 2 5 LZlxh 7 xh7 26 JiLb6

17 CDc6 cS is equal according to

Epishin. Here White's control of the
c-file looks impressive, but with some
fine moves, Epishin gradually pushes

ge6 27 JiLa5 JiLa3 ! 28 f 1

b4 29

e2 ge5 30 JiLb6 g b 5 31 JiLd4 b3

32 gxe4 b2 3 3 JiLxb2 gxb2+ 34 f3
JiLb4 0- 1
91

Th e Sla v

Instead of 6 b3, sometimes White

plays 6 cxd5 .

e2 e5 33 h3 e4 V2 - V2

Game 55
Karpov-Short

Dortmund 1 995
1 d4 d5 2 c4 c6 3 tZJc3 tZJf6 4 e3 a6
5 tZJf3 b5 6 cxd 5 cxd 5 7 tZJe5

In the next game we see a tricky al

ternative move-order from White.

Game 56
Sadler-Hodgson

Hastings 1 995/96
1 d4 d 5 2 c4 c6 3 tZJc3

A different plan: White prevents

the black bishop from developing
outside the pawn chain and tries to
prove that Black's queenside is weak.

Of course playing 3 ttJc3 first gives

Black the extra possibility of playing
3 . . . dxc4 (see Chapter 10) .
3 . . . tZJf6 4 e 3 a 6 5 c2 ! ?

7 . . . e6 8 .1i.d2 .1i.e 7 9 .1i.e2 0-0 1 0 0-0

.1i.b 7 1 1 tZJd3

A typical manoeuvre by White,

bringing the knight in contact with
the weak c5-square.
1 1 . . . tZJ b d 7 1 2 b4 tZJ b 6 !

White has weak squares too!

1 3 a 4 tZJe4!
see

follo wing diagram

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

tZJc5 .i,c6 1 7 .1lxa8 xa8 1 8 .1i.d3
a2 1 9 h 5 g 6 20 e5 a7 21
ga 1

b8 22 x b8 gxb8 23 ga5

.1i.d8 24 .1i.e1 ga8 2 5 !!xa8 tZJxa8 26

g4 tZJb6 2 7 f3 tZJc4 28 .i,f2 .i,g5 29
f4 .i,e7 3 0 g 2 f6 31 f3 f7 3 2
92

5 ttJf3 would o f course simply

transpose to the games we have al
ready seen in this chapter. On c2 the
white queen prevents Black from
developing his bishop to f5, which

Th e 4

suggests that it would be natural for

Black to put his bishop on g4 instead.
However, after 5 . . . b5 6 b3 g4,
White's idea is to play 7 tiJge2, and if
7 . . . tiJbd7 then 8 h3 h5 9 tiJf4,
picking up the bishop pair.
Question 3: I thought you said that
Black wanted to exchange off his
light-squared bishop for White's
knight! Aren't you contradicting
yourself?
A nswer: It is a conflict of ideas - as
Black you say, ' The bishop on c8 was
my problem piece and I'm glad I've
exchanged it,' whereas with White
you say 'Yes, I've won the bishop
pair!' Frankly I would be happy to
play either colour! It is clear, how
ever, that in comparison with the line
5 tiJf3 b5 6 b3 g4 7 h3 xf3 , White
has gained the two bishops at a much
lower cost: he has not had to either
weaken his kingside pawn structure,
or misplace his queen on the kingside.
So basically White has got a good ver
sion of this typical sort of position.
And that is the point of waiting with
5 'lic2.
A similar idea for White is 5 d3 ,
preventing . . . f5 . Personally, I would
grab this opportunity to transpose
into a Queen's Gambit Accepted with
5 . . . dxc4 6 xc4 e6 7 tiJf3 c5, but I
know that not everyone feels the
same way! 5 . . . g4!? 6 'lib3 Ma7 is
possible, however, as neither 7 f3 dxc4
(7 h5 8 cxd5 cxd5 9 g4 g6 10
xg6 hxg6 11 g5 tiJh5 12 tiJge2 [12
tLlxd5 tiJg3 !J 1 2 ... e6 1 3 f4 is better for
White due to the offside knight on
h5) 8 xc4 h5, intending . . . tiJbd7
and . . . e7-e5, nor 7 cxd5 cxd5 8 f3
. . .

a 6 Sla v : Wh i t e p la y s 5 e 3

c8 !?, with . . . tiJc6 and . . . e7-e5 to fol

low, is advantageous for White.
Right, back to the game!
5

. . .

e6 6 iDf3 e 5 !

A very imaginative idea. Black to

tally changes his plan; he no longer
plays for . . . b7-b5, but strikes in the
centre with . . . c6-c5 .
Question 4: Hasn't Black just
wasted a move, since he's played first
. . . c7-c6 and now . . . c6-c5?
A nswer: This is true of course.
White is playing the variation 1 d4
tiJf6 2 c4 e6 3 tiJf3 c5 4 e3 d5 5 tiJc3
a6, with the move 'Iic2 added in for
free. Black's contention is that this
variation is not advantageous for
White normally, and that the move
'lic2 does not make any difference in
White's favour.
I felt that the best way to try to
make use of 'lic2 was to play 7 cxd5,
so that after 7 . . . exd5 I could attack a
clear target on d5 by bringing my
rook to d 1 , exploiting the fact that
my queen has already vacated this
square. Moreover, my queen could be
very useful on c2 to attack a bishop
on c5 after d4xc5 xc5.
7 exd 5 exd 5 8 e2 iDe6 9 0-0 e6 !
93

Th e Sla v

of this defensive course, White plays

for the initiative.
1 3 a 3 ! exf3 1 4 iLxf3 ctJc6 !

Julian did not like the look of

14 . . . ttJdS 15 dxcs ttJc7 16 iLxb7 with
three pawns and an initiative for the
piece.
1 5 d5 ctJe5 1 6 dxe6 ctJxf3+ 1 7 gxf3
fxe6 1 8 b3 xd 2 1 9 iLxd2

The start of a superb tactical plan.

Normal development could have eas
ily ended in disaster. For example, if
9 . . . iLe7 10 Md1 0-0 then 1 1 dxcS iLxcs
12 ttJxds wins a pawn, since the
bishop on cS hangs. Therefore Black
begins a complicated tactical manoeu
vre that seeks to exploit the exposed
position of the queen on c2 .
1 0 gd 1 ctJ b4 ! 1 1 d 2

1 1 'iVb 1 'iV cS , threatening . . . iLfS,

could be embarrassing.
1 1 . . . ctJe4! 1 2 ctJxe4 dxe4

Black's queenside pawn maJonty

even gives him a slight pull, but after a
few adventures the game was eventu
ally drawn.
1 9 . . . d 7 !

20 iLc3+ c6

21

ga2

iLe7 22 gad2 gad8 23 gxd8 gxd8

24 gxd8 iLxd8 2 5 iLxg7 c4!

Sacrificing a pawn to activate the

king.
26 bxc4 c5 2 7 f4 xc4 28 e4
iLc7 29 f5 exf5 30 exf5 d 5 3 1 a4
b 5 32 axb5 axb5 33 iLc3 iLe5 34
iLd2 e4 35 f 1 xf5 36 h3 iLf4
37 iLc3 e4 38 e2 iLd6 39 iLd2
d4 40 iLe3+ c4 41 iLd2 iLb4 42
iLxb4 !

White's pieces are not coordinating

well, and if White were to continue
routinely with 13 ttJeS, then 13 . . . cxd4
14 exd4 MCS , threatening . . . ttJc2,
would be extremely annoying. Instead
94

I spent a while just checking that

the pawn ending was drawn. Remem
ber that 4 1 . . .b4 just leads to a draw
after 42 iLxb4 as Black has the wrong
coloured bishop for the rook's pawn!

Th e 4

a 6 Sla v : Wh i t e p la y s 5 e 3

42 . . . <;t>xb4 43 <;t>d3 <;t>a3 44 <;t>c2 b4

7 . . . d 7 S ttJf3 e 6 9 ttJe5 cS 1 0 f3

45 <;t>b 1 <;t>b3 46 h4 <;t>c3 47 h5 -

ttJfd7 1 1 ttJxd 7 ttJxd7

Since 47 . . . h6 48 f4 d4 49 b2
e4 50 b3 xf4 51 xb4 g4 52
cj;c3 xh5 53 d2 g4 54 e1 f3
55 f1 draws Gust) for White.

Also possible was 1 1 . . .'i'xd7! ?

Another tricky move for White is 5

a4 .

Gam? 57
Atali k-Miles

Hastings 1995/96
1 d4 d 5 2 c4 c6 3 ttJc3 ttJf6 4 e3 a6
5 a4

An unusual move in this position

when White has already committed
himself to e2-e3, shutting in his dark
squared bishop.
5 . . . f5 6 b3 ga7 7 a 5 !

Suddenly, t o his horror, Miles real

ised that after the natural 7 . . . e6, 8
'{!Vb6! is extremely strong, as after the
forced 8 . . . 'i'xb6 9 axb6 a8 10 c5 ,
intending b2-b4 and b4-b5 breaking
through, White has a magnificent end
ing. Tony, practical as ever, just
played a few necessary defensive
moves and got on with the game!

1 2 cxd 5 cxd 5 1 3 e4 dxe4 1 4 fxe4

g6 1 5 e 5 ! Le7 1 6 Le2 0-0 1 7 0-0
gaS 1 S Le3 .l:!.bS 1 9 a4 dS 20
.l:!.ad 1 MCS 2 1 d5 c5 22 Lf2 xf2+
2 3 gxf2 ttJc5 24 b4 g 5 25 dxe6
fxe6 2 6 gxfS+ <;t>xfS 27 d4 <;t>gS
2S c4 Wie7 2 9 b4 ttJ d 7 30 Lb3
ttJfS 31 d6 eS 3 2 La4 f7 3 3
Mf 1 f5 3 4 ttJe4 g 6 3 5 ttJ c 5 h 6 3 6
d 1 e4 3 7 B f 2 Ld 5 3S L c 2 Wig 5
39 h3 e3 40 e7 <;t>hS 4 1 f7
ttJh7 42 f4 e 1 + 43 <;t> h 2 .l:!.gS 44
g3 xb4 45 ttJd3 a3 46 ttJf4
xg 3+ 47 <;t>xg3 gdS 4S ttJg6+ <;t>gS
49 ttJe 7 + <;t>hS 50 ttJg6+ <;t>gS 5 1
95

Th e Sla v

%-%
After 5 a4 Black can simply play
5 . . . e6, aiming to put a bishop on the
hole on b4, but this leads more to a
Semi-Slav type of position, so for Slav
devotees, I offer a few other ideas:
a) 5 . . . g6 is interesting, leading to a
sort of Schlecter Slav (see Chapter 1 1) .
ClJe7+ h8 52 ClJg6+

96

b) 5 . . . 'ii c 7 is not as stupid as it

seems: Black defends b7 in advance. If
6 ct'Jf3 then 6 . . . g4 7 'ii b 3 xf3 8
gxf3 e6!
c) 5 . . . Ma7!? with the same idea. If 6
as then maybe 6 . . . e6!? 7 ct'Jf3 (7 'ii b 3
dxc4!) 7 . . . dxc4 8 ct'Jg5 g4 9 f3 h5
with a completely unclear position.

Th e 4

a 6 Sla v : Wh i t e p la y s 5 e 3

S u m m ary

In general Black is doing fine in these lines, but since the 4 . . . a6 Slav is such a
recent development, there is still scope for improvements for both colours. 4
Qjf3 a6 5 e3 bS 6 b3 oltg4 7 h3 xf3 8 'i'xf3 e6 9 d3 b4 10 d2, as in
Sadler-Levitt, is worth further tests, and the game Sadler-Hodgson is certainly
crazy enough to be worth analysing!
1 d4 d5 2 c4 c6 3 tUf3 tUf6 4 tUc3

4 e3 a6 (D)
5 'i'c2 game 56
5 a4 game 57
-

4 . . . a6 5 e3 b 5 6 b3

6 cxdS game 55
-

6 . . . iLg4 (D) 7 h3

7 e2 e6
8 0-0 - game 53
8 h3 game 54
-

7 . . . iLxf3 (D)

'i'xf3 game 5 1
8 gxf3 game 52

. . .

a6

. . .

iLg4

. . .

iLxf3

97

The 4

a6 Slav:

Aggressive o ptions for White

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

a6

In this chapter we deal with the

more attacking systems against the
4 . . . a6 Slav, in which White delays e2e3 in order to develop the bishop on
c 1 to an active square outside the
pawn chain. We shall first examine S
cS, which has virtually become the
main line.
Question 1 : S cS looks like a begin
ner's move. Isn't it bad to release the
tension in the centre so early?
A nswer: White is aiming for a 'big
clamp' on the centre: S cS stops Black
from breaking with . . . c6-cS, and f4
will prevent Black from achieving
. . . e7-eS . White will then either launch
a queenside offensive with b2-b4, a2a4 and b4-bS, or he will organise a
central break with e2-e4. Overall, this
is a very ambitious plan.
The queenside break . . . b7-b6 is not
great for Black, since cSxb6, . . . ik'xb6
leaves him with weak dark squares on
cS, b6 and as, and a backward c-pawn.
Question 2: Black can't play on the
queenside, and he can't get in any of
his breaks. What can he do?
A nswer: . . . c6-cS is not on the
agenda, but . . . e7-eS should not be im98

possible to achieve if Black can organ

ise his pieces properly. White has
closed the centre, so Black has more
time to achieve his plan, as he never
has to worry about a sudden central
breakthrough. Moreover, Black can
spare a lot more pieces than usual to
prepare . . . e7-eS, since he doesn't need
any to defend his own centre.
Game 58
K ra m n i k-Shirov

Vienna 1 996
1 lLlf3 d 5 2 d4 c6 3 c4 lLlf6 4 lLlc3
a6 5 c 5 .iLf5 6 b3 .l:!.a7 !

I couldn't help laughing when Jul

ian Hodgson told me that he played
. . . a7-a6 in order to facilitate . . . .l::!. a 7,

Th e 4

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

but the idea is a good one. It really is

not clear whether the queen on b3 is
any better placed than the rook on a7:
on b3 the queen attacks very little and
gets in the way of one of White's
main plans, the queenside pawn storm
with b4, a4 and bs. If the queen
moves away from b3, then the rook
can simply return to as !
The more conventional 6 . :cS 1S
discussed in the following game.

White positionally. White had to an

ticipate . . . e7-eS and either prevent it
with 17 eS or play 17 iYc4, intending
17 . . . eS 1S dS .
1 7 . . . e 5 ! 1 8 c4 f6 1 9 dxe5 lLlxe5

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

The pos1tlOn is now horrible for

White: Black can just gang up on his
weak pawns.
20 d4 gaa8 !

Having done its job, the rook

comes back into play.
2 1 e 3 J::!. a d8 2 2 .lte2 e 7 23 J:Ihe 1

This move shows a good under

standing of the position. Black leaves
his weakened queenside alone and
grabs space on the kingside, where
White has little going for him. The
move . . . g7-gS also allows the bishop
to come to g7, supporting the . . . e7-eS
break.

lLld7 24 d4+ lLlf6 25 e3 e5 26

.ltf3 h 5 !

Preparing . . . gS-g4, driving away a

defender of e4.
2 7 a3 gfe8 28 nxd8 nxd8 29 .ltd 1
J::!. d 4 30 .ltc2 J::!. c4 3 1 Wb 1

1 0 .lth2 .lt g 7 1 1 lLle5 0-0 1 2 f3

Unwilling to let Black have things

his own way, White prepares the e2e4 break. This raises the stakes, as
... e7-eS will become doubly effective
against an expanded white centre.
1 2 . . . lLlxe5 1 3 .ltxe5 lLl d 7 1 4 .ltxg7
xg 7 1 5 e4 dxe4 1 6 fxe4 .ltg6 1 7
O-O-O?
A

careless move that condemns

99

Th e Sla v

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

3 1 . . . 1:!.xc5

The first pawn falls . . .

3 2 d2 1:!.c4 3 3 dS ctJxe4 3 4 3Lxe4

3LdS 1 2 3Le2 ctJxe5 1 3 3Lxe5 3Lc7

14 ctJa4

3Lxe4+ 35 ctJxe4 .!:!xe4

. . . and now the second.

0- 1

A really good exposition of the

ideas behind the 4 . . . a6 Slav.

Game 59
I . S okolov-Shirov

Erevan Olympiad 1 996

1 d4 d 5 2 c4 c6 3 ctJf3 ctJf6 4 ctJc3
a6 5 c 5 3Lf5 6 b3 cS ! ?

This is a normal plan, trying to

gain the two bishops by exchanging
the knight for the dark-squared
bishop on ds . However, there is a
tactical problem on this occasion.
1 4 . . . 3La5+! 1 5 ctJc3

Or 15 Wfl CDe4 16 Wg2 f6, intending . . . e6-e5 .

1 5 . . 0-0 1 6 h4 h6 1 7 0-0-0 ctJd7 1 8
.

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

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

I'm not sure why Shirov did not

want to repeat his previous game.
Perhaps it was nothing objective, just
the desire not to be too predictable.
The idea behind this move is ex
tremely neat: Black will develop normally with . . . e7-e6, . . . CDbd7, . . . :ie7
and . . 0-0, and will then prepare . . . e6e5 by means of . . . MeS and . . . :idS-c7,
making use of the fact that the queen
has vacated dS !
.

7 3Lf4 ctJ b d 7 S h 3 ! ?

Black i s going t o castle kingside, so

White wants to have his pawn storm
ready on that side of the board.
100

I think that Black IS Just better

here: White is having to make all sort
of positional concessions and his at
tack just isn't getting anywhere.

Th e 4

a 6 Sla v : A g g r e s s i v e o p tio n s fo r Wh i t e

2 3 gxh6 g x h 6 2 4 e 4 dxe4 2 5 tLia4

b 5 e2+ 60 c3 f4 61 gd8 f3

d8 26 tLic5 tLixc5 2 7 dxc5 f6 28

6 2 .l:!.d2+ e 1 0 - 1

c3 xc3+ 29 bxc3 Uab8 30 gd4

Obviously Black is rather happier

than White in this line at the mo
ment!

a 5 31 d 1 b6 3 2 cxb6 gxb6 33
jLc2 .!:!.d8 34 d7 .!:!.b5 3 5 .!:!.gd 1 '!:!'xh 5

Another very popular idea for

White is to play 5 a4.
Question 4: What is the point of
this move?
A nswer: 5 a4 merely aims to pre
vent Black's idea of . . . b7-b5, and force
him to find another plan. Though it
severely weakens the b4-square, this
move does give White the possibility
of a4-a5, cramping the black queen
side.
Black wins yet another pawn.
Though Sokolov fights hard, the re
sult is never in doubt.

Pushkov-Epishin

36 c4 g7 37 b2 f8 3 8 .ita4

Russian Championship 1 995

Game 60

nc5 3 9 \t>c3 e7 40 .!:!.d6 e3 4 1

fxe3 e4 42 .!:!.6d4 .!:!. e 5 4 3 gf 1 g2

1 tLif3 tLif6 2 c4 c6 3 tLic3 d 5 4 d4

44 gg 1 gxe3+ 45 gd3 .!:!.xd3+ 46

a6 5 a4 e6

\ii x d3 '!:!'xd 7 + 47 e3 h3 48 xc6

nc7 49 b 5 e 5 50 Uh 1

.ite6 5 1

\ii d 3

\ii f4 54 .!:!.h4+ g4 55 gh8 f5 56

The most natural and overwhelm

ingly the most popular choice. The
game loses a little of its Slav character
I'm afraid, but Black does get to oc
cupy that weak b4-square.

ne8 g c 5 5 7 a4 f3 58 a6 f2 59

6 g5

51 . . . f6

52 .!:!.xh6+ g 5

53 gh8

101

Th e Sla v

The fianchetto 6 g3 is dealt with in

the following game.

g7 38 ba3 .i:txb2 39 xa6 .i:tf4 40

6 . . . tiJ b d 7 7 e3 iLe7 8 iLd3 0-0 9 0-0

.i:td3 43 na5 Wf6 44 na6+ f5 45

dxc4 1 0 iLxc4 c 5 1 1 a 5 cxd4 1 2

ga7 e6 46 .l:!a6+ .i:td6 47 ga5 f5

exd4 b 5 1 3 axb6 tiJxb6

6a4 xa4 4 1 nxa4 nd2 42 g 3

48 a8 Wd5 49 f2 nb6 50 l:IdS+

e6 51 g3 nb1

52 h 5 gxh 5 53

l:Ih8 h4+ 54 xh4 d5 5 5 XlfS f4

56 .i:td8+ e6 57 ne8+ f6 58 .l:!fS+
e7 59 h8 gg 1 60 nh7+ f6 6 1
l:Ixh6+ g 7 62 n e 6 .l::i. x g2 6 3 gxe5
f6 64 J::!. a 5 .l::!. g 1 % - %

Let us see what happens if White

opts to fianchetto .
Game 61
Cifuentes-Dreev

Black has already equalised.

Wijk aan Zee 1 995

1 4 iLe2 iL b 7 1 5 tiJe5 h 6 1 6 iLh4

tiJfd 7 1 7 iLxe7 xe7 1 8 tiJxd 7 xd 7

1 c4 c6 2 d4 d 5 3 tiJc3 tiJf6 4 CUf3

1 9 tiJ a 4 tiJxa4 20 l::l x a4 l::l f d8 2 1

a6 5 a4 e6 6 g3 tiJbd7 7 iLg2 iLb4 8

a 1

0-0 0-0 9 b3 a5 1 0 gd 1 b6

c6 2 2 iLf3 b6 2 3 iLxb7

xb 7 24 c 1 d6 2 5 h 3 gad8 26
gcc4 e 5 !

Black wins a pawn, but can't quite

convert it into a win.

gg6 30 .i:tg4 l:I d d 6 3 1 gb8+ h7 3 2

White will find it hard to achieve

e2-e4, since Black has his bishop en
trenched on b4 and can always play
. . . xc3 , removing a defender of e4.
Without this idea, however, White
may struggle to find a plan.

b 3 xb3 3 3 g x b 3 gd 1 + 34 h 2

1 1 tiJe5 tiJxe 5 1 2 dxe5 tiJd7 1 3 cxd 5

g f 6 3 5 f3 g d 2 3 6 h4 g 6 3 7 ga4

exd 5 1 4 iLf4 e7

27 gab4 e4 28 d5 xd 5 29 a3

1 02

1 5 e4 d4!

16

Th e 4

a 6 Sla v : A g g r e s s i v e o p tio n s fo r Wh i t e

gxd4 lZJxe 5 1 7 J. x e 5 \'ixe5

swift attack.
Game 62
Ward-Levitt

British Championship 1 995

1 d4 d 5 2 e4 e6 3 lZJe3 lZJf6 4 lZJf3
a6 5 J.g5 dxe4

Black's two bishops and queenside

pawn majority give him a clear advan
tage.
1 8 .l:!.d 2 J.e6 1 9 \'i c2 gad8 20 gad 1
gxd2 2 1 gxd2 b 5 22 h 3 h 5 23 h4
bxa4 24 \'ixa4 J.xc3 2 5 bxe3 \'ixe3
26 ge2 \'ib4 2 7 \'ixe6 a4 28 \'ie3
Vi1Vb3 29 f 1 ge8 3 0 \'ix b3 axb3 3 1

6 a4

gd2 g e 2 3 2 gd8+ h 7 3 3 e 5 J.f5

Also possible is the immediate 6

e4!?

0-1

6 . . . h 6 7 J. h 4 lZJd5 8 e 4 lZJ x e 3 9

This does all seem fine for Black

but if you cannot live happily with
out developing your bishop outside
the pawn chain, then S . . . fs 6 iVb3
Ma7 seems interesting. After 7 as,
threatening 8 iVb6, 7 . . . dxc4 8 iVxc4
CLlbd7 is not so stupid. For example, 9
gS h6 1 0 i.. xf6 (10 i.. h 4 gS!? 1 1 i.. g 3
i.. g7 seems fine for Black) 10 . . . exf6
(Please don't even think of 10 . . . CLlxf6,
allowing 1 1 iV cS ! , winning a piece) 1 1
e4 i.. g4 (1 1 . . .i.. h 7!?; 1 1 . . .i.. e 6!?) with a
very murky position.
S i.. g S is a sharp continuation that
is only occasionally seen. White de
velops h is bishop to its most aggres
sive square and is willing to sacrifice
the pawn on c4 for the chance of a

bxc3 b 5

Black has gone about this in a

strange way. S . . . Lbe4 6 i.. h4 (6 h4 has
been played but I don't believe it!
Even 6 . . . h6 is not stupid, when I don't
see the value of h2-h4!) 6 . . . Lbxc3 7
bxc3 dxc4 8 e4 bs is more natural if
Black wants this type of position,
since the omission of a2-a4 and h7-h6
must help Black a little. After S . . . dxc4
6 a4 another idea is to play 6 . . . fS (a
more natural Slav move) , aiming for a
little trap: 7 e3 Lbe4!? 8 Lbxe4 i.. x e4 9
i.. x c4 i.. xf3 ! 1 0 iVxf3 aS+! 1 1 e2
xgS 12 xf7+ d8 , when White
does not have sufficient compensation
for the piece.
1 0 lZJe5 \'ie7 1 1 J.g3 \'ib7 1 2 g b 1
1 03

Th e Sla v

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

The opening has not been a success

for Black, but somehow he hangs on
and almost turns the tables com
pletely.
1 5 . . . g h 7 1 6 f3 3l. b 7 1 7 0-0 tiJd7
1 8 tiJxd7 xd7

1 9 f4 e8

After this freeing break, Black has

no problems.
11

tiJxd 7 tiJxd7

1 2 3l.e3 3l.e 7

13

dxc5 tiJxc5 1 4 Wiix d8+ gxd8 1 5 gd 1

0-0 1 6 a3 gxd 1 + 1 7 xd 1 M.c8 1 8
3l.e2 f8 1 9 c2 tiJa4 20 3l.d2 3l.d6

20

21 h 3 e7 22 Bd 1 tiJxc3 23 3l.xc3

axb5 axb5 21 ga 1 ! Wii b 6 22 :!::i. x a8+

b4 24 axb4 3l.xb4 25 gd3 a5 26

3l.xa8 23 g a 1 Wii b 7 24 d5 cxd 5 25

b3 3l.xc3 2 7 Bxc3 .l:i.xc3+ 28 xc3

exd 5 3l.e7 26 e3 3l. g 5 2 7 f4 3l.e7

d6

28 3l.f3 Wii c 8 29 d 6 3l.xf3 30 Wiix f3

c5+ 31 3l.f2 Wiix d6 3 2 .!:!.a8+ 3l.d8
3 3 ga7 Wii d 5 34 Wiix d 5 exd 5 35 g b 7
h 5 36 xb5 f5 3 7 M.xd 5 .!:!. d 7 38
ge5+ f7 3 9 g 3 3l. f6 40 M.c5 V2 - V2

Next, the adventurous 5 CLJeS . I

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

Game 63

tiJxc4 b5 8 tiJe5 3l.b 7 9 e4 e6 1 0 f3

You might have expected the play

ers to have called it a day here, espe
cially when you consider that this was
the last round of a gruelling Olym
piad for both players: Beliavsky was
playing his 14th game on board 1 and
Shirov his 1 3th!

c5!

29 d4 e 5+ 30 c3 c5 31 3l.c4 f6

Beliavsky-Shirov

Erevan Olympiad 1 996

1 d4 d5 2 c4 c6 3 tiJf3 tiJf6 4 tiJc3
a6

1 04

5 tiJe5 tiJ b d 7 6 3l.f4 dxc4!

Th e 4

a 6 Sla v : A g g r e s s i v e o p tio n s fo r Wh i t e

32 h4 c6 3 3 g8 h 6 3 4 c 4 d 7
3 5 h 5 f5 3 6 e x f 5 x f 5 3 7 d3 e6
38 e4 f7 3 9 g4 c4 40 b7
b 5 41 a8 c4 42 e4 d5 43
xd 5 xd 5 44 b3 e4 45 f4 e3 46
d3 e 2 4 7 xe2 e4 48 g5 xf4
49 gxh6 gxh6 50 d 3 g4 5 1 c4
xh 5 5 2 b5 g4 5 3 xa 5 h5 54

b4 h4 55 b5 h3 56 b6 h2 57 b7
h 1 58 b8 a 1 + 59 b6 b 1 +
60 rtJc7 x b8+ 6 1 xb8 V2 - V2

In the end even these two battlers

cannot continue the game! A really
impressive display of fighting spirit
from both players.
The final idea to be considered is 5
'i'b3 , preventing the development of
the light-squared bishop by attacking
b7. The advantage of this move is that
Black cannot counter in normal Slav
fashion, but the Semi-Slav approach is
easy to understand, and nice for
Black.

6 1Lg 5 ? !

6 cxd5 cxd5 7 iLg5 iLe7 S e 3 h 6 9

iLxf6 iLxf6 10 iLd3 ctJc6 1 1 0-0 0-0 12
Mac 1 was played in Piket-Shirov,
Aruba 1995, and now Shirov suggests
12 . . . iLd7, as 13 xb7 ctJa5 14 b4
e7 snares the queen.
6 . . . dxc4 7 xc4 b 5 8 d 3 c 5 9 a4
cxd4! 1 0 LiJxd4 b4 1 1 LiJe4 1Lb7 1 2
xf6 gxf6 1 3 ld.d 1 1Le7 1 4 f3 b 3 !

Game 64
Lautier-Ba reev

Linares 1994
1 d4 d 5 2 c4 c6 3 LiJf3 LiJf6 4 LiJc3
a6 5 b3 e6

The most popular move for Black.

6 xc4 iLf5 7 g3 is the 4 b3
dxc4 5 xc4 iLf5 line with the extra
moves 4 ctJc3 a6, which favour White
more than Black, while 5 . . . Ma7 6 iLf4!
is awkward. I briefly considered 5 . . . b5
6 cxd5 cxd5 7 a4 b4!? S xb4 ctJc6 9
'i'b3 Mbs 10 dl iLf5, intending
. . . ctJc6-b4, but 1 1 ctJh4 chases away
the bishop and just leaves White a
pawn up.
5 . . . dxc4

To meet 1 5 ctJxe6 with 1 5 ... a5+!

1 5 e3 a5+ 1 6 e2

Watch this king!

1 6 . . . rtJf8 1 7 f4 e5 1 8 f3 xf4
1 9 exf4 g7 20 f5 LiJc6 21

fxe6

LiJxd4+ 22 .i;l;,xd4 ghd8 23 gd7 gxd 7

24 exd 7 J:i.d8 25 rtJe3 f5 26 LiJg3
iLc5+ 27 f4 !
1 05

Th e Sla v

g4xfS-f6-f7-f8-g8 ! !

27 . . . 1Ld6+ 28 xf5 gxd 7 29 ctJe4

1Lc7 30 1Lc4 g e 7 3 1 ctJc3 ge5+ 32

40 . . . 1Lb4 4 1

g4 h 6 33 f4 ge3 34 1L d 5 f5+ 3 5

gxc3 43 ge 1 .l:i.c2 44 1Ld5 .l:i.d2 45

xf5 1Lc8+ 3 6 f6 1Lxf4 3 7 1Lxb3

1Le4 1Lh5 46 gf 1 1Lg6 47 1Lb7 gd4

1Lg5+ 38 f7 1Lg4 3 9 f8 ! ! 1Le7+

48 a 5 na4 49 gf6 gxa5 50 nxa6

40 g8 ! !

g h 5 5 1 h3 ge5 52 ga8 g 5 53 1Lf3

The white king completes a remarkable J ourney:

e2-e3-f4xfs-

h 5 12 - 12

1 06

1Lf7 1Lxc3 4 2 bxc3

A brilliant game!

Th e 4

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

S u m m a ry

If you had asked me a few months ago, I would have told you that Black had
a few problems in the 4 . . . a6 Slav, but now I am not so sure! 5 cS should be
avoided for the time being, while 5 a4, 5 ctJeS and 5 'iYb3 don't really seem to
promise a great deal, though 5 a4 and 5 'iYb3 can be good weapons if you
know that your opponent does not like to play systems with . . . e7-e6. 5 gS is
aggressive and deserves further tests.
1 d4 d 5 2 c4 c6 3 CLJf3 CLJf6 4 CLJc3 a6
5 c 5 (D)

5 a4 e6 (D)

6 gS - game 60
6 g3 game 61
-

5 gS game 62
5 ctJeS game 63
5 'iYb3 game 64
-

. . .

jLf5 6 b3 (D)

6 . . . .l:"!.a7 game 58
6 . . . 'iYc8 game 59
-

5 c5

. . .

e6

6 b 3

107

CHA PTER NINE

The Exchange Variation

1 d4 d 5 2 c4 c6 3 cxd 5 cxd 5

The Exchange variation is often

dismissed as just a dead draw, but sev
eral top players, Yusupov and Vaiser
in particular, play it to win. What do
they see in this variation?
Although the pawn structure is
symmetrical, there is the open c-file to
play with and we know from experi
ence the annoying pressure that
White can exert on b7. White also has
the advantage of moving first in this
position. Indeed, if Black does not
know what he is doing, his position
can very easily become highly un
pleasant.
Question 1 : What plans has Black
tried in this position?
A nswer: Black has tried two ap
proaches:
a) To put the bishop on c8 outside
the pawn chain, on either f5 or g4.
b) To unbalance the pawn structure
with . . . ctJe4xc3 .
Question 2: And what does White
do?
A nswer: White also has two ideas:
He can put his king's knight on either
f3 or e2 (after d3) .
Question 3: What is the difference
between them?
1 08

A nswer:

a) ctJf3 allows White to increase his

pressure on the c-file by playing ctJe5,
attacking a black knight on c6. If
Black exchanges knights with . . . ctJxe5,
then White has removed the main
barrier to invasion on the c-file.
b) With d3 and ctJge2 White aims,
not to attack on the c-file, but rather
to keep Black passive by preventing
the light-squared bishop from devel
oping outside the pawn chain. The
white bishop on d3 controls f5, while
the absence of a knight from f3 means
that . . . g4 can be met by f2-f3 .

Game 65
M i l ov-Sadler

Isle ofMan 1994

1 d4 d 5 2 c4 c6 3 cxd 5 cxd 5 4 iLf4

White must be wary of move

orders here as 4 ctJc3 e5 !? 5 dxe5 d4 6
ctJe4 ctJc6 7 ctJf3 f5 8 ctJg3 g6 9 a3
c5 10 "i'b3 ctJge7, as in Tozer-Levitt,
London (Lloyds Bank) 1993, gives
Black good counterplay for the pawn.
4 . . . lLlc6 5 e3 lLlf6 6 lLlc3

White could try 6 d3 ! ? here to

prevent 6 . . . f5. Black's best is 6 . . . i,g4
7 ctJe2 h5 ! , intending . . . g6 to swap

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

off bishops. 7 b3 ctJaS ! 8 bs+ d7

also poses few problems.
6 . . f5
.

The old main line. The modern

6 . . a6 is considered in the next game.
.

the main line, which does not seem

bad for Black, but at the board I got a
little carried away. The positional idea
is correct: since White has released the
tension in the centre very early, Black
feels much more able to take action
on the wings. Indeed this idea was
seen to great effect in Kramnik-Shirov
from Chapter 8, but in this case
White has the open c-file to help his
queenside play.
1 1 g3 h5 1 2 h3 g4 1 3 hxg4 hxg4
1 4 tZJd2 h4 1 5 tZJe2 a5 ! ?

7 ]L b 5 e6 8 tZJf3

4 ctJf3 ctJf6 S ctJc3 ctJc6 6 f4 fS 7

e3 e6 8 bs is the normal way to
reach this position. White is threaten
ing ctJeS, ganging up on the knight on
c6.

S o far so good. With my last move,

attacking the knight on d2, I was hop
ing for the reply 16 'sfd 1 , when
16 . . . c2! 17 xc2 xb5 is reasonable
for Black.
1 6 e4 ! !

8 . . . tZJd7

Breaking the pin on the knight and

thus dealing with the annoying threat
of ctJeS .
9 0-0 e7 1 0 b3 g 5 ! ?

After a great deal of thought, my

opponent found a very strong idea,
blowing the centre open.
1 6 . . . xd 2 1 7 exf5 ]Lxg3 1 8 tZJxg3
xd4 1 9 t!.fe 1 d8 20 fxe6 fxe6 2 1
l:!xe6

Bold but probably not the best.

1O . . . ,Sc8 1 1 xc6 bxc6 12 b7 as is

Well, this is not great for Black, but

at least I'm still alive! My opponent
was already in serious time-trouble,
and after the game I was really upset
that I had made things so easy for him
1 09

Th e Sla v

in the rest of the game.

a6 5 cxd5 cxd5 6 .tf4 ttJc6.

!!xd4

Question 4: Why is . . . a7-a6 useful in

4:Jxb3 24 xd 5 4:Jbd4 2 5 e4 4:Jxb5

this position?
A nswer: 6 . . . a6 is a constructive
waiting move: Black keeps the white
pieces from occupying the b5 square,
which means he no longer has to
worry about .tb5, pinning his knight
on c6.
Question 5: What if White just
plays 7 e3?
A nswer: Then Black plays 7 . . . g4 8
h3 .txf3 9 iVxf3 e6. This is another
example of Black giving up his light
squared bishop for White's king's
knight in the Slav. In this case, since
White has played cxd5 so early, Black
has been able to play his knight to c6,
its most natural and best square. The
manoeuvre . . . .tg4xf3 also removes
the attacking idea ttJe5 .
Question 6: So what does White do?
A nswer: The only way that White
can go for advantage is to avoid play
ing e2-e3 too early and thus sidestep
. . . .txf3 .

21 . . . c7

22

d 1

4:Jc5

23

26 x b 5 ad8 2 7 .l::i. e 1 d 2 28 4:Je4

.l::i. e 8 29 e3 c2 30 a4 .l::i. e 7 3 1 4:Jf6
I!.xe3 32 4:Jd5+ c8 3 3 4:Jxe3 :!:!.d2
34 4:Jx94 d4

3 5 4:Je5 :!:!.xa4 36

4:Jxc6 bxc6 3 7 b3 c7 38 f3 c 5 39
f2 c6 40 94 c4 41 b8 a 5 42
e3 c7 43 b5 c6 44 :!:!.b8 Wc7
45 .l:!a8 I!.a2 46 95 b7 47 96! c3
48 9 7 c 2

Everything else is also hopeless, e.g.

4S ... cxb2 49 .l:IbS+ xbS 50 gSiV +
a7 5 1 'ixa2 and wins.
49 d2 1 -0

Now let us take a look at the fash

ionable 6 . . . a6.

Game 66
H odgson-Sadler

Ischia 1 996
1 c4 c6 2 4:Jf3 d5 3 cxd 5 cxd 5 4 d4
4:Jf6 5 4:Jc3 4:Jc6 6 iL.f4 a6

The modern main line, which will

be of particular interest to 4 . . . a6 Slav
players, as this line can occur by
transposition after 3 ttJf3 ttJf6 4 ttJc3
1 10

7 4:Je 5 ! ? e6

Black does not fear S ttJxc6 bxc6, as

White is not sufficiently active to pre
vent Black from freeing himself with

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

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

7 . . . iLf5

An important new idea. 9 . . . bs was

also possible, but that does give White
a bit of a target to attack on the
queenside. 9 . . . d7 develops the bish
op to the as-h 1 diagonal (preventing
White from ever achieving e3-e4) ,
while still allowing Black to cover the
cS-square with the advance . . . b7-b6 if
necessary.

7 . . . g4 S ctJ eS ! (the point) is annoy

ing, but now S ctJ eS is met by s . . . ncs .
8 e3 e6 9 LDe5 LDxe5 1 0 iLxe5 CLld7
1 1 iLg3 .!:!.c8 1 2 iLd3 iLxd3 1 3 xd 3
iL e 7 1 4 0-0 0-0 1 5 '!::!' c 2 a 5 Y:z - Y:z

Despairing o f making anything

against this plan of . . . g4xf3 , white
players turned to 'Plan B ' : d3 and
ctJe2.

1 0 iLd3 iLc6 1 1 f3 CLld7 1 2 iLg3

Game 68

Y:z - Y:z

A real no-holds-barred classic! In

mitigation, England were playing
Switzerland that day in 'Euro 96' so
we did have other things on our
minds. (Although after watching the
match for half an hour we began to
think that even our game might have
been more exciting!)

Yusupov-Sh i rov

Zurich 19941%

1 d4 d 5 2 c4 c6 3 CLlc3 CLlf6 4 cxd 5

cxd 5 5 iLf4 CLlc6 6 e3 iLg4

Game 67
Andersson-Epishin

Ter Apel 1995

1 CLlf3 af6 2 c4 c6 3 d4 d5 4 cxd 5
cxd 5 5 LDc3 CLlc6 6 iLf4 a6 7 .1:Ic 1

Here we see another attempt to de

lay e2-e3 .
1 1 1

Th e Sla v

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

A nswer: After 6 . . . a6 White can play

7 i,d3 i,g4 8 tLJge2 (but not 8 f3 hs

and Black can exchange bishops
with . . . i,g6) .
Question 8: What's the problem?
Can't I just go 8 . . i,xe2?
A nswer: You certainly can, but in
comparison with the line 6 tLJf3 a6 7
e3 i,g4 8 h3 i,xf3 9 'iVxf3 , we see that
Black has an inferior version: White
has played i,d3 instead of h2-h3 , and
his queen is better placed on e2 than
on f3 . This position is perfectly play
able for Black, but there just seems no
point in going in for a worse version
of something if there is no real need
to.
Question 9: What is the point of the
6 . . . i,g4 7 f3 i,d7 manoeuvre?
A nswer: By forcing f2-f3 , Black
weakens a dark square on e3 in
White's position. Moreover, Black has
more chance of achieving the . . . e7-eS
break, since the white knight can no
longer go to f3 . Finally, the bishop is
not too badly placed on d7, as it sup
ports Black's counterplay with . . . b7bS.
Question 1 0: Wait a minute, if I get
this via a transposition from the . . . a7a6 Slav, then I won't be able to play it
in this way will I?
A nswer: Don't worry! White can't
use this subtle move-order if he trans
poses to the Exchange variation via
the 4 . . . a6 Slav, since he will either
have shut in his dark-squared bishop
(3 tLJc3 tLJf6 4 e3 a6) or will have al
ready committed his knight to f3 (3
tLJf3 tLJf6 4 tLJc3 a6) .
.

7 f3 Si. d 7 8 Si. d 3 e6
1 12

Also playable is 8 . . . g6! ? , intending

9 . . i,g7 to support the . . . e7-eS break.
.

9 Si.g3

The aggressive 9 g4 is considered in

the next game, while 9 tLJge2 allows
Black to win the bishop pair with
9 . . . tLJhS .
9

. . .

Si.e 7 1 0 tLJge2 0-0 1 1 0-0 a6

Black's main counterplay is to ex

pand on the queenside, placing his
knight on the c4 outpost and advanc
ing the a- and b-pawns.
1 2 tLJc 1 ! ? tLJa5 1 3 tLJ b 3 !

An imaginative idea: on b3, the

knight protects d4, allowing White to
expand in the centre with e3-e4, while
eyeing the cS-square, which will be
weakened by . . . b7-bs. The one

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

problem is that the knight will be in

the line of fire when Black plays . . . a6as-a4.
1 3 . . . lLle4 1 4 e2 J:i.e8 1 5 e4 b5 1 6
e5 lLlh 5 ! 1 7 f4 lLlxg3 1 8 hxg3 a 5 1 9
2Lxh7+ xh7 20 h5+ g8 2 1 f2
f5 Y, - y,

After 22 h l White will force per

petual with iVhs+ and iVhs+.
Game 69
Vaiser-Nalbandian

Erevan Open 1 996

1 d4 d 5 2 e4 e6 3 lLlc3 lLlf6 4 exd 5

Although White's position is a little

loose, his extra pawn will count in the
end.

exd 5 5 .itf4 lLle6 6 e3 .itg4 7 f3 2Ld7

20 . . . b6 21 lLld2 0-0 2 2 g5 h5 23

8 .itd3 e 6 9 g4

e3 a 5 24 a 3 2Lb7 2 5 .l:!.b1 a6 26
ge7 2Le6 27 .l:!.e1

lhe7 28 .itxe7

.it b 5 29 lLlg3 ge8 3 0 b2 g 6 3 1

lLlge4 a4 3 2 l:!e3 2Le6 3 3 .itg3 .it d 5
3 4 .l:!.xe8+ xe8 3 5 lLle3 .ite6 3 6 e4
a6

This aggressive thrust is V aiser' s

trademark: White advances on the
kingside, while Black counter-attacks
on the queenside.
9 . . . a6

9 . . . i, b4 is interesting, so that after

10 a3 i,aS 1 1 b4, the bishop returns to
c7 and supports the central break
. . . e6-eS .
1 0 g e 1 h 6 1 1 h4 ge8 1 2 lLlge2 lLla5

37 d5 .ite5+ 38 g2 exd 5 39 exd5

2Lxd 5 40 lLlde4 .itxe4 4 1 lLlxe4 iU8
42 d4 e2+ 43 .itf2 f5 44 gxf6
b5 45 lLlg 5 lLle5 46 e4 .ith6 47
f7+ 1 -0

1 3 f2 lLl e4 1 4 ge2 b 5 1 5 lLlb 1

2Le6 1 6 e 1 .ite7 1 7 b3 lLld7 1 8
bxe4 bxe4 1 9 2Lxe4 dxe4 20 .l:!.xe4

Finally, we take a look at the most

recent attempt from Black, 6 .. .cDe4.
1 13

Th e Sla v

This game brought the whole idea to

popular attention.

Game 70
Portisch-Kra m n i k

Biel lnterzonal 1993

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

The key to this concept. Black an

ticipates that the centre will open
with c3-c4 or e3-e4, and places his
bishop on an influential diagonal.
9 .lte2

An idea o f Iosif Dorfman.

Question 1 1 : What is the point of
the . . . lZJe4xc3 manoeuvre?
A nswer: With . . . lZJe4xc3 , Black
closes the c-file, blocking one of
White's major assets. Moreover, he
unbalances the pawn structure, isolat
ing the white a-pawn but giving
White the chance to play a pawn to c4
for the second time!
Question 12: What if White had
played one of the subtle move-orders,
such as 3 cxd5 cxd5 4 f4 lZJc6 5 e3
lZJf6 6 lZJc3?
A nswer: Well, 6 . . . lZJe4 is still possi
ble and leads to unclear play after 7
lZJxe4 (or 7 lZJge2! ? , intending f2-f3 to
recapture on c3 with a knight)
7 . . . dxe4, intending a quick . . "e7-e5 and
. . . b4+.
7 e3 ct:Jxc3 8 bxc3 g 6 !
1 14

9 c4 g7 10 cxd5 'i'xd5 is a posi

tion from the 4 g5 Griinfeld (1 d4
lZJf6 2 c4 g6 3 lZJc3 d5 4 g5 lZJe4 5
f4 lZJxc3 6 bxc3 g7 7 e3 c5 8 cxds
cxd4 9 cxd4 'i'xd5 , etc.) , which is fine
for Black, while Kramnik shows that
9 d3 g7 10 0-0 0-0 1 1 e4 g4 12 h3
xf3 1 3 'i'xf3 e5! 14 dxe5 (14 exdS
lZJxd4! 1 5 cxd4 exf4 equalises) 14 . . . d4!,
intending . . . lZJxe5, is also fine. Finally,
9 lZJe5 is dealt with in the next game.
9 . . . g7 1 0 0-0 0-0 1 1 c4 dxc4 1 2
xc4 f5

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

Kramnik gives 12 . . . g4 13 h3 xf3

14 xf3 es 1 5 dxes xes 16 xes
xes as equal: 17 xb7 xh2+ 1 8
xh2 h4+ 19 gl 'iVxc4 keeps the
balance. He also suggests 12 . . . a6!?,
intending ... b7-bs.
1 3 Rc 1 l;l c 8 1 4 e 2 a6 1 5 h 3 ?

1 5 ds bs 16 b3 leads to a slight
advantage for White according to
Kramnik.
1 5 . . . ttJa5 1 6 Ji.d3 Ji.xd3 1 7 xd3
d 7 1 8 .l::!. c 3 b 5 1 9 Rfc 1 ttJc4

A very impressive demonstration

of Black's strategy.
Game 71
Cifuentes-Leyva

Cien/uegos 1 996
1 d4 c5 2 c3 cxd4 3 cxd4 d 5 4 ttJf3
ttJc6 5 ttJc3 ttJf6 6 Ji.f4 ttJe4

After a weird transposition, we are

back to the main position.
7 e3 ttJxc3 8 bxc3 g 6 9 ttJe5 Ji. g 7 ?
20 e2 ttJb6 21 'f1.c7 e6 22 Ji. g 5
4Jd 5 ! 23 'f1. 7 c 5 h 6 2 4 Ji.h4 b4 ! 25
b2 ttJc3 !

I like the way in which Black has

moved his outpost from c4 to c3 !

A careless error. As Lalic has

shown, Black can equalise with
9 . . . 'iVas! 10 b3 g7 1 1 bs xes!
12 xes 0-0.
1 0 ttJxc6 bxc6 1 1 a4 !

26 .l::!. x c8 Rxc8 27 \t>h 1

27 'iVxb4 ctJe2+ wins.

27 . . . \t> h 7 ! 28 'f1.a 1

Or 28 'iVxb4 ctJxa2!
28 . . . a5 29 b3 xb3 30 axb3 g 5
3 1 Ji. g 3 a4
see

follo wing diagram

32 ttJd2 a3 33 l;l c 1 e5 34 d 5 a2 3 5
'f1.a 1 e 4 3 6 d 6 'f1. a 8 3 7 ttJc4 ttJb5 3 8
.Jt e 5 ttJxd6 39 l:!xa2 .l::!. x a2 4 0 Ji.xd6
gxf2 41
0-1

Ji.xb4 l:!f 1 + 42 \t>h2 l;lb 1

A very awkward move: Black

1 15

Th e Sla v

suddenly finds himself in big trouble.

This is a good illustration of what can
happen to Black if he does not take
enough care.

e8 22 VJlib3 b8 23 VJlia2 h8 24
.!:!:fd 1 VJlia8 25 h4 h 5 26 '!:!: d e 1 exd4
27 exd4 .!:!:e4 28 e2 h 7 29 J::t d 1
a7 3 0 iLe7 .!:!:a6 3 1 a 5 .!:!:xd4

1 1 . . . SL d 7 1 2 iLa6 ! e5 1 3 a3 exd4
1 4 exd4 0-0 1 5 0-0 iLe8

3 2 SLb6

Black has to shed a pawn to meet

the threat of iLb7, winning the ex
change.

Now the inevitable advance of the

white a-pawn swiftly proves to be
decisive.
32 . . . .!:!:xd 1 + 33 .!:!:xd 1 VJlib7 34 c5

1 6 iLxe8 .!:!:xe8 1 7 VJlixa7 f6 1 8 a4 e 5

a8 3 5 xd 5 VJlixd 5 36 xd 5 f5 37

1 9 SL g 3 a 8 20 VJlie5 a5 2 1 VJlie3

d8 .!:!:a6 38 d7 g8 39 a7 1 -0

1 16

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

S u m m ary

All three major continuations of the Exchange variation seem to be doing fine
for Black. Personally I would recommend either 6 . . . a6 or 6 . . . .iiJ5 , as 6 . . .liJe4
leads to positions that are more typical of the Griinfeld than the Slav.
1 d4 d5 2 c4 c6 3 cxd 5 cxd 5
4 11. f4 CLlc6 5 e3

5 liJf3 liJf6 6 liJc3

6 . . . a6 (D)
7 liJe5 game 66
7 c 1 game 67
6 . . . liJe4 7 e3 liJxc3 8 bxc3 g6 (D)
9 e2 game 70
9 liJe5 game 71
-

5 . . . CLlf6 6 CLlc3 1Lg4

6 . . . f5 game 65
-

7 f3 lL d 7 8 lL d 3 e6 (D)

9 g3 game 68
9 g4 game 69
-

6. . . a 6

. . .

g6

. . .

e6

1 17

CHA PTER TEN

Move-Orders an d
T ranspos itions

1 d4 d5 2 c4 c6

This chapter is extremely impor

tant both for white players, and for
black players who wish to play lines
with . . . d5xc4. Until recently, it was
thought that White had no need to be
too accurate with his move-order.
However, due to the efforts of Ivan
Sokolov, this is no longer true.
We shall first consider 3 e3 . With
this move White aims to remove the
force from . . . d5xc4 (White can simply
recapture with the bishop and has no
need to play a2-a4, preventing . . . b7b5) , and thus to prevent Black from
entering the Slav. Usually, the game
continues 3 . . .cbf6 4 e3 e6, leading to
the Semi-Slav, a very interesting open
ing but not the one we want to play.
The following game shows the way to
meet 3 e3 .

Game 72
K rasen kov-I . S o kolov

Malmo

1 995

1 d4 d 5 2 c4 c6 3 e3 jLf5 !

I n the introduction, I said that

. . . f5 can only be good for Black if
he can defend the b-pawn with his
queen. Sokolov noticed that after 4
1 18

cxd5 cxd5 5 iYb3 , 5 . . :Vllic 7 is possible

as 6 iYxd5 loses to 6 . . . iYxc 1+. 3 . . . ilfs
is not the end of the world for White,
but he has only very small chances of
gaining an advantage once Black has
developed his queen's bishop outside
the pawn chain, while White has shut
his inside.

4 b3 c7 5 cxd 5 cxd 5 6 CiJc3

6 ,b5+ is considered in the next

game.
6 . . . e6 7 jLd2 CiJc6 8 CiJf3 CiJf6 9 jLe2
a6 1 0 0-0 jLd6 1 1 gfc 1 0-0

Black has developed very naturally,

and has equalised comfortably.
1 2 a3 e7 1 3 jLe 1 h6

Ivan
Sokolov
suggests
that
13 . . . MadS, aiming for . . . e6-e5, was
most accurate.

M o v e - O r d e rs a n d Tr a n sp o s i tio n s

1 4 ctJa2 gac8 1 5 ctJb4 ctJa5 1 6 a2

iLc7 1 7 ctJe5 ctJe4 1 8 ctJbd3

attack continues despite Black's deficit

of a rook. However, White had to try
this as 25 iiLd3 ? , trying to deflect the
bishop on e4 from its protection of
the d5-pawn, simply loses a piece.
Sokolov finishes very efficiently.
2 5 . . :xd8 26 dxe5 ctJdc4 27 iLxe4
dxe4 28 .l:!.c 1 r;t>h8 29 b 1 ctJxe5 30
.l:!.xc8 xc8 3 1 xe4 ctJac6 32 h 3
ctJxb4 33 axb4 ctJ c 6 3 4 h4 d 8 3 5
h 5 e7 36 f 5 xb4 3 7 c8+ r;t> h 7
38 ,, 5 hxg5 39 f 5+ r;t>g8 40 e6+
r;t>f8 41 f5+ r;t>e8 42 g6+ r;t>d8 43
xg 7 g4+ 44 r;t>f 1 xh 5 4 5 x b7

Somewhere around here, White of

fered a draw, but Black was having
none of it!
1 8 . . .f6 1 9 iL b4 ctJd6 20 ctJf4 gfd 8 !
21

g4 iL e 4

22 .l:!. x c 7 ?

xc7

23

ctJxe6 e 7 2 4 ctJ x d 8 fxe 5

h 1 + 0-1

46 e2 t2Jd4+ wins the queen on

a8 . A very important game: Sokolov
won his last four games in the tour
nament to pip Krasenkov by half a
point!

Game 73
N ogueiras-I . Sokolov

Erevan Olympiad 1 996

1 d4 d 5 2 c4 c6 3 e3 2Lf5 4 b3
c7 5 cxd 5 cxd 5 6 2L b 5+ ctJd 7 ! ?

2 5 iL d 3 ?

After the natural 25 iiLxa5, winning

a piece, 25 . . . nc2 26 iiLdl h4 gives
Black a very dangerous attack: 27
iiLxc2 xg4+ 28 f1 g2+!
(28 . . . iiLxc2 29 xd5+!) 29 e2 g4+
30 e l gl+ leads to a draw by per
petual, while Ivan Sokolov also men
tions 29 . . . iiLxc2!? when, with threats
of . . . t2Je4 or . . . t2Jc4 and . . . g4+, the

7 ctJc3 e 6 8 2Ld2 ctJe7 ! ?

An interesting development o f the

knight.
1 19

Th e Sla v

9 J:rc 1 ctJc6 1 0 ctJa4 J::t c 8 1 1 ctJe2

square, away from the centre.

iLe 7 1 2 0-0 0-0 1 3 ctJf4 ctJdb8 ! ?

4 e4

There is no way that White is going

to get through on c6!
1 4 iL d 3 iLxd 3 1 5 ctJxd3 iLd6 1 6 93
e7 1 7 ctJ c 3 a 6 1 8 ctJa4 c7 1 9
Mc2 Y2 - Y2

Exchanges will follow on the open

c-file.
The next move that we shall exam
ine is 3 tLlc3, which has exactly the
same idea as 3 e3: after 3 ... CLJf6, 4 e3
would prevent . . . dSxc4 and lead to the
highly theoretical pastures of the
Semi-Slav after 4 ... e6. This was the
main move-order for several years
until that man Ivan Sokolov again got
cracking.

Game 74
LautieH .So kolov
Groningen 1995

The most natural continuation:

White gets in his desired e2-e4 as soon
as possible. If White plays 4 a4 in 0[
der to prevent . . . b7-b5, then Black can
exploit the fact that White has not
played CLJf3 by playing 4 ... e5! 5 e3 (5
dxe5 xd1 + is very nice for Black)
5 . . . exd4 6 exd4 e6! Instead 4 e3 b5 5
a4 b4 6 CLJe4!? is the critical test of this
idea, when 6 ... dS (not 6 . . . a6 7
CLJc5!; but 6 . . . CLJf6!? 7 CLJxf6 + exf6 8
xc4 d6 9 'Wic2 0-0 1 0 d3 g6 led
to a quick draw in Epishin-Sakaev, 5t
Petersburg 1997) 7 CLJg3 CLJf6 (7 ... e 5!?)
8 CLJf3 a6 (8 ... h5!?) is very unclear.
4 . . . b5 5 a4 b4 6 ctJa2 ctJf6 7 f3

Advancing with 7 e5 is possible, al

though 7 ... CLJdS 8 xc4 e6 9 CLJf3 ile7
10 d2 as 1 1 CLJc1 CLJdl 12 CLJb3 ilb7
13 'Wie2 c5! was absolutely fine for
Black in Yakovich-Sadler, European
Team Championship, Pula 1997.

1 d4 d 5 2 c4 c6 3 ctJc3 dxc4!

7 . . . e 5 8 iLxc4?

Once White has already committed

his knight to c3, the sequence . . . b7-b5,
a2-a4 (to regain the pawn) ... b5-b4
gains a tempo on the knight on c3,
which must then move to an inferior

White sacrifices a pawn, but he

must have missed something, as he
never gets anything for it. The correct
move is 8 dxe5, which is considered in
the next game.

1 20

M o v e - O r d e rs a n d Tra n sp o s i t io n s

ing path for White.

Game 75
Hjartarson-Gulko

Reykjavik 1 996
1 d4 d5 2 c4 c6 3 liJc3 dxc4 4 e4
b5 5 a4 b4 6 liJa2 liJf6 7 f3 e5 8
dxe 5 ! xd 1 + 9 xd 1 liJfd 7 1 0 e6 ! ?

8 . . .'Vxd4 9 c2

9 'i'b3 , attacking f7, is strongly met

by 9 . . . iilcS ! 10 iilxf7+ cJ;; e 7, when
Black's threat of . . . 'i'f2+ gives him an
overwhelming position.
9 . . . it c 5 1 0 it g 5 ita6 ! 1 1 itb3 ite7
1 2 liJe2 b6 1 3 liJg3 h 6 1 4 itd2 g6
1 5 c 1 liJfd 7 1 6 liJf 1 liJ c 5 !

Hjartarson suggests that 10 f4 iila6

1 1 iile3 f6 is unclear, but 10 . . . CLJcS
may be stronger, meeting 1 1 iilxc4 by
1 1 . . .iila6! 12 iilxa6 CLJbxa6 with ideas
of . . . CLJb3 and . . . CLJxa4 as well as
. . . CLJxe4.
1 0 . . . fxe6 1 1 itxc4

1 7 a 5 b7 1 8 liJxb4 liJxb3 1 9 xb3

White has the more attractive pawn

structure and hence the better long
term prospects, but Black's piece ac
tivity and the slightly open position
of the white king should give him
equally good chances.

c5!

1 1 . . . ita6

This wins a piece due to the pin on

the knight.

1 2 itxa6 liJxa6 1 3 ite3 it c 5 1 4 e2

White's position is truly horrible.

20 d 5 cxb4 2 1 xe 5 f6 0 - 1

White players will obviously not

want to repeat this experience. A re
cent game has shown a more interest-

Also possible was l 1 . . .CLJeS! ?

see

follo wing diagram

1 4 . . 0-0-0?
.

Hjartarson says that Black should

have taken this opportunity to swap
12 1

Th e Sla v

off bishops by 14 . . . ii,xe3 1 5 xe3

0-0-0 16 ctJh3 ctJe5 17 Mac 1 with mu
tual chances. In the game, Black fails
to make the most of his chances and
falls into an unpleasant ending.

4 1 e5 c6 42 f2 e4 43 ne2 ng4
44 e6 1 -0

Nice technique from Hjartarson,

and a critical new idea for White.
In view of Ivan Sokolov's success
with 3 . . . dxc4 against 3 ctJc3 , you may
be wondering whether Black can play
3 . . . dxc4 against 3 ctJf3 but, amongst
others, the following game has put the
line under a cloud. It is so unbalanced,
however, that I would not be at all
surprised if there is a sneaky resource
for Black!

Game 76
Miles-Hodgson

Hastings 1 995/96

1 5 .l1i. g 5 ! ctJf6 1 6 b3 .l1i.d4 1 7 n c 1

b7 1 8 ctJ h 3 e 5 1 9 hd 1 c5 20 .l1i.e3
.l1i.xe3 21 nxd8 xd8 22 xe3 h 6 23

1 ctJf3 d 5 2 d4 c6 3 c4 dxc4

ctJf2 gd6 24 gc2 ctJd7 25 ctJc 1 h5

3 ctJf3 ii,f5 4 cxd5 cxd5 5 'i'b3 'i'c7

is tactically possible as 6 'i'xd5 allows
mate after 6 . . . 'i'xc 1+. The crucial dif
ference with the 3 e3 line is that
White has not blocked in his dark
squared bishop, which means that
after 6 ctJc3 e6, 7 ii,f4! is extremely
nasty: 7 . . . 'i'xf4 loses the rook on as
after 8 'i'xb7, while 7 . . . 'i'b6 8 'i'xb6
axb6 9 e3 , intending ii,b5+, e2 and
then Mhc 1 to invade on the c-file,
gives Black a very depressing ending
to defend.

26 ctJfd3 g5 27 ctJ b 2 g4 28 ctJcd 3 !

gxf3 29 gxf3 g g 6 30 ctJc4!

4 e3 b 5 5 a4!

White's pieces are ideally placed,

and Black can no longer hold his
pawn weaknesses.
30 . . . ng 1

31

nb2 ctJc7 3 2 ctJcxe5

ctJxe5 3 3 ctJxe5 ne 1 + 34 f2 gh 1
3 5 g2 nc 1 36 f4 ctJe6 37 f5 ctJg5
38 f6 c7 3 9 f7 ctJ h 7 40 ctJg6 .!;i.e 1
1 22

see

fol/o wing diagram

The
and 3
vance
tempo
White
them.

main difference between 3 ctJc3

ctJf3 is that Black cannot ad
his queenside pawns with
against 3 ctJf3 . This gives
plenty of time to undermine

M o v e - O r d e rs a n d Tra n sp o s i tio n s

1 5 0-0 "IlIc7 16 f3 ! (16 ttJb3? ttJe4!,

aiming for c3 , is to be avoided at all
costs) 16 . . . a7 17 cS is very nice for
White.
1 2 . . . CLlfd7 1 3 c 5 ! 0-0 1 4 CLlec4!

5 . . . e6 6 axb5 cx b 5 7 b3 iL b4+ 8
iLd2 iLxd 2+ 9 CLlbxd 2 a5 1 0 bxc4 b4

A very confusing situation: Black

has two passed queenside pawns while
White has more central control. The
essential conflict is whether Black can
get his pawns moving or whether
White can blockade them so that they
will become weak.

ttJc6 20 0-0 Wifc7 21 Bfc 1 iLxc4 22

1 1 CLle 5 !

CLlxc4 CLlce7 23 b3 Ba7 24 g 3 !

Exchanges help Black free himself.

1 4 . . . CLlf6 1 5 iLe2 CLl d 5 1 6 iLf3 iLd7
1 7 c2 iLb5 1 8 CLld6 iLa6 1 9 CLl2c4

R b 8 2 5 g 2 !

Very instructive: Black's pawns are

going nowhere, so White quietly im
proves his position, removing the
possibility of back-rank mates and
making sure that if Black does queen,
it won't be with check.
25 . . . h6 26 Rc2 CL'lc3?

Threatening 12 ttJxf7! 'it>xf7 1 3

"Wf3+, winning the rook o n a8 .
1 1 . . . CLlf6 1 2 a4+ !

This is the key idea, forcing Black's

pieces into a nasty tangle. 12 . . . ttJbd7
loses a pawn to 13 ttJc6, while
12 . . . d7 1 3 ttJxd7 is exactly what
White wants: 13 . . . ttJbxd7 14 e2 0-0
1 23

Th e Sla v

Allowing a combination, but Black

was struggling to find anything to do.

b3 0-0 1 2 e4?

27 nxa 5 ! xa 5 28 tLlxa5 xa 5 29
xc3 bxc3 3 0 xb8+ \t>h7 3 1 b7
f5 3 2 xe 7 c 2 3 3 xe6 c 1 34
xf5+ g 6 35 f7+ \t>h8 3 6 f8+
\t>h7 37 .li d 5 !

Mate cannot be averted.

3 7 . . . h 1 + ! ? 3 8 xh 1 1 -0

The next line is 3 tLlf3 tLlf6 4 e3. I

am surprised at how popular this
variation is, as it really doesn't prom
ise White very much at all, and his
results have not really been that good.

Game 77
Kozu l - l il escas

Erevan Olympiad 1 996

1 2 . . . tLlc5 ! ! 1 3 dxc5 dxe4 ! 1 4 xd8

fxd8

The point: Black regains his piece

with a vastly superior position, as the
knight on f3 is tied to the defence of
the bishop on d2.
1 5 tLla4 exf3 1 6 fd 1 d3 1 7 itc3

1 d4 d 5 2 c4 c6 3 tLlf3 tLlf6 4 e3

:ad8 1 8 ne1

.lif 5 !

20 tLlb2 f3 21 \t>xg2 z:i.f5 2 2 t2Ja4

I don't think that this needs any

comment!

d3 23 .l:!.cd 1 .l:i:df3 24 e2 h 5 ! 25

:8d7 1 9 :ac 1 fxg2

h3 g 5 !

This kingside advance finishes off

the game.
26 b4 axb4 27 axb4 g4 28 hxg4
tLlxg4

5 .lid 3

5 cxd5 cxd5 6 iVb3 is the other way

to play, leading to positions very simi
lar to Game 72.
5 . . . .lixd3 6 xd3 e 6 7 0-0 tLlbd7 8
tLlc3 1I.. b 4 9 1I.. d 2 a5 1 0 a3 .lie7 1 1
1 24

Now f2 must fall.

29 .l:i:h 1

:xf2+ 30 nxf2 tLlxf2 3 1

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

J:!.a 1 h4 3 2 tZ:lb6 h3+ 33 h2 J::!. g 5 34

J:!.g 1 .l:!.xg 1

3 5 xg 1 tZ:lg4 36 tZ:lc8

f8 0 - 1

And finally, 4 'iVb3 and 4 'iVc2. In

both case White's queen protects c4
and prevents the light-squared bishop
from developing safely: 4 'iVb3 attacks
b7, while 4 'iVc2 covers the f5-square.
However, these moves do nothing to
further White's development, while
exposing the white queen to attack by
Black's minor pieces. Although nei
ther line promises much, they are
both popular with positional players
who wish to avoid any sharp options.
Recently, black players have been
trying a Semi-Slav approach, 4 . . . e6 5
g3 dxc4 6 'iVxc4 b5 7 'iVc2 j,b7, with
. . .tDbd7 and . . . c6-c5 to follow, but as
befits a book on the Slav, I will con
centrate on our beloved plan of de
veloping the light-squared bishop out
side the pawn chain.

1 1 'iVxb4 attacks b7 and d4) 1 1 'iVxb4

e6 12 'iVc3 l:!.c8 ! , winning the queen, as
13 'iVd3 allows 1 3 . . . l:!.xc1+ mate! Instead of 6 . . . cxd5, 6 . . . CtJxd5! ? (hitting
the bishop on f4) 7 j,g3 e6!?, intend
ing a quick . . . c6-c5, is also interesting.
Note that 5 CtJc3 transposes to 4 CtJf3
a6 5 'iVb3 .
5 xc4 .itf5 6 tZ:lc3 tZ:lbd7 7 g 3 e6 8
.itg2 .ite7 9 0-0 0-0 1 0 e 3

Be warned! The 'natural' 10 l:!.d1

loses the exchange to 10 ... j,c2!, as 1 1
l:!.d2 allows 1 1 . . .CtJb6! , trapping the
queen!

Game 78
Akopian-Shirov

Wijk aan Zee Open 1993

1 d4 d 5 2 tZ:lf3 tZ:lf6 3 c4 c6 4 b3
dxc4

The most solid and reliable option.

Black diverts the queen from its attack
on b7, thereby allowing the light
squared bishop to develop to f5. I
suppose that 4 . . . a6 is possible here, as
5 j,f4 (preventing . . . l:!.a7!) 5 . . . b5 6
cxd5 cxd5 7 a4 b4! 8 CtJbd2 (8 'iVxb4
e5! wins a piece) 8 . . . CtJc6 9 l:!.c 1 j,b7
10 CtJe5 (to remove the knight on c6,
which both guards b4 and blocks the
c-file) is met by 10 . . . CtJa5! (10 . . . CtJxd4

1 0 . . . tZ:le4 1 1 e2

A novelty. 1 1 CtJd2 is the old move,

when theory recommends 1 1 . . .CtJxd2
12 j,xd2 e5! 13 e4 cxd4 14 exf5 dxc3
15 j,xc3 j,f6! 16 l:!.ad 1 ? ! j,xc3 1 7
bxc3 'iVc7 with equality. White must
consider 16 j,b4!?, retaining the
bishop pair.
1 1 . . . tZ:lxc3 1 2 bxc3 .ite4 !

A typical manoeuvre, preventing

e3-e4 and neutralising White's bishop
on g2.
1 3 c4 c5 1 4 ld:.d 1 wtic7
see

fol/o wing diagram

1 5 .itb2 tZ:l b6 ! ?
1 25

Th e Sla v

White i s material up, after all!

Game 79
Razuvaev-Sturua

Erevan Open 1 996

1 d4 ltJf6 2 c4 c6 3 ltJf3 d 5 4 c2
g6

The start of an eccentric plan from

Shirov. 1 5 . . . MfdS , intending 16 . . . MacS
and then . . . ct:Jd7-bs-c6 to pressurise c4
and d4, was also interesting.
1 6 gac 1 jfLf6 1 7 jfLf 1 !

To chase the bishop from e4 with

out allowing the exchange of bishops.
1 7 . . . ltJa4 ! ?

1 8 jfLa 1

.l::i. f d8 1 9 ltJd2

A very sensible idea. Black allows

his bishop to develop to g7 whilst
supporting . . . cS-f5 , attacking the
queen. 4 . . . a6 5 f4 b5?! 6 cxb5 cxb5 7
ct:Jbd2! (intending ct:Jb3) is best
avoided, as White's pieces are well
placed to exploit the queenside dark
square weaknesses, so Black should
try 5 . . . dxc4 6 'iVxc4 e6 7 e3 b5 S 'iVc2
b7, intending . . . ct:Jbs-d7 and . . . c6-c5.
Finally, 4 . . . dxc4 5 'iVxc4 transposes to
the previous game.

jfLc6 20 ltJ b 3 b 6 21 d5

5 f4 g7
21 . . . jfLxa 1

2 2 dxc6 jfLf6

23 jfLg2

ltJc3 24 .l::i. x d8+ gxd8 2 5 c2 gd6

26 .!de 1

gxc6 27 jfLxc6 xc6 28

ltJd2 a6 29 ltJ b 1 ltJe4 30 h4 b 5 3 1

cx b 5 axb5 3 2 ltJ d 2 ltJc3 3 3 ltJ b 1

%-%
I am a little surprised that White
accepted the draw here. Although his
opponent has good counterplay,

ltJe4 3 4 ltJ d 2 ltJc3

126

5 . . . ct:Ja6!? is considered in the next

game and 5 . . . dxc4 is also good: 6
'iVxc4 g7 7 e3 (7 ct:Jc3 0-0 S e4 b5
leads to a typical Griinfeld position
which, though reasonable for Black,
may not appeal to pure Slav players)
7 . . . 0-0 S e2 e6 9 iVc1 ct:Jbd7 10 0-0
c5! 1 1 ct:Jc3 ct:Jd5 ! 12 Mdl MCS 13 ct:Jxd5
xd5 14 dxc5 ct:Jxc5 15 c4 ct:Jd3 ! 16

M o v e - O r d e rs a n d Tra n sp o s i tio n s

g,xd3 g,xc4 1 7 'iVd2 e6 gave Black no

problems in Goldin-Yusupov, Tilburg
1992. However, s . . . iiJs 6 'iVb3 'iVb6 7
cS 'iVxb3 S axb3 is a touch better for
White, and not very exciting for
Black.

axb3 are improved for Black, as the

white knight is misplaced on gS,
which makes . . . e7-eS easier to achieve.

6 e3 0-0 7 tLlc3 ilLe6

And finally, an amazing sacrificial

line in this most solid of openings!

1 1 . . . ilLg4 1 2 cxd 5 tLlxd 5 1 3 tLlxd 5

cxd5 1 4 xb6 tLlxb6 Y2 - Y2

Game 80
Alburt-S habalov

USA

Championship 1996

1 d4 d 5 2 c4 c6 3 tLlf3 tLlf6 4 c2
g 6 5 ilLf4 tLla6 ! ? 6 e3 ilLf5 7 b3
tLlb4 ! !

S tLlg 5

S 'iVb3 dxc4 9 xc4 (9 'iVxb7 ctJdS ! ? ,

intending to trap the queen i n the
corner after 10 'iVxaS 'iVb6, is ex
tremely murky but not worse for
Black) 9 . . . xc4 10 'Vi'xc4 ctJds 1 1 g3
ctJa6, intending . . . ctJdS-b6 and . . . c6-cS
is about equal.
S . . . ilLf5 9 b3 b6 1 0 ilLe2 tLlbd7
1 1 tLlf3

A typically inventive idea of Ku

preichik.
S xb4 e 5 9 x b7

9 cS cxf4 10 exf4 b6 1 1 ctJeS bxcs 12

'iVb7 d7 13 ctJxd7 ctJxd7 14 'iVxc6
g,cS 15 'iVxds cxd4 16 bs b4+
(16 . . . g,c1+ 17 d2 g,xh 1 1S 'Vi'eS+!
'iV e7 19 'Vi'xhS+ 'Vi'fs 20 'Vi' eS+ leads to
a draw by repetition) gave Black a
powerful lmtlatlve in Epishin
Kupreichik, Russia 1989.
9 . g b S 1 0 xc6+ ilLd7
. .

11

xf 6 !

xf6 1 2 ilL x e 5 b 6 1 3 b3 ilL b4+ 1 4

The endings with 1 1 cS 'iVxb3 12

tLlbd2 0-0 1 5 ilLxbS J::!. x bS

127

Th e Sla v

wmmng.
1 6 . . . 'ii' a 5 1 7 c4 g4 1 8 0-0 xd2

White has some pawns, but Black

has the big guys!
1 9 CLJe5 f5 20 a3 'ii' c 3 21 d 6 g7
22 CLJxf7 .l::!. b 6 23 e4 c8 24 .l:i:fd 1
f4 25 e 5

1 6 cxd 5

A strange move to play since

Shabalov had already won a convinc
ing game in this line. 16 i,d3 is the
theoretical recommendation, meeting
16 . . . i,g4 with
17
e2 ik'a5
(threatening . . . i,b4xd2)
18
hd l ,
which i s a bit of a mess. 17 O-O!? seems
more natural, to counter 17 . . . ik'a5
with 18 h3 ! i,h5 (18 . . . i,xd2 19 hxg4
i,c3 20 ac 1 i,b2 2 1 c2 ik'xa2 22
4::l d2! [intending b l] 22 . . . xb3 23
4::lx b3 ik'xb3 24 b l ik'xd3 25 cxb2
dxc4 26 c 1 , intending bc2 with an
advantage) 19 a3 ! i,xd2 20 b4! i,xb4
21 axb4 ik'xb4 22 ab l ik'f8 23 g4! ,

1 28

25 . . Jbb3 26 x b 3 'ii'x b3 27 CLJd8

'ii' b 6 28 g 3 'ii' x d8 29 g xf4 'ii' h 4 30
f3 'ii'x f4 3 1 f2 b7 3 2 .l:i:d3 a6
3 3 d7 'ii'x h2+ 34 e3 'ii' h 6+ 35 e2
'ii' h 4 3 6 d2 f7 37 d5 e7 38 e6
'ii'f 2+ 39 c3 'ii' c 5+ 40 d2 xd3
41 .l::!. c 1 'ii' d 4 42 .l::!. c 8 b 5+ 43 c2
xd7 0- 1

A fascinating game.

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

S u m mary

3 e3 JLfs equalises for Black; 3 CDc3 dxc4 is a very interesting line and only
Hjartarson-Gulko (Game 7S) is a possible attempt to play with White; 3 CDf3
dxc4 is very unbalanced but seems to be good for White; 3 CDf3 CDf6 4 e3 JLfs
is nothing for White; while 3 CDf3 CDf6 4 'iVc2 and 4 'iVb3 are also nothing spe
cial. Hence 3 CDf3 CDf6 4 CDc3 dxc4 is the most accurate order for both sides.
1 d4 d 5 2 c4 c6
3 e3

3 CDc3 dxc4 4 e4 bs S a4 b4 6 CDa2 CDf6 7 f3 eS (D)

8 JLxc4 game 74
8 dxeS game 75
3 CDf3
3 . . . dxc4 game 76
3 . . . CDf6
4 e3 JLfs game 77
4 'iVb3 game 78
4 'iVc2 g6 S JLf4 (D)
S . . . JLg7 game 79
S . . . CDa6 game 80
-

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

6 CDc3 game 72
6 JLbS + game 73
-

. . .

e5

5 f4

5 . . . cxd5

1 29

CHA PTER ELEVEN

Odd s and Ends

1 d4 d5 2 c4 c6

In this chapter, we take a brief look

at all the sidelines that I couldn't fit
into the main chapters!
A)

T h e Winawer Cou nter

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

This gambit was all the rage four or

five years ago, but the following game
somewhat dampened the ardour of
the black players.

Game 81
. Kasparov-N ikolic

Manila Olympiad 1 992

7 . . . \'l;Vd 6

Black has also tried 7 . . . cS, but after

8 CL'lf3 Vic7 9 e3 dxe3 1 0 xe3 CL'lf6 1 1
d3 White i s slightly better due t o his
lead in development.

1 d4 d 5 2 c4 c 6 3 ltJc3 e 5 4 dxe5

8 ltJf3 ltJf6 9 \'l;Vc2 e7 1 0 O-O-O!

d4 5 ltJe4 \'l;Va5+ 6 d 2 !

0-0

This natural move had been practi

cally ignored prior to this game (6
CL'ld2 had been the main line) .
6 . . . . \'l;Vxe 5 7 ltJg 3 !

This move looks obvious, but it

was Kasparov's new idea. Of course,
the genius is not in the move itself,
but in the astonishing attacking plan
that flows from it. It is not easy to
guess that White is gearing up for
kings ide attack!
1 30

O d ds a n d En ds

1 1 e3

Azmaiparashvili, a long-time ana

lyst for Kasparov, played 1 1 jLc3
agamst Eslon in Seville 1994.
1 1 . . .dxc3 !? 12 xd6 cxb2+ is given as
unclear by ' Gazza' , but 'Azmai'
claims an advantage after 1 3 'iVxb2
jLxd6 14 e4 (threatening e4-e5)
14 . . . jLf4+ 15 'It>b 1 . This assessment is
objectively correct, as White's mate
rial advantage, coupled with the threat
of e4-e5, driving away the knight on
f6 and exposing the bishop on f4 to
attack by White's pieces ('iVb2-d4,
ctJg3-h5) , should tell in the end. How
ever, in a practical game White's dark
square weaknesses and exposed king
give Black definite counter-chances.
Eslon played 1 1 . . .'iVf4+, unpinning,
and White sacrificed a pawn with 12
e3 dxe3 13 fxe3 (13 d4!? looks very
strong, meeting 13 . . . 'iVh6 with 14
h4! 'iVg6 1 5 jLd3 ! , trapping the
queen, and 13 . . . 'iVc7 with 14 jLd3 ! ? or
just 14 fxe3 with very dangerous at
tacking play) 1 3 . . . 'iVxe3+ 14 'It>b 1 ctJa6
15 ctJd4 (aiming to put a knight on f5
and threatening de 1 , winning the
bishop on e7) .

17 ctJxf5 'iVe6 18 jLd3 White had a

superb attacking position for the
pawn. This looks very smooth and, in
his annotations, Azmaiparashvili
comments that if Black tries to im
prove with 15 . . . ctJb4 (instead of
. . . jLb4) then 16 'iVb3 c5 17 e 1 wins
after 17 . . . 'iVgS 18 xe7 cxd4 19 jLxb4,
but in fact 16 . . . jLcS! is extremely good
for Black, as 17 jLxb4 'iVxb3 18 ctJxb3
jLxb4 and 17 e 1 'iVf4! (threatening
. . .jLxd4) 18 jLxb4 'iVxd4 both leave
Black simply a pawn up . The position
is extremely risky for Black, of
course, and I would not recommend
this sort of position against a good
attacking player, but Black's tactical
chances must not be underestimated.
1 1 . . . dxe3 1 2 fxe 3 !

This is the key to the whole idea as

12 jLxe3 'iVc7 promises nothing for
White. The text allows the bishop on
d2 to move with tempo to the attack
ing diagonal a1-h8.
1 2 . . .'c7

12 . . . ctJa6 is considered in the next

game.
1 3 SLc3 SLg4? !

1 4 SLd3 tLlbd7 1 5

SLf5 !

Now after 1 5 . . . jLb4 16 ctJdf5 jLxf5

Naturally, White wants to occupy

131

Th e Sla v

the f5-square with his knight.

1 5 . . . 1l.xf5 1 6 CLlxf5 ZUeS 1 7 CLlxg7 ! !
xg7 1 S f5 CLlfS 1 9 h4!

If Black wishes to play this line, he

needs to find an improvement on
12 . . . c7.

Game 82
Rogozenko-Bets

Moldovan Championship 1 994

1 d4 d 5 2 c4 c6 3 CLlc3 e 5 4 dxe5
d4 5 CLle4 a5+ 6 1l.d2 xe 5 7 CLlg3
CLlf6 S CLlf3 d 6 9 c2 1l.e7

10

0-0-0 0-0 1 1 e3 dxe3 1 2 fxe3

Following Kasparov's example.

In order to chase away the knight
on fs if it should surface on g6.
1 9 . . . h 6 20 g4? !

20 g4! iDg6 2 1 h5 was even

stronger according to Kasparov.
20 . . :cS 21 xcS .!:!:axcS 22 g 5 !

1 2 . . . CLla6

12 . . . .sdS is possible, preventing 13

Jtc3 as 13 . . . xd1+ 14 xd1 .sxd1+ 15
'ii;>x d1 exchanges queens, destroying
White's attacking possibilities.
1 3 1l.c3 c7

The ending i s still very unpleasant

for Black, and Kasparov powers
through with his customary energy.

Since this turns out so badly, Black

must consider 1 3 . . . e6, keeping the
queen close to the kingside to help
with defence, while threatening
. . . iDb4 and . . . iDg4 as well as . . . xe3+.

22 . . . CLlSd7 23 e4 cdS 24 d f 1 fS

14 a 3 !

25 gxf6 1l.xf6 26 e5 1l. g 7 27 .i:i.hg 1

c5 2S c2 .!:!:e6 29 .!:!:g4 1l.hS 30 b4

see

follo wing diagram

.!:!:b2 1l. g 7 34 b7 .i:i.xa2+ 3 5 b 3

Preventing 14 . . . iDb4, activating the

knight.

1;!a6 3 6 e6 .l::!. x e6 3 7 .!:!: x g 7 1 -0

1 4 . . . CLlg4 ! ? 1 5 '!:!:e 1 1l.f6

b6 3 1 bxc 5 bxc5 32 .!:!:b 1 a6 3 3

1 32

O dds a n d En ds

A nswer: This is another Smyslov

favourite. Black accepts a slight space
disadvantage and develops his king's
bishop on g7, avoiding . . .e7-e6 in or
der to retain the option of developing
his bishop on c8 outside the pawn
chain if he wishes.

Game 83
G u l ko-Salov

Reykjavik (World Cup) 1991

This manoeuvre, exchanging the
dangerous bishop on c3, briefly gave
black players hope in this line.
1 6 h 3 ! jLxc3 1 7 hxg4

Obviously 17 . . . .txe l loses to 1 8

iYxh7+ mate.

1 l2Jf3 d 5 2 c4 c6 3 e3 l2Jf6 4 l2Jc3

g 6 5 d4 jLg7 6 jLe2 0-0 7 0-0 b6

A solid move, developing the

bishop to b7 to support Black's cen
tre. The more active 7 . . . dxc4 8 .txc4
.tg4 is dealt with in the next game.

1 7 . . . h 6 1 8 xc3 xg3 1 9 g 5 ! hxg 5

20 jL d 3 d8 2 1 l:!.h7 f6 22 l:Ieh 1
c7 23 l2Jxg 5 f8 24 c 5 ! jLg4 25
xf6+! 1 -0

8 cxd 5 cxd 5 9 l2Je 5 !

25 . . . gxf6 26 h8+ cJi;e7 27 lh7+ is

mate!
B)

The S c h leeter S l a v :
3 CD c 3 CD f 6 4 e3 9 6

Question 1 : What is the idea behind

this system?

When the central pawn structure

becomes fixed, the first side to gain a
central space advantage will be able to
claim the initiative.
9 . . . jLb7 1 0 jLd2 l2Jfd 7 ? !

10 . . . CDc6 was stronger, meeting 1 1

f4 with 1 1 . . .CDe8 ! , intending to de
velop the knight to d6, when Black
only stands a little worse.
1 1 f4 ! f6 1 2 l2Jf3 ! ?

1 2 CDd3 i s also possible. Black has

1 33

Th e Sla v

not developed his pieces harmoni

ously: the king's knight stops the
queen's knight from developing to d7
and has no moves of its own.

iLxe7+

CL:Jxe7

34

W'h6+

CL:Jg6

35

W'xg6+ e 7 36 W' h 7 + d6 3 7 CL:Jf7+

rt;e7 3 8 iLf5 1 -0

Game 84

1 2 . . JH7 1 3 iLd3 CL:Jf8

D reev-Piket

Dortmund 1994
1 d4 CL:Jf6 2 e4 e6 3 CL:Jf3 d 5 4 e3 g6
5 CL:Je3 iLg7 6 iLe2

White can try 6 d3 (preventing

. . . f5) 6 . . 0-0 7 h3 (preventing . . . il,g4)
if he wants to prevent the plan in the
game, although 7 . . . c5 is an interesting
reply. The game transposes to a quiet
variation of the Griinfeld, in which
White has played the useful, though
hardly earth-shattering, extra move
h2-h3 .
.

1 4 f5? !

A little hasty according to Gulko,

who prefers 14 g4! e6 15 f5 .
1 4 . . . gxf5 1 5 iLxf5 e6 1 6 iLd3 CL:Je6
1 7 CL:Je2 d6 1 8 CL:Jg3 CL:Jg6 1 9 CL:Jh5
iLh8 20 W'e2 gaf8 21 a 3 e 5 22 dxe5

6 . . . 0-0 7 0-0 dxe4 ! ? 8 jLxe4 iLg4

This is a much more active idea:

Black will follow up with . . . 4Jbd7 and
a quick . . . e7-e5.

fxe 5 23 CL:Jg5 gxf 1 + 24 .i:l:.xf 1 iLe8?

2 5 .i:l:.xf8+ W'xf8 26 iLe4!

9 h 3 iLxf3 1 0 W'xf3 CL:Jbd7 1 1 iLb3

A very nice move: 26 . . . dxe4 loses to

27 "ik'c4+.
26 . . . CL:Jee7 2 7 iLb4!

iLb7 28 W'g4

\i"c8 29 W'f3 f8 3 0 h3 e8 3 1
CL:Jf6+
1 34

g7

32

W'xh 7+

xf6

33

The critical line is 1 1 gd1 e5 12 d5

e4!? 13 4Jxe4 4Jxe4 14 "ik'xe4 4Jb6 15
b3 (threatening d5xc6; Bareev sug
gests 15 gb 1 !?, protecting b2 and seek
ing to avoid the time-loss with c4b3xd5 as in the game) 15 . . . cxd5 16

O dds a n d En ds

,xds CiJxds 17 Mxds 'i'b6.

ity and gains space.

1 8 'fie2 tLl b 6 ! 1 9 gxd8 gxd8 20 ge 1
tLlc4 !

Imperceptibly, White has drifted

into big trouble: Black controls the
central dark squares and White's
queenside is an easy target.

Question 2: What is going on here?

A nswer: Black has sacrificed a pawn

for a lead in development and pres

sure against the b2-pawn. However,
this play is sufficient to regain the
pawn, but not to achieve complete
equality: i s 'i'd3 MadS 19 e4! itd4 20
Wh l itxf2 2 1 ith6 Mxds 22 exdS Mds
23 Mdl 'i'd6 24 'i'c3 , intending 'i'g7+
mate, when White's passed d-pawn
and Black's weak kingside gave White
a slight advantage in Bareev-Kramnik,
Novgorod 1994.

2 1 jLc 1 tLle6 22 'fif 1 'fic5 23 jLb3

tLld2 24 jLxd2 gxd2 2 5 jLxe6 fxe6
26 e 5 gxb2 27 tLle4 'fixe 5 28 'fid3
IWd 5 29 IWg3 jLe5 3 0 f4 jLd4+ 3 1
h2 c 5 3 2 \\!,Vh4 gxa2 3 3 \\!'ve7 h 5 !

1 1 . . . e 5 1 2 gd 1 'fi e 7 1 3 e4? ! exd4

1 4 gxd4 gad8 1 5 jLe3 tLlc5 1 6 jLc2
tLlfd7 1 7 gdd 1 b 5 !

3 4 \\!'ve8+ g 7 3 5 \\!'ve7+ h 6 3 6 h4
IWf5 3 7 tLlg3 \\!'vxf4 38 gxe6 g a 1

Here White lost on time.

Black activates his queenside major-

What I like about the Schleeter Slav

is its flexibility: on move 7, Black has
1 35

Th e Sla v

a huge range of plans. We have seen:

a) 7 . . . b6 reinforcing the centre by
fianchettoing the other bishop; and
b) 7 . . . dxc4 8 i1Lxc4 i1Lg4 to break
quickly in the centre by means of
. . . tLlbd7, . . . i1Lxf3 and . . . e7-e5; but
Black can also try
c) 7 . . . a6 to expand on the queenside
with . . . b7-b5, before or after . . . d5xc4;
and
d) 7 . . . tLle4!? to unbalance the pawn
structure with . . . tLlxc3 . Bates-Sadler,
British Championship 1995, contin
ued 8 'iVb3 b6 9 tLlxe4 (9 cxd5 tLlxc3 !
10 bxc3 cxd5 intending . . . tLlc6-a5, hit
ting the queen) 9 . . . dxe4 10 tLld2 f5 1 1
f3 e5 1 2 dxe5 exf3 ! 1 3 tLlxf3 tLld7, re
gaining the pawn with a better pawn
structure;
. . . and the list continues! There
really is huge scope for personal ideas.
C)

3 tLlf3 tLlf6 4 tLlc3 dxc4

5 e3

5 . . . b5 6 a4 b4 7 CLJb 1

7 tLla2, attacking the b4-pawn, re

gains the pawn by force, but seriously
misplaces the knight. The text is more
ambitious.
7 . . . 1La6 8 1Le2

8 tLlbd2 regains the pawn, but after

8 . . . c3 9 bxc3 i1Lxf1 10 tLlxfl bxc3
Black will break with . . . c6-c5 and
equalise.
8 . . . c 5 ! 9 0-0 CLJc6 1 0 dxc5 CLJ a 5 ! 1 1
CLJbd2 e6 1 2 b3 1Lxc5 1 3 1Lb2 c3 1 4
1Lxa6 cxb2 1 5 1L b 5+ cI;e7

This was one of my favourites when I

was little, but I'm not quite sure why
I decided to play it against Bareev! It is
actually not a very promlSlng con
tinuation.

Game 85
Sadler-Bareev

Hastings 1992/93
1 d4 d 5 2 c4 c6 3 CLJf3 CLJf6 4 CLJc3
dxc4 5 e3
see

fol/o wing diagram

Trying to manage without the

standard 5 a4, which prevents the ad
vance . . . b7-b5.
1 36

The king is very safe in the centre.

Bareev rapidly outplays me, but
around the time control begins to ask
too much of his slight advantage.
1 6 .i:t b 1 a6 1 7 1Le2 b6 1 8 .i:txb2

O d ds a n d En ds

l2Jd 5 1 9 l2Jc4 l2Jxc4 20 .ltxc4 l2Jc3

21 c2 c7 2 2 c 1 Ithd8 23 .l::i. d 2
a 5 2 4 c 2 g 6 2 5 .l::i. c 1 .ltd6 26 g 3
.ite5 27 .itf 1 h 5 28 .it c 4 c 5 29 h 4
.lt f 6 30 .itf 1 l:! a c 8 3 1 .l::i. x d8 l:!xd8 3 2
l2Jd2 .l::i. c 8 3 3 l2J c 4 g 5 34 hxg 5 xg 5

As far as I am concerned, 5 e4 just

loses a pawn, but some die-hards just
keep on playing it! Foremost amongst
them is the attacking Swedish player
Tiger Hillarp-Persson. So here is one
of his typical efforts.

3 5 h 7 h 8 3 6 d 3 .i:!. d 8 3 7 h7

Game

.l::i. h 8 3 8 d3 c5 39 J:!c2 l:!d8 40

86

h7 h4 41 gxh4 Rh8 42 d3 lad8

Hillarp-Persson - Acs

43 h 7 .l::i. h 8 44 d 3 .i:!.g8+ 45 .ltg2

Budapest 1996

.l::i. d 8 46 h7 .l::i. h 8 47 d3 :ad8 48

h7 l2Jxa4?

1 c4 c6 2 l2Jf3 l2Jf6 3 d4 d 5 4 l2Jc3

Too risky. After the tactics, Black

only just manages to hold on to the
draw.

l2Jg 5 ! ?

dxc4 5 e4 b 5 6 e 5 l2Jd 5 7 a4 e6 8

The modern line. 8 axb5 tL'lxc3 9

bxc3 cxb5 10 tL'lg5 .i,b7 1 1 'i'h5 g6 1 2
'i'g4 .i,e7 i s the old continuation,
when Black will follow up with
. . . tL'ld7-b6-d5, while White plays for
tricks!

49 .l::i. a 2 ! l2J b 2 50 l2Jxa5 c 1 + 5 1

c;t>h2 lah8 5 2 l2Jc6+ xc6 5 3 l:!a7+
c;t>d6 54 xh 8 .itxh8 5 5 .ltxc6 c;t>xc6
5 6 .l::i. x f7 l2Jd3 5 7 h5 l2Jc5 58 h6
l2Jxb3 5 9 Itf8 .ite5+ 60 f4 .itb2 6 1
e4 l2Jc5 6 2 e 5 l2Jd7 6 3 h 7 l2Jxe5 64

8 . . . .lte7 9 h4 ! ? h6 1 0 l2Jge4 b4

fxe 5 .itxe5+ 65

10 . . . .i,a6 immediately is interesting,

to avoid weakening the queenside too
early. The text leads to absolute
chaos, though I think that Black is
fine.

c;t>g2 b3 66 h 8

.itxh8 67 J:!xh8 c;t>c5 68 c;t>f2 c;t>c4 69

c;t>e2 b 2 7 0 lab8 c;t>c3 71

c;t>d 1

e5

YZ - YZ
D)

T h e S l a v G a m bit :
3 tL'lf3 tL'lf6 4 tL'lc3 dxc4
5 e4 b 5 6 e 5 tL'ld5 7 a4 e6

1 1 l2J b 1 .lta6 1 2 g4 g6 1 3 l2Jbd2

c3 14 l2Jc4 .itxc4 1 5 .itxc4 a 5 1 6 h 5
g 5 1 7 0-0 l2Jd7 1 8 bxc3 l2J7b6 1 9
.ltb3 l2Jxc3 20 l2Jxc3 bxc3 2 1 e4
137

Th e Sla v

tiJ d 5 22 f4 gb8 23 ll.a2 gb4 24

fxg 5 b6 ! 25 ll.xd 5 exd 5 26 f3
IWxd4+ 27 ll.e3 xe 5 28 IWxf7+

Here 6 . . . c8 !? 7 tLl3 takes us back to

square one, but 7 e3 e5 8 ilxc4 exd4 9
exd4 ile7 10 0-0 0-0 1 1 h3 tLla6!? 12
ct'lf3 ct'lc7!, intending ... ile6, leads to
an unclear position. In the next game
we see Black allowing the exchange
on f5.

Game 87
Savchenko-N inov

Cappelle La Grande Open 1 994

1 d4 d 5 2 e4 e6 3 tiJf3 tiJf6 4 tiJe3
dxe4 5 a4 .liLf5 6 tiJh4 e6 7 tiJxf5
exf5 8 e3 .liLd6
28 . . . d8

29 gae 1

l:!e4 30

b6+

d 7 3 1 l:!xe4 xe4 3 2 ll.e5 J:!. h 7 33

f2 ll.xg 5 34 J:!.e1 f 5 3 5 e2 e5
3 6 b5+ e8 3 7 e6+ ge7 38
IWa8+ d 7 39 IWxd 5+ e8 40 a8+
d7 41 gd 1 + ll.d2 42 d5+ e8
43 gf 1 .liLf4 44 g8+ d7 4 5 d5+

1h - 1h

An amazing game!
E)

3 tLl f 3 tLl f 6 4 tLl c 3 dxc4

5 a4 f5 6 tLlh4 ! ?

This i s an interesting little idea that

has even been tried by Kasparov.
1 38

This cannot be bad, but I wonder

whether Black cannot develop more
effectively. White's basic idea is to
play iVf3 , h2-h3 and g2-g4, removing
the f5-pawn and thus undermining
Black's control of e4. In order to pre
vent this plan, I would therefore sug
gest protecting the f5-pawn with
. . . iVd7, putting the bishop sensibly on
e7 and then developing the queen's
knight to b4 via a6. This line is of
course very similar to Yusupov
Kramnik from the 6 e3 main lines,
but since White has taken on 5 so
early, allowing Black to prevent e3-e4

O d ds a n d En ds

with a pawn on fS , rather than his

pieces, Black has more flexibility with
his piece placement: S . . . aS 9 iLxc4
CUa6 10 'i'f3 'i'd7 1 1 h3 CUb4 12 0-0
,YLe7 is fine for Black.

ga7 40 gb8 ge7 4 1

9 1Lxe4 0-0 1 0 0-0 CL.lbd7 1 1 a 5 a6

gxb5 ga7 43 Ub4+ We5 44 a4

28 . . . Wf6 29 gb4 gd7 30 gb6+ We5

31 Wf2 ge7 3 2 b4 we4 3 3 Ud6 h4
34 We2 g 5 3 5 Wd2 ge7 36 e6 bxe6
3 7 gxe6 ga7 38 gd6 gb7 39 gb6
b 5 axb5 42

1 2 h 3 h5 1 3 f3 g 6 1 4 e4! CL.lxe4

Wd6 45 a6 We5 46 We3 f 5 47 ga5+

1 5 CL.lxe4 fxe4 1 6 xe4

Wb6 48 gxd 5 Wxa6 49 gxf5 g7 50

Wd3 Wb6 51 We2 g4 5 2 gh 5 gxh3
5 3 gxh3 a7 54 Wd3 We6 5 5 gxh4
Wd5 56 Uh5+ We6 5 7 We4 ga2 58
gh6+ Wf7 59 Wf4 1 -0

3 iLf4

F)

With 3 iLf4 White intends e2-e3,

when he will have safely developed
his queen's bishop outside the pawn
chain.
The move . . . h7-hs has rather weak
ened the black kingside and now
'i'xg6+ is threatened.

Game 88
Psakhis-Sadler

Megeve (peA rapidplay) 1 994

1 6 . . . wg7 1 7 1Ld2 CL.lf6 1 8 f3 CL.l d 5

1 9 gfe 1 1L e 7 20 1Lxd 5 exd 5 2 1 1L b4
1L d 6 22 e 5 xe5 23 dxe5 d7 24
gad 1 gad8 25 gd4 gfe8 26 Uxe8
xe8 27 e3 xe3 28 fxe3

1 d4 d 5 2 e4 e6 3 f4 dxe4!

Basically winning a pawn.

4 CL.le3 e6 5 e3 b5 6 f3 a5 7 g4
b4 8 CL.le4 b3+ 9 CL.le3 1Lb4 1 0 CL.lge2
xa2 !

White's space advantage gives him a

very pleasant rook ending.

Now after 11 Mxa2 bxa2 the pawn

1 39

Th e Sla v

queens!

ttJa4 26 e 1 xc3+ 27 O .:gxg7

1 1 g c 1 xb 2 1 2 e4 ttJf6 1 3 g 5 e 5

28 xe6+ fxe6 29 e4 exd 5 30

1 4 g x f 6 exf4 1 5 xf4 a 3 1 6 e3

e6+ b8 3 1

b2 1 7 .:g b 1 a2

e4+ !

ttJg3 c2 3 2 g2

Threatening 1 8 . . . 'iVxb l+.

1 8 d 1 b3+ 1 9 d2 ttJa6 20 fxg 7
gg8 2 1 h6 e6 2 2 h 3 ttJ c 5 !

Making things safe!

33 xe4 dxe4 34 .:ghd 1 gxd 1
gxd 1 e5 0 - 1
23 xh 7 0-0-0 2 4 d 5 cxd 5 2 5 exd 5

1 40

Not a good advert for 3 f4!

35

O d ds a n d En ds

S u m m a ry

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

However, note that Black can only play this line after e2-e3 by White, as 3
CDf3 CDf6 4 CDc3 g6 5 cxds cxdS 6 itf4! leads to a superior version of the Ex
change variation, which is rather depressing for Black.
White players looking for an offbeat alternative might care to examine 6
CDh4!?
1 d4 d 5 2 c 4 c 6
3 i2J c 3

3 CDf3 CDf6
4 CDc3 dxc4 (D)
5 e3 game 85
5 e4 game 86
5 a4 itfS 6 CDh4 game 87
3 itf4 dxc4 - game 88
-

3 . . .e5

3 . . . CDf6 4 e3 g6 5 CDc3 itg7 6 ite2 0-0 7 0-0 (D)

7 . . . b6 - game 83
7 . . . dxc4 game 84
-

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

1 2 . . . iVc7 - game 81
1 2 . . . CDa6 - game 82

4 . . . dxc4

7 0-0

1 2 fxe 3

1 41

INDEX OF COMPLETE GAMES

Adianto-Kramnik,

London (Intel Grand Prix) 1994

Wijk aan Zee Open 1993
Alburt-Shabalov, USA Championship 1996
Andersson-Epishin, Ter Apel 1 995
Atalik-Miles, Hastings 1 995/96
Bacrot-Smyslov, A lbert {sixth match game} 1 996
Beliavsky-Shirov, Erevan Olympiad 1996
Benz-Gretarsson, Oberwart Open 1 996
Cifuentes-Dreev, Wijk aan Zee 1 995
Cifuentes-Leyva, Cienfuegos 1 996
Dautov-Nikolic, Ter Apel 1 994
Dreev-Piket, Dortmund 1 994
Ehlvest-Schwartzman, New York Open 1996
Epishin-Pomes, Manresa 1 995
Gelfand-Lautier, Zurich 1 994
Gelfand-Nikolic, Manila Interzonal 1 990
Gelfand-Shirov, Dortmund 1 996
Gofshtein-Sadler, Ischia 1 996
Gulko-Salov, Reykjavik (World Cup) 1991
Hillarp-Persson - Acs, Budapest 1996
Hjartarson-Gulko, Reykjavik 1 996
Hodgson-Sadler, Ischia 1 996
Hiibner-Beliavsky, Munich 1 994
Illescas-Gelfand, Dos Hermanas 1 996
Ivanchuk-Bareev, Dortmund 1 995
Ivanchuk-Lautier, Linares 1994
Ivanchuk-Smyslov, Tallinn (rapidplay) 1996
Karpov-Hj artarson, Tilburg 1988
Karpov-Kramnik, Monte Carlo (blindfold) 1995
Akopian-Shirov,

1 42

49
125
127
111
95
69
1 04
64
1 02
115
77
134
23
75
30
82
55
27
133
137
121
110
24
56
18
22
62
45
16

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

Karpov-Short,

Dortmund 1995
Novgorod 1994
Kasparov-Nikolic, Manila Olympiad 1992
Kasparov-Shirov, Dos Hermanas 1996
Khalifman-Georgiev.Kir, Elenite 1994
Klarenbeek-Rogers, Dutch Team Championship 1996
Kozul-Illescas, Erevan Olympiad 1996
Kramnik-Damljanovic, Moscow Olympiad 1994
Kramnik-Ivanchuk, Linares 1994
Kramnik-Ivanchuk, Monte Carlo (blindfold) 1996
Kramnik-Lautier, Linares 1994
Kramnik-Shirov, Dortmund 1996
Kramnik-Shirov, Vienna 1996
Kramnik-Short, Moscow (Intel Grand Prix) 1996
Kramnik-Short, Novgorod 1994
Krasenkov-Epishin, Bmo 1994
Krasenkov-Sapis, Polish Championship 1995
Krasenkov-Sokolov.I, Malmo 1995
Lalic-Sadler, Hastings 1995/96
Lautier-Bareev, Linares 1994
Lautier-Sokolov.l, Groningen 1995
Leitao-Beliavsky, Erevan Olympiad 1996
Miles-Hodgson, Hastings 1995/96
Milov-Sadler, Isle ofMan 1994
Nesterov-Imanaliev, Bishkek Zonal 1993
Nogueiras-Sokolov.l, Erevan Olympiad 1996
Novikov-Gretarsson, Berlin Open 1995
Oll-Anand, Biel Interzonal 1993
Parker-Hellsten, Copenhagen 1996
Piket-Ge1fand, Wijk aan Zee 1996
Portisch-Kramnik, Biel Interzonal 1993
Psakhis-Sadler, Megeve (PCA rapidplay) 1994
Pushkov-Epishin, Russian Championship 1995
Razuvaev-Sturua, Erevan Open 1996
Richardson-Sadler, Islington Open 1995
Rogozenko-Bets, Moldovan Championship 1994
Ruze1e-Thorsteins, Lyon (European Club Cup) 1994
Sadler-Bareev, Hastings 1992/93
Sadler-Ferguson, British Championship 1996
Sadler-Hodgson, Hastings 1995/96
Sadler-Miles, British Championship 1998
Savchenko-Ninov, Cappelle la Grande Open 1994

Kasparov-Bareev,

92
35
130
54
35
84
124
72

47
66
43
47
98
67
59
91
81
118
39
105
120
84
1 22
108
79
119
63
87
76
53
1 14
139
101
1 26
13
132
60
136
42
92
34
138
1 43

Th e Sla v

Schandorff-Hellsten,

Copenhagen 1996
Biel 1 991
Shirov-Nikolic, Wijk aan Zee 1 993
Sokolov.I-Bareev, Leon 1 995
Sokolov.I-Hellsten, Malmo 1 995
Sokolov.I-Shirov, Erevan Olympiad 1996
Strauss.D-Lakdawala, USA 1992
Topalov-Gelfand, Belgrade 1 995
Topalov-Gelfand, Dos Hermanas 1996
Vaiser-Nalbandian, Erevan Open 1996
Van der Sterren-Petursson, San Bernardino Open 1992
Van der Sterren-Shirov, Biel lnterzonal 1993
Ward-Levitt, British Championship 1995
Wells-Flear, Oakham 1 994
Xu Jun-Akopian, Moscow Olympiad 1994
Yusupov-Kramnik, Riga 1 995
Yusupov-Shirov, Zurich 1 994
Shirov-Bareev,

1 44

75
46
78
50
80
1 00
36
21
58
1 13
57
89
103
90
32
17
111