Jan Pinski

The
Benko
Gambit
The Benko Gambit

Jan Pinski
First published in Sweden 2005 by Quality Chess Europe AB

Copyright © Jan Pinski 2005

All rights reserved. No part of this publication may be reproduced, stored in a re-
trieval system or transmitted in any form or by any means, electronic, electrostatic,
magnetic tape, photocopying, recording or otherwise, without prior permission of
the publisher.

ISBN 91-975243-8-7

All sales or enquiries should be directed to Quality Chess Europe

Vegagatan 18, SE-413 09 Gothenburg, Sweden
tel: +46-31-24 47 90 fax: +46-31-24 47 14
e-mail: info@qualitychessbooks.com
website: www.qualitychessbooks.com

Edited by John Shaw & Jacob Aagaard

Typeset: Ari Ziegler
Proofreading: Danny Kristiansen and Colin McNab
Cover Design: Carole Dunlop
Drawings: Claus Quist Jessen
Printed in Estonia by Tallinna Raamatutriikikoja LLC
 CONTENTS

Introduction to the Classical Benko Structures 7

The Benko Gambit Accepted | 31
The Benko Gambit Accepted 11 41
W b

The Benko Gambit Accepted 111 57

The Positional 5.b6 71
The Ambitious 5.e3 83
0 N OV W

5.f3 91
5.9c3 — The Zaitsev Variation 105
Five Sidelines 111
Index of variations 121
Index of games 124
 List of symbols

T
Check
A strong move
-

A brilliant move
An interesting move
A dubious move
A mistake
A blunder

White has a winning position

+

White is better
|+

White is slightly better

=+

The position is equal

I

The position is unclear

8

Compensation for the sacrificed material

+l +i 8l

Black is slightly better

Black is better
Black has a winning position
+
|

1-0 White won

Va—Ys The game was drawn
0-1 Black won
(x) x'th match game
corr. Correspondence game
Introduction to the Classical
Benko Structures

In the Benko Gambit there is no pawn structure Black’s ideal development

more common than the classical structure:

///% /‘/}
> /M 7%}
/ //%
//,,,
// s /7 )

, | //// ,,,,, 7
/ Z.
//

Black is usually striving towards a position not

This structure is the most important structure
you will face when you are playing the Benko
Gambit. Whether a bishop or a king is on g2
(after ¢4, &xf1, ©xf1 and $g2) does not matter
too much. Also the e-pawn could be on €2 as well
as ¢4, and the white pawns on the queenside can
be on b3 and a4, or (weaker) on a3 and b2. In
this chapter we shall look at the most common
decisions and motifs in this structure, which will
make it easier for us to investigate the theoretical
lines in the coming chapters. We will do so by
starting looking at the main strategic ideas in a
kind of express summary.
too dissimilar to this. The a6-bishop might be
exchanged, the knight on d7 has not yet found
its place in the world, and the same goes for the
f6-knight, but in general this is the standard way
of placing the pieces for Black in the Benko.
White’s ideal world
For White the story is slighty different.
Having won a pawn as early as move 4, he is in
a different situation from Biack. Black wants to
achieve pressure on the queenside against the
two weak white pawns, while White is behind in
development and basically just wants to have time
8 The Benko Gambit

to catch up. For this reason White does not often

have an ideal position to be heading towards. His
moves are usually designed to soften the black
initiative and to avoid creating weaknesses in his
own camp.
Often this means that White will head towards "y
something not too far away from this position:

7/7/
Al AR B
Z, % 7, .., A 2
5 7 A A%

/ //z/ e N
It is probably for this reason that White
players have started to develop their bishop
to g2, where it looks somewhat inferior to its
mighty opponent at a6. Also, it is not that easy
% // to land a knight on d3 when you are Black.
As indicated above, White will usually strive
towards putting his pawns on b3 and a4.
Obviously it could also prove beneficial for
Never mind the rest of the pieces. In a position White to play a3+b4 if it does not lose a pawn
like this White has managed to block the - or something else for that matter!
queenside completely. He can now consider if he However the move a2-a3 is usually a very bad
wants to play b3-b4 or if he should focus on a idea, as it weakens a lot of light squares on the
kingside attack with h2-h4-h5xg6 and so on, or queenside. Of course b2-b3 weakens the dark
the most common advance, e4-e5 in the centre, squares, but they are easier for White to control.
trying to create weaknesses. He has a bishop at c1 for that job, and a knight
at ¢4, or more likely b5, will also take control of
So, whenever we see a game with this standard some dark squares. If White plays a2-a3 he can
Benko structure, we can expect to see Black easily end up with a weak b3-square. Let us look
being a little ahead in development, and at some examples that show what this can lead
White somehow relating to Black’s actions in to.
a strategy of containment.
Game 1
The bishop on a6 Ehlvest — Fedorowicz
A natural part of the Benko Gambit Accepted is a New York 1989
black bishop on a6. Either White will advance his
e-pawn and give up the right to castle, or he will 1.d4 D6 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.bxa6 g6
face a well-placed bishop for quite some time. 6.9c3 8xa6 7.g3 d6 8.8g2 g7 9.2f3 Dbd7
However modern practice seems to suggest that 10.0-0 Db6 11.RDel??
the weakness of the light squares in the white This is a somewhat slow plan, which did not
camp is more important than the strong bishop, find many followers. These days people try to
and most importantly, the bishop on a6 is also a focus on getting their queenside developed
little bit in the way. before thinking about this kind of manoeuvre. As
 Introduction to the Classical Benko Structures
we shall see on the very next move, this quickly
becomes more difficult.
11...8c¢4 12.d3
12.b3?! is usually dubious when the knight is
at c4. Here Black has a cunning blow: 12...22d7!
13. %22 WasS! wicth a clear edge for Black is
worth noting. 13.bxc4 with almost equality is
better, but still White should not go there to
start with.
12...d7 13.¥c2 0-0 14.h4?
As discussed below this is part of an erroneous
plan. White should instead try to catch up with
Black’s lead in development.
14.%e4!? is suggested by Fedorowicz. But after
14...2db6! it is not easy to see any advantage
for White at all. A natural move such as 15.2d1
is answered with 15..20a3 16.%d2 HxdS! after
which Black is doing fine in all the complications.
This is 2 good example of how Black can gain the
advantage through a lead in development, a lead
that was expanded because of the slow manoeuvre
N f3-e1-d3.
Another game from this position continued
14.24 W25 15.822 Bfb8 16.5)e4 Hde5 17.Dxe5
Hxe5 18.8d2 Wb6 19.8c3 &4
.~ AhLa
4%%21%w%
gy
E%>
Now we have:
a) 20.8aal &xd5 21.20661 &xf6 22.8xd5 Eao,
which would lead to an interesting position.
Here White exchanged his extra pawn for a
position with two bishops and a passed pawn.
Usually this would sound very good, but Black’s
knight is by no means bad, and the f6-bishop will
eventually be exchanged for the one on ¢3, and
the black pawns are very likely to shut out the
light-squared bishop eventually. If I were to give
9
an evaluation here, I would probably guess that
Black has slightly better chances, as the b-pawn
is still backward, and White has not solved the
problems of having weak squares at b3 and b4.
b) 20.a5 ¥b5 21.Haal fxe2 22.8fel £d3
Here a draw was agreed in Lyrberg — Lonborg,
Copenhagen 1998. Black probably offered a
draw because he was lower rated and had too
much respect for his IM opponent, rather than
because he thought the position held no promise
for him. In my eyes play mighe still be close to
equal, but the trend is certainly with Black, and
White cannot be said to have complete and
eternal control over his queenside.
14...%a5 15.a3
T/%g/
Ee
¥
%Z %/
//
w A /7/%@/
47@%4 // , , , :
pxXBan A
,,,,,,
White is distressed about the threat of ©a3.
15.2b1?! allows this trick, based on the weak c3-
knight: 15...80a3 16.bxa3 ¥xc3 17.Wxc3 2xc3
and Black has a good position. Butalso 15...2fb8!
seems to be good. It is not clear how White shall
improve his position, and the ©a3 threat is still
hovering over his head.
15...Hab8!?
Black is instantly targeting the weakened b3-
square. | am sure thac at least one reader will
wonder why the queen’s rook goes to b8, and
not the king’s rook as in the diagram of ideal
development above.
There are two reasons for this.
1) White has made his intentions clear. He
wants to play b2-b4. This will open the c-file and
10 The Benko Gambit

Bfc8 will then be a good active response.

2) After 15.a3 the a-pawn is not an immediate
target and does not require instant attention, so
] 7
» oy
the rook is not too active on a8 anymore. 7 74 & VA
16.8a2!?
Played with the idea of b2-b4.
16...8ce5! 17.8xe5
17.b4 does nor work now. The endgame after
17...8xd3! 18.bxa5 &xc2 19.8xc2 Ea8 is torture
for White.
17...0xe5
%
Fedorowicz thinks Black is better here, which is
probably too optimistic. One thing is that Black
/ ,,,
//
is not under any pressure and White’s pieces
are slightly uncoordinated. Another thing is to
interpret a solving of all one’s opening problems
Black wants to occupy the b3-square with a
as achieving an advantage. That the position is
heavy piece and keep the a6-bishop where its
easier to play for Black is however true, and that
influence is highest.
is not too unfair an argument for saying “a slight
19.We42!
advantage for Black”.
White is continuing his kingside attack, but it
18.h5?
is not really dangerous.
White is preoccupied with his own agenda,
19...¥b3 20.8al
but this is against the spirit of the opening. As
20.Wh4? just loses because of 20..&xe2
explained above, White’s strategy should often be
21.&5xe2 ££6! and Whire will not get any kind of
that of containment. Here White is playing as if
compensation for the exchange.
he is the better-developed player, who therefore
20...
¥ c4!
has the right to attack, rather than being the
An easy move for a Benko-expert like
player with a pawn more, who wishes for nothing
Fedorowicz. The endgame would be better for
to happen.
Black here, something we will discuss further
18.82¢5!? was interesting, but also sensible was
below.
18.b4 ¥c7 19.bxc5 (19.8e4 £c4! also gives Black
21.%c2
good counterplay) 19..%xc5. Here White has
Slightly resigned, but what should he do?
managed to exchange his weak b-pawn for the
21.Wxc4? Dxcd 22.8a2 8b3 is equal according
c5-pawn. But this does not mean that his problems
to most chess-playing programs, but actually
are solved. Black’s pieces are still far more active
Black is a whole lot better. Eventually the white
and better coordinated. White will struggle to
queenside will collapse, and Black will have a
keep what he has together and the position level.
better endgame.
Still this is what he should have done.
21...2b3!
18...%b6!
Black is in total control now. The white minor
pieces have little influence on the queenside.
22.hxg6 hxg6 23.F42!
Ehlvest was one of the truly great players of the
1980s, and he is still a very strong grandmaster,
but in this game it seems as if he was a little
hung-over and allowed Black to move the pieces
for both sides.
 Introduction to the Classical Benko Structures 11

figy
23..Dgh 24.813 8d4t 25.c0g2 Df6!
3 /// ,,,,, /////
Now the d5-pawn is targeted.
26.%d12
White is continuing his plan and not offering

7. a lot of resistance. After the more natural

26.£d2 Black has two plausible way to play for
an advantage. 26..Zfb8 (26...8xc3% 27 .&xc3
Hxds 28.8xd5 Wxd5+ 29.8f3 &b7 30.8h1 5
is more double edged, but the white king is in
much more trouble than its black counterpart)
/,,,,A
27.2abl In this position only one black piece
is currently redundant, but it can easily find a
better spot. 27...8c8! 28.e4 £g4 and White is
still struggling.
26...2fb8 27.8h1
Though his move banishes the black knight 27.Ha2 8xc3 28.bxc3 Ded! looks close to
from its prominent outpost on €5 momentarily, decisive as well.
it does not tame the wild horse. Rather it opens
White's king's position for an attack.
K_= =
Z ks
Obviously Ehlvest is more than capable of
understanding all these things - so why did he play

s & @x%
the move all the same, you might ask. T believe this
is a common phenomenon. His plan has been to
play 23, 22 and Bh1 in order to start an attack
on the h-file for some time now. This was not a A/ 53/ .
2D =i
A 7&K 7/
good idea when he started it on move 14, and it g// ,,,,, »
is still not a good idea. But, instead of preparing
for a daunting defence, White carries on as if his
wishful thinking represents the realities of the /%/ 78
position. Not because he believes it, but because
the alternatives are (psychologically) even more
unpleasant.
23.8d 11 wasanalternative. Play could continue 27...Bxb2!
along the lines of Fedorowicz” imaginative line: A not too difficult sacrifice. The dark-squared
23.. 88 24.e4 D311 25.8xf3 Exc3 26.bxc3 bishop on d4 easily equals a rook in value.
Hxc3 27.Wa2 Bxf3 28.Wxc4 fxcd and, though 27..2b7 would also win eventually, but this is
we have opposite-coloured bishops, the position strong and the most thematic.
is far from drawish. The only problem is that after 28.8xb2 Exb2 29.9a4
29.%e3 the f3-rook is somewhat uncomfortable, 29.8b1 £ xd5 followed by £e37 also decides.
but 29...£5!? should solve this. Black is a little 29...Bb3!
better. He would also be so after the simpler Traps the knighr. In this way games of chess
24.. H#3b6!?, which might be better. are always decided by tactics eventually...
23.8¢5 Bfb8 24.8fb]1 H8b7 would also be 30.2b1 Exb1 31.¥xb1 Wxa4 32.15 g5 33.¥c1
better for Black, though this is probably what Hh7 34.¥b1 &fc 35.9b8t D8 36.g4 c4
would have made most sense from White’s 37.8b1 c3 38.%d8 Lxe2!
perspective. 0-1
12 The Benko Gambit

It is because of games like this that White players However a quick c¢5-c4 and ©£d3 is not an
have started to play b3+a4 and then put a knight instant recipe for success for Black. First of
on b5. Black’s counterplay down the two open files all White might play b2-b3 and be ready to
is well supported by his minor pieces and by his exchange the black pawn. This could prove to
pawn structure, so it should not be taken lightly. achieve very little for Black (though we shall
also look at situations where this benefits Black
The d3-square highly). Secondly, we should not forget the old
To play in the centre White will eventually have saying that moving a piece from one square to
to advance his e-pawn. Moving the e-pawn has another will always leave a new square less well
one clear drawback: The d3-square becomes very guarded than on the previous move. All moves
exposed in some lines, and Black will try to place in chess have drawbacks, and with the advance
a knight there, cither with the support of the a6- c5-c4 White will definitely spot an avenue for his
bishop, and (or) with the support of the c-pawn: knight:

/7 /M/ 4
B By
g/ x /}/
0.
'y
\\\\
\\\

7 7 7
///47,,/ %,////%//,,/ /&57//
N iy 2
4

e 2w
From d3 the black knight is a potential A white knight on c6 is potentially at least as
destroyer. It can attack the soft spots b2 and f2, damaging for Black as a black knight on d3 is for
which the white pieces cannot defend naturally. White. €7 needs protection, and both queen and
rook are prohibited from using their favourite

0, / .
squares. So the ¢5-c4 advance has advantages and
disadvantages, and to understand them correctly
\\\\

/ // , , , %} in the very dynamical setting of the Benko

s & /}%
Gambit is the challenge.
[ have selected two examples to show how this
87 7 can be done.

%&@;%W% %Y
Game 2
Kraidman — Halkias
7. 7.
Antalya 2001

1.d4 D6 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.bxa6 g6

6.9c3 £xa6 7.g3 g7 8.82g2 d6 9.2f3 Dbd7
10.0-0 Db6!?
 Introduction to the Classical Benko Structures 13

Another way to play the position. There is no
law stating that the queen must be on a5.
11.Eel 0-0 12.2d22!
This is not too impressive. The knight is
preventing &c4, but besides this it is not clear
where the knight is going.
12..%c7 13.8c2 b7 14.¢4 Dg4 15.Dd1
A non-move like 15.h3 would not only
be a pleasant tempo for Black, who wants to
play 15...20¢5, but could also allow Black to
spectacularly sacrifice a knight with 15...2xf2!
16.50xf2 8d4t
has positioned himself in such a way that he
would be able to neutralise the white centre with
e7-e6, rather than suffer annoyance from the
knight's presence. 16...2xf1!? 17.Hxf1 c4 was the
alternative. The young Greek grandmaster seems
to have decided thar the exchange of the bishops
would somehow benefic White. This could very
well be the case short-term. The c4-pawn will
need some reinforcement, and the bishop on f1
would be blocked by the knight on d3, while the
same is not the case for the a6-bishop.
On the other hand the light squares would be
weakened in the white position. Both options

// Edo make sense.

17.8c3 Bfc8 18.8d1 Dbd7 19.52g2
///
EM/%/é
2 The white position is already very difficult to
/ defend.
After something like 19.2bl He5 20.b32,
which seems absolutely reasonable, White is too
passive to be able to allow a knight to hibernate
at d3. But the sad thing is that he also seems
completely unable to kick the knight away:
20...cxb3 21.8xb3 311 22. %92 Wd71 23.2xa6
The only move. (23.55xf32! &xf1 24.Hxfl Qa4
when play will have to continue 17.&e3 {5 with and White's position seems indefensible. There
massive complications. e.g. 18.2f3 €5 19.dxe6 are simply too many threats to the heavy white
fxe4 20.)dxed D4 21.9b3 We7 22.20d5 Wxe6 pieces.) 23.. Hxa6 24.8b2 fxc3 25.%8xf3
23.5\f4 Wes and Black has a very strong, maybe (25.8xc3 ©b4 and Black wins) 25..8b4
irresistible, attack. 26.%xc3 Bxc3t 27.8xc3 ©xa2 and Black should
15...5e5 16.8f1 c4! be able to turn his slight material advantage into
a full point, though the game is not completely
over yet.

LR
19...2¢5 20.8)£3 Ded3 21.Ebl

/%
,,,,,,,,,,,

%fiV
g%/
TN

,,,,,,,,,,,

7,
__ %
////

/3/&7/&//
// /
,,,,,,,

/
, , ,%@“ / %&@%
Now the white knights have no avenue to d4-
¢6 (and even if a white knight did reach ¢6, Black / Z.@:z/
n
14 The Benko Gambit

E . &
Black has quite clearly achieved his strategic
plan of domination, and now needs to rethink
the position, so that he does not continue to
play it on the conditions of what it was, but
/%’/ 2
what it has become. Wilhelm Steinitz was
the first to state it accurately: When you have
the advantage you need to attack, or you risk
losing your advantage. The problem for Black
here is to find another point of attack. Halkias
correctly sees that the b7-g2 diagonal is very
treacherous for White, and decides to use it to
his advantage.
Here a program like Fritz 8 wants to play e7-
e6 in various forms, which is very sound advice
indeed. However, the grandmaster is able to get 24...Bxd5! 25.sbgl
even deeper to the core of the position. Maybe White’s idea was 25.8xd5 Wxdst
21...£5! 22.exf5 Bf8! 26.%2¢1 and now all debates are cancelled by
This move is what the high-tech wizard does 26...£b7 with marerial advantage and positional
not comprehend in advance. Black does not domination; and then 25...2xd4 might be even
count the pieces, but rather counts the kings! stronger...
In the early days of Man vs. Machine matches 25...Exd4!
it was thought that the computer could only be 25...8xd4?? 26.2¢2 and suddenly it is White
beaten with positional means - now we know who is coming out on top.
that computers are bad at understanding long 26.8g2 Wa7 27.0e2 Db4 28.Wcl Dxa2
term compensation, and humans therefore have 2982 Dbs 30.8c1 Da2 31.¥2 Db
an advantage in tactics. 32.¥c1 b3
Maybe White should have a go at 23.b41? and Black won the queen and later the game.
here, but no matter what, his position is not too
hot. The close to paralysing effect on the white
23.8e3 position the knight had from d3 is surely
23.fxg6? loses quite easily to 23..Exf3! something worth remembering. As already
24.50xf3 &xc3 25.Wxc3 (otherwise it is just an stated, we can see exactly the same function from
extra piece) 25...¥xd57 26.%¢2 Ef8 and all the a white knight at ¢6:
black pieces will tear the white king apart. e.g.
27.8e3 Ded and there is no defence against
Whst.
23..8xf5 24.9)d4?
This loses to a not too difficult tactic. After
24.8xc5 Black gets a dominating position with
24.. 8xc3 25.0h4 Wxdst 26.3 Bf7 27.Wxc3
g5 28.8xd3 cxd3 29.2e3 gxh4 and White has
no defence because of 30.8f1 h37!.
24.9h4! was the best try, with the idea
24.. 8Bf7 25.@;;1 and White is not completely
lost, though Black has strong pressure after ideas
like 25...g5!? 26.%)g2 Eaf8.
 Introduction to the Classical Benko Structures 15

Yusupov — Vitolinsh 16.9¢6

Frunze 1979

% W Ed
/m/ :/x
,,,,,,,,,,,

a4 A 7
//////

/7

Py
/‘//&
%7///%/

/& / o
iy //
M %//g/B;3 7 //; % ////
H AW
,,,,,,,,,,,,

We enter a spectacularly bad version of the
Benko Gambit, where Black has “been allowed”
to sacrifice the pawn, but not gained a lot of
counterplay down the a- and b-files yer, nor
received a great lead in development. Yusupov
(at this time a teenager on his way up) correctly
decides to put a knight on b5.
12.a4! Hfd7 13.2b5 c4??
Black is lacking in good moves, and therefore
decides to allow a white invasion in return for a
knight on d3. Objectively this is not particularly
bad (it is the position which is the problem, not
the move), but a matrer of style. Vitolinsh was a
very aggressive player and a gifted tactician, who
preferred to solve his problems dynamically.
Now White can, of course, not decline the nice
invitation to c0.
14.9\bd4 & c5 15.a5!
Though the a-pawn becomes a little loose with
this move, it is definitely worth it.
15...2bd7
A major problem of having an enemy knight
stationed at c6 becomes clear after 15...4c8
16.%¢c2! where Black’s inability to protect c4
is clearly marked. Black now has to cook up
something fanciful, which can prove difhculr,
as after 16..80d3 17.&xd3 cxd3 18.¥c6! Ha7
19.£2¢3 Black is completely paralysed
without hope for the future.
16...%c7!
Black needs the e-pawn much less than he
needs to be active. Sometimes our position
becomes so objectively bad that correct defence
cannot save it. We can then choose between
activity and passive defence. The nature of the
position decides our strategy as much as it does
normally - only now we will care less about the
objective correctness of the strategies.
The alternative 16...%¥e8 17.6d2 looks very
bleak for Black.
17.9xe7t
Yusupov was always very happy to accept
material, but also other moves make sense
here.
17.¥¢c21? putting pressure on ¢4 makes a
lot of sense. One line continues 17...2b8 (Or
17..6b3 18.8a3 b8 19.&xcd Dxcl 20.Excl
Axch 21.8xa6 Exab 22.b4! would leave White
with two passed pawns, where at least one is
likely to claim material.) 18.%cd4! and Black
has continuing problems with the c4-pawn.
17...2h8 18.8)c6!
Of course the knight should return to its
wonderful position.
18..2d3 19.Ba2
Though not to my taste, there is nothing
wrong with this slightly passive move.
and 19..27¢5
16 The Benko Gambit

22.8xcl cxb2 23.9d3 ©xe4 Black would win.

T 7
N B
7 d
Now he wins a piece, but White
pawns to put Black under pressure.
keeps enough

. 21...cxd2 22.8xa6 dxcl¥ 23.¥xcl Hxa6

EX A ¢ 24.b4
The game was very unclear (though still better
for White) and eventually ended in a draw.

Before we move on from this somewhat narrow

subject of ¢5-c4, I would like to show another
two examples where Black is unsuccessful when
advancing the c-pawn. This is not because White
,,,,,, wins most of the games where this happens,
but because I feel they display the various ideas
surrounding the ©d3-jump. Basically I prefer to
20.d2?
let the games organise the ideas, and not the other
But this is simply bad. Actually White loses a
way around. In the next two games two 2600-
piece by force! Luckily for him he will get more
players choose to make long knight manoeuvres
than enough compensation for it, butstill it is very
with the only idea being ©d3. However, as
unlikely that this was a grand scheme thought up
we shall see, even a strong grandmaster can
by the young Russian! In positions like this it is
overestimate the damage a knight on d3 can do
important to combine all your desires into one.
all by itself. Both of them lose rather pitifully, but
White would like to develop, and at the same
the truth behind these games is, as so often, to be
time also calm down his opponent’s counterplay,
found in the annotations.
so he can be ensured that the material advantage
eventually will count.
Stefansson — Felgaer
20.8e3! was the right move, as it solves all
Calvia (ol) 2004
White’s inconveniences at the cost of just one
pawn. Now Black could try various moves, but
probably without any luck:
/ 2 z ,
/
20..5xb2 21.¥c2 &£bd3 22.8xc5 dxcS
23.2xd3 cxd3 24.¥xc5 and White wins.
20..£5 21.8xc5 dxc5 22.e5! Hixe5 23.8fxe5
£xe5 24.W¢2 simply wins.
20...8x%e4 is the only chance. One way for
White to continue now is 21.2b6 ¥d7 22.£xd3
cxd3 23.8el £c4 24.Ha3 &xb2 25.Hxd3 &xd3
26.¥xd3 5 27.9g5! and White has exchanged
his material advantage to a sort of material
,,,,, 7, =
equilibrium, where the active white picces 7y 77 % 7/7
ot 7
completely dominate their black counterparts. %g-z & 43
20...c3!
20..£51? also gave Black counterplay and
12.8d1
therefore could be considered. However after
Now after 12..58d7 we would have a
21.e5! White would keep a plus.
normal position, but Felgaer somehow becomes
21.8xd3!
hypnotised by the tempo gained with &b4 in a
After both 21.bxc3 &xcl! and 21.8b3 &xcl
 Introduction to the Classical Benko Structures 17

few moves, but in the process wastes a few of his
own.
12..8b72 13.e4
White decides wisely to play normal moves.
13...0a6 14.8f4 Db4 15.¥d2
White is preparing for e4-e5 and 2h6. His
queen is excellently placed on d2 and since none
of the other black pieces are in place, the knight
will be quite ineffective on d3. Basically Black
has lost his lead in development and has no way
to gain anything close to equality here.
15...2a6 16.2h6 Dd7
A clear symptom that Black has failed. We can
honestly ask now: what did the knight achieve
on b6?
16..8c4 17.%c1 8xh6 18.¥xh6 Dxb2 would
21.Exd3 £xd3 22.¥xd3 ¥b4 23.2b2
And White converted the extra piece to a full
point.
And finally, 1 promise, the last game concerned
with this theme.
Game 3
Grischuk — Tregubov
Noyon 2005
1.d4 D6 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.bxa6
g6 6.0c3 £xa6 7.g3 d6 8.8g2 Rg7 9.0f3
Hbd7 10.Eb1 0-0 11.0-0 De8 12.¥c2 Hc7
13.24 £b7 14.8d1 Db6 15.e4 Da6 16.b3 Db4
17.%d2 £26 18.82a3

CREN /
of course lose to 19.9g5 with mate to come.
17.8xg7 oxg7

% %,%w%
/ oa
/ slo /////é
,,,,,, /////
/ //

A/// %}% /2/ )

/ / %
A 7, ,,,,,,,,,,,,,,,,,,,,

/ @,/ /
& /fi/ g/g-
Black has again used a long time getting a
knight to b4, and now comes the probably
18.b31? decisive mistake - moving it to d3! But his
White is not afraid of the knight on d3 at all, position was not necessarily bad. The white pieces
but 18.8e1! was maybe even stronger. Now after are not too active where they are (containment’s
18...2b7 19.a3 Ha6 20.b4 White is strategically disadvantage) and Black can harass them if so
winning. inclined. Probably White would have been better
18...2d3 19.2a4 ¥b8? off playing 18.2b2!? with a slight edge.
Black had to find a way to bring the knight 18...2d3?
home with something like 19...£6!? where his 18...&d3!? was best.
position is unenviable. Now White just wins. 19.8b2! is now the best shot for an advantage,
20.¥c31 f6 bur Black still has his chances. Black should
20...93e5 21.8xe5 dxeS 22.9xc5 would probably go 19..8d7 (19...8a6 20.8xb4 cxb4
just be two pawns up for White, so Black tries 21.5a2 &xb2 22.¥xb2 would probably favour
another form of material unbalance. White. The two passed pawns give a lot of play,
18 The Benko Gambit

and now the g7-bishop is gone.) 20.2xb4% I do 20...cxd3 21.2el £5 22.9xd3 Dd7
not see another way to play for an advantage. Black’s last chance to create some counterplay
20...cxb4 21.%xd3 &c5 22.Wc4 bxc3% 1 believe was 22...fxed 23.23b4 £b7 24.8Dxed Ef5, but
that if Black plays very actively, he should be OK. White should be able to play for an advantage
A line 1 have toyed with is 23.8Ec2 Eb8 24.8b] with close to every possible move here.
Was 25.e5 dxe5 26.8el e4 27.b4 ¥a7 28.a5 23.exf5 Bxf5 24.9b4 De5 25.52g2 W8 26.£4!
Hfe8 29.%e2 5 with chances for both sides. of3
The real points behind 18..2d3 are only Why not? 26...23f7 27.2b2 would also leave
displayed in the less important lines, where White in complete control.
White does not react correctly to the challenge. 27.8xf3 fxc3
19.%a1 allows the bishop to invade with ...2c2.
Now after 20.82xb4 £xd1 21.8xd1 cxb4 22.8a2
&d7, with the g7-bishop still in action, Black is E/g///@/
doing fairly well.
0 %7&/%/
However it is 19.8bcl which allows Black to
display his idea: 19...&c4!! iy 4 %x@
. /é
8@ % %
,,,,,,,,,,,,, X

% % %/%
3 8,
28.4)c6!
Decisive.
28..8xd2 29.9xe7t @g7 30.0xc8 Exd5
A stunning move, which finally brings life to 31.0xd6 Ed3t 32.58f2 Bb8 33.0b5 He8
the knight on b6. The chances are probably level 34.8¢1 b7 35.8xd2
after 20.bxcd ©Dxcd 21. W4 Hxa3 22.e5 where 1-0
both players have their fun. Of course White can
also repeat the position with 20.E5b1. Long knight manoeuvres
Now, let us return to 18...8d3.
19.8f1! In the last two games we saw great grandmasters
Now White strips Black of any counterplay. play a close to uncountable number of knight
19...c4?! moves. Tosomeone unaccustomed with the Benko
19..5e5 20.8xe5 £xe5 21.8xa6 Exal Gambit this can maybe seem unfathomable, but
22.2)b5 was a standard advantage for White, so to the insiders these kinds of manoeuvres are
Black decided to sacrifice another pawn. This everyday practice.
was probably the best try. Here are some standard manoeuvres:
20.8xd3
Simple chess.
 Introduction to the Classical Benko Structures 19

The characteristicofall these knight manoeuvres

is that the knight is generally not very well placed
u/ on f6. The g7-bishop needs to be able to actively
influence the queenside and the knight is seeking
,,,,,
greener pastures. The long manoeuvres to a4 or
”//// b5 are often based on an exchange of the best

/
piece in the white position, the knight on ¢3, but
///&// it can also go all the way to d4 as seen in the next

///
v //
/////
game. From d4 the knight is able to completely
dominate the white position and create all kinds
of tactical threats.

/”/% //
,,,,, A

// Game 4
Sjodahl — Adams
Qakham 1992

1.d4 56 2.c4 ¢S5 3.d5 b5 4.cxb5 a6 5.bxa6 g6

6.5 c3 £xa6 7.3 d6 8.8g2 £g7 9.8h3 Dbd7
/
/
Y 10.0-0 0-0 11.50f4 Wa5 12.2d2 EfbS 13.¥c2
/4

;f}?///
8.8 14.b3 a6 15.0a4 De5 16.8c3
V\\\\

Q\\

// /

//
Vaxerd

7'% 7/
//

» /,&
~ aia
Z / ’ 2., " e

5 //// 2///
,,,,,,,,,,,

%
/ // /‘/ ////// A

/4 7
//
// We enter the game somewhere between the

/ / / s)
opening and the middlegame, where neither

7 WG%Z&
player has been able to do exactly what he wanted.
Now Black starts a long knight manoeuvre from

B _E B 6 to d4. This in itself should not be enough

% e n
,,,,, to decide the game in Black’s favour, but the
talented English junior just cannot seem to put a
foot wrong in this game, and with simple moves
he increases the pressure on the white position,
something his Swedish opponent does not seem
to be able to deal with.
20 The Benko Gambit

16...20e8 17.2b2 D7 18.5£d32! 22.e3

Already this is a slight mistake. The white 22.8fd1 is met strongly with 22..£&xd3!
knight is not really good on d3 and Black will 23.¥xd3 $1d4 and the white pawns will drop as
be able to create real pressure on the white easily as anything.
pawns from now on. Maybe White thought the 22.89b2 5 d423.8 c4isbestaccordingtoFritz8,
coming exchanges would make things easier for but a human being quickly spots 23...8xb3!
him? Instead it was another knight move that 24.axb3 ¥xal with advantage because of the
would have helped to keep White in a fighting back rank mate. White has something close to
mood, rather than leaving himself with a difficult a defence in 25.g4! (25.8xal?! Hxalt 26.&f1
defensive task. Ah3 27.8e3 Bxb3 28.f3 &xfl and Black is
18.8¢c4! was better and now we have two close to winning) 25..8xg4 26.h3 but after
different paths Black could take: 26...Wa2! 27 hxg4 8xb3 the rooks dominate the
minor pieces and Black has the advantage in the
EEe
7 o/ E
/@% endgame. White still has drawing chances after
. %/‘/ 28.&f3! but certainly Black is not unhappy about
the turn of events.
22...¥a5!
Black continues to improve his position with
each exchange.
.
%97
LY =

//l
a) I do not like 18...8xc4 too much. After flz% ,}/i%?
,,
19.8xg7 @Xg7 20.bxc4 Bb4 21.8fcl White is
keeping his material advantage, and although = /%7%
P
the pawn on a2 is weak, the real essence of the
position is the restrictions presented to the black
»
7
@7////
/M
minor pieces by the c4-pawn. I think White is a

little better here.
b) 18...2b5! The active approach is most often
the best for Black in the Benko. Now play should
probably continue: 19.2xe5 dxe5 20.2d3
Nd4 21.Wd2 e4 22.8xed Hxb3 23.axb3 fxal
24.9xc5, where White possibly is a little better,
but the position is immensely complicated,
providing both players with practical chances.
18...20xd3!
Black does not mind the exchanges. True, he
does lose his powerful g7-bishop, but this was
already to some extent cancelled out by the c3-
bishop. What is more important is that Black is
able to exchange all the active white pieces, and
at the same time activate his own pieces.
19.9xd3 &xc3 20.¥xc3 Db5 21.¥d2 &f5!
Suddenly White is facing all kinds of tactics.
80 W
/%78 //;%//8%/ ////é/ <
. 7///,
Y
23.Wxa5?!
After this the black advantage is certain,
but here White had his last chance to offer
some resistance in a commonly used defensive
measure. He could exchange his two weak pawns
for Black’s pawn on c5 with 23.b4! cxb4 24.e4
£d7 25.%xb4, though after 25...8c3 26.%xa5
Hxa5 the position feels more comfortable for
Black. Still White has one or two moves to make
before a2 falls and should have excellent chances
of maintaining equality. Also 25.2{b1!? needs to
be investigated.
 Introduction to the Classical Benko Structures 21

23...8xa5 24.e4 8d7 25.8fcl 34.%xd2 Exd2 and the rook endgame is over
25.£41? ©1d4 26.%cl is the digital suggestion. before it has begun.
Here I like both 26...c4 followed by c4-c3 and 31..0e2+ 32.§§2g2 Dxcl 33.exd6 exd6
26...€6, both with something that could look 34.9xd6 Hxb3
like an advantage for Black. But the position 0-1
remains complex and maybe it is not as bad for
White as I imagine. However it is clear thar all In the next game as well we shall see how
the black pieces are more active than their white precisely targeted exchanges can give Black a
counterparts. better endgame and a full point with elegance
25...50d4 26.£f1 Ba3! and style, apparently without any resistance.
Preparing 2ba8 and b3 winning material.
‘The white position seems to be beyond salvation, Game 5
though attempts can still be made. V. Perez — Magem Badals
27.9b2 Terrassa 1995
27.%8el to exchange the strong black knight
seemns logical, but it is too litdle too late. Black 1.d4 D6 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.8c3
just needs to attack the white centre and activate axb5 6.Dxb5 d6 7.2f3 g6 8.2c3 &a6 9.g3
his bishop for the white position to fall apart. £g7 10.8g2 Hbd7 11.0-0 Hb6 12.2d2 0-0
e.g. 27...e6! 28.8)c2 Hixc2 29.8xc2 exdS 30.exdS 13.Bel
Bbxb3 31.8c4 Hb4 32.8el La4 33.Hccl &b5 We will not go deeply into this game, as the
34.&xbS Hxb5 35.Ec2 Hd3 and Black has a amateur is no real opposition for the grandmaster.
winning rook endgame. This line should, of Rather we will see what Black’s main plan looks
course, not be seen as forced, but the conclusion like when completely unopposed.
is that White finds it very difficult to hold his 13...80¢8 14.e4 Dc7 15.¥c2 Hbs

IE R z@//
position together.
27...8ba8
Now the a-pawn falls, and with it White’s
entire position.
28.%c4 Exa2 29.8xa2 Bxa2 30.e5?
Desperately seeking some kind of counterplay. /////////// 7//
Ifit were not for the reply this would have been the
best course of action. As we can see the endgame
, , , 8/////
a pawn down is hopeless as well: 30.2b1 &3}
%
. ////&//
Z /
3].@g2 Hg5! 32.bg1 &h3 33.4xh3 Hxh3t
3y b
LA RS,
%, )T iin 2 ST ,, >
34.%g2 Dxf2 35.8f3 ©1d3 and, with all the
pawn weaknesses, good technique should secure 7Ly /e g 7Y, 707
Black the full point. This would most likely start
with @d3-b4-c2-d4 and everything is under
o 2 2 &
control.
30...2h3! The ¢3-knight is exchanged and the queenside
A wonderful blow. White now loses the becomes harder for White to prorect.
exchange. 16.2xb5 &xbs 17.8f1
31.8xh3 17.2b1"? with the idea of 18.83¢3 with chances
White has no good moves anymore. 31.exd6 for both sides deserves investigation from those
9131 32.%0h1 Bxf2 and mate on the next move. seriously analysing this game.
31.5%6h1 &xb3 32.8bl Lxfl 33.8xf1 Hd2
22 The Benko Gambit

17..¥d7 18.9¢4 Bfb8 19.a3?
After this the endgame is simply untenable for
White. Better was something like 19.8e3!? 2xf1
20.8xf1 with the idea 20...2a4 21.2c4 Db6
22.5e3 and the unrated player would not be
first to avoid the repetition of moves. However,
Black has several moves to provide him with
chances of a full point, one being 22...e6%, but
none of them are a clear avenue to an advantage.
The chances are even.
19.8xb6 Exb6 20.2xb5 ¥xb5 21.a4 would
give an endgame where Black is by no means
worse, but maybe it is still too soon to describe
his position as preferable.
19...8xc4 20.8xcd Dxcd 21.¥xcd
22.8xb2 Bxb2 23.8e2 Exe2 24.%xe2 Was
now starts to weaken the white kingside. Strong
players know how to accumulate advantages
in this way in positions where the opponent is
passive, so that when they finally try to force
their will on the opponent there are as many
points of attack as possible.
29.h4 c3 30.%f3
It would be easy to criticise this move, but
passive play would also not be able to save White:
30.8al W4 31.8a2 £5! 32.£3 fxed 33.fxed Wh5!
34.2h3 Bb2 35.Exb2 ¥xb2 36.Wa4 &f7 and
the black pawn queens.
Lxb2 /%/@/
///

// 7.7
,,,,,,,,,, /,

/ gz
,,,,, s// 0 30...%b5!
/ //g//
This wins because ..Hb2 now cannot be
prevented by Hal with the idea of ¥xc3, as the
king is on the third rank.
The endgame might not look that bad for 31.a4 ¥b4 32.a5 Eb2 33.Waq Wcs!
White at first glance, but a short list of Black’s White would not be in trouble if Black did not
advantages should help us see how bad things have this move - but he does!
really are. Both e4 and a3 are weak, Black will 34.8xb2 oxb2 35.We8t g7 36.8b8 Wc2
occupy the only fully open file, and the black 37.a6 b1¥ 38.a7 Wxe4 mate!
pieces are all better placed than the white ones. 0-1
Finally, Black’s activity allows us to imagine that
the white king is in much more danger than its In the next game we have a clear motif: the
black enemy. All in all, a full point should be a exchange of the c3-knight. It is because of games
picnic for Black. like these that White hurries to play a2-a4 and
25.%d3 Eb8 26.2a2 Bb3 27.Wc2 c4 28.¢g2 b2-b3, keeping control over the light squares on
h5! the queenside.
Black has full control on the queenside and
 Introduction to the Classical Benko Scructures 23

Game 6 White is depending on having his rooks on

Herraiz Lopez — Illescas Cordoba light squares, so Black decides to keep the bishop
Albacete 1999 for the task of harassment.
23.b4
1.d4 D16 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.bxa6 g6 White tries to give up the pawn, but with close
6.g3 to identical pawn structures, Black will be better
White is declaring his intentions a litde because of piece activity. White would also find
bir early, which allows Black to expand his no joy in lines like this: 23.&2h3 £xh3 24.2xh3
possibilities (see move 10). More standard A cd! 2553 Dxa3 26.bxad Dxbl 27 Hcl Bb4
would have been 6.8c3 and then 7.g3, when 28.a5 a3 29.2al Ha4 and Black wins a pawn,
the threat of 7.e4 would have to be answered and probably later the game.
with 6...2xa6, or Black could risk ending up 23...8f5 24.2al xb4 25.axb4 Haxb4 26.8£g2
in a version of the Benko similar to Yusupov b2 27.13
— Vitolinsh above.
6...d6 7.2g2 8g7 8.Dh3 0-0 9.2c3 Dbd7
10.0-0 Exa6 11.%c2 Hb6 12.8d1 £b7 13.Eb1
Wd7 14.064 E //? 1D
White cannot prevent the exchange of knights.
14.b3? & bxd5 and Black picks up a great pawn
. / zxf@’x
for nothing.
W /8@/.&/5//2
P 5 5D
//3/4,”
/.é.x .Z /238/
42 /, L, o S 2

14..Da4! 15.Dxa4
Positions like that after 15.e4 &xc3 16.bxc3
Bfa8 17.a3 £c8 18.h3 Wad 19.Wxad Exad 20.13
Ad7 21.89e2 ©e5 might be a pawn up for White,
bur he s still struggling 1o keep equaliry.
15..%xa4 16.¥xa4 Exad 17.a3 Bb8 18.2d2
£a619.8¢3
From here the bishop protects the queenside,
but then it is rather easy to remove it.
19..2d7 20.8xg7 xg7 21.8f1 De5 22.¢3
£c8!
27.. .flxg21‘!
Strictly speaking this was not necessary, but
since the front has become rather short (d- to h-
file) the minor pieces are increasing in value and
the heavy artillery becomes less effective - so why
not have some pawns for an exchange?
28.0xg2
28.dxg2 Bb2t 29.%f1 Hixf3 30.h4 £d7 and
Black wins with ...&b57.
28...0xf31 29.2£2 Dxh2 30.2a2 Eb3 31.8d4
Dgat 32.%e2 g5
Dominating the g2-knight. Black is winning,
33.Raa4
Being active with 33.Ba7 was possible, but
after 33..9f6 34.Hxe7 Sgdt 35.8f2 Eb2T
36.82g1 Black will simply play 36...2d7! and
after 37.Del 28 38.8xd7 Hxd7 39.9f3 ho6 he
24 The Benko Gambirt

has a very good chance of winning the game with Bishop surprises
his extra pawn and better structure.
33...h5 34.Bab4 Ea3 35.2a4 Eb3 36.Eab4 There are two simple bishop manoeuvres/moves
Ec3 37.Eb7 that often surprise White players in the Benko
37 8bcd £d37! Gambit, ...&c4 and ..&xc3, often winning a
37..kf6 38.82b8 Hc2t 39.Bd2 Bc3 40.Bb4 pawn. There is not really a lot to explain or
&es understand about them, so let us see how they
look when executed in practice.
The first of the two bishop surprises is basically
a double threat on d5 and a2. A classic example
7 / /7 /////%7/
/% // Ak went like this:

Game 7
I. Farago — Knaak
Novi Sad 1979
P n
.. BE A& & ; z K
A
1.d4 D6 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.bxa6
£xa6 6.Dc3 d6 7.2f3 g6 8.g3 Lg7 9.82h3
), BT 5) Not the most standard, but still seen from time

/mE .
o

to time.
9...2bd7 10.0-0 b6 11.8el 0-0 12.Eb12!
White wants to play b3 and £b2, but Black
It seems that White resigned right after the has a surprise ready for him.
time control, probably demoralised by the

E %7%@/
coming ...&e4 when he will be unable to defend

/ ///é%
e3 and g3. The following endgame is easy to win
for Black.

7/// , , , // /// , , , //////

0-1

Though this passage has been on long knight

manoeuvres there is another fact that has made
/%/ i
itself apparent in these games. Black is usually
doing OKin the endgame in the Benko Gambit. It is
a funny gambit in the way that the pawn sacrifice
% A% I
is not only connected to a lead in development 3 »
and attack, but just as much to achieving the
= =
desired pawn structure. White s a pawn ahead,
but he gave up his ¢-pawn for the less important
a- and b-pawns. However, if White is successful in 12...8c4!
blockading the queenside without Black gaining Here it is - double threat on a2 and d5. Now
any kind of active play, he is usually better. The Ivan Farago is a grandmaster and not one to lose
idea is basically that whether or not the queens are faith over such minor details as losing the pawn
on the board is not that important for the evaluation he won on move 4.
of the position. Basically, the difference from a lot 13.e4!
of other gambits is that Black is not mainly playing But of course now Black has a better pawn
for ractical tricks. His compensation is positional. structure, so White needs to be active to make up
 Introduction to the Classical Benko Structures 25

for it. 13.b3 Dfxd5! 14.5xd5 £xd5 with a slight will win a piece for White, who then is at least
advantage for Black is a weaker option. no worse.
13...8xa2 14.9xa2 Bxa2 15.2d2 b) 22...2a8 is unfortunate, based on 23.d6!
White is maybe using a couple of moves to when the position suddenly is very perilous for
exchange the knight on 6, bur this ensures his Black. e.g. 23.. .83 24.0xf6T exf6 25.d7 Df8
breakthrough in the centre. 26.2b3! Hc2 27.8g2 2d8 28.8¢6 He2 29.0f1
15.e52 is too early and just drops a pawn to Hxelt 30.%xel and White wins based on
15..22fxd5, but interesting was 15.b4!? c4. 30...5a2 31.8e3 followed by Ze87.
Generally [ would favour the passed c-pawn, but ¢) 22..2d8 is probably best, though still met
there is the case of the bishop suddenly being well with 23.d6!. Now the best is probably 23...Hxd7!
placed on h3 and then White has two bishops. | 24.8xd7 exd6 and Black seems to have a slight
think “chances for both sides” is a fair description edge in the endgame, as after 23...8¢3 24.2b7
of the position. Of course 15...cxb4 16.2xb4 He2 25.80xf61 exf6 26.8c1 there is hardly any
#bd7 should equalise pretty ecasily, if this is what advantage for Black to be found.
Black wants. 19...2xb2!
15...%a8 16.8c3 A not too hard combination for the East
16.b41? was interesting again, but it is not clear German grandmaster.
that the bishop is better on d2 then. 20.Exb2 Ba3 21.¥b6 Exf3 22.¥xc5 e4 23.8b5
16..0Da4 Wa2 24.8g22?
Here I like 16...Bb8! based on 17.8xf6 &xf6 A blunder, buc the situation already looked
18.e5 dxe5 19.6)xe5 WxdS! and Black is doing bad after 24.8xed Zc3! 25.Wbd Eclt 26.8g2
very well. So White would have to do something Wal and the white king is in danger.
else, which is not so easy to see. 24...e3!
17.8xf6 £xf6 18.e5!
Otherwise White has no counterplay.
18...dxe5 19.¥b3?
This fails to a fairly simple combination. The
right approach was to believe in the counterplay
in the centre with 19.8xe5 when after 19...2xb2
20.%c2 Wad 21.Wxad Hxad 22.80d7! Whire has
decent counterplay, as can be seen in a number
of illustrative lines:
.e / AE
/ 2 /,,W/V/,//,,,/,A

. B &
é Now f2 cannot be defended and White will
simply be two pawns down.
0-1

[ will not go further into the whole &c4-thing. It

should be clear enough that Black usually wins
back his pawn like this, and White then struggles
a) 22...2d4 23.5)xf8 £xf2+ 24.%f1 &xel does to keep equality, but is by no means lost.
not really work. After 25.8b8! the threat of mate
26 The Benko Gambit

The other main bishop trick is ...&xc3.

%
/ 2/
//,,,,,;&////-
/ @ » //E
_ . M\g/ ///%
Z 7= A

& 7// % / /,,,,, % %:///,7/{ ,,,,,,

19...8xc3!
Often this exchange can come as a surprise 19...2d4%% first could be marginally more
to amateurs, as they somehow believe that the precise, as indicated by Fritz 8, but as both moves
weakness of the dark squares around the king is remove the defence of the white queenside it is
much more important than it often is. But please rather academic.
do not misunderstand me. Certainly giving up 20.%¥xc3
the bishop is a concession. The knight is from the 20.8xc3 ©a3! wins the exchange.
start a weaker piece and the g7-bishop is very well 20...Exa2 21.¥xb4 Exb4 22.b3 Da5
placed. But many times this is the move Black White had had enough and quite respectfully
needs to break open the white queenside or to resigned.
win a vital pawn. In these positions Black should 0-1
not fear that he will be easily mated on g7. This
is very unlikely to happen. The basic example of Aseev — Khalifman
this theme is something like this: St Petersburg 1995
Game 8
Asanov — Rashkovsky
Pavlodar 1982
,,,,,,,,,,,

1.d4 D6 2.913 g6 3.g3 £g7 4.8¢2 0-0 5.0-0 /

PO
d6 6.c4 c5 7.d5 b5 8.cxb5 a6 9.bxa6 £xa6
10.5c3 Dbd7 11.¥c2 Was 12.Hel Efb8 13.e4 =
g= 8/
Dgs 1481 &xfl 15.8xfl Dge5 16.Dxe5
Dxe5 17.f4 D4 18.2f3 ¥b4 19.8b12
Everything has gone wrong for White, who is
clearly out of his league, and now he commits his
final blunder. After 19.b3 ©a3 20.2xa3 ¥xa3
Black has the advantage. The dark squares in ,,,,,,,,,,,,

White’s position are very tender and the pressure

on the queenside pawns is horrible. Still, this was The white queenside is often held together by the
the best option. knight on ¢3, and the bishop on g7 can, as said,
 Introduction to the Classical Benko Structures 27

be exchanged advantageously under the right 29...9xb1? 30.8xb4 cxb4 31.Ed1! does not
circumstances. Here is another example where seem to promise Black any real advantage.
a future FIDE World Champion shows how a 30.Bal
grandmaster can be beaten with this theme: 30.Hc1 Exb2 and Black will eventually win the
23...0e5 24.8d1 ¥b4 a-pawn.
First Black puts the queenside under as much 30...8c2 31.8b1 Hxa2 32.8c3 Ha3 33.8cl Dcd
pressure as possible. 34.g5 Ba7 35.8b1 Hab7
25.8b1 Dc4 26.2d3 and Black should win the endgame, but still
26.a3 ¥b3 27.Wxb3 Exb3 would give Black needs to display good technique.
typical Benko-compensation at its best. White 28.. Wxedt!
would have very few chances of saving the game. Maybe this move surprised Aseev.
26.£d2 is also questionable, because of 29.23
26..5%xd2 27.¥xd2 &xc3 28.bxc3 ¥xedt 29.3 As so often in modern chess, the real beauty
Wxbl 30.Exbl Hxbl 31.%kg3 Hal and the is hidden in the sub-variarions rather than
endgame holds little promise for White. Still this displayed on the board. 29.f3 We2t 30.%¢3
could very well be the best practical chance. Hxbl 31.He3 ¥fl
26...8xc3! 27.¥xc3
7
27.8xc3 a3 just wins the exchange. /% L

7, - 0

,,,,,

ER W
32.82b2 (32.Eel Hxa3!! is the stunning point of
the whole tactical operation) 32..%g1+ 33.%h4
Wt 34.5g5 h6T 35.404 Whot 36.50e4 Badt
Now Black can win his pawn back, but the and Black wins.
possible knight check on el (!) gives him a 29...8xb1
stronger tactical option. Now Black is cleatly on top.
27...0a3! 30.2b2 Wd4! 31.¥xd4 xd4 32.8xd4 BbS
A very subtle move that uses all the undefended 33.8c3 Bxd5 34.8¢3 Bda5
pieces in the white position. 27..8xa2 is less 0-1
convincing. After 28.%xb4 Exb4 29.b3 White is
ready with £d2, and Black will have to accept But, of course, it is not always the a2-pawn that
only a slight edge after 29..8e5 30.Ee3. eg. is won with the bishop exchange. At times ic is
30...c4?! 31.8c3! and White is alright. the d5-pawn that will have to go back in the box
28.bxa3?! after the bishop exchange. An example where
This loses by force, but by no means trivially. Black was happy to make this exchange is the
White had a better defence in: following:
28 Wixb4 Hxb4 29.£d2!
29.8al? &c2 30.2b1 HelT wins a rook.
29...8bb8!
28 The Benko Gambit

Oblitas Guerrero — B, Lalic 15.a4!

Manila (ol) 1992 White is aiming for the a4+b3 structure, when
he could hope for a slight advantage. In doing so
he is disregarding the d5-pawn.
15...%b6 16.b3!
7 7 7 y/
,,,,,,,,

h A S / ,,,,,,
White has two problems in the position,
something the Classical World Champion
-9-/ i )//@‘ 0 nacurally had appreciated in advance. After
. A 8 /¢ ) 16.e42! Wb3! Black’s compensation is easy to

/m/%;4
understand.
16...8xc3
Black has been heading this way, but maybe
,,,,,,,
16...82b7!? was a better choice - see page 22.
17.8xc3 Dxd5 18.9e5!
This is what White was intending. This
//////
combination leads to a position with a better
pawn structure.
17...8xc3! 18.¥xc3 ¥xd5 19.b3 Wes! 18...0xe5
The main tactical point behind Black’s play. This seems to be the lesser evil. After 18.. 2716
20.8d2 19.5¢c4 Wa7 20.8xf6 and 18...dxe5 19.£xd5
Here the dartabase says White played 20.2b2, Black’s position is rubbish.
but, as 20...8xb2 wins on the spot, I prefer to 19.8xd5 &b7
believe the best. Black had to choose between two evils. Should
20...¥xc3 21.8xc3 Da3 he allow a dangerous kingside attack or should
Black had a slightly better endgame, which he he try to hold a worse endgame? By choosing
won comfortably. the latter Topalov is playing in his opponents
backyard, bur still he was very close to actually
However, at times taking the d5-pawn can be
drawing the game. 19...8)¢6 20.We4 b7 21.Wf4
dubious. In the following example Kramnik
?d8 22.a5 and White is better - Igor Stohl.
forces Topalov to take the pawn, as Black is
20.8xb7 Exb7 21.8xe5 dxe5 22.Eecl
running out of sensible alternatives:
Kramnik - Topalov
Wijk aan Zee 2003 z/ / /@%
1% 7 Fel
T aasi
g_/) 5 W’W
% 3
1 21 ///%
z/&% _ ///
/ 0 /M%
By i / E // é
&/@/”&%g%
,,,,,,,,,,,,,,,,,,,,,,
 Introduction to the Classical Benko Structures 29

White has a lasting plus, but hardly enough Ezat — Sergey Kasparov
to win the game against the best defence... Abu Dhabi 2003
...1-0
(For more detailed coverage of this game and the
concluding moves, see chapter 3, page 60.) g,
//.ém//7 , ,/
,
There are other kinds of tactical tricks related
to the “bishop surprises”. I could try to explain
them with words, but in this case I find truth
//
in the old saying: A picture paints a thousand
words...

Ballesteros Gonzalez — J. Polgar

San Sebastian 1991

= o /
///
13.b3? &xe2!!
The strength
tactically.
of the two bishops is felt

14.¥xe2 8xc3 15.8b1 £g7

Black is a little better and went on to win.

The last bishop surprise is a surprise sprung

on the bishop, and not the other way around.

/ / /
Generically it looks like this:
/

’////z@
,,/é

//////

19...50xd2 20.%xd2 Bxb2!

/ ,,,,,,,,,,,
Utter destruction of the dark squares.
21.%xb2 &xc3 22.¥c1 243
7
,,,,,,

Black won comfortably:

23.8d1 £xbl 24.¥xbl ¥Wxa2 25.¥Wc1 £d4
//
7
26.%d2 b3 27.%d3 Wixd3 28.8Bxd3 EHa2 /8%
29.h4 Exf2 30.5h2 Le5 31.50g1 Bb2 32.8f1
¢4 33.8a3 ¢3 34.8a1 £d4t 35.2h1 2 36.82d3 é//8&
£e3
0-1

Here White is threatening the e7-pawn,

Or like this:
or is he? After 1...h6! 2.8xe7 He8 3.2h4 g5
the white bishop is trapped. So it is not always
necessary to protect the e-pawn. The following
example from a high level game is a good way of
finishing this chapter.
30 The Benko Gambit

Gelfand — Ivanchuk 22.9\e4 fixal 23.8xal £xe2 24.8cl

Monaco (blindfold) 2002 24 2xa8 Welt with mate to come.
24...8b5 25.0f61 g7 26.De4 Wd42!

E/ E g
26...81d7 was simpler.
27.50xd6 £d3 28.Wd2 He2 29.Wf4 Wixf4
30.gxf4 Hea2 31.h3 Dd7 32.Ec3 Bd2 33.50h2
Haa2 34.b4 oxb4 35.8Bc7 b3 36.Bxd7 b2
37.8xf7+ Lg8 38.8e7
/4/8/// / ,,,,,,,,,,,
7 7 T
///////

Though thematically this example is a good

illustration, it is a bit more complex. To see
examples in their simple form and then see how
world-class players use them to defeat each other
in a more complicated form is very instructive.
Of course sometimes ideas such as the one by
Gelfand in this game work, but still this is only 38...He2!!
because he understands the theme, and finds an 38...H2a8?! 39.8e4! was a great trap. Now
exception to it. A fancy word for this is “pattern 39..2e2! sdll favours Black, while 39..b1%?
recognition’”. 40.0f61 2f8 41.d6! seems to give White
16...h6! enough counterplay to draw, all based on the
16...Bfb8!? was also strong. The idea is that following murky line. 41...g5!? 42.d7t @gS
after 17.&xe7 ho! the bishop is still in trouble 43.8xa8 igG 44 82d57! Bxd5 45.9(67 and Black
on e7/. should just take the draw with 45...€0f8, as after
17.8xe7? 45...0h8 46.9xd5! gxh4 47.d7 b8 48.5c7!!
17.8c1 was best, but this was not what White Wxc7 49.He8t g7 50.d8W Wxf4t 51.cbg2
had intended with 16.2g5. £xe8 52.Wxe8 only Black can be said to be worse
17...EBfe8 18.b3 (though it is a draw still, of course).
This was Whites idea, but it is not really 39.8xe2 fxe2 40.8e4 Ea3!
tenable. But then we should not be too harsh on An important move.
the players here. After all, it is a blindfold game. 41.f5 £d3 42.8xd3 Exd3 43.f6 b1¥ 44.f7%
18...9xa3? g7 45.0xg6 BExh31! 46.%xh3 ¥xg6
18..8¢cb6 19.82xd6 Exe2 wins instantly, so 0-1
does 18...5)a5!.
19.8xa3 Bxe7 20.8c32
20.8f3! was the last chance, though after
20...Hae8 Black is dominating the game.
20...%b4 21.Efal b8!
Now Black wins material.
Chapter 1

The Benko Gambit Accepted I

1.d4 &6 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.bxa6 g6 11 none other than Kasparov makes a guest star
appearance in the book, opposing a slightly odd
manoeuvre from Evgeny Bareev.
,,,,,,,,,,,,,,,,,
Game 9
Radziewicz — Pinski
Jaroslaviec 1995

//f/%/
» 1.d4 D6 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.bxa6 g6
5...e62! is a move from a fairy tale. After 6.8)c3
Hxd5 7.89xd5 exd5 8.¥xdS it is not realistic
that Black will get sufhicient compensation.
A
// (Already now he can lose the game in one move
,//,
/8// i
,,,,,,,,,,,,,, with 8...Bxa6? 9.¥e5T winning a piece.) 8..2¢6
2=)
/////////////////
9.3 &e7 10.£d2! White immediately contests
Black’s activity on the long diagonal. 10...0-0
11.4c3 Wb6 12.8d1 d6 13.2d3 £e6 14.We4!
This is the starting position of the Benko This is better than 14.%h5 h6 15.a3 De5 with
Gambit Accepted, the most fashionable way to some counterplay, Knaak — Pytel, Zabrze 1977.
battle against Black opening in the beginning of 14...g6 15.8f3 White has a clear advantage due
the 21% century. White has won a pawn and Black to his extra pawns. However, it is not clear that
is developing rapidly, depending on his better 12.8d1 leaving a2 unprotected, was necessary.
pawn structure and two open files to support his 5...g6 is in my opinion the best move order.
slight lead in development. Black should delay playing ...&xa6 until White
Black’s last move probably requires some has played ©b1-c3. If White should attempt to
explanation. Basically g6 is the standard do without this move, Black can take on a6 with
development, but against some White set-ups the knight. The history behind 5...g6 becoming
Black can advantageously recapture with the the “industry standard” in this position is based
knight or rook at a6. To understand perfectly on the following games:
why Black chooses this move order, see game 20 5..&xa6 6.g3 d6 7.8g2 g6 8.b3! Kg7 9.8b2
in chapter 3. 0-0 10.5h3 &bd7 11.0-0 ¥b8 12.8c3 Ec8
In this chapter we shall look at less fortunate 13.8el Ha7 14.9t4 Bb7 15.£a3 and White
systems for White. In game 9 he will try to attack is better, Portisch — Geller, Biel 1976. Portisch
(me!) on the kingside without any development, won this game and some people believed that the
a strategy that is continued in game 10. [n game Benko Gambit was done for. First of all Black
32 "The Benko Gambit

can play better than Geller. Secondly, the 5...g6
move order would prevent Portisch’s plan. But
Black can also improve on move 11:
a) 11..2a71 12.8e1 Wa8 13.5f4 Bb8 14.h4
b6 15.8d2 Hab7 16.82c3 and according
Ernst the position is unclear.
b)11...%b612.£c3Eb813.0d2De8 14.8xg7
Dxg7 15.0f4 fe5 gave Black compensation for
the material, Kovaliov — Vetemaa, Minsk 1981.
So the 5...8xa6 move order is playable,
why give White this extra option?
In chapter 3 we shall look at what happens if
White tries to play this set-up against 5...g0.
6.0c3 £xa6 7.4 £xf1 8.8xf1 d6 9.g42!
Black’s interest with the queens still on the board,
though an immediate mate is hard to see.
The game continued 16..%xcl 17.Ehxcl!
where White is slightly better. (There is nothing
to for White on the kingside but unfavourable
exchanges, therefore the recapture with the h-
rook.)
10.gxh5?
'This makes one wonder why White played g2-
but g4. Now White is left with a weak pawn on h2
and a fragile king’s position.
10.g5 was better. Now after 10...2fd7 1 1.shg2
de7 12.f4 £a6 13.8f3 0-0 14.We2 Wc7
15.8¢3 Efb8 Black has excellent compensation
for the pawn. In Sakovich — Lanka, Riga 1980,
White played 16.8ad1?! when after 16..%a5
17.8c1 &b4 18.a3 Da2! the white queenside is

A /
crumbling. This is an unorthodox version of the
exchange of the c3-knight, but definitely viable.

/t/&/&/
Better was therefore 16.2hcl a5 17.0d1 Wa4
where Black has sufficient compensation.
10...2xh5
Black is already slightly better.
11.cbg2 87 12.8ge2?!
This seems a little irrational. The natural square
for the knight is 3. 12.2f3 &d7 and Black is
slightly better.
12..8d7 13.9¢g3 ¥c8!
////////////

This little move order is worth remembering.

A move like this is, in my opinion, only Here I deliberately played ...&3d7 first and only
dangerous for White. But of course Black will later .. W8, as | wanted to hide my threats
need to play well to prove this! against the white king.
9...h5! 14.£4 Db6 15.913 D4
This cancels all White’s chances for an attack Perhaps stronger is 15...2xg3 16.hxg3 Exhl
on kingside once and forever. Other moves to 17.8xh1 Wh3t 18.g1 £5! with an attack. But
take a note of are: [ could not help myself from playing for a mean
9.%c8 10.h3 &g7 11.8g2 Dab 12.8ge2 trap.
0-0 13.2f4 ©d7 14.¥d2 He5 15.8xe5! Lxe5 16.b322
16.8acl and White is slightly better. Haba — White gracefully falls for the trap. Bur even
Palkovi, Austria 1997. after something like 16.%ge2 f5! 17.Wd3 fxe4
9..8g7 10.g5 Dhs 11.0ge2 W8 12.0g2 18.Wxed Wodt 19.8f1 &xc3! 20.bxc3 216
Wedt 13.0¢3 D4t 14.2xf4 Wxf4 15.h4 h6 21. g2 (21.Wxcd?? Wf3t 22.0gl Ded and
16.¥c1 Komljenovic — K. Berg, Copenhagen White cannot protect the 2-square) 21 . Wrxg2t
1989. Now after 16..hxg5 17.hxgs Wg4 (21...%15!? with a continuing initiative is probably
18.2xh8+ £xh8 19.f3 ¥c8 20.¥d2 White has even stronger. White’s king’s position is very
a slight edge. The opening of the h-file is not in weak.) 22.@xg2 &xd5 and Black is much better.
 The Benko Gambit Accepted 1 33

Best was maybe 16.h3 Wa6 and Black is at Game conclusions: White started the opening
least slightly better. White is trying to control the normally, but then went for a counterintuitive
black pieces, but they are roaming everywhere on artack with 9.g4?!. 'This move has been played
the board, so White will have a busy job closing even by strong grandmasters, but violates the
the holes in the dike with his fingers. basic rules of chess and, not surprisingly, Black
should get good chances with correct play.

z/y/ E Game 10

_ / {,,.9,3 Andruet — Fedorowicz

/// Y
Wijk aan Zee 1989

. & %/}m 1.d4 D6 2.¢4 ¢5 3.d5 b5 4.cxb5 a6 5.bxa6 g6

/ /135y
6.2¢3 8xa6 7.f42!

%% i 4 ’g@/ /z
7 &K
B 4 % /////
16...¥%h3+1!
Now White is mated or suffering decisive
material losses. In his continued grace White
,,,,,,,

/éa /&/ %&/

allowed himself to be mated. 7/’// /
17.$bxh3
17.@g1 £d471 and Black wins.
,,,,,,,,,,,,
17..Oxf4t 18.2gs Dest! 19.xf4
19.¢0g5 ©h3 mate!
19...2h6 mate! Similarly to g4 in the previous game, this
move is too much too soon. White wants to

I e & play &3 and e4, to recapture on f1 with the

g / /z,,,,{/
rook. If he is successful doing so his position
will be good. However, according to my classical
understanding of chess this does not seem like
/ an appropriate way to play chess. White is

% ?” behind in development so opening up the centre

does not seem sound. So it is quite logical that
Black’s strongest continuation against this line is

//852 very concrete and based on disturbing White’s

development.
7..8g7
7..Wa5 is also possible. After 8.2d2 £g7 we
would transpose to the game.
8.2f3
8.e4 makes little sense here, as White wants to
recapture with the rook.
34 The Benko Gambit

8..Was! and White is slightly better in the endgame. But

This is played solely to prevent e4, though of not 16.8xg7? @xg7 when we have a standard
course the queen is a little more active on a5 than Benko position, with lots of play for Black.
d8. But can Black really prevent e4? No of course e.g. 17.8cl &c7! 18.a3 5 19.exf5 &xd5 20.g3
not, but he can win time by annoying White. gxf5 and Black was much better in Rivas Pastor
8...d6 is slightly dubious, as White gets — Matamoras Franco, Orense 1995.
everything he wishes for - and a pawn! 9.e4 13.¥xa6 Bxa6
£xfl 10.8xfl1 0-0 11.e5! The direct assault is This is the position Black has been heading
the best, though 11.%2!? was playable as well. for. Practically all endings with rooks on the
11..80e8 12.5f2 ©1d7 13.%e2 b6 14.8d1 Hc7 board are good for Black in the Benko Gambit.
15.2e3 and Whice is slightly better, Naumkin Especially when White has a weak pawn on e4
— Vaisser, Budapest 1989. e.g. 15...Ha7 16.8d2! as a consequence of the f4-move. Here Black’s
(After the sharp 16.b4?' ¥Wa8 17.¥d3 (17. position is, in my personal opinion, a teeny tiny
bxcS Dbxds 18.8xd5 &xd5 19.Bxd5 ¥xdsS bit better.
20.cxd6 8d7 is just unclear) 17...cxb4 18.8xb6 14.52d112
Ea3 19.8xc7 Bxc3 20.¥d4 Hxc7 21.Wxb4 Hcs The king is usually not a good defender in open
we have a typical position where Black has full positions where lots of pieces roam the board,
compensation, if not even a bit more.) 16...%a8 and the same goes for this position, queens or no
17.H2adl and White stays in control. The black queens. But still it seems that this plan is not as
counterplay seems to be diminishing. insane as it looks. At least Black was not able to
9.8d2 0-0 10.e4 d6! punish it without assistance in the game. 14.%e2
Black placed his queen on a5 in order to 8b8 15.8hb1 was the alternative.
prevent White from castling. So 10...2xf1?!
does not really make sense: 11.8xf1 d6 12.e5%? @
(This is better than 12.8f22! ¥b6. Another plan
behind putting the queen on a5. Now b2 is no
/.Q:t7
longer well protected and Black is preparing
.4t 13.50gl Wxb2 14.%e2 Wby 15.8abl
Wa6 16.¥xa6 ©xa6 17.8b7 Afb8 18.Efbl
@g4 19.h3 De3 20.8xe3 &xc3 21.Exb8t
Hxb8 22.8xb8% & xb8 and Black is slightly 77
76
better because of the plan ...Ad7 followed by
..f5, when d5 becomes a problem child for
White, Tozer — Tseidin, Hastings 1991.) 12...
dxe5 13.fxe5 Dg4 14.%e2 Wb6 15.0-0-0 Hd7 Now Black has a few interesting options:
16.£f4 In this interesting and very unclear a) 15...8e8 isabirslow. Whire can respond with
16.g4!. Not the beginning of a kingside attack,
position the players agreed a draw in Kopylov
— Pinski, Warsaw 1997. The decision might
burt prophylaxis against Black’s idea of ...¢7 and
seem odd, but in my defence I can say that it ...f5, undermining the d5-pawn. 16..2¢7 17.g5
was the second round that day. Here the position is rather unclear.
b) 15...%xe4? is bad because Black wins only
11.8xa6 ¥xa6 12.We2 Hbd7
Black needs to develop or his lead in two weak pawns and a passive rook for two of his
development will easily disappear. Worse was most “excellent pieces”. 16.9xed &xb2 17.4)c3
12...0¢d72! 13.3d1!. Now, when the e4-pawn &xal 18.8xal Ha3 19.8el D6 20.05c2 HaG
is not under actack, this becomes possible. 21.%e3 and White is better.
13..Wxe2t 14.85xe2 £1ab 15.8c3 Bfb8 16.a4!
¢) 15...c4! The most energetic - which is quite
often also the best in the Benko Gambit. 16.h3
 The Benko Gambit Accepted I 35

55 17.e5 D7 18.63d4 dxeS 19.8c6 Ebal that are protecting each other usually do not
20.0xe7t Bf8 21.8¢6 e4 and Black retains a function well, but there are exceptions, as here.
dangerous inirtiative. 19.a3! seems necessary, or b4t will continue
14..8g4 15.8c2 c4® to spook White. 19..Eb8 20.Ebl Hab6 (the
As you have probably seen in the Introduction, entertaining 20...83h1?! does not really work
it is very important to be very careful when you after 21.e5 Dg3 22.8e3! Hxf4 23.)d4 and
play this move. It is hard to remember, even for White is a lictle bit better) 21.9ad! a6 22.8¢3
strong players, that you cannot move the pawn (22.b3? Bba8 23.2el cxb3t 24.Hxb3 Hxad
backwards later on. Black creates a weakness in 25.6Vxd3 Bcdt 26.80b1 &ixed and Black is better)
his own camp, but at the same time he has the 22...Hab6 with repetition of moves as the logical
chance to attack the white weaknesses on e4, b2, outcome of the game.
d3 and, not least, the king. 19.9el £c5 20.¢5
16.Bhel Hc5 17.8e2 After 20.g4!? normal play with 20...8b8 should
17.b4? cxb3t 18.axb3 Hxal 19.8xal Hxed! offer Black sufficient compensation for the pawn.
and Black has a clear edge. Bur not 20...0b3? 21.axb3 Hxal 22.bxc4 Ec8
17..2d3 23.b3 and White is better.
Here my sources say that Black played 20...0fd7 21.8e3 Bb8 22.8xc52
17...5)d7, but this move make little sense, and Very odd. The bishop is quite useful and,
seen in connection with the 19cth moves, this what is more important, Black has two knights
must be the move actually played in the game. that want to occupy one good square (c5). Now
18.h3 White has solved this problem for him.
18.¢5 does not really work, as after 18...dxe5 Better was 22.8d4 g51? 23.g3! (23.8xc5 sill
19.h3 &6 the d5-pawn is hanging after ...Ab4t does not seem right. After 23..89%xc5 24.exd6
next. exd6 25.fxg5 Dad 26.Dxad Zxad 27.sbcl Has
28.0f3 #xdS Black’s compensation is more
than enough.) Now after 23...dxe5 24 fxe5 Bh6
25.8d1 Bxh3 26.e6 £xd4 27.8xd4 fxe6 28.dxe6
H8 29.8xc4 Dexe6 30.8e4 White is slightly
berter placed in the endgame.
22...0xc5 23.8d1

/ R e
///Z‘% 7
,,,,,,,,,,, %x@x
s o=¢ Jiy8/
JiY i Zfig//
/éé/
E 7// A //
Fuihom
18...22f612
A litde passive, and maybe something that
N oo #a
g
could have landed Black in a slightly inconvenient
situation. | am quite attracted to the adventurous
18..4)gf2!. 'The knight is not immediately
threatening anything, bur the knight is closer to
the action. Of course, it is known that knights 23...0a4!
36 "The Benko Gambit

The c3-knight is thematically removed. (34.Bgl1? would repeat the same passive strategy,
24.9xa4 Bxa4 25.8al?! to Black’s great pleasure. 34...g5! (34...8xe5?!
This is too passive. The rook is the strongest 35.fxe5 He2 36.50b5! and Whice is at least OK)
piece on the board, it cannot be packed away in 35.g3 He2 36.8c1 &xe5 37.fxeS Hxe5 38.sbb4
a corner. Hedt! 39.6b5 (39.8cd Hxcdt 40.hxcd Bf8
25.a3 would allow Black to undermine the 41.0c5 de7 42.a4 5 43.a5 £d7 and Black
white centre with 25...g5! 26.8f3 gxf4 27.8d4 wins) 39...2e3 40.a4 8xg3 41.a5 Bxh3 42.Hal
a5 28.8xf4 BxdS 29.8xc4 dxe5 30.b4 £5 31.a4 2b3+ 43.%c6 Hb8 44.a6 h5 45.27 Ba8 46.%2b7
e4 when Black is slightly better. Hxa7t 47.8xa7 h4 48.%c6 h3 49.8al g4 and
25.$bb 1! was interesting. Black could continue Black wins) 34...8xe5 35.fxe5 Hxg2 36.59b4 and
with 25...c3 26.b3 Exf4 27.exd6 exd6 with a White should be able to draw with the a-pawn.
definite advantage. 31...85b7!
25...8a5 Not the only winning move, but the most
25...g5"? was strong too. direct.
26.%2c3 0-1
26.9\f3 Hxd5 27.%c3 Bc8 and Black is better.
For example: 28.24 &h6! Game conclusions: 7.£42! is a sharp and treacherous
26...8Bxd5 27.xc4 variation, but I have my doubts about its
White is playing a risky game. Objectively it soundness. In the game Fedorowicz proved that
is sound, but as we shall see, the risks had a very control over the centre is important in this line,
real practical nature. and that Black with active play can get a very
27...¢6 28.0)f3 B8t 29.82b3 b5+ good game indeed.
29...2b8+% with a draw was of course a
Game 11
possibility.
Bareev — Kasparov
30.0a4 Bcb8 Linares 1994

1.d4 &6 2.c4 c5 3.d5 b5 4.cxb5 a6 5.bxab g6

o ///xz,éx 6.3 £xa6 7.9c3 d6 8.2d2

fif ’\!@f/ E
7,,,,,, A a{//}

// ///
%/
/ Zfi’é/"""" 0
// / //
% 28 B
31.exd6??
mww
///////////
A horrible blunder that loses on the spot.
31.a3! was the only truly tenable move here to
prepare b4 and prevent £b4t, 31...dxe5 32.8)xeS This is not an especially dangerous variation.
Hxb2 33.8xb2 Exb2 Now White has to choose. White wants to retain the possibility of castling,
34.8cl! to create counterplay is the best strategy. but this is only achieved at the expense of a long
 The Benko Gambit Accepted 1 37

journey for the knight, which afterwards is out are counselling on the a-fi le. 24...f6 More or
less forced, even though it does encourage the
of contact with the eS-square, which is where
White would hope for a breakthrough. Having coming blow. 25.e5!!
pointed out so many flipsides to a single coin, it
is of course important to stress that the move is
not a mistake, just not the most dangerous.
8..8g7 9.e4 8xf1 10.Dxf1 ¥a5!
When the light-squared bishops disappear it is
favourable to control the a6-f1 diagonal.
11.8d2
White is placing all his pieces very

harmoniously, but they still seem to be pretty

inactive in the game. Maybe they just look nice?
The other plausible move 11.5d2 gives some
other options:
A typical trick for White in the Benko Gambit.
Black has apparently obtained complete control
over the e5-square, but White can advance
unhesitatingly because of the weak black king’s
position. 25...dxe5 26. Hxe5 HxdS (White
wins after 26...fxe5 27.%xe5 De8 28.2e2! Wd8
29.Wh8t Bf7 30.¥xh71 ©f8 31.8e6! with the
simple plan to capture the gb-pawn) 27. %3
Hab4 28.5)c6! Wad? (28..Wa6 was better, after
29 %Bxc5 ©f8! Black seems to hold his own)
29.6Vxe71! The decisive breakthrough. 29...0xe7
30.¥9xf6 afs 1-0 Kuzmin - Averjanov,
a) 11...bd7 12.0-0 0-0 13.h3 BbH8 14.0f3 Simferopol 2003.
He8§ 15.8el D7 16.2e2 Db6 17.8c2 Here b) 11..Wa6k 12.We2 0-0 13.8c4 Dbd7
Black has the typical compensation for the 14.2d2 Bfb8 15.8cl 9e8 with the standard
pawn. 17..&c4 (also interesting is 17...%20a4 compensation, Navrotescu — Plachetka, Haifa
18.50xa4 ®ixad 19.Wd3 Eb4 with sufficient 1989.
compensation) 18.%¥e2 Ha3? An instructive 11...0-0 12.9e3 Dbd7 13.0-0 a6 14.¥c2
inaccuracy, which gives White the initiative. If Nes
Black had no other strong option here 1 would Also interesting is 14...c4 15.8e2 &c5 16.5¢3
not complain about giving the initiative away Bfc8 17.8c3 Wa4! with standard compensation,
like this, but he has. (18...2xb2! was a great Spassov — Trigonov, Bulgaria 1985.
variation of a combination considered on page 15.b3
to Kasparov's analysis Black
29. 19.&xb2 Hxb2 20.Exb2 £xc3 21.Habl According
&xb2 22.8xb2 Wa6! with at least equality in is doing fine after 15.a4! ¥d3 16.b3 Hfb8
the endgame. White has more pawn islands and 17.8b1 Wxc2 18.6xc2 Dd3 19.3 @d7
his pawns seems to be easier to attack.) 19.2d2 20.f1 £5! 21.55e2 (quite a common king
8xc3 20.bxc3 Bbl 21.Hxbl &xbl 22.Ec2 of mistake is 21.exf5?! gxf5 22.%e2 &xc3
Hxc3 23.We3 Hxa2 24.£b2 Though Black at 23 &xc3 Hf4t and Black is much better)

times can exchange his dark-squared bishop, 21..fxed 22.fxed Of4t 23.8xf4 &xc3 24.8a3
he should not let his king sing blues songs ££6 with decent play.
like “All by Myself”, while the black officers
38 The Benko Gambit

E%%37¢ 7%@7%

w4 A,
/ /é 7 Y

///y/%%%‘@/
%7} .
. . 7 4/
53 7 7
% /é/ )
éa,é@ .
ATWE% AR
//&%%
B . cEg /g/é
%z7
,,,,,,

15...2d7! 19...c4!
It is tempting to play 2fb8 or a similar standard The Benko Gambit suits Kasparov’s active style
developing move, but Kasparov has seen that quite well, and it is a shame he did not play it
Black might not necessarily want to place the more often.
rook on the b-file. Therefore he makes the move 20.b4
he will have to make no matter what. White has no choice but to accept the strong
16.f4 Hd3 17.a3 £d4 18.52h1 Efc8! pawn on c4. And he can forget all dreams of
The rook tries to spot the queen through all landing a knight on c6 in the foreseeable future.
the trees berween them, well, pieces of wood at 20.90xchd DRt 21.8Bxf2 2xf2 22.a4 Hxc4!
least... 23.bxcd Wxc4 and Black is better. Or 20.bxc4
19.8ab1?! &xe3 21.Wxd3 fxd2 22.Wxd2 Wixc4 23.8fcl
Slightly inaccurate. White should have created &5 and Kasparov has little time for the white
an air hole for the king, as the following analysis position.
shows: 19.h3! ¢4 And now: 20.&9xc4 (20.bxc4d 20...2f6 21.Ded1 Ecb8?!
A7c5 21.8a2 (21.15 Lxe3 22.8xe3 Wxcd and 21..Hab8 22.Wa2 Wa7 and with his strong
Black is at least equal) 21...8xe3 22.82xe3 Wxc4 threats Black is slightly better. 21..Wb6!? was
23.8d1 Ba5 24.%h2 Hca8 and Black has great also strong.
compensation for the pawn) 20...407¢5 21.8abl 22.9e3
Dxb3 22.8xb3 (22.¥xd3 Hixd2 23.¥xd2 Wxc4 22.8e3? is punished with 22..¥b6 23.8a4
24.8b5 £c5 25.8fel &xa3 looks absolutely Bxad! 24.9xd4 Wxd4 25.%xad Dxed 26.¥c2
acceptable for Black as well) 22..Wxc4 23.8f3 Hb5 and Black is much better - Kasparov. White
A2+ 24.58h2 W5 25.%a2 Wed 26.9b1 &£xc3 is of course an exchange up, but the white rooks
27.8bxc3 We2 (27..Wxed?? 28.8xc8T Hxc8 are not playing at the same decibels as the black
29.%f1! traps the knight) 28.Exc81 Hxc8 29.8¢3 rooks. Coming up after ...Exd5 is the threat of
& d3 and Black has compensation for the pawn, ..&\g3T and ...Bh5 mate, which will force White
which means the same here as: it is all a mess. to make further concessions.
22...8c8 23.b5
White did not believe for a second that
Black really wanted to draw by a repetition of
moves. Other alternartives are: 23.2f3?!, which
is met strongly with 23..8xc3 24.8xc3 Dxed
 The Benko Gambit Accepted 1 39

25.2d4 b5 and Black is much better. 23.2ed1 the endgame. However this is not necessary at
transposing to the note on move 21. all.
23.. b6 24.Dxc4 D211 b) But actually the best move was: 30...#a7! A
This wins time, and time is everything the way surprising solution to Black problems. 31.Wxa7
Kasparov plays the Benko Gambit here. (BL.Yb2 Hxf3 32.We2 4c5! 33.¥xcd Dd2
25.8xf2 Exc4 26.8f3 Exa3? 3456 Wb7 35.%d3 Hixbl 36.¥xb1 ¥xb6 and
Kasparov, as so often before, chooses the most Black is just a pawn up and therefore much
energetic solution, but much stronger was the better) 31...82xa7 32.8al (32.b6? £xb6 33.2xb6
prosaic 26...8xc3! 27.8xc3 Dxed 28.Eb3 Hac8 &xf3 34.gxf3 Hcl and Black wins the rook
29.h3 Wc5 30.b6 ¥xd5 31.b7 Eb8 and Black is endgame) 32..0xf3 33.Hxa7 and Black can
much better - Kasparov. choose between ...Ec1 and ...2d4, both Jeading
27.9b2 Hxe4! 28.Wxa3 Hxd2 29.Wa2 to very advantageous endgames.
No better was 29.Wa8t drg7 30.Wa2 &xbl The computers 30.8e2!? Ec8 31.%a6 HbS
31.¥xc4 Dd2 32.8b4 Hxf3 33.gxf3 216 and was maybe the best chance. White is surely
Black is much better. struggling, but at least he is struggling.
29...8¢e3 30.8cl2? 30...
¥ d4!
This is a horrible blunder for a player of Now it is all over. Many other moves would
Bareev’s standard. White walks straight into a have won, but this is the most accurate.
losing pin. 30.23d1 was given as the best defence 31.b6 Exc3
by Kasparov, but this is not necessarily true. 0-1

Game conclusions: The long knight manoeuvre

fails to impress. White might retain the chance
to castle kingside, bur it takes just as long as it
would to simply walk over there with the king!
Kasparov proved in this game that Black has
excellent counter chances.

A Chapter conclusions: In the beginning of the 21*

2

AN century the most popular way to fight against

the Benko Gambit is accepting and holding the
pawn — however the alternative ways of playing
We should look at the following options: this line, which we have investigated in this
a) According to Kasparov the best move was chapter, should only bring joy to Black players
30...%c5 31.b6 Hxb1 32.b7 Hb4 33. a8t de7 when they face them. White is not pushing for
34 b8 Exb8 35.Wxb8 Wcl 36.9h27 Wxb2 an advantage, and can end up in trouble quite
37.8xb2 £c5 and though White is slightly casily if he plays just one or two extra dubious
worse he will have reasonable chances to draw moves.
Chapter 2

The Benko Gambit Accepted 11

1.d4 &6 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.bxa6 g6 10.82g2 0-0 11.Dge21?

6.9c3 £xa6 7.e4

TR) Usd E
. /xfif/t/t 7%M% mz/
,,,,,,
7,
Y

//% , , , / % /x / /
///a/ /
//go//////////y///
//@/ é%)/é
= I NG T

Practice shows that this move is rather harmless.

In this chapter we shall look at the variations White strengthens his control over ¢3, but at the
of the Benko Gambit Accepted where White cost of fighting for vital squares in the centre.
exchanges the light-squared bishops to start 11..bd7
with and quickly walks with his king to the 11..¥b6!? is in my opinion strongest.
kingside. We shall begin this chapter by looking Also 11...%9a6!? is interesting. Here we can
at two oddities from White, which have still been consider:
played by very strong players. They are not really 12.8b1
dubious, like those inspected in the previous 12.%¥ 22 is dubious, as Black can play 12...8a06!
chapter, but they also cannot be considered when the queen on c¢2 stands awkwardly.
dangerous for Black. 13.8d1 Hb4 14.9b1 Bfb8 15.2¢5 c4!? with
excellent compensation.
Game 12 12...%8a6 13.a3
Korchnoi — Borik Generally this cannot be recommended, but
Vienna 1986 White is quick with b3, so there is no real
pawn weakness.
1.d4 D6 2.c4 5 3.d5 b5 4.cxb5 a6 5.bxab g6
13...8ab8 14.b3
6.5c3 £xa6 7.e4 £xf1 8.0xf1 d6 9.g3 Lg7
42 "The Benko Gambit

Forced. White cannot allow Black to play the central push 18...e6! 19.dxe6 (Or 19.h5%
Wib3. exd5 20.hxg6 hxg6 21.exd5 Ee8 22.2h6 £f6
14...e6 15.dxe6?! 23.84f4 Eb4 with standard compensation.
This should only be played if Black has The options are many and complicated.)
problems keeping the pawn chain intact. Better 19...fxeG 20.h5 De5 21.8e7 De8 22.f4 &7
was 15.%d3 9c7 16.8g5 Dg4 and Black has 23.hxg6 hxg6 24.¥e3 Eb4 and Black has good
counterplay - Karpov. Now after: compensation.
15...fxe6 16.2f4 Bbd8 17.b4 Dg4 19.8xc3 6 20.8f4 Eb4 21.h5!
Black had good compensation for the pawn in, Suddenly White has a strong presence on the
Nowak — Sznapik, Poland 1986. kingside.
12.%c2 Ha6 21..g5 22.8e32
In my opinion this is not the best option. I Very cautious, and unnecessarily so: correct
prefer to use the kings presence on the g2-a8 was 22.8xg5!. Moves like this can be played
diagonal with ...¥b6-b7. However, on a6 the practically without calculation. White will
rook has other useful functions, among others get three pawns and a strong initiative for the
protecting the d6-pawn. A completely different piece. But what is more important, the board is
alternative was: cut into two, and the black queen is shut out of
12.. %2512 13.8b1 Wa6 the game. 22...fxg5 23.Wxg5t BHf7 24.h6 Hg8

T EeT
259151 D6 26.8hel! He8 (26...2g6 27.5
dxe5 28.8xc5 Hxh6 29.8bel Wd6 30.Hed
% /m/ //7‘ Hxed 31.H5xed gives White a slightly better
endgame) 27.e5 DfxdS 28.8e4 with good
\\\
\fllé

// /” a] // compensation for the piece.

/%/ // %
o//
3
22..8cb8
with compensation for the pawn in Lagumina
,,,,,,,
/
/
— Sermek, Castellaneta 1999.

with standard play on the light squares.

14.b3 Bfc8 15.a4 De8 16.82¢5 Hc7 17.¥d2
T %/
7.
// | 76n BB e
7
7
Z
17.8xe71? was not impossible, just very
complicated after 17...f6 18.2f4 &f7 19.80e6
£e8 20.£d8 and it is not clear that White
has benefited from the change in the material
balance.
17...Hab8 18.h4 £xc3?! / ///
This is not bad just because Black exchanges
his good bishop, but also because the white
/E?
Grwaa /A /
) 7

knights are on the same circuit, and only one

of them can occupy the dream square ¢3. So 13.a4!
with this move Black solves the problem of 13.8d1 Wa8 14.b3 6 15.2b2 (15.dxe6? fxe6
overpopulation in White’s position. I should 16.f3 d5 and Black was already slightly better
probably also say something general abour the in Schmidt — Conquest, Poznan 1986, but this
suicidal tendencies behind Black’s idea, but has little refevance for a critical assessment of the
I guess it will be pretty evident. Correct was
 The Benko Gambit Accepted 11 43

position) 15...exd5 16.8xd5 ?xd5 17.8xg7 And 17.83 Db6?

here we have two interesting options: Probably Black made a mistake in assessing the
The lesser option is probably 17..@b4 endings after the 25th move. Black should have
18.Wb2 Wxe4t (18..He8 19.f3 d5¢ is an even tried 17...8c8 18.8a3 Hgd when he has some
more dubious plan. After 20.2f4! Bxa2 (20...d4 compensation.
21.8xd4 cxd4 22.8xd4 ¥b7 23.¥d2 and White 18.2xb6! Txb6 19.8d4! Bc8
wins back the piece with interest) 21.5xa2 Wxa2 19.. %52 20.80b1 Bc8 21.5) 1a3 Wb4 22.Exc4
22.9xd5 Wxb2t 23.8xb2 and White should win and White wins.
this ending.) 19.f3 ¥c2 20.2d2 ¥xb2 21.8xb2 20.9d1 Dg4 21.Bxcd Bxcd 22.¥xcd Wxf21
Nxa2 22.80c3 Hfa8 23.9ed Hb4 24.8xa6 Hxal 23.0xf2 De3t 24.2F3 Dxc4 25.91d3 £xb2
25.8xd6 and White has exchanged his extra Whirte exchanges his material advantage for a
pawn for an unpleasant initiative. positional one. This position is very difficult for
17...e37! It is important to wreck the white Black to play.
pawn structure. 18.fxe3 dxg7 19.5c3 Hf6 26.8a2 £a3
20.2f1 Hgd! (20...He8?! 21.8xf6! Lxf6 22.8d51 Black must keep control of the b4-square.
g7 23.5c7 Wc6 24.Dxe81 Wxe8 25.Wc3t 26..816 27.5b4 Bb6 28.9c6 DeST 29.8)xe5
and White is better in this ending) 21.We2 &e5 fxes 30.0e3 e6 31.%0d3 &f8 32.%9c4 and the
22.4\d5 Ha3 23.2g1 b8 with full compensation a-pawn decides.
for the pawn. 27.9c3
13...1%a82! 27.8xa3 Bxad 28.9xcd Hxa2 29.h3 gives
It is better to play in the centre as soon as it is White a slight plus.
possible: 27...8a5 28.%2e2 &5 29.Dxc5 Bxc5
13...9b6 14.b3 e6 15.¥d3 Wa8 16.2g5 exd5
17.8xf6 4xf6 18.exd5 Hc8! (18...He8 19.Badl
Ba7 20.9)b5 Hae7 21.9ec3 £xc3 22.8xc3 He5
23.8d2 White is slightly better, Cmilyte — Vajda,
Yerevan 1999) 19.¥f3 (19.8hel ¢4 20.bxc4 Bxcd
21.8ad1 Ha5 with equality, because Black takes
back a pawn) 19...2g7 20.Bhel c4 21.8ab1 cxb3
22.Hxb3 HHxad with equality.
14.2b5 ¥b7 15.2d1
T

e7-¢6 does not work and Black has a problem:

how to find another sensible plan.
15...2fa8
15..8c8 16.£3 Db6 17.9ec3 ©fd7 18.b3
£e5 19.8e3 §a8 20.Habl &c7 21.0xc7 Wxc7
22.6b5 ¥d7 23.h3 Wb7 24.9a3 Eb8 25.8d2
e6 26.b4! White is much better, Cmilyte — 30.d1Y
Rohonyan, Aviles 2000. After a move like this you know White is a
16.Dec3 c4?! world class chess player. This was the only way
A better alternative is 16...9g41? 17.8f4 (this for Korchnoi to break Black’s defence.
is better than the sharp 17.h3 &ge5 18.f4 &cd 30...f52
19.b3 &cb6 20.2a2 Bc8 when after ...H0a8-c7 Black had more chances for a draw after
Black has good chances to hold the position) 30...2a5 31.5bd3 De5t 32.¢hc3 Bes5t 33.80d4
17...4ge5 18.b3 White is better, but it is not an (33.80b42? Nd3t 34.%0a3 Bel and Black is much
easy job to break through if Black just waits. better) 33..Hcdt 34.0e3 Hcs 35.%0e2 £aS
44 "The Benko Gambit

(35...5)c4? 36.8c2! and White wins) 36.2)e3 e6 The beginning of an interesting plan. White
37.85d2 £5 38.exf5 exf5 39.%c3 HeSt 40.8b3 wants to avoid weakening the h1-a8 diagonal.
White should win. 'This move also prevents Black’s standard plan
31.8c2! fxed 32.02 Db6 ...\ f6-g4-e5. If we compare the h2-h3 plan to
32..e3 33.50d3 Hc8 34.a5 6 35.dxe6 A8 the most popular way (with g2-g3 and tg2) it
36.a6 he7 37.a7 d5 38.£f4 and White wins. is worth mentioning that White will have similar
33.Bxc5 dxc5 34.a5 Dxd5 35.Dxed c4 problems — only they will be on the h2-b8
36.0d2! diagonal (instead of h1-a8).
Now White wins rather comfortably because 10...0-0 11.skg1 Hbd7
of the distant passed pawn. 11..50a6 12.82h2 ¥Wb6 13.8el Dd7 14.He2
36...c3 37.8\e4 c2 38.82d2 Db4 39.5c5 &7 Db4 15.8g5 Hfe8 16.23 Wa6! 17.%d2 Hd3
40.26 D6 41.8xc2 Bf6 42.50c3 Df5 43.0c4 18.2h6 £h8 and Black has full compensation,
€5 44.8b5 a7t 45.82b6 D8 46.2b7 Dd6t Forintos — Bukic, Vrnajacka Banja 1973.
47.2b8 Db5 12.%h2 Wa5 13.2e1 b8 14.%c2
1-0 14.%e2 is a viable alternative. White has a
standard plan here in 8e2-c2, which I could very
Game conclusions: White played fantastic chess in well have included in the Introduction. Black
this game, but this does not mean that 11.4ge2 is should prevent it with 14...2b4! as 14...6)e8?!
a big threat to Black. There is less reason for him 15.8c2 &ic7 16.8d2 &b5 17.a4 Dxc3 18.8xc3
to focus on the centre when White so clearly has £xc3 19.bxc3 leads to a typical position where
no control there. If Black improves his play on White is better, Horr — Jimenez, Manila 1974.
move 12 1 feel certain that he will gain as much
compensation for his pawn as he can realistically
expect to get in the Benko Gambit.

Game 13
Haba — Nun
Czech Republic 2003

1.d4 D6 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.bxa6

8xa6 6.9c3 d6 7.3 g6 8.4 £xf1 9.52xf1
g7 10.h3%

i wEo X Now after 14...Zb4 we have following options:

C T aasa
a) 15.%Wc1? Eab8 (15..8e8 16.450d1 Had
17.£d2 ¥c7 18.a3 Bb8 19.8c3 &xc3 20.8xc3
Ha6 with standard compensation) 16.e5 dxe5
/// ‘ m‘// 17.8xe5 HixeS 18.8xe5 Wc7 19.f4 and according

A ?/8%//////
to Garcia Palermo the position is unclear.
b) 15.2b1 Eab8 16.b3 Wa6! (16...40e8 17.82d2
Wa6 18.8cl &eS 19.5)xe5 &xeST 20.@g1 N6
21.%e1 and White is slightly better, Kosyrev —
Stojanovski, Skopje 2002) 17.£d2 c4 and Black
has enough compensation to balance out the
invested material.
o) 15.5g1 De8 16.5Hc2 Bab8 17.We2 Hc7
18.8g5 Exb2 19.Exb2 Hxb2 20.¥xb2 fxc3
 The Benko Gambit Accepted 11 45

21.9b7 &xal 22.8xe7? (or 22.¥c8t D8 21...cxd4?

23.8xe7 fg7) 22..20f8! 23.Wc8 Lg7 24.2d8 You can only play an anti-positional move like
W3 25.8xc7 Welt 26.89h2 &e5T 27.g3 Wil this in the Benko Gambit when it wins by force.
28.8xe5 Wxf2t with equality. After 21..6xd4 22.6c3 Wb4 23.8bl &Hb5
So after 14...Eb4 Black is not worse. 24.5xb5 ¥xb5 25.a3 ¥Wd3 Black has the usual
14..5e8 compensation.
14...53b6!? is an equally playable option, but 22.¥d2 Bc8 23.8e2 Bc7 24.9d3 €52
only if Black knows exactly what he is doing. After this Black will have yet another problem:
15.50d1 c4! (15..%a6 16.82d2 Dc4 17.8¢3 and a weak king’s position. After 24...Hac8 White is
White is slightly better) 16.2¢3 @fd7 17.He2 better, but Black still has some chances.
N5 18.50d2 Hb3! 19.5xb3 cxb3 20.%xb3 25.dxe6! fxe6 26.¢5 d5
a4 21.%d3 Hxb2 22.Dxb2 Exb2 Va-2 Haba Now the pawn structure on the kingside
— Prusikin, Germany 2003. looks a bit like a French Defence. Positions like
15.2d2 Dc7 16.2d1 Wa4 17.¥c1! this remind me of a possible product from the
After 17.%xa4 Hxad 18.8c3 DbsS 19.8xg7 research of the famous Russian scientist Ivan
doxg7 20.g4 Bba8 21.a3 Dd4 22.80xd4 Exd4 Michurin. His subject was interbreeding of
23.5g3 4 24.f4 ©c5 Black has excellent different species. There is a famous joke that says
compensation for the pawn. One thing you that Michurin bred a dog with an apple tree.
should remember in positions like this one: As From this he got an apple tree thac barks at apple
long as all four rooks remain on the board, Black thieves and provides itself with fertilizer. Here
will have sufhcient compensation. the interbreeding of the two structures makes a
17..8e5 weak impression.
17...5b5 18.8)¢3 Wa6 19.6xb5 Wxb5 20.&c3 27.8d2 Hac8 28.a3 Wa5 29.2h2 Bb7
29..Wa4? 30.b3 Wa5 31.a4 Hc3 32.¥xd4
&xc321.bxc3 Wb2 (21...8d312 22.82e3 Wb5 with
the idea ...5Yb6-c4 is interesting) 22.Wxb2 Exb2 and White wins.
23.a4 Ba5 24.%0g3 Db6 25.Heb1 Exb1 26.5xbl1 30.8¢2 Hxc2 31.Wxc2 Bc7 32.%d3 Wbe
B\ xa4 27.Hal Ha8 28.5f4 £b6 29.Hxa8t Dxa8 33.%d2 W5 34.a4
30.5d2 & b6 White has some plus here, but it Now the infantry is set up for a frontal attack.
obviously is not enough to get the full point. 34...50a7 35.b4 Wc4 36.b5 D8 37.2b2 ¥b3
18.5xe5 fxe5t 19.%2g1 £d4 20.8c3 Db5 38.Wxd4 Bc3 39.0d1 Bc7 40.¥b2 ¥4 41.b6
21.8xd4 Hxb6 42.¥xb6 Wit 43.g3 Wxe5 44.¥b8Y
7 45.8cl

//@/
/ 1-0

/ Game conclusions: The idea with h3 and a king

march to h2 seems logical in many ways. Black
loses access to the g4-square and the king is not
in potential danger on the a8-g2 diagonal. On

@// the other hand it is also rather time consuming,.

So though it certainly is a playable alternative
to the main lines, it should not cause Black any
serious headaches.
/8/
46 The Benko Gambit

Game 14 understanding the ideas behind moves and

Gyimesi ~ Sermek the typical reactions in different positions is
Bled (ol) 2002 much more important than precise theoretical
1.d4 D6 2.¢4 ¢5 3.d5 b5 4.cxb5 a6 5.bxa6 g6 knowledge.
6.8\c3 8xa6 7.4 £xf1 8.2xf1 d6 9.DF3 Hbd7 13.Eb1
10.g3 £g7 11.52g2 0-0 13.9c2 Wd7 14.b3 6 15.dxe6 fxe6 16.8b2
Now White has to decide: Either he will play d5 17.a4 Ea5! and Black has good play for
with h2-h3 or without this prophylactic move. the pawn. 13.He2 ¥d7 14.%d3 Da4d 15.Eb1
In my opinion it is better to play without. The Bfc8 16.h3 c4 17.¥c2 &c5 18.8e3 6 19.dxe6
standard manoeuvre .2 f6-g4-83e5 is not fxe6 20.8xc5 BxcS 21.e5 dxed 22.9ed Dxed
dangerous for White hete as far as I can see, and 23.Bxed Wc6 and the position is unclear, Salov
the pawn move (though it is popular) would — Tseshkovsky, Minsk 1987.
appear to be a waste of time. 13..%d7
12.2el! ECO considers only 13...8a6, which in my
opinion is less good, as the b6-knight seems
to have little use then. The moves should be
W oo connected likes peatls on a string.

/ /@)j’t%t 14.b3 Dg4 15.8b2 512

7%
/// /‘/%‘
//
. ///%g
/ ///&/ %%
a

5
7 7 ///W&é//.

//////

This position is as standard for the Benko

I stole this idea from Kaidanov (see game 17).
Gambit as the position after 1.e4 e5 2. D3 &c6
16.24 Wb7 17. 92 H1f6 18.exf5 Hfxd5 19.8xd5
3, £b5 a6 4. 2a4 D16 5.0-0 Re7 6. Bel b5 7.
£b3 d6 8.¢3 0-0 9.h3 a5 10. Rc2 c5 is for the Hxd5 20.fxg6 Dc3 21.gxh7t ©h8 22.HExe7
Ruy Lopez.
Wxf3t 23.dhgl &d4 24.2e3! fxe3 25.fxe3
Wxe3t 26.@g2 W3+ 27.@?g1 15-15 Ehrenfeucht
Inferior is 12.2d2 Wa5 13.Hc4 Wb4 14.We2
— Pinski, Wile (training match) 1997.
Ab6 15.9xb6 Wxb6 16.8b1 Hd7 17.82d2 Wao!
Another idea to consider is:
18.%¥xa6 Bxa6 and according to Tukmakov Black
12..8 ¢4
has full compensation.
There was a time when, because of this move,
12...%a52!
The most natural way to play, but not enough White would play 12.h3.
13.%e2 Hge5 14.2d2!
for full equality.
Now Black has some problems with the
12..83b612 is a very interesting and vyet not
knight. The exchange on e5 would solve
too popular move. Black’s idea is to play ...%¥d7
Black’s problem, which is that the two knights
and ...9a4. You will no doubt recognize this idea
from the Introduction. This again demonstrates
are basically fighting for the same square.
14...0b6 15.f4 Ded7 16.24 £xc3 17.bxc3 Exa4d
my belief that in the Benko Gambit Accepted
18.8xa4 xad 19.c4
 The Benko Gambit Accepted 11 47

According to grandmaster Viktor Gavrikov, Black’s defensive task is rather difficult:

White is slightly better. White’s plan is racher 22.. b5 23.bxcd Hxcd 24.¥b3 Wxb3 25.axb3
simple: break with e4-e5 as soon as possible. Bxcl 26.5xcl Bal 27.60d3! (the inferior 27.2)e2
13.e5! @g4 was played in Gyimesi — van der Weide, Cappelle
After 13..dxe5?! 14.8xe5 HxeS 15.Exe5 la Grande 2002) 27..2d4 28.9Hb2! &xb2
White is much better, because of the weakness 29.8xb2 2a5 30.b4 Exd5 31.b5 Hc5 32.b6 Ec8
of ¢7 and because the black queen does not have 33.g4 16 34.03 &7 35.0ed De6 36.b7 b8
any direct influence on d5. 37.t4 and White wins.
14.exd6 exd6 15.2f4 Efb8 23.axb3 Wb7
15...2)ge512 16.%Dxe5 Hxe317 Bed! b6 18.We2 23..Wbs!? 24.8d3 and White keeps the
£5 19.2a4 Hab8 20.b3 According to grandmaster pressure. But after the sharp 24.142! £g7 25.8d3
Dautov this position is much better for White. In He8 26.f5 Hbed 27.0)c3 fxc3 28.Edxc3 He2t
my opinion Black has some compensation here, Black appears to be safe.
which probably means that White is only slightly 24.8d3 Bb5 25.2Dgl Haa5 26.¥13 Bxd5 27.b4
better. I honestly find it easy to decide which Habs 28.Bxd5 ¥xd5 29.¥xd5 Exd5 30.Ebl
position I want to defend, the one after 15... @geS Bb5 31.8 €2
or after 15...Bfb8: I definitely choose the first. For Even here White retains a plus.
example: 20...¥b7 21.Wd2 Bfe8 22.8d1 £\ 7 and 31...d5 32.8c1 ©f8 33.f4 26 34.2f3 Be7
it is not that bad, is it? 35.%0¢3 ©d6 36.De2 h5 37.52d3 Eb7 38.d4
16.Be2 ¥a6 Eb6 39.8b3 h4 40.2f3 hxg3 41.hxg3 L6
16...4ge5 17.5xe5 ®xe5 18.8xe5 HfxeS 42.0d4t Rd6 43.9D2 g5 44.8a3 gxf4 45.gxf4
19.8c1 with a similar position. the6 46.8a5 Rd62?
17.8c1 Dge5 18.Dxe5 DxeS 19.8xe5 Bxe5 Fatigue sets in! After 46..8g7 47451 &Hxfs
20.b3 48.8xd57 the position is equal.
47.8e3 b2 48.Hxd5t

E_EwE Now White wins.

Tz
48..52e6 49.b5 Eb8 50.%2c4 fcl 51.EeSt
&d7 52.Bd5t de6 53.f51 f6 54.Dgdt

wo & tal g5 55.0e5 Be8t 56.0c6 Le3 57.82d3 £b6

/// z%//
58.2d7 ££659.8d61 g5 60.Ed7 6 61.Ses
Be8t 62.864 Re31 63.2g4 8¢5 64.2d5 Eedt
65.963 He3t 66.%4 Bc3 67.9e5 Sa3 68.%e4
// %% 7, 7 Bb3 69.2Dc4 Le7 70.80d4 268 71.2d2 Eb4t
%4 72.52¢3 Bh4 73.9b3 £b4t 74.2d3 Eh3t
75.%2c4 Lel 76.8¢5 Bhét 77.82d3 Eb4 78.b6
/// s g7 79.b7 L¢3 80.Hd4
. %Z%f/// // 1-0

Game conclusions: After 12...%a5?! Black has a

On positions like this experienced players say:
Black has to beg for a draw until the end of the
world, which would maybe not be that bad, if
it were not for the many deaf white players out
there!
20...Bb4 21.8d2 c4 22.9e2 cxb3
difficult defensive task ahead of him, which is
not just dangerous, but also demoralising because
of the passive nature of the position. Instead
12...5b6! followed by 13...¥d7! with the idea
of ...&3b6-a4 to exchange the strong white knight
seems like an interesting alternative for Black.
48 The Benko Gambit

Game 15 14...Eb8
Gyimesi — Stamenkov 14...e62! would be too early: 15.8b5 ¥b7
Nova Gorica 2004 16.£f4 and White is better.
15.b3
1.d4 96 2.c4 c5 3.d5 b5 4.cxb5 a6 5.bxa6 g6 15.8b1 c4)? 16.8e3 Dgd 17.2d4 Dges
6.0)c3 £xa6 7.e4 £xf1 8.8xf1 d6 9.g3 Lg7 18.8xe5 Dxe5 19.8xe5 &xe5 20.b4 cxb3
10.%2g2 0-0 11.5f3 Dbd7 12.Eel! Eab!? 21.Bxb3 Hc8!' 22.Wd3 Ha5 23.Hebl Eac5
12..%¢7 is one of my late coach Ehrenfeucht’s 24.8e2 Wxad with equality in Banusz —
ideas: Black keeps control over the e5-square. Predojevic, Budapest 2003.
Unfortunately White has a good answer in 15...e6
13.8e2 Hfb8 14.8g5! h6 15.8xf6 £xf6 16.8cl Now Black needs to strike against the centre.
and White keeps a small and stable advantage After 15...8e8 16.82g5 Hab6 17.8abl (17.8xe7?
forever, Kaszowski — Pinski, Poraj 1997. Hxb3 18.8e3 Wa5 and Black is better) 17...e6
18.dxe6 fxe6 19.50b5! White is better, because

P Wy Hes
d6-d5 does not work now the e7-square is
weakened.

/ Y yr
//////

16.dxe6 fxe6 17.8b2

4 ax
,,,,,,

7
/// P /
) D
//% ) //%fl% 0
z% %xmx%
27 : //”/ 2 A '7/ /

sy
e o

This move is Black’s way of combining the

compensation of play down the open a- and b-
//
// g Z //’/'/ 7
NN
\\\§
NS

files with attacking the centre.

13.%c2
13.h3 transposes to a position that can also 17...8f8!
arise after 12.h3 BaG 13.8el. Somehow this is too slow. Black needed to act
13.e512 dxeS (13...0g42! 14.8e2 a8 15.83h4! with greater vigour: 17..%b7 18.83b5 and now
and White is better because after 15...8gxe5? choose between two options:
16.f4 and Black has problems with the knight) a) 18...d5 19.exd5 exdS 20.He7 £f8 21.8e2
14.9)xe5 Dxe5 15.8xe5 Ed6 gives Black good He8 22.8xe8 Hxe8 23.Wd3 He6 24.9g5 He7
compensation. Also good is 15..%d7 16.We2 25.@?g1 ef6 26.a5 ¥a6 Black has compensation
Ba7 17.¥c4 ¥b7 18.5kgl Ed8 19.8¢5 Stohl here. The problem for White is that after some
— Zueger, Prague 1996. Now Black should play very possible exchanges the endgame with three
19..h6 20.&cl &gd 21.8el £d4 with some pawns against two on the kingside will not be
initiative. enough to win.
13..Wa8 14.24 b) 18..&xe4!? is very interesting: 19.Wxe4
14.b3 e6 15.dxe6 fxe6 16.Eb1 d5 and Black Wxed 20.Hxed Sxb2 21.8d1 d5 (21...e52!
has perfect compensation. 22 8xd6 Bxd6 23.50xd6 Exb3 24.4d2! Ed3
 The Benko Gambit Accepted 11 49

25.502¢4 and White is much better) 22.8e2 &6 29.5xb7 EBb6 30.2c5 Db2 31.8Bd2 Ec8
23507 B 24.8)xe6 dd 25.8d2 df7 26.2del 32.5)e5! HxcS 33.8d7 Bxb3 34.Dxc5
& e5 27.8\f4 and Black’s compensation should be The rest is played in pure execution style.
enough to get a draw. 34..2b8 35.2d7 Ha8 36.8d4 Hc8 37.Eb4
18.5b5 ¥b7 19.8ad1 d5 20.exd5 exd5 21.8e7 Hdl 38.5e4 Bc2 39.Eb8t &f7 40.82b7t
d4? the6 41.Bxh7 Ha2 42.8b7 &f5 43.2b4
Preferable was 21..8b8 22.8gl £f8 23.2¢2 Be3t 44.8f3 Dd5 45.8b3 Hxad 46.0c3
He8 24.8xe8 Hxe8 25.8e5 though White is still Bb4 47.8xb4 Hxb4 48.h4 Dc6 49.g41 De5
much better. 50.50¢3 she6 51.f4 De7 52.%ed DB 53.0b5
He7 54.0)d4t 66 55.g51 Bf7 56.65 D8
57.86 Db6 58.9f3

7 7 Bl 1-0

%//éwfi%/ Game conclusions: 12...5a6? is certainly an

interesting alternative in this highly critical
/// ,,,,,, / position for the Benko. 12...¥b0 is a clearer way

poa B % to equality, but, as we shall see in the next game, it

also “risks” allowing an early draw. So objectively
&/ / 12...Ea6 is probably slightly inferior, but for the
practical player it is an interesting alternative all
the same. Especially for those games when a win
is essential.

Game 16
Hjartarson - Alburt
22.
W 412 Reykjavik (Nordic — USA) 1986
Here White had a fancy win with 22.@\bxd4
cxd4 23 .Bxd4 Heb (23...8c8 24.¥d3 Bc7 25.Ec4 1.d4 96 2.c4 c5 3.d5 b5 4.cxb5 a6 5.bxa6 g6
a7 26.%d6 and wins: the pin is o strong) 6.0)c3 &xa6 72063 d6 8.e4 Lxf1 9.5oxf1 Lg7
24 Wedll Hixed 25.Exg7t Sxg7 26.Exd7T &g8 10.g3 0-0 11.52g2 Dbd7 12.Eel ¥b6

¥ 7
27 Hxb7 Bc2 28.8d4.
22..0h8 23.8d3 Dg4??
Here Black misses his chance in turn: 23...Ee8
24 Bxe8T Hxe8 25.Wf7 Def6 26.§i?g1
27.Wc4 Be6! 28.Wc2 &gd and Black rakes over
We4
z%w¢%g%z;zz:
4 Ak
L

the initiative.
24.Bxg7?! /” 4,,,,//7 _
Still good enough to clinch the full
after 24.89bxd4! Black is just dead
point, but
lost: 24... 87
cxdd 25.8xd4 &xd4 26.¥xd4t Def6 (26...Dge5 D ///@/8”
27.%g1 and White wins) 27 Wh4 He8
Wxd7 29.8xd7 and White is winning.
28 Hdxd7
&/&é7/ ///g//&/
24...@xg7 25.8\bxd4 Hde5 ////////////
Or 25...cxd4 26.¥xd4t Ddfo 27.@xg4 and
White wins.
26.9e6% @gS 27.8xe5 Dxe5 28.Dxc5T Dxc4
50 The Benko Gambit

13.e5% is an easy way to a draw. When Here White should maybe start to change his
the black queen can take control over the line of thought. Possible was 23.8Bxb3 Wxb3
a8-hl diagonal this is not a good move. But 24.¥xb3 Hxb3 25.8xa8 with equality.
now this would not happen as smoothly, and 23...2¢5 24.8xc52
therefore White can afford this immediate push. White is too optimistic. 24.2b1 &xa8 25.e5
13...dxe5 Why not? (13...@g4 14.exd6 exd6 ®c4 26.8c1 Wad 27.8b1 with a draw was the
15.We2 Qgf6 16.2f4 8fe8 17.¥d2 and White correct decision. Note that after 27.%d2?! ¥a2!
is slightly better) 14.5xe5 @ixe5 15.8xe5 Wb7! 28.%d1 b2 29.2b1 Eb8 White is only playing
16.%f3 Had8 17.8¢5 Wxb2 18.8b1 Wa3 19.8xe7 for two results, and winning is not one of
Hixd5 20.0xd5 Wxf31 21.8xf3 ExdS 22.8e3 c4 them.
23.Hc7 3 24.a4 with complete equality, Muir 24...dxc5 25.0¢7 c4
— Malakhov, Budapest 1995. A secret and probably untrue principle in
13...Efb8 14.h3 chess is that black pawns always advance more
After 14.8c2 Wa6 15.We2 Wxe2 16.Exe2 Sigd quickly...
17.50d1 HigeS 18.8xe5 Dxe5 19.82d2 D4 26.€5
20.8¢3 £xc3 21.bxc3 Ha3 224 6 23.12 The best defensive try was probably 26.d6
Bba8 Black also achieved equality in Hort — Wd7 27.3d5! (27.e5? exd6 28.exd6 b2 29.¥b1
Damljanovic, Zagreb 1972. c3 30.8el Wxd6 31.89e8 WcS and Black wins)
14...¥a6! 27..¥xd6 28.93d2 Wc5 29.8Bel £d4 30.2e3
Often the queen is very well placed here when £xe3 31.8xe3 Wd4 though Black is still much
the bishops are exchanged, which is probably the better.
main reason why the lines presented in chapter 26..%a5 27.d6 b2 28.¥b1 3 29.e6 fxe6
xxx (on the accepted Benko with the bishop on 30.2el
g2) are more popular these days. 30.2xe6 ¢2 and Black wins.
15.2g5 h6 16.8e3 @b6 17.Ecl D4 18.b3 30...exd6 31.2Dxe6 152
Dxe3t 19.8xe3 Dd7 20.a4 c4 21.83b5 31...c232.Wxc2 Wixel 33.Wc7 Wxe6 34. Wxb8 T
21.bxcd ¥xc4 22.8d2 Wd4 and Black is ©h7 35.%b4 ¥d5 and Black wins.
probably already better. 32.0ed4 ¥xbl 33.Bxbl He8 34.%2f1 £xd4
35.0xd4 Re5

E¥ | 7
A
7 The position is of course still winning for
Black, but White at least has a reason to fight on
\\\\

. akas for a litcle while.

% 4 i
36.%g2 Be4 37.0c2 &7 38.f3 Be5 39.h4
f6 40.8d1 Te6 41.g4 Bd5 42.Belt d7
43.9a3 Bd3t 44.0ed Bd2 45.64 2 46.Dxc2

/&/// Bxc2 47.8b1 %e6 48.h5 gxh5 49.gxh5 &f6

50.52d5 f5 51.2xd6 Dxf4
0-1
Z % ,,,,,,,,,,

/% O
% fi@%Ew Game conclusions: 12... b6 is a very safe option
that leads to even chances. White can more or
less force a drawish position with 13.e5, but
sometimes this is a risk you must take when you
21...cxb3! 22.8¢7 ¥xa4 23.5xa8 face strong opposition. ..
 The Benko Gambit Accepted 11 51

Game 17 like this: 17...8d7 (17...Eb4) 18.Ee2 ¥a8 19.£4

Titenko — Kaidanov Ebd8 20.£g5 h6 21.8xf6 &xf6 22.8cl £xc3?!
Kislovodsk 1984 (22...c4 givesmore chances to draw) 23.8xc3
Wxa? 24.8xc5 Hxd5 25.8xd5 Wxds 26.%xd5
1.d4 56 2.c4 c5 3.d5 b5 4.cxb5 a6 5.bxab gb £xd5 27.8xe7 with a close to winning endgame
6.5c3 £xa6 7.e4 Lxfl 8.0xfl d6 9.g3 g7 for White, Kuzmin — Libeau, Biel 2002. After
10.52g2 0-0 11.2f3 Dbd7 27..8b5 28.8e2 White’s plan is simple: The king
will come to the b2-pawn’s aid and, with any
luck, promote it.
E W %’jfi@@// 13.Bel De8 14.82e2

%//m%w4x Another
Wh7 16.8adl
game
B8
went
17.e5
14.%e2
(17.8c1?
&7 15.8f4
is well
/ answered by 17...2b5 18.%¥c2 Wa6 with perfect

/ ,,, compensation) 17..39xb2 18.exd6 exd6 19.9e4

\\
\\

Wixed 20.8xe2 He§ 21.Beel {5 22.8xd6 Exel

23.5xel BM6 24.9xf5 DexdS 25.80xg7 Dxfat
26.gxt4 thxg7 with even chances in Ehlvest
— Hodgson, Rakvere 1993.
7
14..20b6
Better than the slow 14..9¢7 15.82g5 He8
//////
16.8¢1 and Whirte has some advantage. The rook
has no business on e8.
12.h3 15.8g5 ¥b7 16.¥c1

I aEe
As I said earlier: 1 do not think this move is
necessary. Now Black has plenty of good moves to
answer it. Most popular are 12..%a5, 12..¥b6
and finally 12... a6. Bur let us first consider a /W ds 7y
very interesting, but not very popular, plan.
12..9b812 , , , A ’/x/
This usually forces the opponent to play the //,ié/// i
game on his own. Help! - he shouts - I am out

///%
of book!
Bur Black has some other interesting options
at his disposal:
12..50b6!? 13.8e1 Wd7 14.Wc2 (14.8g5 Dad /
15.Wd2 Bfb8 16.€5 Dxc3 17.bxc3 dxe5 18.Hxed ,,,,,,
H#d8 with compensation) 14...e6! (14...Da4
is less effective here: 15.82d2 Efb8 16.b3 ©xc3
17.8xc3 and White is slightly betcer) 15.dxe6 16...15!
fre6 16.8g5 Dh5 17.8adl Ef7 18.23h2 d5 The black pieces are naturally placed for this
19.8¢g4 d4 20.b3 ¥c6 and Black has good strike against the white centre.
compensation, Korchnoi — Alterman, Dresden 17.We3
1998. White could also consider 17.%¥¢2 h6 18.&d2
After 12..Wa5 13.8el Hfb8 14.e5! dxed Hf6 19.8h4 fxed 20.9xg6 Ef7 21.8f4 e3
22 &xe3 HbxdS 23.56xd5 Hxd5 24.¥ed €6
15.86xc5 HxeS 16.8xe5 Ha7 17.¥f3 White is
slightly better. The game could very well go on with an unclear game.
52 "The Benko Gambit

But he should stay away from 17.exf5?! when Game conclusions: 12...¥b8!? seems to be a
after 17...9xd5 18.8xd5 ¥xd5 19.W¥d2 W6 perfectly viable option. If 12.h3 really is a simple
20.8xe7 Bxf5 Black is slightly better. loss of a tempo, as I claim, then it is not surprising
17..8¢4 18.¥cl b6 19.¥e3 fxes?! that Black can play in this way. As far as | can
This is not very dynamic. 19..%c4 with see Black is perfectly fine and should have even
equality was better. chances.
20.Wxe4 Bf7 21.8d12
Here White misses his chance to fight for an Game 18
advantage. After 21.8xe7! &xc3 22.bxc3 Had Bystrov — Leconte
23.Md3 Bxe7 (23...5xd5 24.2d8! and White is Correspondence 1999
much better) 24.8xe7 ¥xe7 25.%bs Wd8 26.8b1
N6 27.%xb6 Wxb6 28.8xb6 DHxd5 29.Hxd6 1.d4 DF6 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.bxa6 g6
Axc3 30.20g5! and the threats against the black 6.2c3 £xa6 7.2f3 d6 8.e4 Sxf1 9.bxfl Lg7
kingside secure White a substantial advantage. 10.g3 0-0 11.52g2 Hbd7 12.Eel Ba6 13.h3
21..80¢7 Wa8
Threatening ...£&c3 against which White has
no defence.
W& 2 ER&
~ miasa
22.8cl
Also after 22.8dd2 &xc3 23.bxc3 Had 24.¥d3
,,,,,,,,,,,,
A cxds Black would be better.
22...8xc3!23.bxc3 Ea4 24.¥c2 Nexd5 25.2d3
c4 26.8d4
A typical (k)nightmare for White.
/%%,,,,/4///47& /%7,%/ 0
V%,O%%&
/S f % 7

This position is fairly safe for Black. The

////// counterplay in the centre with ¢7-¢6 gives him
close to even chances.
14.¥c2
In my opinion this is not the best. Black has
//// made his plan perfectly clear: He will play ...e6
as soon as possible. After Wc2 White will have
to give up the centre with dxe6 or play 15.¥d1,
26...8xf3! which looks more than a little stupid. Therefore,
This final tactic is not too surprising. I recommend White to play:
27.xf3 Dxc3t 14.8e2! b6
Also good is 27...5xa2!? 28.%xa2 Hxc3T and 14...e6!? The obvious move. 15.8f4! (15.82g5
Black wins. exdS 16.exd5 ©b6 17.8xf6 &xf6 18.8e4
28.@g4 Nxe2 29.¥ixe2 €5 30.Exd6 h5t £07 19.5xd6 ©xdS 20.%c4 Had and Black
0-1 has good compensation. Or 15.dxe6?! fxe6
16.8g5 d5 17.exd5 exdS 18.82¢g1 d4 19.2b5
 The Benko Gambit Accepted 11 53

& d5 with excellent play for Black in Zugic 20...%xd51 21.5egst

— Lawson, Toronto 2004.) 15...exd5
Hd8!? Black defends the d6-pawn.
17 8xe8t Dxe8 18.%e2 and White is slightly
better) 17.¥9d2 ©b6 18.8g5! (inferior is
18.8d1 Dad 19.5xad4 Hxad 20.b3 Eed)
18.. He8 19.8d1 Ha4d 20.Exe8T Wxe8 21.8el
Wag 22.8xf6 £x{6 23.Dxad Exad 24.a3 and
White keeps a slight advantage.
15.a4 Dfd7 16.82g5 He8 17.h4 £xc3 18.bxc3
Exad 19.8xa4 Wxad 20 Wb1 £6 21.8h6 Wc4
with unclear play, Illescas Cordoba — Topalov,
Alcobendas 1994.
Also interesting are:
14.8f4 9b6 152 Had! 16.9xad
17.a3 Bb8 and Black has full compensation for
16.exd5 After 21.50¢2 WixgS 22.¥d3 Ha7 23.50f1 Gig7
(16...Ee8 the position is rather unclear as well.
21..2df6t! 22.d2h422
Not very good defence. After 22.8xf6 DHxf6T
23.¢%f4 Ba3 White needs to have another think:
a) 24.g47 Wc4t 25.50g5 He7! 26.Wd2 Hxc3
27.%0h4 Bd3 28.Wc1 Wd4 29.He3 Wf4l and
Black wins.
b) 24.5%0e3! and Black has nothing better than
24, ¥1d4t 25.Be2 Hxc3 26.Wb1 Wedt 27.c0d2
Wd4t with a perpetual check. Notice that after
27...%g7 28.Ee3! White is suddenly better.
22...5kg7 23.Wd2 Badt 24.g4 Wg2!!
Exad

%/// // /;
the pawn.
14.805 h6 15.2d2 €6 16.dxe6 fxe6 17.24 g5
18.2a3 d5 19.Wc2 Bb6 20.8c1 Bf7 21.5d2
8b4 with unclear play, Ljubojevic — Topalov,
Linares 1995.

E% // -
,,,,,,,,,,,
/’V//// o
14...e6!
Mikhail Gurevich’s idea. Bur it is not too
surprising at this point.
/é%/// o
//// 7w ,
15.dxe6 fie6 16.8f4 Dh5! 17.8g5 d5 18.exd5
After 18.e5 d4 19.8e4 Bxa? 20.Bxa2 W¥xa2 K
21.8xc5 WdS Black is slightly better. But Black
also takes over the initiative in the game.
En
w2 ~ E% Whice resigned in despair, probably because

/ //'/.@/
he understood that lines like the following

/// would come: 25.c4 Excd 26.8h6T Hf7 27.We2

g5t 28.@.xg5 (28.@xg5 Hedt 29.2h4 %hgB
and Black wins) 28.Bxgdt 29.hxgd Wh2

/f /8%// z.ém mate!

0-1

e
/%y/,/////
,,,,,,% / 5? ,,,,, ?////
Game conclusions: 12...8a6 and 13...Wa8 is a

gg&z@% p. /%)//% very logical plan and 1 think the counterplay

18...Bxf3! 19.2xf3 £xc3 20.bxc3
Or 20.¥xc3 ¥xdst 21.%e2 @(XgS 22.%d3
Za7 23.55f1 @g7 and the position is unclear.
against the centre should be able to provide Black
with a fine game. But, as can be scen from the
notes to move 14, it is not really clear how this
should happen. As it stands right now, White
seems to have a slight edge.
 54 The Benko Gambit

Game 19 Some alternatives should also be considered:

Piket — Topalov 15.82¢5 h6 16.2xf6 £xf6 17.8cl De5 18.Hxe5
Monte Carlo (rapid) 1997 &xe5 19.%d2 £g7 20.b3 c4 21.bxcd Wxcd
22.53d1 and here Black's compensation was
1.d4 D6 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.bxa6 g6 enough for a draw in Schandorff — Khalifman,
6.20c3 £xa6 7.e4 £xf1 8.xf1 d6 9.g3 Lg7 Aarhus 1997.
10.s2g2 0-0 11.2£3 Hbd7 12.h3 15.8b1 &He8 16.8d2 ©c7 A new move
(previously 16...c4?! had been played, which can
be met with 17.2e3 2b7 18.2d4 and White is
E/ o Ee
%//m%
FY
much better). 17.b3 e6 18.dxe6 fxe6 19.Ecl
@De5 20.Dxe5 &xe5 21.f4 2g7 22.8e3 W6 with
compensation.
15.e5! dxeS 16.9xe5 Dxe5 17.Exe5 b7
18.%f3 and White is slightly better, however it is
very hard to convert an advantage like this into

% , , , //
a whole point. 18...c4 19.8f4 Bd8 20.Hael Ed7
21.8g5 Ea5 22.81e2 h6 23.8xf6 &xf6 24.585¢4

=
=
%f //// A/
,,,,,
/ g/
/
£xc3 25.bxc3 Wxd5 26.Bxe7 Wxf3t 27.dxf3
Hxe7 28.Hxe7 Ha3! with equality.
15..90e8 16.8£4 &xc312
16..0¢7 17.8cl &3b5 18.80xb5 ¥Wxb5 19.b3
¥h7 and White is slightly better.
12...%b6 13.Hel Efb8 17.bxc3 £6 18.a4 HeS 19.8xe5
13...83€8 is an interesting alternative to this 19.9xe5? dxe5 20.2e3 Wed and Black is
very logical move. Now we should consider: probably better because White’s bishop is not
a) 14.8e2 £¢7 15.8¢5 h6! This trick should be useful here.
well known to the readers from the Introduction. 19...dxe5! 20.Bea2 ¥d3 21.We3 ¥ixe3 22.fxe3
16.8f4 (16.8xe7?! Bfe8 17.8h4 g5 18.82xg5 Nd6 23.9d2 Ha5!
hxg5 19.8xg5 ©b5 20.8xbs Wxb5 21.8c2 Hf6 Black is building a fortress.
and Black is at least slightly betrer) 16...0b5 24.g4 Bf7 25.%f3 h5 26.0e2 c4 27.8f3 Bbas
17.8cl ©xc3! (better than 17..5d4 18.5xd4 28.52g3 hxg4 29.hxg4 g5
cxd4 19.9a4 b5 20.b3 £5 21.f3 and White

77/7/
is slightly better) 18.bxc3 Ha3 19.c4 Wa6! with
excellent compensation for the pawn, D. Bekker
Jensen — Pinski, Rimavska Sobota 1996.

»
b) 14.%d2! Dc7 15.b3 4 16.2b1 (16.bxc4
Wa6 17.¥c2 Wxc4 with compensation) 16...cxb3 " 7e ' // &
17.8xb3 Wc5 18.¥d3 Efb8 19.£b2 Exb3
20.axb3 and here, according to GM Aualik,
White is slightly better.
14.8e2 Wa6
14..9e8 15.8f4 Ac7 16.8c] Wa6 is rather
slow. White can strike with 17.e5 dxe5 18.8)xe5
Dxe5 19.8xe5 2xe5 20.8xe5 and be slightly
better.
15.%e1
 The Benko Gambit Accepted 11 55

Done! alternative. My prediction/guess about games in

30.5263 Be8 31.0f1 Rd7 32.2¢3 b8 33.0f5 this line is that White will have the easier side
Hxf5 34.exf5 Eb3 35.Ba3 Bb8 36.e4 ©d6 of a draw.
37.83a2 £b3 38.8a3 Rc5 39.e3 Chapter conclusions: The critical position of this
39.8xb32? cxb3 40.%e3 Pc4 41.Ld2 b2 and chapter is reached after 1.d4 &6 2.c4 ¢5 3.d5
Black wins. bS 4.cxb5 a6 5.bxa6 g6 6.8c3 £xal 7.e4 fxfl
39...82b8 40.2d1 Bba8 41.Eb1 Bxa4 42.8Bxa4 8.50xf1 d6 9.0f3 Hibd7 10.g3 £g7 11.g2
Hxa4 43.2b7 Ha3 44.2c7T b5 45.8b71 Bc5 0-0 12.Eel! (I stand by my belief that 12.h3 is
46.Bc7t b5 47.8b71 Rc5 unnecessary). Black has a range of interesting
1h-1/ replies: 12..%b6, 12..H2a6 and 12..&b6. In
choosing berween them Black must also decide
Game conclusions: The plan 12...¥b6 and how much he is willing to risk to avoid a drawish
13....Bfb8 brings Black very close to equal position: if you do not mind a draw then
chances, while 13..2e¢8 is =z reasonable 12..Wb6 is excellent.
Chapter 3

The Benko Gambit Accepted I1I

1.d4 56 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.bxa6 gb! Game 20

6.g3 Sosnicki — Pinski
Barlinek 1996

,,,,,,,,,,,,,,,,,, 1.d4 56 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.bxab

g06!
This game is a good example of how to fight

// , ,
against the Portisch plan when you do not play

/////;/
5...8xal.
6.g3
6.b32 g7 7.8b2 0-0 8.3 Dixab 9.8g2 £b7
10.20h3 e6! and Black is slightly better.
6...d6 7.b3?!
7.8g2 8g7 8.00h3 Dixa6 9.0-0 ¥h6 10.2¢3
0-0 11.5f4 When White chooses a plan with
/////////////////
&\f4 Black automatically has an immortal knight
one5. 11..8g4! 12.h3 De5 13.b3 8d7 14.Ebl
In this chapter we shall look at what is currently Wa5 15.8d2 Wa3 with compensation for the
considered the strongest threat to the Benko material, Alexandria — Levitina, Dubna 1983.
Gambit by most experts, including yours truly. 7...8g7 8.8b2 Dxat!
White takes the pawn and develops sensibly by
g3 and £g2. Compared to the previous chapter
the bishop can have a tendency to be in the way E i2le E
on a6, especially if White does not advance his /% s.g./x /// ///// 2///4 ////// /,7/
e-pawn. As we shall see Black often uses (loses)
time to force White to do exactly this, so he can //*4
my‘/é
attack the weakened light squares.
The reader is advised to pay extra attention
to the currently popular Epishin Variation with
10.£2b1, which has been analysed in greater detail
than any other line in this book.
58 The Benko Gambit

This is one of the ideas of 5...g6. Black wants

By =
e 4
to attack the d5-pawn as soon as possible.
%¢3Z%t
%

9.8g2 8b7 10.5h3 Db4 11.50f4 Exa2

EE
/,%
12.8xa2 Hxa2
Black is at least slightly better because of his
better pawn structure and more active pieces.
13.0-0 0-0 14.e4 Hb4?!
The knight has no support on this square and
just covers White’s weaknesses. Much stronger
was 14..Wb6 15.9d2 £a6 with a suong
initiative.
15.8e1 ¥Wag?!
The black bishop should go to a6, therefore
it makes no sense to keep the queen on the
same line as the white bishop (h1-a8). Instead 23...@3g4! 24.e5 Dxes5 25.%xb4 Hxc4 26.bxcd
15..8a6! 16.¢5! Dd7 17.exd6 exd6 18.&xg7 a6 27.8c1 ¥Wd4 28.201 W5 29.Wad &b5
txg7 and Black is slightly better. 30.%a5 £d7 31.%xc5 Exc5 32.8b12!
16.8c3 Da2 17.8b2 Db4 32.80d3 Ea5 33.¢5 dxc5 34.8xc5 Hal 35.@?g2
Just look how badly placed the black queen ££5 and Black is better, however it is hard to win
is. 17..8b8 18.0d2 Ra62! 19.e5! with the this position.
initiative. 32...216 33.Eb7 &5 34.2g2 h5 35.2e3 Lc8
18.8.c3 Bc8 19.¥Wd22! 36.8b8 g7 37.8d3 £d7 38.2b7 £h3 39.£32
After 19.8xb4 cxbd 20.¥d2! White has The key to playing positions like this is the skill
equality (Black is forced to sacrifice the b4-pawn). of waiting. In the game White cannot stand the
20...8a6 21.¥xb4 ¥a7 with good compensation, pressure and loses quickly. Instead after 39.2a7
but nothing more. Black is better, but White has excellent drawing
19..9a2 20.8a5 ¥Wa7? chances.
Both players were having an off-day. 39...2a5! 40.8b12?
20..5b4! 21.8xb4 cxbd 22.¥xb4 a6 with Mistakes, like Liverpool fans, never walk alone.
compensation. 40.6f2 Ba2t 41.8¢2 £d4! and Black is much
21.9a3? better, because the pawn is poisoned: 42.Exe7
21.¥xa2! Ba8 22.e5 De8 (22...80g42? 23.exd6 D! 43.Bed £c5 44.0el 2d7 45.9g2 g5 46.g4
exd6 24.8e7 and White wins) 23.exd6 £xd6 hxg4 47 .fxg4 f5 and the rook is trapped!
24 8xe7 WxaS 25.Wxa5 Hxa5 26.8d2 Halt 40...8d4 41.0F2 Ba2t 42.8¢e2 Ha3
27.401 £d4 28.5e2 £.c8 29.9xd4 cxd4 30.g2 The end is nigh.
£f5 Black’s compensation should be enough for 43.g4 Bxe3 44.2d1 Hd3?
draw, but nothing else. 0-1
21...20b4 22.8xb4?!
Better was 22.8c4 8a6 23.8xb4 cxb4 Game conclusions: This game shows the benefits of
24.Wixb4 g4 25.9d3 &xcd 26.bxcd Wd4 with 5...g6: by delaying taking on a6 with the bishop
good compensation. Black keeps the option of recapturing favourably
22...cxb4 23.8 c4? with the knight (if White messes up the move
23.Wxb4 Dgd 24.Bf1 £a6 25.8c4 He5 with order).
some initiative (or 25..%xc4”? 26.bxcd DeS
27.8b1 &2f6 28.9b7 ¥Wxb7 29.8xb7 Exc4 with
equality).
 The Benko Gambir Accepted 111 59

Game 21 game, Szabo — Bronstein, Moscow 1971.

Oleshkevich — Kotenko
Correspondence 1997
1.d4 &6 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.bxab
8xa6 6.8)c3 d6 7.2f3 g6 8.g3 Lg7
8..56bd7 9.2h3 &b6!? This move order was
introduced by the Hungarian grandmaster Laszlo
Szabo. Black puts pressure on the d5-pawn as
quickly as possible. 10.0-0 &c4 and now:
a) Polugayevsky was convinced that White is
much better after 11.b3 @bxd5 12.bxcd Dxc3
13.%c2 but here 13..0a4 is not necessary!
13..8g7! 14.8b2 (there is no compensation
after 14.%xc3? HOhS 15.Wb3 &xal) 14..80a4
and Black is fine.
b) 11.5d2! &xd5 12.6)xdS & fxd5 13.e4 Db4
14.e5 2g7 15.exd6 Wxd6 16.a3 0-0 17.8g2
a7 18.5e4 Wxdl 19.8xdl Hc2 20.Ebl and
White is slightly better, Polugayevsky — Szabo,
Budapest 1975.
9.8h3"
7 // 774
@ //
/x .e:x
§\\
11.8e1 ®c7 12.4f4 Zfb8 13.Wd2 Hb4
14.8acl h5! with the obvious idea ...5\f6-g4.
Black had full compensation for the pawn in
Beliavsky — Vaganian, USSR 1977.
11.8b1 ©b6 12.b3 &b7 with the usual
compensation, because White is forced to play
e2-e4.
11...Ba7
11..5b612 12.8d1 W7 13.e4 Dfd7 148144
Hed 15.8xd7 (15.b3 Dce5 with standard
compensation) 15..%xd7 16.b3 Ha3 17.%9d2
This position has been played in a few games, for
example: Sorokin — Malbran, Boca 1997. Black
should play very aggressively. 17...%g4 18. B2
£¢8! Black’s plan is to play ...%h5, ..&h3 with
an attack. Now 19.9g1 leads to unclear play.
(Not 19.Hac1?! Wh5 20.h4 &£g4 with the idea
...f5, when Black is better. 19.e5?! also does not
work, because after 19...¥h5! Black is better.)
11...8c42 12.8d1 8a7 13.e4 Wa8 14,042 A
new idea (14.a3 Eb8 15.9d2 £a6 16.8f1 Hab7
17 822 &xf1 18.8xf1 He8 19.b3 Wab gave Black
great compensation in Furman — Geller, USSR
1976). 14..Eb8! 15.g5 &h5 with unclear play,

,,,,,,,,,,,,

because 16.b3?! &xb3! 17.axb3 Hxal 18.8xd7

///////////
W¥h7 19.8a4 Wc8 gives Black better chances.
\

\\

12.8d1 ¥a8 13.e4 2b8 14.8b1 h5

E\\\
\\\\
\\\

14..5e81 15.b3 De5 16.5)xe5 £xe5 17.8b2

\

¢4 with standard compensation.

Y
&

15.82¢5
15.b3 Dg4d 16.2b2 Dge5 17.8xe5 &ixeS with
compensation.

@ W@% @g
,,,,,,,,,,,

w0 /
7

The idea of this move is simple: White wants to

take away the g4-square from the black knight.
In addition, the bishop occupies an important y//

diagonal (in contrast, on g2 it is very passive).

‘// /g// /
s\\\\

However, after 9.£h3 the d5-pawn is weak and

D@x\

White must soon play e2-e4, which gives Black /

good counterplay.
9...5bd7 10.0-0 0-0 11.¥c2
%@%”%/@
11.5d2 Was 1282 c4 13.2d1 Db6 14.2¢2
£b7 15.b4 cxb3 16.Dxb3 ¥a6 with an unclear
B8 &
60 'The Benko Gambit

15...52f8
E Wd @ E
asisa
By defending e7 Black prepares ... g4. Z 4
16.8g2 Dg4 17.h3 Dge5 18.Dxe5 Dxe5 é‘%é
19.b3 c4 20.a4 //%/ 7

The sharp 20.f4 gives nothing special after 20...

cxb3 21.axb3 £1d7 22.6)a4 £b5 with dangerous
compensation.
77 /
/ @ //@3o
20..cxb3 21.8xb3 Ec7 22.Hxb8t Wxb8
23.%b1
Risky is 23.Bb1?! ¥c8 24.82b3. Now Black
can play safe with 24..%c4 (Or try the more
ambitious 24..h4!"? with a dangerous attack:
25.8xh4?2? (25.%b2 is unclear) 25..8d3! and
Black is much better) 25.2a3 £a6 26.a5 D4
27.8a2 Hxa5 28.8xa5 £xc3 with equality. This is the latest word in fashion. From articles
23...Eb7 24.2b5! and commentary it seems that many grandmasters
White very sensibly wants a draw. 24.%a2 Wc7 believe that this idea by grandmaster Vladimir
25.8d22! (25.80b5 with equality) 25...h4! 26.g4 Epinshi kills the Benko Gambit. But, as you will
&\c4 with an attack. see in the next game, it is not as dangerous as the
24...8xb5 25.axb5 Hxb5 26.¥c1 Hd3 27.¥e3 merchants of doom would have you believe.
Hb2 28.Ecl 6 29.8h6 Dad 30.h4 D5 10...0-0
31.8f4 Wb7 This may seem automatic but, as the next game
-1/ shows, perhaps it is not the most accurate.
11.0-0 ¥a5 12.8d2
Game conclusions: 9.2h31? is an interesting idea 12.8c2 Efb8 13.8d1 Dg4d (13..0e8 14.8d2
but Black can gain sufficient counterplay by A7 15.24 Wb 16.b3 &6 17.8¢1 6 18.dxe6
putting pressure on d5 and thus forcing e2-e4. Hxe6 19.e3 and White keep slight pressure,
Yusupov —~ Ghaem Maghami, Yerevan 2001)
Game 22 14.8d2 Dge5 15.Dxe5 Hixe5 16.b3 Wa3 17.h3
Kramnik - Topalov White is slightly better Ibragimov — Gurevich,
Wijk aan Zee 2003 Fort Lauderdale 2004.
12...Bfb8?!
1.d4 D6 2.¢4 ¢5 3.d5 b5 4.cxb5 a6 5.bxat g6 12...8b7! Black should force White to play
6.9c3 8xa6 7.3 d6 8.8g2 Dbd7 9.2f3 g7 e2-e4. 13.Wc2 Wa6 (Instead 13..%a7 14.e4 £a6
Now Black has the following choice: Play 15.8fd1 Hga 16.h3 Dges 17.2el Hicd 18.4cl
against the d5-pawn or play the standard black &b7 19.2f1 Ddbe 20.2b51 ¥Wxa2 21.8g5
attack (...¥a5, ...Bfb8 etc.). In my opinion, the left White clearly better in B. Lalic — L. Milov,
second way is fine against a strong club player, Sauzal 2004, especially since 21...0xb2? fails
but if your opponent is titled (say, Elo over 2300) to 22.5d2.) 14.2h4 6 15.dxe6 (15.e4?! exd5
then the first option is safest. It is very easy to lose 16.exd5 ©b6 and Black is slightly better) 15...
the initiative in the standard attack and end up fxe6 16.a4 &xg2 17.9xg2 d5 18.b3 c4 19.80b5
fighting only for a draw. cxb3 ¥—1 Kasimdzhanov — Ramirez Alvarez,
10.8b11? Tripoli (2) 2004. It is strange that Black settled
for a draw, as he is already better. For example:
20.%xb3 Wxad 21.¥xa4 Bxad Of course it is hard
to win, but he could have tried without risks!
 The Benko Gambit Accepted 111 61

13.¥c2 §\e8 14.8fcl D7 15.a4
15.¢4?! This move is almost always bad in this
variation. 15...9b5 16.5xb5 ®xbs 17.2c3 ¥d3
18.8bcl ¥xc2 19.8xc2 £xc3 20.bxc3 &£c4 and
Black is at least equal.
15..%b6 16.b3
White is slightly better.
16...8xc3?!
In legal jargon this is called
accepting punishment”. 16...82b71? 17.¢4 £a6 and
Black has some compensation. (Instead 17...e6?!
is just weak: 18.dxe6 fxe6 19.&f4 B8 20.8bd1 e5
21.8e3 De6 22.8h3) Dd4 (22...Eae8? 23.2b5!
and White wins) 23.9xd4 exd4 24.8xd7 dxc3
25.%d3 Had8 26.Wc4+ Hh8 27.8b5! and
White is much better.) An example of Black’s
compensation: 18.9a2 ¥a7 19.b4 £c8 20.a5
£b5 21.Wc1 Bb7 22.8g5 Dd4 23.5xd4 &xd4
38.g5! hxg5 39.hxg5 5 40.¢4 6 41.8a5 Bc7
42.0e3 Bb8 43.Exe5 Exab 44.8c5!
White wins because the black king is cut off
from the kingside.
44..58b7 45.9f4 2b6 46.8c8 fxed 47.Lxed
Db7 48.8g8 Rc6 49.8xg6 B d6 50.£42
According to Kramnik White wins after
50.2g7! Hadt S1.0f3 €5 52.5f7 e6 53.2f6T
“voluntarily e7 54.5g3 Bb4 55.%h3!. The king makes
room for rook, and the gb-pawn is unstoppable.
For example: 55..Ha4 56.8f3 Bb4 57.¢6 8bs
58.g7 b8 59.8g3 ©f7 60.g8%T Hxe8 61.Exg8
begS 62. @g/} and White wins.
50...2a4t 51.0f3 e7 52.Hg7t &8 53.8b7
//’//V//

\
17.8xc3
20.8xb7
Hxd5
Exb7
18.0e5
21.8xe5
Dxe5
dxe5
19.8xd5
22.Becl
&b7
f6 //////
23.We4 ////
,/, %
I admire Topalov’s courage. In this position
Black should have excellent drawing chances...

//7/
but not against Kramnik.
23..sg7 24.8Bc3 Bab8 25.Wc2 We6 26.a5
soer
Better was 26..8b5 27.8xc5 Hxb3 28.Exb3
Wxb3 29.%g2! and White is much better
s ans
according to Kramnik.
27.8al Bxb3 28.2xb3 ¥xb3 29.¥xc5 ¥Wb2 53...Bc4?
30.Wc4t g7 31.Wa2 Wxa2 32.8xa2 In the majority of rook endings the rook should
This ending should be lost for Black. be behind the enemy pawns; this is no exception:
32...2a8 33.a6 Ba7 34.52g2 Bf72! 53...2al! 54.@g4 (54.50e4 Belt 55.¢d4 @gS
I am deeply convinced that Black’s best chance and Black should get a draw, although it will
for a draw is 34...h5!. The point is that now after clearly be after a hard fight) 54.. 8gl+ 55.%h5
a later g3-g4 (after White prepares it of course) 21 56.5h6 (56.2b4 tg7 and Black holds on)
Black will be able to reduce the material. It is 56...Bxf4 57.¢6 Bh4t ‘38.@g5 Zh1 and it is a
the old story: every pawn exchange brings the draw.
defender closer to a draw. 54.g6 Bcl 55.8g4 Ec5 56.8f7 ke8 57.65
35.963 &e6 36.g4 d6 37.h4 h6 1-0
Perhaps a better try was 37...h5 38.gxh5 gxh5
39.he4d e6, but White is probably winning here Game conclusions: The position after 10.8b112 is
as well: 40.f4! exfd 41.hxf4 HcS 42.¢4 b5 both fashionable and critical. After the natural
43.e5! fxeSt 44.50g5! cd 45.8xh5 e 46.0g4 10...0-0, as played in the game, White is generally
e3 4794 &d3 48.8a3t Bd2 49.Hxe3 Hxab a little better. The next game shows how Black
50.h5 and White wins. should play.
62 The Benko Gambit

Game 23 Ed8 with equality) 18...c4 with compensation.

P. Varga — M. Pap ¢) 13.a3 c4 (13..0-0"? 14.b4 cxbd 15.axb4
Bajmok 2001 HfxdS 16.8)xd5 Bxd5 17.8xd5 Dxd5 18.8b2 €5
and Black has compensation for the exchange)
1.d4 D6 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.bxa6 14b4 Dfxd5 15.0xd5 Exd5 16.Wc2 Wa8
8xa6 6.Dc3 d6 7.g3 g6 8.82g2 Rg7 9.3 17.8xd5 £xdS5 18.&b2 &5 19.f3 5 with full
bd7 10.8b1 Db6! compensation for the material.
11.b3

£ %?@/ E
///‘/Vt & /

% &/
/// 7//
/
7

% 7

///,,, / & /
7 g// & %E

a) 11...0-0 12.£b2 Ha7 13.0-0 ¥a8 14.5h4

Hb8 15.Hel £c8 16.h3 and White is slighty
better, Epishin — Kopylov, Werther 2001.
11.0-0
11.b421 0-0! 12.0-0 £c4 and Black is at least
b) 11...2b7 12.65h4 0-0 13.2b2 He8 14.9d2
Ne7 15.0-00 &xc3?! (15..¥d7 16.e4 e6 and
equal.
White is slightly better) 16.Wxc3 6 17.c4
11.5h4! £b7! 12.0-0 Ba5!?
Bxa2 18.2al and White is much better, Tunik
— Fominyh, St Petersburg 2001.
0 E@/ ¢) 11...2c8! And now:
/i/ /t,,f/,z/ cl) 12.0-02! &5 13.5b2 Ded 14.8xed Lxed
A x% 15.8d2 8c3 16.8b2 8xd2 17.8xh8 £h6 and
Black is at least slightly better.
c2) 12.e4?! 2a6 and Black has powerful
compensation.
c3) 12.%7d3? 0-0 13.0-0 (13.£b2 &f5 14.e4
£c8 with more than enough compensation for
the pawn, mainly because the white king is in
the centre) 13..Exa2! 14.8e5! £a6 (14...8a6
Now White has a choice: 15.8¢6 We8 16.b4! and White has the initiative)
a) 13.c4 Dfd7 14503 0-0 15.8el &cd 15.8¢4 (15.913? &xe2 16.8xe2 dxe5 and Black
16.00d2 £a6 17.Dxcd &xcd 18.Wd2 HeS and is better) 15...%a5 16.£d2 &fd7! and Black is at
Black has full compensation for the pawn. least equal.
b) 13.2d2 Wa8 14.e4 (14.8el Dcd 15.8f4 c4) 12.8d2? &f5! 13.e4 £c8 with standard
0-0 with compensation) 14...8246 15.Bel Hcd compensation.
16.9013 Hxd2 17.89xd2 Hd7 18.8al (18.2b3 ¢5) 12.8h4! is the critical line and had been
£xc3 19.bxc3 Hxa2 20.Eal £c4 21.5d2 &d3 assessed as an edge for White, however 12...ho!
22.8%xa2 Wxa2 23.Wa] Wixal 24.8xal 0-025.2a7 with the obvious idea of g6-g5 is very interesting.
 The Benko Gambit Accepted 111
But first let us look at 12...8)g4. The copyright to
this new idea belongs to Rafal Przedmojski, who is
a strong Polish master (vice champion of NATO)
and also my friend. It seems that after 12..%¢4
White is still better, but there are plenty of
interesting and important ideas in this variation,
so it is useful to study this line. 13.%c2 &6
(13...50h6 14.0-0 M5 15.0xf5 8xf5 16.e4 &d7
17.a4 and White is slightly better. Unfortunately
this kind of advantage gives White the chance to
play for a win without any risk.) 14.0-0 White
still has that edge.
Let us return to 12...h6!. 13.¥c2 ¥d7 14.f4!
(14.8b22 g5 15.8f3 &b7 16.8d2 0-0 gives
Black a good game, because White must either
lose the d5-pawn or leave the king in the centre)
///&///
A,,,
//
,,,,,,,
v %%@@J&
i o g
51) 14...0-0 15.0-0 Hgd (15...%a7 16.%h1
Was 17.4d2 Wa3 18.e4 £a6 19.Bf3 Williams
— Hertneck, Salzburg 2004. Here after 19...£c8,
with the idea £g4, Black has some counterplay.)
16.e4 and it seems that White is better.
¢52) 14...8¢4 15.h3 gains nothing for Black.
c53) 14..2b7! 15.£b2 0-0 16.¥9d2 Barsov —
Mannion, Port Erin 2003. 16...Efe8! This is the
new idea. Black wants to play ¢7-¢6 as soon as
possible. 17.0-0 Dg4!? 18.833 ¢6 19.dxe6 Wxe6
20.¢4 £5 21.exf5 (21.83b52! Bxa2 22.8)c7 Wxed
23.5xe8 Wxe8 24.8fel Ze4 and Black is better
because after 25.h3 he has 25..%c6l) 21..We3+
22.Wxe3 §xe3 with excellent counterplay.
11...8c4
11...0—0?" is less accurate: 12.b3! (12.He12! &c4
13.¢4 &xa2 14.5)xa2 Exa2 15.%b3 Ba7 16.2d2
W7 and Black is fine, Petran — Rovid, Budapest
2004) 12...8b7 13.£h4!? and White is slightly
63
better. The & h4-idea belongs to grandmaster
Vladimir Malaniuk, who played it against me in
a rapid game in 2003 in Warsaw.
12.5el
12.23 &fxd5! 13.6d2 Dxc3 14.8c61 Hf8
15.bxc3 B8 16.8¢2 £d5 17.e4 &c6 18.c4 Dad
Black is OK. g7 is the best place for the king.
12...8xa2 13.8)xa2 Exa2 14.2)c2 Ha8
14...%d7' looks better: 15.b4 (15.e4 0-0 and
Black is at least equal. 15.£a3?! Wa4! and Black is
already slightly better.) 15...0-0 16.bxc5 &bxd5
17.8xd5 Dxd5 18.Wxd5 Bxc2 19.£e3 Wgd
20.cxd6 exd6 21.¥xd6 Wxe2 with equality.
15.b4 Das 16.¥d3 Qd7 17.bxc5 Daxc5
18.%e3 0-0 19.2b2 £xb2 20.Exb2 He5
T
%%147
nin
A
A 4
//%//zg
A
White has some plus here, but it is clearly not
enough to force a win.
21.8Bb4 Had 22.Bxad Dxad 23.Eal Wbe
24.¥xb6 Dxb6 25.20d4 Hc8 26.f4 Ec4
27 .fxe5
27.80b5 D4 28.8b1 d7 29.43 Higf6 30.0g2
&ic5 and perhaps Black is a tiny amount betrer.
27...2xd4 28.8a6 Dc4 29.exd6 Dxd6
Interesting was 29...exd6 30.%£2 f512 31.2a7
b6 32.8a5 & cd with equality, because Black is
safe after 33.5ad &p7 34.8f1 Bed.
30.2a7 2f8 31.262 £5 32.%e3 Eb4 33.h4 h5
34.8f3 Bf7 35.8c72
This is a reason why you should keep playing in
drawish positions like this: there are plenty of ways
to make a mistake. Instead 35.9f2 was equal.
64 The Benko Gambit

0 7 7
¥
//%/
m//% | o
// A B
P E _wariAg
7
%/w///
) 7 g{g/%
"y/// 7 AR,,,,, {ӎ@/
///ggx
85
Ae s ,,,,,,,

35...0f6?
I am convinced that White would have lost
after 35..f4T! 3G.gxt4 D5t 37.0f2
Objectively, Black is better, but White still has
great drawing chances. The point is that in a real
game it is hard to force yourself to fight for a
draw after making a scupid mistake.
36.22 Eb3 37.8a7 Bc3 38.82d7 Bc2 39.2a7
2d2 40.Ba4 Bb2 41.2a7 Bb7 42.8xb7 Hxb7
43.e4 Dd6 44.exf5 oxf5 45.%0e3 Dokt
46.%2d4 9d6 47.8.e2 2f6 48.g4 hxgs 49.8xg4
N7 50.%e4 Hd6T
-1,
Game conclusions: This game covers one of the most
important variations in the book. 10...2b6! is
the correct reply to the popular 10.5b1 variation.
The analysis contained within this game (much
of it original) shows that 10.8b1 is not quite as
fearsome as many have claimed.
11.%c2
The following game is very weak as far as
Bxf4 White’s play is concerned, but it shows Black’s
ideas very clearly:
11.H2el Wa5 12.e4? This is a typical mistake
for club players (Elo up to 2000). Of course,
all plans like this should be prepared by 12.h3.
12..9g4 13.¥c2 HgeS 14.Dxe5 Hxe5 15.8d1
Efb8 16.a3? White creates another weakness.
After 16. h3 White would have chances for a
defence, but now Black is much better. 16...8c4
17.¥d2 ©d3* (17..8b3 and White will be
strangled soon) 18.2f1 &xb2?! (18...4¢5 goes
back to the good old plan) 19.£xb2 Exb2
20.Wxb2 &xc3 21.Wb7 @xal 22.Hxal £xfl
23.hxf1 Ha7 24.Wb8T g7 25.¥b27 6
%//
,,//,//7/////7/
%E/////;& @
,,,,,,

-, .

Game 24
van der Sterren — Adams
Ter Apel 1992

1.d4 @16 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.bxa6 g6

6.2c3 £xa6 7.013 d6 8.g3 £g7 9.£¢2 0-0
The move order in the Benko Gambit is
26.g4? 'There is one thing you should know
extremely important. Black’s last move is probably
abour this typical ending: If Black exchanges
a small mistake, because White can now obrain
queens and White’s a-pawn can be stopped on
an excellent version of the b1 plan.
a3 (or a2) then the rook ending is lost! Black
10.0-0 &\bd7
 The Benko Gambit Accepted 111 65

winning plan is clear: play &g7-f6-e5 and f7- 18.5xb5 Wixb5 19.8cl Wc4! with perfect
£5. It wins a pawn and the game. (26.a4! a6t compensation for Black, Taimanov — Browne,
27 We2 Wivedt 28 ¢hxe2 Ba5 This is not enough, Skopje 1970.
because Black cannot attack the e4-pawn. 29.t4 12.8b1 Efb8 13.8d2 He8 (13..20g4"? 14.h3
with a draw.) 26...¥a4! and Black is much better, Hge5 15.8xe5 Dxe5 16.b3 Wa3 17.¥cl £c8!
K. Michalski — Pinski, Warsaw 1994. with compensation: Black wants to force White
11..%a5 to play e2-e4) 14.Efcl &xc3 15.8xc3 Wxa2
11...¥b62! The queen is just useless here. [ have 16.¢4 Hic7 17.8al ¥b3 18.%d2 Hb5 19.¥g5
proved this at my own expense. 12.2d1 (usually £xc3 20.bxc3 ¥b2! and Black holds the position
White plays 12.2b1 2fb8 13.b3 &He8 14.5d2 because the €7-pawn is poisoned: 21.Wxe7??
Wa5 15.2b2 when he is slightly better) 12...8fb8 £d3! and Black wins.
13.8b1 ¥b7?! I saw one game that Black won 12.8d2 Dgd (12..2b8 13.h3 De8 14.2fel
using this move, so I decided to try it myself. Now N7 15.b3 H1b5! (15...c4 16.Eacl cxb3 17.axb3
10 years later I never play moves like this. There Abs 18.9xb5 ¥Wxb5 19.¥e4 and White keep
are no real advantages of attacking the d5-pawn some pressure) 16.2xb5 Wxb5 17.8c3 8xc3
with the queen: the bishop on g2 is very happy 18.%¥xc3%Wb4 19.Wxb4 Hxb4 with compensation)
about such clever ideas. (13...@g4 14.h3 Hged 13.h3 Sige5 14.b3 Efb8 with compensation.
15.8xe5 @xe5 gives Black some of the standard 12..9g4%
compensation) 14.b3 Bb6 15.64 He8 16.2f4 12..Bb8 13.8b1 Hgd 14.8d2 Dge5 15.8xe5
Nd7 17.Hel ¥c7 18.h4! and White is better, Hxe5 16.b3 and now there is a famous draw:
Chernuschevich — Pinski, Jaroslaviec 1995. 16..Wa3 17.8c1 ¥as 18.8d2
11..%c720 This has the same problem as 13.£.d2 Efb8 14.b3 ®b6 15.h3 Dge5 16.Dxe5
11..%b6. There is no sensible reason to play a Hxes5 17.8abl
move like this (of course it is not a losing mistake,
but White has easy play). 12.8d1 Zfb8 13.h3
EE 7/ /@/
7 / / o Aded
b6 14.50d2 Hfd7 15.8b1 and White is at least
slightly better.
11...5b6?! Now this move looks strange. If
Black wants to play ...22bG6 then it is better to do E / ///i// _
it before White castles, so as to keep the queen
on d1. 12.8d1 Ha7 13.23d2 Wa8 14.8f1 b8
15.b3! and White is slightly better.
7
11...Ha7 12.h3 Wa8?! After White has castled /8/
this plan looks odd. 13.2d1 fc4? (13..Eb8
with some compensation) 14.c4 Axa2 15.e5! ,,,,,,,,,,,

dxe5 16.d6 £b3 (16...e4 17.dxe7 He8 18.8¢g5

Bxe7 19.2gxed!? Dxed 20.8xd7 Hexd7 21.8xe4
We8 22.Hxa2 Hxa2 23.9xa2 We6 24.9¢3
Wxh3 25.2e3 and White is better) 17.Exa7 Now is the hardest moment for a Benko
8xc2 18.Exa8 Bxa8 19.Del!! 2xdl 20.2xa8 Gambit fan. White has played the opening
£b3 21.dxe7 and White is much better, Ilincic correctly. What next? As usual Black should force
— Sekulic, Yugoslavia 1993. White to play e2-e4!
12.8d1 17...2c8! 18.a4 £f5 19.¢4 £d7 20.2e3 ¥a5
12.6)d2 B8 13.8d1 &e8! Here this standard 21.f42!
manoeuvre gives Black excellent play. 14.Ebl 21.8dcl ¢4 22.bxcd &xcd 23.Bxb8T Exb8
Bc7 15.5b3 Wbd 16.2d2 Hes 17.8el ObS 24.£p5 BEc8 with compensation.
66 The Benko Gambit

21..0c4! 22bxcd £xc3 23.Exb8t Exb8 in combining this plan with h2-h4, because after
24.8b1 Bb4 25.8f1 Wxad 26.¥xad Lxad h7-h5 White’s weakness on g4 is much bigger
27.2d3 £b3 28.%f2 £a2 29.Bc122 than Black’s on g5.
29.8xb4 &xb4 and the position is more or 9...0-0 10.Eb1
less equal, but maybe White has some tiny plus 10.0-0 &bd7 11.Wc2 Wask (11..8a7
because of his space advantage. Instead 29...cxb4 12.b3 a8 13.2b1 8b8 14.8d1 Hab7 15.2d2
30.c5! dxc5 31.8xc5 would leave White better, c4 16.8dcl Dc5 17.bxcd 8xcd 18.9f4 Wa6
because the b-pawn is under control (after ...b3 19.8xb7 Bxb7 20.2b1 Hxblf 21.¥xbl Qfd7
White will play £c4). with compensation for the material in Aseev
29...£b2 30.Ec2 £b1 31.8d2 £xd3 32.8xd3 — Ponomariov, Ohrid 2001.) 12.2d2 Efb8
Bxcd 13.0f4 &c8 14.b3 Wa6 15.Da4 He5 16.&c3
Put simply: Black wins. He8 17.60b2 He7 18.0fd3 ©xd3 19.Hxd3
33.8b3 Bb4 34.Exb4 cxb4 35.%¢2 b3 36.52d3 £xc3 20.¥xc3 b5 21.%Wcl &f5 with excellent
2£637.f5 218 38.g4 Le8 39.2d2 £d7 40.8a5 compensation, Sjodahl — Adams, Oakham 1992.
g5 41.8d2 ®c7 42.8e3 h6 43.82d2 b7 For more of this game, see the Introduction.
0-1 10...2bd7 11.0-0 ¥a5 12.8d2 Bfb8 13.¥c2
De8 14.b3 Wa3l 15.0f4
Game conclusions: This game shows that even 15.8c1 Was 16.2b2 c4? (16..8c7 17.0f4
when White achieves a decent version of the &b1- Hbs 18.5xb5 Wxb5 19.4xg7 xg7 20.a3 HeS
plan Black still has reasonable practical chances. 21.%c3 g8 with compensation) 17.2f4 &c5
18.&a1 cxb3 19.axb3 &Hf6! with compensation.
Game 25 Black’s last move prevented £e4, which would
Korchnoi — Baklan give White some initiative.
Ter Apel 2000 15...80c7 16.£h3 Des5 17.Dg222
A terrible blunder. Better was 17.2cl Wa5
1.d4 D16 2.c4 c5 3.d5 b5 4.cxb5 a6 5.bxa6 g6 18.£b2 ©b5 19.5xb5 ¥xb5 when Black has
6.9c3 8xa6 7.g3 d6 8.2g2 £g7 9.20h3 compensation.

zéfim/zfiz
o L 3 ////@/
/ /fi”i
'gf/
,,,,,,
//

//7/// »
/
/%/
) é/%&
//
1
san
\\\

Ms %&/&g/
,,,,,,, ////// /E/ /E;
This plan is not dangerous for Black. White 17...8xe2!
keeps control of the d5-pawn, but loses the This is a simple motf, but well worth
possibility of fighting for the e5-square (it will remembering.
be impossible to play f2-f4). There is no sense 18.Dxe2
 The Benko Gambit Accepted 111 67

18.&c1 £d3 and Black wins. pieces. (14..2fb8 15.8xc4 &xcd 16.82d2 &ad7
18..¥xa2 19.¥xa2 Bxa2 20.f4 Exd2 21.2£2 17.b3 226 18.8e4 ¥b5S 19.8c3 £xc3 20.¥xc3
Ad3t 22.50e3 Bb2! 23.8xb2 and White is slightly better, Stempin — Kumaran,
23.%xd3 B8xb3t and Black wins. Lyngby 1990) 15.¥c2 c4 16.Hedl Bfc8 17.e4
23..50xb2 24.5c1 Dxd5t 25.2F3 c4 26.bxcd £bd7 18.£h3 h6 19.8e3 Hcb8 20.5f3 &8
Sxcd 27.8d1 D3 28.8el €5 29.fxe5 dxe5 21.50d4 De5 with excellent compensation for
0-1 the pawn, Hiltunen — Nilsson, corr. 1998.
13.8g5 h6 14.8c1 2f6 15.8b1 &c4 16.e4
Game conclusions: The plan with 9.2h3 is not $£xa2 17.5)xa2 Bxa2
dangerous for Black. White’s knight will look

//% Eg//
good on f4, but Black’s knight will be even
prettier on e5.

Game 26 ///////

Bacrot — Ponomariov ,,,,,, 7- 7 7

Lausanne 1999 % Al 7 2
//A,/ //l/ /53 /% /% »
1.d4 &6 2.c4 5 3.d5 b5 4.cxb5 a6 5.bxa6 g6
6.9c3 &xa6 7.013 =Qg7 8.g3 d6 9.§g2 Hbd7
. 8 )
/ 7 7 %
10.0-0 Ab6! /%//wy B| 57, & Y
X 7 e
E A
BAWEH
T
&
//,,,74
g.m 18.e5¢!
In my opinion this is too sharp.
18.b4!? This move is close to being a draw
offer. 18...cxb4 19.Bxb4 White may claim some
plus here, but it is very unlikely to turn into a
/ whole point. 19..9fd7 20.¥b3 Had 21.2e3
/&/? /&% Hxb4 22.%xb4 Wa8 23.00d4 2b8 24.¥b5 Wad
L = O e S s, e ]
25.Wxa4 Dxad with equality, Hulak — Sermek,
////// Pruj 1995.
18...2#xd5 19.¢6 Ba4
After 19...83b4 20.exf77 Bxf7 21.2£h3 White
This is the 100% guaranteed strongest move.
Black attacks the d5-pawn and prevents White has compensation. For example: 21..%2h7
22.8e6 W8 23 . &xf7 Wxf7 24.8Bed ¥dS 25.We2
from playing Wc2 followed by Bd1. The position
after 10.0—-0 2bG was a main line in the 1990s.
€526.8e3 W4 27 Wxcd Dxcd 28.8c3d529.8d2
as in P Nikolic — Fedorowicz, Luzern 1989.
11.8el
Black is safe after 11.8el 0-0 12.83c2 &c4 29...%xd2! (instead weak is 29...£d6? 30.Exc5
and White is much better) 30.8xd2 c4 31.8xc4!
13.8b1 Hd7 14.9ed Wa5 15.a3 BfbS with full
compensation.
(Very risky is 31.5a3?! Hxa3 32.bxa3 @d3. My
old “rule” that Black’s pawns are faster in endings
11...0-0 12.8f4 Dh5
12.80c4 13.%c1 Was 14.Hd2 &be! works perfectly here.) 31...dxc4 32.2xb4 with
Remember this useful trick: when you have equality.
20.exf7t
a space advantage there no reason to exchange
68 "The Benko Gambit

20.h4 8d4 21.exf7T Bxf7 22.¥c2 g5 23.863 32.g4 ¥b5 33.g5 Wxe2 34.Hxe2 hxg5 35.8xg5
(23.6065?! Hcd 24.Wxcd! Hxcd 25.8xd5 DeS Zff8:
26.8al and White has compensation for the Black is beginning to falter. 35...2d3! 36.2xf6
queen, but only enough for a draw, van der Sterren 8xf6 37.8d2 e4 38.9g2 8xb2 39.8bd1 c4 and
— Hertneck, Germany 1994) 23..Ec4 24.We2 Black wins.
According to Hertneck, White has compensation 36.8h2 Df6 37.0g2 Dbd5 38.2d1 Efc8
here; in my opinion, Black is better. 24...e5! 39.8xf6 Dxf6 40.82xd6 Bxb2 41.£h5 c442.8d1
(24..9b4 25.5d2 Hd4 26.&£h3 ©d3 27.8f1 b7 43.9e3 Bb1 44.52f1 28 45.8e6 Bd8
Bh8 28.8e6 Hf6 29.0b3 Hixcl 30.Ebxcl with 45..8¢5 46.5 xc4 Bxd11 47.be2 Bd5 48.2h6
full compensation) 25.Ed1 @b4 26.5d2 Hd4 Dd7 49.8h71 &¢8 50.2h5 £f8 and Black wins.
and Black’s pawn chain is very dangerous. 46.%e2 c3 47.8c2 Bd2t 48.%8f3 e4t 49.52g3
20...Bxf7 21.¥c22! Hglt 50.2h3
Better was 21.83h4 Bd4 22.Wc2 - see 20.6)h4.
21...8£6 22.9h4 Db4 23.We2 g5?
After this move White has compensation o0 %
because of the weakness of the white squares.
Instead 23...e5 24.2d1 W8 25.%b5 d5 and
////
Black is much berter.
24.9)f3 €5 25.h4?
After this mistake White loses the initiative
/
and Black is close to winning. 25.8xg5! hxg5
26.8xg5 c4 (risky is 26..Wf8 27.4xf6 Wxf6
28.f4! (28.¥b52! d5 and Black is better) 28...8a7
29.fxe5 ®xe5 30.%b5 and, because of his
better king position, White has the advantage)
27.f4 £d3 28.b3! with an unclear game. After
28..80xel (28..Ha5 is unclear) 29.bxad Hxg2
30.Wxg2 exfd 31l.gxfd @xad 32.9d5T Lhs 50...8h6?
33.8xf6 Wxf6 34.Whs5t Wh6 35.%xh6t &£xh6 This is a good example that even if you are a
36.8b4 the game is equal. GM it is hard to win a winning position. 50...Eel
25...gxh4 26.9xh4 a7 27.£4 Ee7 28.£5 51.89g4 Hxh2t 52.5xh2 He3t 53.80h4 Ee2 and
28.¢4 looks better c.g. 28...Ee8 (28..Bee6? Black wins.
29.f5 He8 30.g5 Ef7 31.WhS and White is better) 51.9)c4 £452.9xd2 cxd2 53.8e2 Ecl 54.82a4
29.5f5¥d7 30.8d2 Dc2 31.8ecl Hd4 32.8xd4 Bel?!
exd4 33.Wd1 Wf7 and Black is much better. 54..8h1t 55.5g2 Hel 56.8f2 &dS 57.%2h3
28...2e8 e3 and Black wins.
28...d5"7 29.g4 The only reasonable chance 55.8b3 Bb1 56.8c2 Eb2?!
(29.8e3 Wd6 30.g4 He8 31.00g6 4 32.84 Wco 56..Bhit 57.€@g2 Hel and Black wins.
33.%e5 Wc8 34.8bd1 B8 35.9¢6 Ef7 36.Wb5 57.82d1 Bb1 58.8c2 Hcl 59.8a4 Ec31?
&b7 and Black is much better). 29...8c6 30.g5 ‘The last chance is missed. Black could still have
£f7 31.20g6 e4 and Black is better because of won with 59..Bh1¥ 60.@g2 Eel.
32.06 hxg5 33.0¢5 xf6 34.0x7 hxf7 35 £d1 60.52g2 €3 61.2f3 &g5
Hd3 when White is in trouble. 61..00d5 62.8ed Hcl 63.2d4 Bf1t 64.52g2
29.2f3 ¥d7 30.2h5 Eb8 31.82g6 H6d5 Hel 65.Hexd2 exd2 66.2xd5 with equality.
31...c41? 32.b3 £d3 33.2d1 d5 34.bxcd dxc4 62.22xe3 £xe3 63.8xe3
35.9f3 W6 and Black is much berter. V-1
 The Benko Gambit Accepted 111 69

Game conclusions: 'This game demonstrates 19.8d1 D5 20.e3 Exd1 21.¥xd1 €6 22.b4 exd5
Black’s accurate move order 10.0-0 £b6! when 23.8xd5 $xd5 24.¥xdS He7 Grigore — Banikas,
after 11.8el 0-0 12.£f4 Black seems to have no [stanbul 2002. 25.¥g5! ¢4 26.b5 (26.¥gd W7
serious problems. with compensation) 26..#b7 27.a4 Ea8 with
excellent compensation for Black.
Game 27 12.h3 In my opinion this move is useless
Hiibner — Heinemann now. Black has many good plans. For example:
Germany 2003 12..06d7 1392 Dcd 14.5d2 Was 15.8db1
£cb6 16.8d1 Eb8 17.9h2 Ha7 18.8d2 D4
1.d4 96 2.563 g6 3.c4 ¢5 4.d5 b5 5.cxb5 a6 19.2¢1 Hcb6 20.£d2 with equality, Piket — van
6.bxa6 &g7 7.8c3 £xab 8.g3 d6 9.8g2 Dbd7 Wely, Monte Carlo 1997.
10.0-0 b6 11.8el 0-0 12...0£d7 13.¥c2 D4 14.81 Was5?!
Better is 14..8deS 15.8xe5 Hxe5 16.50g2
Was 17.f4 &xf1+ 18.8xf1 Hcd 19.83 and now:
a) After the natural 19...e6?! White has a strong

% Zt
// /7/
reply: 20.5! De5 21.5f2 and White is better

&fixé/
27
because Black has no good move to prevent f5-
/
f6.

Aé Z S b) 19..15? 20.%e2! Wb4 21.exf5 8xf5 22.a3

2% //,%Z
i,
Wb3 23.0e4 Wa4 24.5¢5 and White is much

e ;
better.

/523& ¢) 19..Bfb8 20.2bl Andresen — Busch,

Germany 1992. Now Black should play 20...#b4!
A é/&afl c% with excellent compensation for the pawn.
=
£, %@/
12.e4
Now White can use the factor that only one
%¢¢7®?f 7
,,,
black knight can go to €5 and save a move (play
e2-e4 without h2-h3).
12.5d2 and now:
/m/é/ /
e // a&
a) 12.%c7 13.8bl Wb7 14.b3 Dfxd5
15.6xd5 Dixd5 16.De4 Had8 17.8¢5! This is a
new idea. 17...5fe8 18.Df6T ext6 19.£xd5 and
White is slightly better. %/7 )
b) 12..5Ha7 As [ have already said, I am not
a big fan of the idea behind this move: battling
B
=l & 2K/ HLE—1
=%

with a queen on a8 against a bishop on g2 looks

very odd to me. 13.2f1! ¥a8 14.8e3 &bS 15.8xc4!
1582 £cd 16.b3 Dxe3 17.8xe3 Dgd 18.82d2 This excellent move must have been a hard
c4 19.h3 He5 20.Hecl Be7 21.%d1 and White decision. By exchanging his bishop White takes
keeps a slight advantage, Tkachiev — Izoria, more control in the centre (e4-e5 will soon be
Istanbul 2003. possible). The vital issue to judge was that Black
) 12...%d7? 13.8b1 b7 14.2f1 Ha4 15.a3 will have no time to exploit the weak light squares
Dgd 16.9xa4 Bxad 17.%c2 #d4 18.h3 &h6 around White’s king.
70 The Benko Gambit

15...8xc4 16.8f4 Db6 30.Dxf6!

16.. W26 17.e5! (17.Badl £xa2 18.8al &xc3! Very simple, but still a nice trick: White wins.
An eye for an eye! 19.%xc3 Efb8 20.2h6 f6 30...9£7 31.8c6 Wxf6 32.8e7+ D7 33.Dxd8
21.80d2 c4 and Black is fine.) 17...2{b8 18.H2e3 ¥xc3 34.bxc3 8xh3 35.Dxf7
and White has a large advantage. 1-0
17.8e3 £.d4 18.Eeel £g7 19.¥d2 Bfd8 Game conclusions: With accurate play Black has
19..£5 20.e5 Bfd8 21.£¢5 He8 22.2ad1 and reasonable prospects against 12.e4, 12.8d2 and
White is better. 12.h3.
20.£h6 £a6?
That is not the safe way to play: Black loses Chapter conclusions: The position after 1.d4 §f6
control of the dark squares in his own camp 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.bxa6 gb 6.8c3
without any compensation. After 20..2h8 &xab 7.3 £g7 8.¢3 d6 9.8g2 Dbd7 is vital
White is only slightly better. for the Benko Gambit. The much feared 10.Eb1
21.8xg7 Pxg7 22.5 D4 2382 W7 line should be met by 10...23b6! when Black
24.exd6 exd6 25.De4 f6 has reasonable chances. Similarly, 10.0-0 should
25...%a5 26.b3 He5 27.8%e5 dxe5 28.Hed! be met by 10...8b6! when Black has decent
£b7 29.d6 and White is much better. prospects.
26.8c3 De5 27.80d4! Ac8 28.f4 Dg4 29.h3
Ah6

E sE .
T4 oS xia
. //8% ;)
) s
ja) ” 7- »
g B2 @
Chapter 4

The Positional 5.b6

1.d4 D6 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.b6!2 5...Wxb6, which is also fully playable. Finaily we
shall look at 5...d6!? in an classic game played by
no less than Pal Benko himself.

,,,,,,,,,,,,,,,,, Game 28
Va. Akobian — Pinski
Yerevan 1999

,,,
/x&7/5 1.d4 &6
a5?!
2.c4 c5 3.d5 b5 4.cxb5 a6 5.b6L

////
This is a tricky move to try to reach an

%///z{@é@
BWEINE
////////////
This is the most popular way to decline the
Benko Gambit. It was taken from obscurity and
popularised by Alexei Shirov in the beginning of
the 1990s, when he won some excellent games
with it. However Black players found a way, and
today it is not considered so dangerous. Still,
many Benko players do not like to play against
this continuation. Instead of giving up a pawn
for an attack, Black is forced to play positional
chess.
In this chapter we shall first look at 5...a5?2],
which does not justify the ! it was given
in Informant, when it was introduced into
tournament play. After this we shall look at
5...¢6, which promises Black equal chances, and
interesting game. However, it has a slight defect:
if White plays the correct moves he will gain a
large advantage.
Worth noting is 5...g6 6.8c3 &g7 7.e4 0-0?!
8.¢5 Ne8 9. b3 and White is much better.
6.2c3 £a6 7.263!
There was a time when people believed that
only the sharp 7.f4 could damage Black’s plans.
However, like the majority of human beliefs, this
turned out to be an illusion:
a) 7..d6 8.83 Nbd7 9.b7 Eb8 10.e4 Lxfl
11.Exf1 £Hb6 12.e5 dxe5 13.fxe5 HfxdS 14.8¢5
Wd7 15.8xf7 and White is much better, Seirawan
— Su. Polgar, Monte Carlo 1994.
b) 7...¥xb6! is the right move. 8.2f3 ¢6 9.¢4
c4!10.Me2 (10.e5 HxdS 11.0xd5 exd5 12.¥xd5
c6 13.8xc4 &b4T with greac compensation)
10..8c¢5 11.0d2 0-0 12.g4 Hc8 13.g5 De8
14.e5 Weber — Nicholls, e-mail 1999. And now
after 14...£d4 the position is terribly unclear.
7...d6 8.e4
72 The Benko Gambit

EaEa E@/
Wee EX
T

zm % % z
/// z,,,/t/%x
A
/ Z

‘%f///’/%
,,,,,,,,,,,,,,,,

%/@ /@/
W //// m i %//&/&//&
,,,,,,,,,,, ,ff////// |
“There are lies, damn lies and then there are 14.Dxd7?2!
statistics.” Nevertheless, the statistics in this White exchanges the perfect knight on €5 for
variation are very good: in a total of 13 games the one on b8: not a great idea. Instead 14.8c4
White has won 9 and lost only twice. Wb4 15.a3 Wxad4 16.2xa4 and White is much
Instead Black has no great problems after 8.g3 better.
g6 9.82g2 £¢7 10.0-0 0-0 11.8el Hbd7 12.£f4 14...0xd7 15.8el
(12.e4? If you play a move like this you should White still keeps some pressure.
be sure that ...4g4 is not dangerous: 12...4)g4! 15...2f6!
13.2¢5 h6 14.2¢3 ¥xb6 15.%b3 2fb8 16.Wxb6 15...f6 16.2c1 ¥b7 17.h4 and White stands
Hxb6 17.&2c1 ©de5S and Black is better, Meier better.
— L. Rogers, Baden 1998) 12...23h5 13.2¢5 h6 16.2h6?!
14.8d2 Dhf6 15.¢4 Hxb6 16.b3 Hh7 17.8cl 16.h4! A typical motif: the rook will join the
£b7 18.a4 and the game is unclear, R. Kempinski game via h3.
— Przedmojski, Suwalki 1999. 16..¥xb2 17.0e4
8...8xf1 9.hxf1 ¥xb6 10.e5! 17.%¢6 &xc3! 18.Wxa8T b8 19.2d1 (19.8cl
Six years ago this innovation closed the debate 2d4 20.8d2 0-0 21.WxaS 6! and Black has
on this variation. I expected only 10. a4 121 &\ fd7 a dangerous initiative) 19..8d4 20.2d2 0-0
11.e5. This was this line thac Hungarian GM 21.Wxa5 W2 22.5el e6! with a strong attack.
Peter Lukacs chose in his game against Zoltan
Medvegy in Budapest 1999. | had prepared a
novelty: 11..%b4! 12.6 fxe6 13.dxe6 ¥xa4 z/ @/ =
14.9xa4 D6
/ ém/ /
m\\x
e

Also no problem is 10.g3 g6 11.5kg2 £g712.h3 /@f

\\

\\»

0-0 13.8el ©fd7 with good play for Black.

x

//////

10...dxe5
Now 10..8fd7 does not look good: 11.e6
fxe6 12.@g5! (After dxeG White is, of course,
also better, but the black knight can go to the
centre immediately b8-c6). 12...e5 13.Wg4 Qa6
///
é
14.%xh7 White is much better. For example:
14..916 15.¥a4t D7 16,551
11.9xe5 g6 12.Wa4t Hfd7 13.8g5! L7
 The Positional 5.b6 73

17...52d8! 18.d62! 26.Wxe2 d5 27.¥b5! and White wins.

Better was 18.h4 with the idea Eh1-h3. b) 24..Wa1? 25.¥e2 and White wins.
18...8e8? c) 24...Bxe4! 25.8xb2 Bxelt 26.¥xel &xb2
18.. b4 19.¥c2 He8 20.8g5 &xg5 21.8xg5 27.¥e7 Now White has time to grab a few
6 22.dxe7 Hxe7 23.8e3 Hc8 and Black is just black pawns. Therefore, he has at least enough
better. counterplay: 27...2f6 28.¥xf7 Eb8
19.h42 24,..Bxed
19.8g5! 2xg5 20.&xg5 f6 21.We4! (in the Now everything is clear.
18...¥b4 line this move was impossible) 21...Eb8 25.8¢3 Rae8 26.Bxed Hxed 27.8gl Des
22.dxe7t shc7 23.2804F &S is unclear. 28.8d2 Wa1
19...exd6 20.8h3 ¥Wd4 21.¥c2 72 0-1
21..d5! 22.8d3 WesS 23.8de3 (23.Wd2 d4
24.8f4 W5 and Black has won the battle) Game conclusions: Despite Black’s success in this
23..Wh2 24 Wxb2 &xb2 25.5xc5 Bxe3 26.8xe3 game, 10.e5! secures White a better game and the
the7 27.8b1 &xc5 28.Hxb2 De6 and in this 5...a5 line simply cannot be recommended.
ending Black has great chances for a win.
22.2d3 ¥b22 Game 29
22.. 8xed! 23.8xd4 Bxelt 24.%hxel fxd4 Vladimirov — V. Mikhalevski
25.8d2 (25.We4? BbS 26.8e3 15 27 . We7 Bb1 T Calcutta 2001
28.¢he2 Hb2t 29.%0el £xe3 30.fxe3 Exa2 and
Black’s advantage is decisive) 25..%2e5 26.t4 1.d4 ££6 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.b6 €6
D6 27.Wb3 He81 and Black has good winning After this move we leave the usual Benko
chances. Gambit structure; now it is more like a Benoni.
23.%d1! £d42? 6.5c3

EaAsUeE K
23...%Wxa2 24 Exd6 Had$ In this rather unclear
position Black is probably still better.

B,
% ,,,,,,, 7 e 7 8 7 s Z

%,/z/ ) A 4>iwfi1%%
o & 7/ AL //
> //
o, LR sy, e T

2 &
& & 7

oo s K
A

, Z.in
Z&;;?¢/Z;;
w7
, , , 4//7, , ,
/%fla/
/% /// ,,,,, /,

s 0 %%/// B SWEORE
- swnel
,,,,,,,,,,,,

6...8b7'2 This is not often played, but it is

24.8f42 interesting. 7.¢4 exdS 8.exd5 d6 9.a4 a5 10.2b5+
Better was 24.8b3! and now: fbd7 11.2g5 (11.8ge2 &fe7 12.0-0 0-0
a) 24..Wxa2? 25 He2 Wxe2t (25..Wal 26.8b1 13.9g31? (13.8f4 ©ixb6 14.Bel with equality)
Wa3 27.8c1 Wb4 28.8xb4 axb4 and Black 13..5xb6 1405 DfxdS 15.Wgd £f6! and
only has nominal compensation for the queen) Black is OK. For example: 16.£h6 £c8 17.8xd5
74 The Benko Gambit

Dxd5 18.8xg7 &xf5 19.9xf5 &xg7 20.Wxd5 threatening ...2a5.) 15..8xc5 16.Hxc5 £xa2
&xb2 and Black is slightly better, Rogozenko and Black is safe.
— Milu, Bucharest 1999.) 11...&¢7 12.2xf6 £xf6 o) 13.e3 a5%? 14.8d2 a4 15.8¢1 &xb6 16.8c4
13.%g4 h5 14.9f5 o6 15.%h3 &xc3t 16.bxc3 $4a617.8xa68xa6 18.8c2 f6 19.50e2 Eb8 20.2d 1
£xdS 17.8e2 with an unclear position. &f7 21.e4 £c7 22.8c1 fe5 23.8d71 g8 24.£4
7.0 xd5 exd5 8.¥xd5 Dc6 9.2f3 EbS £d4 25.8e3 Bab6 26.2xd4 cxd4 27.2xd4 Exb2
Theory promises Black equality after this and Black’s counterplay is enough for a draw, V.
move. Still, in my opinion, White has at least Mikhalevski — Gershon, Biel 2001.
some psychological plus. The point is that the d) White’s best try is probably. 13.e4 &b8
character of this position is completely different 14.8e3 £xb6 15.0-0-0 L6 16.2d6 c4 17.2xb6
from the usual Benko Gambit. 9...&2e71? 10.9e5 Hxb6 18.f4 g6 19.g4!. The only real chance to
0-0 11.5xc6 dxc6 12.Wxd8 £xd8!? play for a win. 19...2xg4! (better than 19...c3
20.b3 @xg4 21.2d3 and White keeps some
pressure, Grabarczyk — Ehrenfeuchtr, Warsaw
1997) 20.8xc4 Re6 21.8xe6 fxe6 22.8f1 e5
23.5 Bb4 24.8g1 Exe4d 25.fxgb Hcdt and a draw
is imminent.
9...£b7?! has a worse reputation, and for good
reasons: 10.9e3! We7 (10...0xe5 11.¥xb7 b8
12.%xa6 Bxb6 13.Wa4 g6 14.¥e4 and White
is much better, according to Shirov) 11.&f4 g5
12.8%c6 8xc6 13.%e5! A new idea. In myopinion
it refutes the 9..&b7? variation. (13.Wxg5?!
Wxg5 14.8xg5 Eb8 15.8f6 Hg8 16.2c3 Exb6
This ending is not as bad as its reputation 17.3 and White is slightly better according
(according to ECO it is almost lost). Black’s to Anatoly Karpov) 13..Hg8 14.Wxe7t £xe7
pawn structure is, of course, hopeless but it also 15.8¢7! £d8 16.¢3 £xc7 17.bxc7 Bc8 18.8xab
gives chances for active play. I use this variation Bxc7 19.f3 and White has one pawn more.
in Internet games and very often I am better 10.8d2%2
after a few more moves. Conclusion: White is This is popular, but not dangerous.
slightly better in this ending, but should play Too slow is 10.g3?! £e7 11.8g2 0-0 12.0-0
very carefully. (Instead 12...52xd8?! 13.8f4 &f6 Hxb6 13.2d1 Hb4 14.¥b3 d5 and Black is
14.8d1! and White is much better, Karpov already better.
— Christiansen, Wijk aan Zee 1993. The idea is 10.8g5 Re7 11.5e5 0-0 12.8xc6 dxc6
that after 14.. Bxd1t 15.¢oxd1 £xb2 16.e3 &6 13.Wxd8 £xd8 14.2xd8 Hxd8§ 15.e3 Exb6 with
17.£c7 White is winning.) counterplay. It is important to know that such
Now after 12...8xd8!? White has several positions are normally OK for Black.
options: 10.8Ye5 W6 11.5xc6 dxc6 12.¥edt (12.%13
a) Black is pretty safe after 13.g3 Eb8 14.2d2 $e7 13.8d2! (13.e32! Wxf3 14.gxf3 Hxb6 15.8c4
Bxb6 15.8c3 &f6! 16.8xf6 gxf6 17.b3 Le6 86 and Black is slightly betcer) 13..¥xb2
18.¢3 &d5 19.Hgl a5! 20.Bcl a4 21.2c4 14.8c3 Wa3l (14..¥c2? 15.8xg7 c4 16.Wc3
(21.bxad4 Ha8! with the initiative) 21...axb3 Wa4 17.8b1 Hg8 18.2e5 and White is much
22.axb3 &xc4 23.8xc4 Exb3 24.¢he2 Bb21 16-14 better) 15.8xg7 Wxf3 16.gxf3 Hg8 17.8c3 Exb6
Urban — Bukhman, Warsaw 1991. 18.e4 £e6 19.h4 c4 20.8d1 Bb5 21.%d2 Eh5
b) 13.8e3 &xb6! 14.8cl Le6 15.8xc5 (15. 22.8b1 Y5 Podzielny — Christensen, England
b3 Hab8! with the initiative. For example, 2000) 12...8e7 13.g3 (13.9£42! Exbo! 14. %577
 The Positional 5.b6 75

£d8! 15.%xc8 0-0 The threat is ...8xb2. It is Or 18.8f4 He6t 19.£e3 Hh8 20.Bdl Hxe3d
hard to find a defence. For example: 16.82d2 21.fxe3 Hd6 and Black is berter) 18...Be6T
{or 16.¥h3 8xb2 17.2d2 Exd2 and Black 19.0f1 £5 20.8e5 ££6 21.8xf6 Bfxf6 and Black
wins) 16...8xb2 17.8d1 Exd2 and Black wins) is better.
13..Bxb6 14.8g2 0-0 15.0-0 Le6 16.2f4 Wd4 18.0-0-0 &b6
17.8e5 Wxed 18.8xed {5 19.2d3 c4 20.8c2

W Ed
2f6 with counterplay, B. Kovacevic — Waitzkin,
Matinhos 1994.
10..8¢7 11.8c3 0-0 12.e4 Hxb6 13.8c4 %A% %x%x

m
Z 2 “y
Ab4! % %
7

/// / / // , , , 5%/2
A,,,,% ///

. . @fi% /z
"
%
,,,,,, %®% 19.8xf71! bxf7?
& // / /i 19..Bxf7 20.9xd81 &xd8 21.Bxd8t Ef8
22.8xf81 &xf8 23.2h4 He6 (23..Hgd?!
24.8el! Pc4 (24...8xh4 25.2d61 ©Ff7 26.He7T
and White is much better) 25.£¢3 &f7 26.£3
14.¥d22! and Whire is better) 24.2el g5 25.f5 &xg2
Now White should run away before it is too and the position is close to a draw, because after
late: 14.8xb4 £b7 15.Wd3 Hxb4 16.0-0 Wa8 26.8g71 ©f7 27.0h61 he7 28.8f81 &d7
17.8fel Bxb2 18.%xd7 and White stays alive Black is fine.
and in decent shape. 20.89f41 he6 21.2f6! gxf6?
14..8b7 15.23 21..8fxf6 22.¥e3t &f7 23.0e5T He8
15.0-0-02! &xed 16.Wxd7 2g51 17.Dxg5 24 Bxd8t &xd8 25.8d1T dhes 26.@xg6 Bxg6
Wxg5T 18.¥d2 We6 with the initiative. 27.f3 and White is better, but Black can hold the
15.We2? &xed! 16.Wxe4 d5 with an arrack. position.
15...d5! 16.exd5 22.Bhelt &f7 23.9e512
16.axb4? dxc4 17.%xd8 Bxd8 18.bxc5 &xc5 23.8xd8 £xd8 24.9\e57 @g7 25.@xg6 hxg6
and Black is much better. 26.Ed1 transposes to the game.
16.2xb4?! cxb4 17.8xdS 8xdS 18.Wxd5 ¥xd5 23...2g7 24.9xg6 hxg6?
19.exd5 bxa3 20.bxa3 Ed8 and again Black is 24..0d5! 25.8xd5 ¥xd5 26.Dxe7 Wgs!
better. (26..¥xg2 27.h4 and White is much better)
16...50xd5 270651 g6 28.WxgSt fxg5 29.g4 £13 30.h3
16..8xd5! 17.2xd5 &xd5 18.0-0 &Dxc3 h5 with enough counterplay.
19.%xc3 &6 and Black is slightly better. 25.H2xd8 £xd8 26.2d1! &d5
17.8¢e5 Eg6? 26...g5 27.9b8 &d5 28.a4 a5 29.Wa7t Ef7
17..f6 18.2g3 (18.0-0-0? fxe5 19.2xd5t 30.Wxa5 £c¢7 31.Wxc5 and White wins.
WedS 20.Wxd5t £xd5 21.8xd5 e4 White wins. 27.¥d6 £b3?
76 "The Benko Gambirt

A blunder, but even after 27...2xg2 28.¥xc5

&7 29.a4 White still wins.
,,,,,,,

//// / / /@
\ //
,,,,,,,

m 7
\\\

7 / / / %// %/g/,
z&/.e///// Now there is a choice:
///53 8 b1) 15.%d2”! Be6! (not 15..2f6¢ 16.e5!
(16.xa5? @xal 17.2a3 2f6 18.8xc5 He6
19.%b4 fxed 20.2xf8 Wxf8 and Black is
slightly better) 16...2xf3 17.exf6 £xe2 18.Wxe2
28.Wxf8+ Wxf6 19.2b2 Wg5 20.8ad1l and White has full
1-0 compensation for the pawn) 16.e5 d6 17.exd6
£xd6 and Black is slightly better, Dokhoian
Game conclusions: 9...Eb8 is surely the main — Hertneck, Germany 1992.
line after 5...e6, but as we have seen 9...2¢7 b2) Best is 15.¥f5! g6 16.%f4 and the position
also appears to be perfectly playable, though the is unclear.
prospect of an endgame that should be drawn 12...d6 13.0-0 Exb6 14.b3
will hardly appeal to diehard fans of the Benko. 14.5g5 &xg5 15.8xg5 Wes 16.2f4 and now:
After 9...8b8 we have looked at White’s less 16..8e61 17.8xd6 (17.£d3% c4 Black is
dangerous options and, surprise, they are not slightly better, Milanovic — Vuckovic, Belgrade
that dangerous. In the next game we shall look at 2001) 17...&xc4 18.¥xc5 £xf1 19.¥xb6 Lxg2!
the main move 10.e4. (19...8e2? 20.8xf8 Hxf8 21.Ecl and White is
close to winning) 20.8xf8 ¥xe4 21.2d6 £h3
Game 30 22.f3 ¥ixf3 23. W2 Wd5 with full compensation
Gleizerov — Tregubov for the material.
Krasnoyarsk 2003 16..20d4 17.8d3 f6 18.¥xe8 Bxe8 - equal.
14..8e6 15.8x¢6 fxe6 16.¥gs ¥c8 17.8b2
1.d4 D6 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.b6 e6 ££6 18.2xf6 Exf6 19.2d2
6.5¢3 Dxd5 7.0 xd5 exd5 8.¥xd5 Nc6 9.3 This is a critical position. The chances are more
Bb8 10.e4 Le7 11.8c4 0-0 12.¥h5 or less equal, so which side to prefer is a martter of
White has two other options: taste. For example, | prefer Black, because of the
a) 12.5g571 We8 13.0-0 1d4! 14.2e3 (14.b4t2 easy plan: d6-d5 with pressure on the f-file.
d6 15.bxc5 dxc5 16.8e3 Bxb6 17.8xd4 cxd4 19..¥18 20.2c4 Bb7 21.f4
18.8)f3 Hb4 19.¥xd4 Wa4 20.8fcl fe6 21.Wa7 21.e520 Hgo! (21..5xe5? 22.8xe5 dxeS
Hxcd 22.Wxe7 Hxed and Black is slightly better) 23 %¥e2 and White has some initiative) 22.%h5
14...8b7 15.We5 &xg5 16.¥xe8 Hfxe8 17.8xg5 ds 23.82d6 Bb4 and Black has some attacking
Bxe4 and Black is slightly better according to chances.
Shirov. Or 21.8ad1 ©d4 with the obvious idea d6-
b) 12.0-0 8xb6 13.b3 Da5 14.8e2 8b7 ds.
 The Positional 5.b6 77

21...8bf7 22.g3 d5! 23.20d2 Hd4 24.Bael 36...59c2 37.¥xc2 Wxe3 38.%b3t 6 39.23
Wes 40.Wxc3 Wxc3 41.Bxc3 Bd2t 42.%63

3y
Exa2 43.g4 Bxh2 44.8c61 Hf7
L,

/// .E/
/ AR
Game conclusions: Black has good play with
5...e6, even after the most natural 10.e4. The
critical position after move 19 is approximarely

AanNY
equal, but in no way simple. I would recommend
7 those wanting to employ this line to take a closer
look at this position, and maybe even play a few
training games against a friend or, if you don’
//%///fi have any, a computer.
/// /
Game 31
Kamsky — Illescas Cordoba
24...c4?! Porto San Giorgio 1994
Black is playing too sharply. Black has some
initiative, but it is hard to get a real advantage: 1.d4 96 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.b6 ¥xb6
24..Md8 25.03 Bg6 26.Wh3 Hixf31 27.Exf3 6.9¢3 d6 7.e4 g6 8.a4"?
Was5 28.8e2 Whs 29.Wo2 and White’s defence This plan was quite popular in the first half of
is solid enough. the 90s. It did not lose popularity because it was
25.exd5 exdS 26.bxc4 dxcd 27.8e4?! found to be bad - fashion just changes.
After 27.He4! Black has some problems with 8..8g7 9.a5

AR
the c4-pawn. For example: 27..%d8! (better
than 27...8g6 28.¥d1 De6 29.Dxc4 and White
is better) 28.80xc4 Wd5 29.84d2 He6 30.Ee2 Eas
& E
etc.
27...8¢6 28.0g5 Ee7 29.¥d1
29.5xh71 We8! 30.2xe7 Wxe7 31.8¢5 c3
32.%h3 (32.8b12 c2! 33.5b8% W8 34.Hxf81
&xf8 and Black wins) 32..g6 33.¥g2! and
White has good drawing chances. For example: // //8// /
33..%kg7 (risky is 33..Ec5? 34.Wed Wxed
35.80xc4 Hcd 36.%f2 and the c-pawn is under
% n // //
control) 34.¥h3 @g8 » /&/ 8 /

29...8d7 30.8e4 ho!?

////////////
Black sacrifices an exchange for the initiative.
Instead 30..8d87 31.8f2 Hcd6 32.¥f1 was
unclear. 9...Wb4!
31.0¢6 Hxe6 32.Exe6 Wc5 33.He8t &f7 Black provokes the white rook into occupying
34.8e3 c3 35.5 the a4-square; this square is not good for the
Or 35.Wa4 £2d8! 36.%¢2 2 37.Bcl Wd5t rook.
38.50f2 S with the initiative. Instead 9..¥b7 10.8c4 0-0 11.2ge2 (11.
35...2d8 36.c2g2! f41 6! (11.0e8 12.063 £g4 13.0-0 D7
White forces a draw. 14.%e1 and White has the initiative) 12.5f3
78 The Benko Gambit

exd5 13.8xd5 Dxed 14.5b6 Ha7 15.0-0 with 13.%e2 &bS! 14.8xb5T axb5 15.¥xb5t
compensation for the material according to transposes to the game.
Shirov) 11..0e8 12.0-0 ©Hc7 13.%d3 &d7 13...8b5 14.8xb51 axb5 15.¥xb5+ Abd7
14.£2¢5 He8 15.f4 2b5 16.b3 &d7 17.¥h3 Also to be considered is 15...¥xb5?! 16.8xb5
and White is slightly better, Li — P Rodriguez, a6 17.f3 b8 18.9c3 b4 19.%e2 0-0. It is
Havana 1992. interesting that Illescas Cordoba wrote that this
10.2a4 Wb7 11.8c4 position is much better for Black (Chess Informant
11.5f3 Qg4 (11..0-0 12.50d2 (12.&e2 60). It is of course much better - but for White!
£d7 13.2a3 £b5 14.53d2 Hbd7 15.0-0 Lxe2 For example: 20.20h3!
16.¥xe2 Bfb8 17.40c4 De8 18.8¢5 £6 19.2d2 16.Ba4 ¥xb5 17.Dxb5 0-0 18.2e2
&7 the game is unclear, Rychagov — Fominyh, 18.2c3 Bfb8 19.9ge2 Bb4 20.2a3 Dxe4 and
Moscow 1994) 12...e6! 13.%c4 exd5 14.exdS Black is slightly better.
(14.0xd6 ¥e7 15.Dxc8 Hxc8 16.8xd5 &xd5 18...2b6 19.8a2 Hc4
17.%xd5 &c6) 14...8g4 15.3 £d7 16.8a3 £b5 19..8xe4 20.50¢7 Ha7 21.9b5 Haa8 with
with an unclear position) 12.8e2 2xf3 13.8xf3 equality.
Hbd7 14.0-0 0-0 15.8e2 BEfb8 16.f4 20.f3 Efb8 21.Dec3 Bxa5 22.Exa5 Dxa5
23.2d1 Db3 24.%0c2 Hxcl 25.Excl Ba8

\\\E/ //%i/
7// i, , , ;,,i”fx/
//%/M /

,,,,,

/// @Jg%
// /////,,,,,// //4/,/,,,/

)=
i & == /é&/&/
11..8d7!
11...0-0 12.f£42! (12.513 &d7 13.Ha3 &b5 Black has the a-file, and bishop against
14.6\d2 Dbd7 15.f4 De8 16.8b3 Hc7 17.0-0 knight. This is too small an advantage to beat a
Bab8 18.¢0h1 is unclear, R. Kempinski
— Mainka, grandmaster.
Lippstadt 1995) 12...2d7 13.2a3 Wb4! 14.¥d3 26.8el Dd7 27.f4 £f6
£b5! 15.2xb5 axb5 16.¢5 (16.¥xb5? Hxed! and Perhaps 27...h6" with the idea ...g5.
Black wins at least a pawn) 16...dxe5 17.fxe5 g4 28.g3 h6 29.9a3 £xc3
18.0f3 (18.Wxb5 &xe5 19.8ge2 Ha6 20.0-0 This look like a draw offer.
A7 21. b6 BbS! 22.Wxc7 Bb7 23.0a2 Bxc7 30.xc3 Db6 31.80c2 Bad 32.5b5 Hb4
24.9xb4 cxb4 25.Eb3 EBc4 and Black should 33.8c3 a4 34.5d1 @g7 35.b3 Db6 36.59c3
win) 18..83d7 19.e6 &deS 20.Dxe5 &xe5 c4
21.Wxb5 fxe6 22.Had Wxb5 23.9xb5 exds 1h 1)
24.5\c7 Bab8 25.a62? §Hd31 26.de2 Bf2T 0-1
Naumkin — Rotstein, Porto San Giorgio 2003. Game conclusions: In this game we have looked at
12.H2a3 ¥b4 13.%4d3 Shirov’s plan of a quick advance by the a-pawn,
 The Positional 5.b6 79

putting a bishop on ¢4, and a wild attack on 11..2¢5 12.52h1

the kingside. If Black recaptures with the queen 12.h3 g5! 13.24 W7 14.Ea3 Eb8 (14..2g6
early on, he needs to know the dangers of this 15.5c4 h6 16.a5 Bb8 17.¥c2 Hd7 18.Da4
system and how to react against it. Especially Hde5 19.0ab6 Dxcd 20.8xcd and White is
noteworthy are 9...%b4! and 11...2d7!. Black slightly better Stempin — Bardeli, Poznan 2003)
should be fine. 1584 Dxcd 16.8xc4 g4! 17.hxgd Dxgd is
unclear.
Game 32 12...g5
Tukmakov — Plachetka
Ostende 1990
5/2/
/
%@%
78
1.d4 D6 2.4 c5 3.d5 b5 4.cxb5 a6 5.b6 ¥xb6
6.80c3 g6 7.e4 d6 8.Df3 2g7 9.2d2 0-0
///////

10.£e2 Hbd7
///8//4 ///// ;7%

NN
@@/@ HAK
% % 5 awB2 &
This is a typical idea: Black protects the knight
8Agé,/é/x
,,,,,,
from f2-f4 and keeps control of the e5-square.

B sw
Va7 13.Dc4 Dxc4 14.8xc4 g4 15.8b1
15.f4 gxf3! 16.gxf3 &h8 17.8g1 Hg8 18.%/d3
Hd7! 19.f4 ¥b7 20.b3 ©b6 21.8b2 Dxc4
s 7>

22.Wxc4 e6 with counterplay, especially due to

11.0-0 the position of the king on hl.
11.8c41? W7 12.0-0 &b6 13.9De3 Now 15...2h8 16.2¢5
the most natural plan for Black is €7-¢6. 13... 16.2¢3 Wb4 17.2d3 ©Hd7 18.a3 Wby
e6! (also interesting is 13..8b7 14.24 Hac8 19.Wxgd Qe5 20.Wd1 5 21.f4 (21.b4?! c4
15.25 Hc8 16.Dc4 e6 17.8f3 exd5 18.exdS 22.8c2 f4 23.8d4 f3 24.Bgl &g4 2593 £h3
Hd7 19.8a3 De5 20.8e2 He7 21.0xe5 £xed5 with an attack) 21...5xd3 22 ¥xd3 fxe4 23.¥d1
with equality, Ipek — Miton, Rethymnon Eb8 and Black is slightly better, Fries Nielsen
2003) 14.a4! exd5 15.25 d4 16.axb6 b8 — Plachetka, Rimavska Sobota 1991.
17.813 &b7 18.5c4 dxc3 19.9xd6 Eingorn 16...2g8 17.¥d2 £d7 18.2e2
— E. Lang, Bad Woerishofen 2002, and now, 18.8xe7 %5 19.8¢2 transposes.
instead of the passive 19...2e8 which leads to 18..8e5 19.8xe7 Ea7
a better position for White, Black should play 19...£5! 20.f3 gxf3 21.gxf3 £d7 is unclear,
19...8xe4! 20.Hxb7 ¥Wxb7 21.Wb3! c4! (21... according to GM Tukmakov.
cxb2 22.8xb2 £xb2 23.Wxb2 5 24.Efb1! and 20.8g5 f5 21.8¢3
White is slightly better) 22.¥xc4 cxb2 23.2xb2 21.We3 He8 22.Hfel Hf7 23.b4 ©d7 with
&xb2 24.8xed Wxb6 25.8a2 Bac8 26.¥xa6 unclear play.
Wxa6 27 Exa6 with equality. 21...%b4 22.a3 ¥b3 23.%d1 ¥xd1 24.Bbxd1
80 The Benko Gambit

- 73 EE 6.3 Dbd7 7.23

/7 _ > @¢‘ 7.a4 a5 Otherwise Black would just end up
with a worse version of the 5...%xb6 line. 8.e4
g6
2
73%@; //
y/‘ //7/%%”?2”/// %7 7

?/,,,é ,,,,,,,,,,,, /

% /// // fi// g
24...B8! 25.exf5 £xf5 26.b4 cxb4! 27.axb4
27.8xa7 bxc3 28.2d4 c2 29.8c1 a5! The point
And now:
of the sacrifice. Black simply holds the position a) 9.0f3 &g7 10.8e2 0-0 11.0-0 Dxb6
and rtakes at least half a point. 12.8g5 h6 13.82¢€3 6 14.dxe6 £xe6 15.0d2 We7
27..8¢7 28.89a4 b8 29.2b6 Ecb7 30.2)c4 16.¥c1 Bfc8 17.4f3 Dfd7 18.8e2 HeS 19.00f3
Hxcd 31.8xc4 Bxb4 32.8xa6 Le5 33.8d3
Hxf31 20.8xf3 h7 21.8d1 Eab8 with equality
£xd3 34.8xd3 g7 Kobylkin —Tseshkovsky, Krasnodar 2002.
Black's pieces are more active: it is enough
b) 9.8b5 2g7 10.24f3 0-0 11.0-0 Dxb6
compensation for the pawn.
12.We2 2b7 13.8f4 §e8 14.8ad1 Hic7 15.8d3
35.g3 g6 36.8g2 Bb3 37.2dd1 EB8bS5 We8 16.h3 £a6 17.8fel &xd3 18.8xd3 is
38.8h1 Za5 39.h3 a2 40.Bhel £c3 41.Eh1 unclear, Shipov — Zubov, Interner 2003. Now
8e5 42.0f1 h5 43.hxgd hxgs 44.8h6T Bf5 interesting is 18..82xc3!1?. Whether to give up
45.8h5t Ded 46.8g5 Bxe3 47.fxe3 Lf3 the dark-squared bishop in positions like this
48.Exe5 dxe5 49.d6 Eh2 50.s2g1 Bh8 51.d7 is always a difficult choice. After 19.2xc3 (or
-1
19.bxc3 Hxad) 19...80xa4 20.8Ba3 Db6 21.2h6
White has full compensation for the pawn.
Game conclusions: In this game the older way to
7..g6 8.e4 2g7 9.8¢2 0-0 10.0-0 Dxb6
battle against a quick recapture with the queen,
11.8f4
the more positional £f3-d2-c4, was investigated. 11.h3 e6 12.dxe6 &xe6 13.82¢5 Wc7! with the
Black has not many problems gaining counterplay
idea d6-d5. (13...h6 14.8xf6 &x16 15.%d2 kh7
in this Benoni-like structure. White’s set-up lacks
16.8adl (16.Efd1?! &e7 17.e5 dxe5 18.We3
sting and White is not positionally better, so
(18.%xd8? Efxd8 19.5xe5 £f6 and Black is
Black should not fear this variation at all. slightly better, Brikov — Krasnov, Donskoj 2002)
18.. W7 19.8\xe5 Hfe8 is unclear) 16..8e7
Game 33
17.8fe1 White has some advantage here) 14.%/d2
Pfleger — Benko
d5 with equality.
Skopije (ol) 1972
11..8g4 12.9d2
12.a4 & e8! Black fights for control of the e5-
1.d4 D6 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.b6 d6
square. 13.%d2 &xf3 14.8xf3 D4 15.Wc2 Eb8
This move (like 5..g6) has independent
16.2abl Hic7 17.8d1 DeS 18.82e2 e6 19.80e3
significance only in connection with the plan exd5 20.8xdS Hxd5 21.exd5 Wb6 with equality,
..2)b8-d7-b6. Gamboa —~ Leko, Yopal 1997.
 The Positional 5.b6 81

12...8xe2 13.¥xe2 Wd7 14.24 15.¥xg4 &xg4 and Black is better because of
the weakness of the b2-pawn.
15...53h5 16.2h6 £xh6 17.¥xh6 W4 18.¥xf4
¥ //%@% Hxf4 19.b3 Hab8 20.g3 Dd3 21.f4 512
A /y/ Black must exchange at least one knight to

ia, , , &4i / mx%

gain the advantage. Here it is impossible.
g/ 22.a5 Dd7 23.Badl fxed 24.Dcxed Db4
25.813 h6 26.Dc3 Hc2
131>

Game conclusions: Like in most of the other games

we have looked at, Black seems to be absolutely
fine. The brutal victories of Shirov will soon
// /
become history, and then myth.
Chapter conclusions: 5.b6 leads to positions that
have just as much in common with the different
14...Wg4! Benoni-systems as with the Benko Gambit. For
This plan is typical for the Benko Gambit. The this reason many Black players feel uncomfortable
simple point is that most rook endings are better with the variation, but this is very subjective. In
for Black. practice a well-prepared player should not fear
15.%e31 playing Black in this variation.
Chapter 5

The Ambitious 5.e3

1.d4 D6 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.3 Game 34

Szczechowicz — Pinski
Trzebinia 1998

fii.&.gf 2 /E ////// 1.d4 D16 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.e3 axb5

Inventions such as 5...e6?! are dangerous only
for Black. 6.9¢3 exd5 7.6xd5 &b7 8.0xf6T
Wxf6 9.60f3 £d6 10.8d2 ¥xb2 11.8c1 ¥f6
12.8¢3 We7 13.a4 axb5 14.axb5 &c7 15.8c4

/ %
0-0 16.£d5 White is much better, Bagirov —
Maculsky, USSR 1975.
//@87
/ Also unconvincing is 5..82b7 6.8c3 Wa5
7.bxa6! (7.2d2 axb5 8.82xb5 Wh6 see 5...axb5)
7..%xd5 8.2d2 £c6 9.63 d6 10.8c4 Hxal
11.0-0 W7 12.¢4 ©b8 13.Zel and White is
,,,,,,
better, 1. Farago — Collutiis, Porto San Giorgio
2004.
This quier move is generally underestimated 6.2xb5 Was5t 7.80c3 £b7 8.2 ge2!
now. It used to be called the Modern Line, but
now it is just referred to as 5.e3, which could
just as well mean the wimp’s line! However, if a A /@% %
player is very well prepared in this line he can
still occasionally win straight out of the opening,
/g/x,w ,,, y
which makes it a dangerous system, especially 4/ % a
\\\\

as most Benko Gambit fans only know this line

,,,,,,,,,, / //
»
superficially.
In this chapter we shall mainly look at
5...axb5, but there has also been left a little % é
)
%////////% ///// //% /%%

B gwe 3
room for 5...g6...
878 %@fl&w
//////
84 The Benko Gambit

8...0xd5 9.0-0 Dxc3

9..82¢7 10.82c4 a6 11.82xa6 Hbxab 12.e4 g6 ,,,,,,,,,,,,
13.$g5 §g7 14.%d2 (Vaisser — Chernin, New
York 1994) 14...8b8
slightly better according
9..5f62! This is just
15.8h6 and White is
to Anatoly Karpov.
weak. If Black wants to
/g /x/
keep this knight on f6 then it is better to take
the d5-pawn with 8...2xd5. 10.f3! e6 11.e4 &£¢7
12.8f4 0-0 13.82d6 and White is better.
10.2xc3 g6 11.8c4 &c6! /%
A great waiting move.
11..%b4 12.%d3 8g7 13.2d2 Wb6 14.20d5
Wd8 15.9b3 £a6 16.2c3 0-0 17.8xg7 xg7
18.Wc3t hg8 (18..f62 19.a3 e6 20.8f4
We7 21.b4! cxb4 22.axb4d Ec8 23.b5 d5 1 had prepared this position before the
24 bxa6 Hxcd 25.%b3! and White is much tournament with my coach, master Wojciech
berter. 25...%0f7 26.e4! Dxa6 27.Dxe6! Wd7 Ehrenfeucht, who was one of the biggest experts
(27.. Wxe6 28.Wb7+ £c7 29.8xa8 and White on the Benko Gambit. He played it for more
wins) 28.exd5 ¥Wxd5 29.8a4 1-0 Lukacs — K than 20 years. In Poland there was a joke that in
Berg, Gausdal 1989) 19.2xa6 Hxab 20.a3 d6 the Benko Gambit he was second in the world
21.b4 and White is better according to GM (after Michael Adams). As we all know, every
Lukacs. joke contains an element of truth...
12.e4 Our conclusion about this line was: it is easier
According to an annotator in Chess Informant to defend this position than it is to attack it.
this move is weak, and White should play: 15.8d5
12.9b3 6 13.2d2 ¥b4 14.8b52! 15.8d3 c4 16.8c2 Ha6 17.8cl WeS 18.8e3
However [ had prepared a novelty against this 0—0 and Black is probably better.
recommendation! 15...6 16.£xc6
Better is 14.e4 £g7 with an unclear position. 16.8el £xdS 17.exd5 0-0 is also not
14...%xd2! impressive.
14..¥xb3? 15.axb3 Hxal 16.Bxal Le7 16...20xc6 17.%d6 Wd4 18.¥c7 0-0
(16..d5> 17.8e2 e5 18.8a7 £e7 19.2c7T 18.. ¥ixed?? 19.2fd1 0-0 20.%xd7 &d4 and
®f8 20.e4 White is close to winning) 17.2a7 Black is better.
White was much better in Ruban — Kalegin, 19.Bad1 Wes?
USSR 1987. This is not home analysis; this is cowardice.
15.9c7t &d8 16.5xa8 To be quite serious: in this game I calculated
16.9b6? Wa5 17.5xa81 Wxb6 18.4)xb6 &c7 variations poorly. Instead 19..Wxed4 20.Hxd7
and Black is better. Bfc8 21.Wb7 2ab8 22.Wa6 ©e5 and Black is
16..%b4 17.5b6 &c7 18.%d3 ¥xbo! much better.
18...8g7 19.8xd7 ©xd7 20.2fd1 is unclear. 20.¥xe5 Dxe5 21.8f4 D4 22.8xd7 Exa2
19.¥c3 2g8 20.Wf6 Hg7 21.e4!? d6! 22.8xe6 23.8c1 Ha4 24.EBc7 Db2 25.8h6 Ed8
£xed 23.8c4 Hd7 26.B7xc5 Exe4
and Black is much better. White is very close to a draw.
12...8g7 13.8g5 &xc3! 27.g32!
This is not courage; this is home analysis. White should have played 27.h4 with the idea
14.bxc3 ¥xc3 h4-h5.
 The Ambitious 5.e3 85

27...Bb4 11...8xb5
Better was 27...50d3 28.5c8 Bed4 29.81c7 €5 Risky is 11..8xe4 12.8xe4 Dxed 13.8)¢3
with the idea ...e4 and ... De5-Af3. Hxc3 14.bxc3 &e7 15.8b2 0-0 16.c4 with full
28.5c7 Ebb829.50g2 £ d3 30.E1c3 De531.h3 compensation. Note that from al the rook has a
Hd7 32.Ha3 Ebc8 33.Eca7 Db6 34.2b3 BEd7 highway to the kingside (al-a3-g3).
35.2a6 ©d5 36.8ba3 De7 37.g4 He8 38.8a8 12.9xb5 Dxes 13.0g3"12
Hdd8 39.8B8a7 Dc6 40.8c7 Ec8 41.8d7 Bed8 Also interesting is 13.8f4 £a6 14.8ec3
11, Hixc3 15.bxc3 d5 16.Bel £e7 17.¥h5 with
compensation.
Game conclusions: The variation with 8...8xd5 is 13..2f6
not the main line, but it is perfectly playable, and 13..80xg3 14.fxg3! with an attack. For
if T had shown just an ounce of real guts [ would example: 14...8¢6 15.2f4! (15.82xf72! This looks
probably have won this game. Still, the line does good, but is not effective. 15...&xf7 16.8xd7+
not lead to an advantage by force, of course. Be6! (16...82e722 17.0d61 66 18.8g51 dxg5
19.¥xe6 with mate in six moves) 17.8£d2 ¥xb5!
Game 35 18.axb5! (18.2c31 €5 19.8xe5T Dxe5 20.Wxb5
Shirov — Adams £d6 21.8f11 he6 22.8el and White has only
Chalkidiki 1993 a draw by perpetual check) 18..8xalt 19.52
&Hd4 20948+ £f7 21.b6 with good winning
1.d4 D6 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.¢3 axb5 chances) 15...8c8 (15...e5 16.8xe5! and White
6.2xb5 Wa5+ 7.80c3 b7 8.Dge2 £xd5 9.0-0 is close to winning) 16.£d6 and Black is in
86 10.a4 6 trouble.
10...g6 11.e4 £xb5 12.8xb5 D6 13.8d2 Wd8 14.8el
14.8c3 &g7 15.¢5 Hed 16.e6 Dxc3 17.ext7T
&xf7 18.Dexc3 and White is slightly better. I
/ 2]

zm /@ z /X/ i
%/

/ »
/’
=7
/
g/
/// ,,,,,,,,,,,
s
// ,,,,,,,,,,
//
5 >
/////

= %/
,,,,,,
14...8a6?!
14...&e7" 15.b4 cxb4d 16.9f5! ext5 17.8xe7T
11.e4 dxe7 18.8d6t ©d8 19.8g5 B8l (19..4¢6
11.¥d3 d5! (in a game against GM Lukacs 20.8xf61 ©c8 21.8xg7 and White's initiative
in Budapest 1996 I played the weak 11..8¢7 looks dangerous) 20.Hclt N6 21.8xc6T dxchd

and after 12.e4 White had a small edge) 12.82d2 155 Weber — Halwick, corr. 1997.
Wb 13.b4 cxbd 14.9a2 &xb5 15.axb5 Ded and 15.2f4 d5 16.2f5 ¥b4
Black is better. 16..%b6 17.¥b3 bd7 18.8adl &2b4 19.40¢3
86 The Benko Gambit

with compensation according to Shirov.
17.8e5 Wg4 18.8xf6 ¥Wxdl 19.Eaxdl gxf6
20.8xd5 £e7 21.Hd2 248"
21..Bd8! 22.8xd8t Axd8 23.Hbd6t &f8
24.5h6 £a5! and Black holds the position -
Shirov.
22.0fd61 Re7 23.9c4 Dbz
23 .84c7 24.8xc7 (24.5c¢3 Db4! with some
counterplay) 24...8xc7 25.a5 Ehb8 26.50f1 Eb4
27.8cl 8a6 and for Black the worst is behind
him.
24.b3
Perhaps better was 24.8ed1! f5 25.52d7
26.b3 and White has the advantage.
24...8a5 25.8ed1 Bhd8 26.52f1 Exd2 27.8xd2
£dse!
Black is playing without a plan. Instead, after
27..h5!a draw is the likely result.
28.cke2:!
33.8xd5?! does not work (better is 33.%e3 with
pressure) 33...exd5+ 34.¢xd5 2h2! 35.8%xc5 &gl
36.c6 8xf2 37.8cd6 Bb6T with counterplay.
33.0a7 Le8 34.5¢6
1-0
Game conclusions: With correct play Black
equalises after 8...£xd5, but the position still
holds some venom and should not be dismissed
as harmless. It is important for Black to be well
prepared if he wants to play like this.
hf6 Game 36
Graf - Felgaer
Calvia (ol) 2004
1.d4 D6 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.e3 £b7
6.5 c3 ¥a5 7.8d2
This move is not at all dangerous for Black.
7...axb5 8.82xb5 ¥b6

=
- 4

/ ‘ V / / * / i / t
7/ T E 77
,,,,,, 7/ ,,,,,,
// /
/c@:
//
,,,,,,
» @i%/i

2,67 e // /
; Y, 4
VA

AlG, . AT 2, @M/
2 e.
@&%//
zz@/@ e
,,,,,,,,,,,,

7,
7//% 9.%£312
9.8c4 6 1004 Dxed! 11.5f3 (11.dxe6?!
fxe6 12.9)xe4 Sxed 13.8h5t 2g6 14.Wf3 Hc6
White secures himself a better pawn structure. and Black is slightly better. Or 11.8ge2 &xd2
Now h7 is a weakness. 12.¥xd2 &e7 with equality.) 11..2xd2 The
29...80d5 30.52f3 £¢7 31.h3 £d8? simplest. 12.%xd2 £e7 13.0-0 0-0 14.Efel Da6
31...h5!32.gxh5 &4 33.Hc2 Eh8 34.8c3 Db4 15.2ad1 £f6 with equality, Meinson — Kogan,
35.He2 £b8 36.8d2 Exh5 with good drawing France 1997.
chances. 9.%b3 ¢6 10.¢4 Dixed 11.80xed £xd5 12.¥d3
32.ed! Hc722 Wb7e (12...65 13.5)g3 &xg2 14.01e2 &xhl
After 32..8c7 Black can fight on because 15.8xh1 &e7 16.82hg3 0-0 17.a4 Dc6 with
 The Ambitious 5.e3 87

unclear play, D. Gurevich — Nicholson, Lugano 23.8xe5! Bxg3 24.8Bxd5 Exg2t 25.@xg2
1993) 13.03 c4 14.8xcd &xcd 15.¥xc4 d5 W56 26.8e3 Db 27.8xc6 dxc6 28.8xc5 Bad
16.¥d4 dxed 17.8c3 Dd7 18.Wxed Wixedt 29.2b8%1 e7 30.20d5t
19.fxed Hc5 20,0063 Dxed 21.8d4 £6 22.8e2 €5 1-0
23.8e3 £d6 is equal according to grandmaster
Patrick Wolff, as seen in Metge — Sarfadi, New Game conclusions: 7.2d2 is the older main line.
Zealand 1987. Black can regain his material tactically, or at least
9...e6 gain a great initiative. As we saw, it is still a line
Perhaps better is another WolfF idea: 9...0a6"? that can cause Black serious problems if he is not
10.5ge2 #\b4 11.0-0 Nbxd5 12.e4 D7 13.a4 ready for it, but theoretically it is not scary at
e6 14.82¢5 fe7 15.Hadl £xbS 16.axb5 Hd8 all.
17.8fel h6 18.£f4 d6 19.8g3 h5 20.8f1 d5
and Black has some advantage, Graf — Halkias, Game 37
Fuerth 2002. Petursson — Fedorowicz
10.¢4 exd5 1l.exd5 Dxd5 12.8xd5 We6t Reykjavik 1990
13.2e2
After 13.We2?! &xd5 14.913 &c6 Black is 1.d4 26 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.e3 g6!?
slightly better. The most natural plan. White keeps the
13..8xd5 14.%g3 £d6 15.%xg7 Le5 16.¥g5 possibility of castling, but it costs one tempo
h6 17.¥e3 &xb2 18.2b1 £d4 (e2-e4 will now take two moves). Black has good
This novelty is not really successful. Instead compensation here.
18...%xe3 19.8xe3 Hxa2 20.0-0 Le4 21.Ebd1 6.9c3
Ha5 and White has full compensation for the 6.bxa6 £xa6 7.8xa6 Wast! A typical check.
pawn according to Wolff. For example: 22.8c4 8.8¢3 Wxa6 with the usual compensation.
Ha4 23.8d5, etc. 6..8g7
19.%¢3 Bf8 20.0-0 Le522
Missed White’s 23rd move. 20...Eg8 21.8xd4!
5%3%@% E
W) 22.0e2 Hxe3 23.fxg3 and White has full
Takitd
i 2 e
compensation for the queen.
21.0f4! ¥d6 22.8fe1 Bg8
Black’s position is hopeless: 22...8¢c6 23.8xc6
Y
%/w/@///
£xc6 24.8xe5 Wxd2 25.8 g6t fxgb 26.¥xg6

= 0

6 o maR
4
7 /,,,,,,,,,,,
/ /
/QM B LAWEHLINE
//////

7.863
7.bxa6 0-0 8.0f3 d6 9.e4 a5 10.2)d2 £xab
11.8xa6 ¥xa6 12.We2 Dfd7 13.a4 Wxe2t
TS 14.¢hxe2 $a6 15.8bl 5 and Black has full
Y, M ’
compensation for the pawn, Portisch — Benko,
== v
Palma de Mallorca 1971.
88 The Benko Gambit

7.e4 0-0! Here we have a similar situation to Black has full compensation for the pawn.
the 5.f3!? variation. Black has compensation for 9.8b1 6 10.8c4 axb5 11.axb5 exd5 12.8xd5
the pawn not only because of his better structure, &Hxd5 13.8xd5 Was5t 14.2d2 ¥xbs 15.8xb7
but also because of his lead in development. Wxb7 16.8c3 £xc3t 17.bxc3 Wa6 with equality,
Therefore, Black has an interesting plan to open Hillarp Persson — Nedobora, Mondariz Balneario
the position as soon as possible. 8.a4 (8.e5 De8 1994.
9.9f3 d6 10.£f4 Lg4! with good compensation
according to the Polish IM Aleksander Sznapik.
For example: 11.exd6 £xd6! 12.h3 £xf3 13.Wxf3
Z
EA W EdEe 7 AN/
axb5) 8...8b7 9.¢5 De8 10.4f3 d6 11.£f4 dxe5
12.8xe5 D6 13.&c4 axb5 14.axb5 (1. Farago
- A.
4 Akd:e
Deze, Novi Sad 1979) 14...8xal 15.%xal £ibd7
and, according to Deze, Black is already slightly
better. The point is that sooner or later Black will
take the d5-pawn and keep the inidative.
H & K6
%,,/,% ¢/////é 7 /// "//

7.24 0-0 8.8c4 d6 9.8 ge2 H\bd7 10.2a3

& RAR
Zrres s e T Y oY

LY 7 A Ay Ay

7
4%%M///
3
74
// %

//8
9...axb5
This is simplest, however Black can also play
9...e6!? with a similar position.
10.8xb5
/z 10.axbS5Exa3 11.bxa3 Was5 12.Wd2Wa813.2c4
e6 14.dxe6 fxe6 15.8e2 d5 with compensation.
10...e6 11.dxe6 fxe6 12.¥d6
10...0b6 (here Black also has an interesting
This prevents d7-d5, which would give Black
way to fight for more than a draw: 10...8e5”? standard compensation. Or 12.0-0 §d5.
11.£a2 (11.b3 ¥a5 12.0-0 Hfd7 13.5a2 &b7 12...8xf3
14.2d2 Wc7 15.e4 ©b6 with compensation) Or12..%e4 13.8)xe4 &xe4 14.0-0 £f5 15.8e2
11..axb5 12.8xb5 £d7 13.8bc3 Wa5 with &c6 with compensation, Petursson — Hodgson,
compensation) 11.82a2 axb5 12.axb5 &d7 Moscow 1987.
13.0-0 De8 14.e4 Dc7 15.Wd3 We8 16.8xa8
13.gxf3 Wc8 14.0-0 De8 15.¥g3
Wxa8 17.8c4 Hxcd 18.Wxcd Was with equality, 15.%d 12 Hc7 (15..d52 16.9xd5! is wrong)
Lukacs — Hertneck, Kecskemet 1988. 16.8.c4 D6 17.e4 £ d4 with the idea ¥b7, and
7...0-0 8.a4 Black’s initiative looks pretty dangerous.
8.8c4 axbs 9.2xb5 Was 10.0d2 Hxds 15...8)c6?!
11.5xd5 Wxbs 12.8xe7+ Hh8 13.0xc8 15...d5 looks more natural. 16.2d2 &c7
Hxc8 14.%e2 ¢4 15.0-0 d5 and Black has 17.8¢2 Wb7 18Ebl &Hc6 with standard
full compensation for the pawn, Razuvaev — compensation for the material.
Vaganian, Moscow 1982.
16.2)e4?!
For 8.8¢2 axb5 9.8xb5 sce 8.8.c4. 16.8d1! d5 17.e4! d4 18.8e2 and White is
8..82b7 9.8a3 slightly better, because Black’s pride and joy, his
9.8¢2 axb5 10.8xb5 ¢6 11.dxe6 fxe6 12.%d6 strong centre, is now weak.
W8 13.0-0 De8 14.Wg3 £xf3 15.gxf3 dS and
 The Ambitious 5.e3
16..20e5 17.f4 D7 18.2g5 d5 19.e42
Now this is too aggressive, because Black can
maintain the tension and keep the pawn on d5.
Instead 19.8d1 ©ed6 20.2d3 was unclear.
19...9Ded6 20.exd5 exd5 21.¥d32!
Now White has some problems with the b5-
bishop. When you do not know what you should
play, just improve one of your badly placed
pieces: 21.8e2 &f5 22.%h3 h6 23.9xf7 Bxf7
and Black is only slightly better.
21..0xg5 22.fxg5 2d4 23.2g2
White wants to play f2-f4, but there is no time
for that.
23.. Wg4t 24 Wo3 Wedt 25.03
25.8g1 &xb5 26.axb5 Zab8 and Black
much better.
25.. %2t 26.5h1 Dxb5 27.axb5
28.bxa3 ¥d3 29.%g2 Wxb5
This ending is essentially lost for White.
30.£4¥d7 31.82b2 £xb2 32.¥xb2 d4 33.¥b3+
4 74
7
/// %//”’//”//
V.
////fi/g
89
34 Wixf7t
Without queens the position is just lost. The
problem is that it would also be lost with queens:
34.Wb5 Wds5t 35.8g1 d3 etc.
34..Bxf7 35.8g2 c4 36.2f3 Be7 37.2d1 d3
38.a4 Ba7 39.%e4 Bxad 40.2d4 2f7 41.8el
d2!
Simplest!
42.8d1 he6 43.8xd2 f5 44.E2
44.5Bc3 hxf4 45.5g2 HaS 46.xc4 Exg5 and
Black wins.
44...c31 45.8xc3 Bxf4 46.8g2
Or 46.8e2 txgS 47.%d3 &gd and Black
wins.
is 46..Hg4 47 .82t dxg5 48.82d2 Bed 49.%2d3
Be7 50.8£3 bh4 51.80d2 g5 52.Ha3 h5 53.8b3
Exa3 He5 54.8a3 g4 55.8b3 Ra5
0-1
Game conclusions: 5...g6 is surely a natural move
for a Benko player. Black seems to receive good
compensation for this pawn investment, but no
guaranteed path to equal play can be found, so
you play this at your own risk!
Chapter conclusions: 5.€3 is an objectively
harmless, but still treacherous, line to face. Be
prepared, my friend, and you will be fine!
Chapter 6

5.13

1.d4 &6 2.c4 c5 3.d5 b5 4.cxb5 a6 5.3 Game 38

Azmaiparashvili — Conquest
Istanbul 2003

1.d4 &6 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.f3 axb5

After this White obtains chances for an opening
advantage. Also insufficient for equality is:
\

/ 5...d6?!
This move is inflexible and limits Black’s
possibilities. Now the ¢7-¢6 break has lost
some of its strength.
6.¢4 g6 7.3
This is better then 7.a4, which is slow and
creates some weaknesses in the white pawn
structure. Remember that pawns cannot move
backwards.
5.f3 was widely popular in the first half of 7..807 8.8€2 axb5
the 1990s, but has since then lost some of its After 8...0-0 9.82¢3 6 (9...2bd7 10.8e2 Hib6
following. Instead of grabbing the pawn, White 11.0-0 e6 12.dxe6 &xe6 13.8¢5 h6 14.2h4
focuses on establishing a strong centre. It is not axb5 15.0axb5 &4 16.Wcl and White is
a surprise that Black’s strategy in this variation slightly better according to Karpov. White
is based on facing White on the open bartlefield is threatening Bd1 and b3.) 10.8c4 axb5
and fighting back for central control. 11.6axb5 &xd5 12.5xd5 exd5 13.8xd5 Ea5?!
Butbefore we look at thestrongest move, 5. .¢6! 14.5%xd6 ¥Wh4t 15.¢3 Wh3 (Solozhenkin —
(game 40-42), we shall look at the alternatives Leykekhman, New York 1994) 16.F4! 8266 17.¢5
5...axb5 and 5...g6. a6 18.%12 there is no real compensation for
the material. Better is 13...2a6! and Black is
just slightly worse.
9.5xb5 0-0 10.5ec3 Hab 11.2e2 N7 12.0-0
Hxb5 13.5xb5 De8 14.24 &7 15.8b1 &d7
16.85xc7 Wixc7 17.b3 Bfb8 18.2d2 £c8 19.Wc2
92 The Benko Gambit

a7 20.2c4 and White is much better, Kolev b) 14.8a5 ¥b8 15.2d3 £h6 16.&c3 0-0
— Marinkovic, Vrnajacka Banka 1990. 17.8e2 £b7 18.60F2 Wa8 19.Wxa8 Bxa8 20.40a5
6.e4 Wa5t 7.8d2 and White is slightly better, Ernst — Ikonnikov,
White is also better after 7.b412. Amsterdam 2001.

LT
10.8d3 g6
After 10...e6 11.dxe6 &xe6 12.8e2 d5 13.exd5
HxdS 14.0-0 £e7 15.Wc2! White is slighty
better, Kasimdzhanov — Tregubov, New Delhi
(4) 2000.
11.9e2 £g7 12.a3 bxa3 13.Exa3 Exa3 14.bxa3
» 5 7 0-0 15.8a5 ¥d7 16.0-0

a:
%%z% Ty
,,,,,,

7... Wxb4t 8.4d2 Wad 9.%c1 Wa7 (9..b4 /// Z

10.¥xc5 Da6 11.¥d4 d6 12.a3 and, according
to Karpov, White is much better) 10.£xb5
and White is slightly better. For example:
10...e6 11.dxe6 (perhaps 11.8c3 as in Ziegler
— D.Cramling, Sweden 1996, is even stronger)
11...fxe6 12.%c4 Wb6 13.8c3 Hc6 14.8b1 &e7
/// ///&/
.%@/z%/
15.8h3 0-0 16.9f4 W7 17.0-0 and 10...9a6
11.9e2 g6 12.83bc3 Dc7 13.8c4 d6 14.0-0.
7...b4 8.9a3 d6 9.2c4 ¥d8
9.. %272 is an old invention by Wojciech
Ehrenfeucht, who started playing it back in the White is slightly better. If you compare the
1980s. Black’s main idea was to fight against strength of the slumbering black pieces to the
White’s standard plan of a2-a3 followed by axb4. freshness and fury of the white ones, you should
So how should White fight against such a simple not be in doubrt.
idea? [t is easy: find a way to defend the al-rook: 16...e6!
10.a3 g6 11.8e2! £g7 12.8cl! 0-0 13.23b3 The only sensible plan.
bxa3 14.Bxa3 ¥Wb7 15.8¢2 Bxa3 16.bxa3 Hfd7 17.dxe6?!
17.0-0 £ b6 18.4bas and White is much better, White does not benefit from making Black’s
Jaworski — Ehrenfeucht, Soczewka 1992. After centre more fluid. After 17.2f4!1? exdS 18.exd5
this game Wojciech decided to abandon 9... £b7 19.2b6 ¥d8 20.%b3 White would have
Wa7. been slightly better.
More relevant is 9...¥c71?. Now after 10.a3 17...fxe6
bxa3 11.Bxa3 Hxa3 12.bxa3 gb 13.%a4+ Hbd7 There is no compensation for the pawn after
Chess Informant will try to convince the innocent 17..¥xe6?! 18.8c7 £a6 19.8xd6 (19.23f4?
reader that this position is equal. T disagree: Wd7 20.8xd6 Bd8 21.e5 ©e8 and Black has
White can get the advantage in several ways: counterplay) 19...c4 20.8c2 We7 21.8b6 De8
a) 14.%c6 Wb 15.8d3! You should nor start 22.8xe8 Hxe8 23.24.
a battle when your soldiers are still snoring in 18.e5 dxe5
the barracks. 15...£g7 16.£e2 0-0 17.0-0 and With this the black pawn structure is wrecked,
White keeps a small advantage. but his only opportunity to fight is through
 93

active play. The weakness of the e3-square gives Black is not up to the task of this middlegame.
Black an additional interesting opportunity: Better was 27...8a6! 28.8xc5 &xcd 29.5xd3
18...50d5" 19.exd6 £a6 20.¥cl fxcd 21.8xc4 &xd3 30.8cl He7 31.8c7 &d5 32.Ha7 and the
Wxd6 22.8e11? &c6 23.893 We7 24.2d1 Hd4 draw is not far away.
with counterplay. 28.2b1! 282
19.5xe5 Wa7 20.Dc4 Dd5 21.8e1 Decb Now Black is collapsing. Necessary was
22.8£2 Bd8 23.¥b3?! 28...8a6 29.8b8% 17 30.8cd6t Re7 31.2e87!
23.%c1 should be answered with 23...2a6! (better than 31.8h472d7 32.2b6 Dxf3+33.gxf3
with unclear play, while after 23..8¢3 24.8xc3! Bd1t 34.8¢1 Bxelt 35.8g2 Be2t 36.%h3 &e5
(24.We3 Hxe2t 25.&xe2 with unclear play) 37.Hxa6 Bxh2t 38.%2g4 Bg2t 39.%h4 2h2t
24...8xd3 25.5b5 Wb8 26.a4 White would be 40.@g5 He2t 41.55h6 Bh2t with equality)
slightly better. 31..0d7 32.5xc5t ©xd6 33.5xd3 2xd3
23..80¢3 24.%c2 34.Hd8+ the5 35.8d7 fe5 36.8xh7 and White
is slightly better according to Azmaiparashvili.
29.h4 Bb3
But not 29..8a6> 30.9e5 Hxa3 31.Eb8
£1b3 32.2d8 dg7 33.8d71 ©h8 34.9f71 dg8
35.0661 &7 36.9e8t dg8 37.84xd4 cxd4
38.21d8 and White wins.

/ff 30.2d1 £a6 31.8xd4 cxd4 32.8xd4 Re7

7/
33.8e5
This position is very unpleasant for Black.

/ / %/
///8/ Even if it is objectively drawn, Black will have to
pay in blood and sweat to get to that outcome.
,,,,,,,,,,, 33...8xh4!
Worse is 33...Hxa3?! 34.8d7 &8 35.5c7 Halt
36.0h2 Ha2 37.5c6! (37.8a7 Lcd 38.8xe7
thxe7 39.8xc4 and White’s winning chances are
24.. Wd7 not too grear) 37..&a3 38.8xh7 £f1 39.8)e5
From here on Black begins to play the game Exg2t 40.2h1 Re7 41.28h87T g7 42.2a8 with
really badly. The exchanges performed after the the idea a7 and White wins.
text only seem to favour White. Stronger was 34.5)d2 Be3 35.8xh4 Hxe5 36.Be4
24..9xe2t! to avoid any monkey business.
25.Wxe2 (25.8xe2 Dd4 26.Wed &b7 with
initiative) 25..%d7 26.8b1 (26.8e4?! Ad4
27%e3 $a6 28.8Hcl? £h6 29.8xh6 De2t //,
30.%f1 Hxcl 31.Wxcl Wad and Black wins)
26...£a6 27.8a2 Dd4! Azmaiparashvili assesses .é./ /
this position as unclear, but I think Black is
already bertter! e.g.For example: 28.Wd3 Wds
29.8cl (29.8e12t Qxf31 30.Wxf3 £xc4 31.¥xd5 /////
/ /
exd5 32.8xc4 dxcd 33.2xc5 Hc8 and Black is
better in this simple ending) 29...20b5 30.¥xd5
exd5 31.89b6 c4 32.5xdS Hxa3 with advantage /%g
,,,
to Black.
25.8xc3 ¥xd3 26.¥xd3 Exd3 27.De4 Dd42!
94 The Benko Gambit

36...Bxe4?
36...8a5! 37.8e3 ®f7 and Black has good ,,,,,,,,,,,
drawing chances.
37.Dxe4 ///Zym 4
This ending is lost as Black has given up his , A »4 74
only active piece and now will be dominated
easily, as usual in such endgames.
37...e5 7 %M//
Black could also try: 37..8c4 38.a4 &f8
39.0f6 hS 40.9d7T &e7 41.80e5 &b3 42.a5 //&/
/
(better than 42.9xg6t hfe 43.0f4 fxa4d
44.5xh5t which gives Black some drawing
>
chances) 42..g5 43.%2f2 2d5 44.a6 &d6 //////

45.9671 $c6 46.9xg5 b6 47.@g3 thxa6

48.%9h4 and White wins. Black is not unnecessarily scared about a white
38.f6t &f7 pawn advance.
Or 38...8¢7 39.80d7 £c8 40.7)xe5 and White 7.8a3
should win. This seems to be the best move. But to reach
39.5xh7 £e640.0) 81 £f5 41.0d7g5 42.2f2 this decision it is necessary to have a look at the
8.8 43.g41 Be6 44.Dc5T Rd5 45.De4 bd4 alternatives:
46.%xg5 After the pushy 7.5 &g8 8.f4 d6 White
1-0 will not be able to keep his strong centre intact.
9.3 9 d7 and we should consider:
Game conclusions: With 5...axb5 Black regains his a) 10.9¢3 Hh6 11.8€2 0-0 12.0-0 &5 13.g4
pawn, but White gains time to set up exactly the Nd4 14.9xd4 cxd4 15.%xd4 dxe5 16.H12 axb5
pawn structure he desires. Especially noticeable 17.9xb5 exfd 18.8xf4 &6 19.c7 Bad 20.b3
is the knight manoeuvre a3-c4, which places Bed 21.Hadl g5 22.8g3 Dxgd Black has the
the knight in control of the e5>-square. Also initiative, Volkov — Baklan, Groningen 1998.
noticeable is that Black does not have play down b) 10.£d2 dxe5 11.2c3 &Db6 12.8xe5
the b-file, and that White can create a passed a- Af6 13.bxab 0-0 14.8c¢6 Wd7 15.8xf6 &x16
pawn. Positional aspects are truly more important 16.8e5 Wf5 and Black is slightly better, Volkov
than the pawn, and White can rely on a slight — Ponomariov, Krasnodar 1997.
advantage in several ways. 7.24?! 0-0 8.0a3 e6 9.dxe6 fxe6 10.0h3

Game 39 //
/
S. Ivanov — Sznapik
Augustow 1996

1.d4 96 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.£3 g6 / 7

This seems relatively logical. Black decides to 7 7
finish his development without wondering about
White’s centre. Though this is not definitely
good enough for equality, it does have a certain
sting,
6.¢4 8g7!
Black has full compensation for the pawn after
 5.63 95

10...d5, and then those who adore chaos will 12...dxe4d 13.9xe8 Wxe8 14.14 e31?
probably consider 10..2xe4!? 11.fxe4 Wh4t 14...9¢6 15.8c3 d5 looks good for Black.
12.$0d2 Wxed4. However Black’s compensation 15.0c3 £b7 16.We2 §ied 17.Dxed
will at most be enough for a draw: 13.¥c2 17 ¥xe3? §a6! and Black is much berter.
Wd4t 14.9d3 Wxad 15.5c2 Wg4 16.We2 17.. ¥xed 18.Wxe3 £xb2 19.%xed &xed 20.8d1
Wxe2t 17.8xe2 &b7 18.8gl1 axb5 19.Exa8 £xa8 Bxa2 21.8d4 @xd4 22.8xd4 &xg2 23.8xg2
20.£xb5 £ c6 with some compensation, Lesiege Hxg2
— Krivoshey, Cappelle la Grande 1999. Here Black is slightly better.
7...0-0 11...2h5 12.¥xd5 2\c6 13.g4
Also interesting is 7...axb5!? 8.8xb5! (8.2xb5? After 13.f4 Black will also reply 13...2a6!.
fails to 8...Hxa3 9.bxa3 Hxe4 10.fxe4 £xal and Now 14.9d6 (14.5ec3 ©b4 15.%d1 Wh4t
Black is much better) 8...2a6 9.24 0-0 10.2¢3 16.g3 Sixg3 17.812 Wxf4 18.8xg3 Wf5 19.Ecl
d6 11.5h3 Obd7 12.82 e8! 13.¥d2 &7 &xeS! and Black is already better) 14...g5 15.%e4
and Black has standard Benko compensation. 1 Wb8 leads to unclear play. Also 15.g3 @b4
do not want to promise equality with this line, 16.%d2 Hd3t 17.8dl1 is interesting, though
but it is a reasonable, playable alternative. most players intuitively would prefer Black.

REY 3 T
8.8 e2 6 9.2¢3

S
9.0¢3 &b7 (worse would be 9..exd5:
10.0xd5 £b7 11.8c4 axb5 12.8xfot Wxf6

/// &
13.9xb5 Wh4t 14.¢3 Wh3 15.8f1 Wh5 16.2d6
86 17.8c4 and White is much better) 10.8c4
axbs
What
11.8axb5 Hxd5! 12.8xd5 (12.exdS?
13.g3 Wxc4 and Black is much better)
12...exd5 13.exd5 He81 14.212 c4 15.g3 Whot
with full compensation for the pawn in Iljushin
fodwr
7
— Ponomariov, Yerevan 1997.
9...axb5
////é%
Also worth considering is 9...%a5t!? 10.¥d2
Wxd2t 11.6bxd2 Dxedt 12.fxed Bxb2 13.8xc5
&xal 14.8xf8 Hxf8 and the position is unclear.
10.2xb5
10.2xc5 d6 11.82d4 exd5 12.exd5 b4 13.8c2 With the knight trapped on the edge Black
Wa5 would give Black excellent counterplay for needs to act with great resolve, something my
his pawn. countryman did not hesitate to display with his
10...exd5! 11.e5 next energetic move.
The sharp 11.8xc5 should also be considered. 13...8a6!!
Black needs to play with great vitality to get the Black simple needs to get moving. What more
most out of his position: can I say?
11...Be8! 14.gxh5?
More convincing than 11..dxe4 12.2xf8 14.5ec3 He8!
£xf8 13.9ec3 &b7 where Black has some This is more testing than 14..8b4 15.%d2
compensation for the exchange, but probably He8 16.0-0-0!. White needs to be careful.
not enough. (16.gxh5 Wh4t and now after both 17.¥f2
12.d6 D2t 18.5d2 Wxf21 19.8x2 Dxal 20.8c4
Or 12.£d6 Was5t 13.9 ec3 dxed 14.9¢7 exf37 $xe5 21.8xal d5! and 17.8£2 Exe5T 18.50d1
15.8xe8 Dxe8 and Black is berter. Wxh5 Black is much better) 16...8xe5
96 The Benko Gambit

17.gxh5 &xb5 18.82xb5 £xc3 19.bxc3 Hxa2 16...Dxe5 17.Dg3 Dxf31 18.82£2 266 19.8g2
20.%xd7 Wixd7 21.8xd7 Balt 22.2d2 Ha2f Bd5 20.¥a3 Des5?
23.6bd1 Bxe3 24.cxbd Hb3 25.%cl oxbd 20..W1f6! 21.8xf3 Ha5 would have won
26.8el Balt 27.%c2 B3t 28.%0d2 Hxel immediately.
29.sbxel Bxf3 30.h6 &8 31.8b7 B4 32.8c6 21.8hd1 ¥f6t 22.%g1 Bxd1t 23.Exd1 £xg2
Bh4 33.2d5 Bxh6 34.8xf71 He8 35.5f2 and The remaining moves were not obligatory.
Black should get a draw. Black is winning no matter what.
15.d6! 24846 ¥f3 25.8c5 Wxa3 26.82xa3 L£a8
White needs active possibilities of his own. 27.8.¢5 Be8 28.b4 5 29.hxg6 hxg6 30.24 218
15.gxh5? Bxe5 16.¥d2 £h6 is problematic for 31.8Bd1 &xc51 32.bxc5 £c6 33.a5 Ha8 34.2al
White. Now after both 17.8e4 £xe3 18.%xe3 A d3 35.a6 Dxc5 36.a7
&xbs and 17.9d1 8xb5 18.2xb5 £ d4 Black 0-1
is much better, while after 17.9d5 ©b4 he
even wins.
Game conclusions: White is quite well placed for
15..Hxe5 16.Wxf71 &h8 17.5)ced the traditional Benko structure in the f3-line, so
White looks better here, but further analysis when Black chooses to develop the bishop to g7
proves that White’s Achilles’ heel, the weak he needs to play e7-e6 and not d7-d6, so that he
position of the king in the centre, is too can open the centre. In this way the 5...g6 line
important a factor to be dismissed so easily. mirrors the 5...¢6! line, with the difference that
17...016 18.¥b3 Be7 19.2xab Black’s development is slightly slower. As I said, I
19.5)F742 Bxf7 20.¥xf7 &e5 and Black wins. believe that White should be able to gain a slight
19...8xe4 20.fxed Wast 21.%0d1 ¥xa6 22.8xc5 advantage in this line, but to this day it is not
with very unclear play. entirely clear to me how.
14...2xb5 15.%xc5
Game 40
15.&xc5 is mer with 15...8e8 16.hxgb hxgb6
Wojtaszek — Harikrishna
17.2d6 Za5! and Black is better. This is a typical
World Junior, Goa 2002
tactical motif in this line.

1.d4 26 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.f3 6!

T EeT
///// , , , / %l/x%x
,,,,,,,,,,,,,,,,,

////y/
,,,,,,,,,,,

7 //7
é
,,,,,
/ /%
S
,///////, V
1a /4
,,,,,,,,,,,,,,,,,

16.¥d6
This is the most natural move and, in my
16.%c2 8xe2 17.8xe2 Exe5 and Black is much
opinion, the strongest. The point behind
better.
developing the bishop on the a3-f8 diagonal
 5.63 97

instead of the al-g8 diagonal is that yet another 8.8b37 &c5 9.He2 0-0 10.¥d3 Dat
diagonal, gl-a7, is seriously weakened after 5.f3. 11.8g5 (11.2e3? exdS! 12.¢5 He8 with a huge
Anotherimportant pointis that the white kingside advantage for Black) 11..%b6! gives Black
pieces cannot be brought into play easily, and it full compensation. For example: 12.2xf6 gxf6
therefore is the most natural choice for Black to 13.50d2 Db4 14.8c3 &b7 15.0f4 Le7!
try to use this by opening the centre. 8..8¢5
6.e4 c4?! Black should play normal moves. Without
A very bold pawn sacrifice indeed - and quite developed pieces he will not be able to break
dubious as well. Objectively it is probably a through the white defence anyway. After
losing mistake, but theory and pracrice are two 8..Wa5t 9.0\c3 &b4 10.¥d3 Da6 11.&e3
different things. For his two pawns Black receives White will be able to finish his development
compensation for one, but the position is very effortessly: 11...80¢5 12.¥¢4 exdS 13.exd5 Db7
sharp. White will need to play very carefully. It is 14.24 0-0 15.8¢7 8xc3t 16.bxc3 Wa3 17.8e2
tempting to compare White’s situation to walking De8 18.0d4 We7t 19.8f2 Hbde 20.2xd6
through a minefield. Every step he takes could be Hxd6 21.¥d3 b7 22.c4 and White has a better
his last. So this line is excellent for blitz and other position and two extra pawns, Carron — Kogan,
fast time controls. You could try it out in games Zurich 2001.
with the classic time control, but you should 9.2c3 £b7!
understand the risk. Remember, the world belongs Black wants open the e-file. This is better
to the brave! Personally I have 2¥2 out of 3 against than:
opponents with an average Elo of about 2300. 9...0-0 10.2ge2 ¥bo
An interesting alternative to 6...exd5, which 10...exd5!? is strongly met with 11.2g5!. When
is the main move here, is 6...2b7'?. Now after you have two extra pawns you can always
7.3 exd5 8.e5 Black has an interesting line return one without regret. White should not go
that leads to a position with good drawing into the complications after 11.£xd5? &xd5
chances. 8...20h5!? (for 8...¥e7 see the next game) 12.¥xds Was5+ 13.8¢3 d6 14.¥d3? (14.2d2
9.5xd5 Wh4t 10.g3 @xg3 11.hxg3 @’ng( Be6 15.9d3 £a6 16.8xa6 Exab 17.8d5 and
12.5he2 Wxe5t 13.9e3 axb5s 14.80£2 c4 15.8g4 White is still better, though this is hardly what
8051 16.863 &xe3t 17.80xe3 ¥Wxb21 18.82e2 he should have been playing for) 14...5a6
£1a6 19.%d6 Wxal 20.£h3 ¥f6 21.2xd71 &d8 15.a3 d5 16.exd5 ©b4 17.¥d1 &d4 18.8bl
22.8e67 with equality according to Notkin. &5 19.axb4 Wxb4 20.8d3 £xc3t 21.bxc3
7.8xc4 Wxc3t 22.012 £xd3 23.Bb2 Wd4t 24.%g3
7.5\¢3? axb5 8.0xb3 £a6 9.8c3 £c5 10.h3 West 25.012 Wd4t 26.2g3 Wxd5 and Black
0-0 11.8g5 d6 12.¥d2 &bd7 with great play is much better, Pinho — Alvarez, Mondariz
for the pawn. 1996.
7...axb5 8.8xb5! 11.2¢5 APTY
The first mine is avoided! It is hard to believe, Black can also try 11..2e8% 12.8cl &f2f
but many players only dare to take the first pawn. 13.56f1 6 14.2f4 5 15.8d2 f5 with some
They do not know that this only means that they compensation.
will have to suffer for nothing. 12.8xf6 gxf6 13.8b1 Hc7 14.8c4 £a6 15.8xa6
8.2d3? 8¢5 9.We2 b6 10.9h3 0-0 11.dxe6 Bxab
Ne6! 12.80¢3 Dd4 13.9d1 Wxe6 14014 Wes A typical position for this variation. White
15.56f1 b4 gave Black deadly compensation in retains the extra pawns and therefore the better
Mikrut — Pinski, Krzeszowice 1998. The point position — however it is not so easy to exchange
is that after 16.£cd5 b3! White is in serious this advantage for a full point.
trouble. 16.f1!
98 The Benko Gambit

Hauchard — Guidarelli, Vichy 2000. Now path to an advantage.

White wants to play g2-g3 and $g2. Most 10...0-0 11.52f1
probably this plan will cost a pawn, but White Just weak is 11.d6?, which gives excellent
will keep one pawn extra and a better pawn prospects for the black knight: 11...8¢6 12.£g5
structure. b6 13.9d2 Bfc8 14.a4 De8 15.8f4 Ba6
10.Dge2?! 16.2b1 & b4! with full compensation, Wojtaszek
The knight is just in the way on this square. — Predojevic, Chalkidiki 2001.
After 10.2h3!? exd5 11.e5 We7 12.8e2 Bh5 11...exd5 12.exd5 £a7!
13.8g5 We6 14.002 &xf21 15.8xf2 Wbt This is the reason why the black queen does not
16.52f1 d4 17.9d1 0-0 18.2d3 £6 19.exf6 Dxf6 go to b6 too soon. The cannon should be loaded
20.%f2 £a6 21.8el Whire is better. the right way so that it is the biggest bombs that
But Black can improve with 10...0-0!. Now we land on the white king’s head.
have the following options to investigate: 13.g3!
After 13.8g5 Wb 14.8h4 He8 15.%b3! We3
16.£c4 £a6 17.89xa6 Hixab 18.8e4 Dxed (but
not 18...20g4? 19.82¢5! and White wins) 19.¥xe3
fxe3 20.fxed D5 21.80¢3 Hxed Black has full
compensation.
13..¥b6 14.c0g2
Of course not 14.Wel? &HxdS and Black is
better.
15
7=
14.. %021 15.2h3 ¥xf3 16.50 84! W2
2
There is no real compensation for the pawn in
the ending after 16...¥xd1 17.8xd1.
17.8f1 ¥c5
a) 11.8¢5 exdS! 12.exd5 He8t 13.%f1 ho6

K ~ E&
14.£h4 ¥b6 with compensation for the pawns.
b) 1181 exdS! 12.e5 He8! (12...5h5 13.g4!
(13.f4? d4 14.9a4 Wa5 15.2d3 ¥Wxa4 16.Wxh5
g6 17.¥e2 {6 and Black has compensation)
44 2 e
,,, = / 4k
13...d4 14.5e4 &xed 15.fxed and White wins)
13.8xd5 &c7 (13...d62 14.2g5! dxe5 15.Wc2
g6 16.9xc5 ¥xdS 17.Wxd5 £xd5 and White is /§, , , AT/ »
much better) 14.8xc7 Wxc7 15.8f4 Hc6 with
2 %%
%7, ,,//"’
//
,,,,,, %
% % ay
compensation for the two pawns. Y,,,,,, » % //4
- %
c) Strongest is 11.¥b3! when after 11...exd5
12.¢5 De8 (12..0h5 13.2e2 £c6 14.82¢5 Wa5
15.82d2!
glorious
and White is better) White has the
13.8g5!" &c6 (13..h6 14.82d3 &6 5 sw z/
B
15.%xb7 8b8 16.&h71! The idea is to create a
path for the queen to come home on. 16...%h8 18.Be1?!
17.%a6 &c7 18.%e2 hxg5 19.£b1 d6 20.h4! White could still have kept a clear edge with
White is better.) 14.2d3 g6 15.%xd5 and White 18.3h5! &ixd5 (18...5xh5 19.%xh5 d6 20.2d3
has taken over the initiative. g6 21. %3 and White retains an extra pawn with
Of course, White should also be better after a better position) 1984 Wd4 20.De4 &%h8
10.£g5", but that is a different story. For us it is 21.8d6 and White is much better.
enough to know that White has at least one clear
 5.63 99

18..¥d6 19.2h5 Dxh5 20.¥xh5 Lfd4 8.8c4 0-0 9.8¢2 ©Dh5 with compensation
21.a4?! for the pawn according to grandmaster Ivan
Better here was 21.2f4! b4 22.We2 and Sokolov.
White is better, based on the fact that 22...8xc3 8...5f8 9.6c3 £b7
23.bxc3 ¥xc3 does not work: 24.2d6 £xd5 Also interesting is 9...axb5! 10.89xb5 £a6
25.8xf8 Le6t 26.@g2 &dsT 27.@g1 Wd4+ with compensation for the material. In the
28912 W21 29.5xf2 dxf8 game where this idea was tried out White
21...8xc3 22.bxc3 £xd5 23.8a3?! went quickly astray with 10.2e3?! ¢4 11.2f2
The last chance to make life difhcult for a5 and Black was much bertter in Bareev —
Black was 23.%We5 ®xe5 (White is better after Khalifman, Russia 1995.
23,8612 24.502 Wh6 25.8e3 Wb7t 26.%g1 10.%d2 fe5 11.bxaéb Dxab 12.8xab £xab
He6 27.8d6) 24.8xe5 Le6t 25.80g2 Hicb 13.8ge2 h5!
26.8el Bfc8 27.8e3 Da5 28.8a3 D4 29.8xcd Here comes the attack (again).
Bxcd 30.a5 Ha6 31.8b6 and White retains some 14.0-0 Wc7 15.f4 £d6 16.%2h1 h4 17.b3 h3
advantage. 18.g3 Wb7
23...8¢61 24.2g2 451 25.2h3 L6t 26.2g2 Black was slightly better in Volkov — Hjushin,
£dst27.52h3 Internet 2004.
a1/ 7..Me7 8.We2 Dg8 9.8¢3 £b7 10.2Dh3 c4
10...d4? 11.9e4 d3 (Gretarsson — Odeev,
Game conclusions: The 6...c4?! pawn sacrifice Moscow (ol) 1994) 12.¥xd3 Wxe5 13.4.f4 ¥Wxb2
is as entertaining as anything can be in chess 14.8b1 Wxa2 15.9d61 £xd6 16.8xd6 &f6
— but basically it is rubbish. White should take 17.82¢e2 and White is much better according to
whatever is offered to him and reinvest one of Gretarsson.
the two pawns at the right moment to take over
the initiative. Against the best line Black has no
chances. ,,,,,,,,,,,

’/%
Game 41
0
///’///
S. Ivanov — Khalifman
St Petersburg 1997 // , , ,
1.d4 Df6 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.f3 €6
6.e4 exd5
2
// 2
Black is leading in development and will find /
it easier to get the remaining pieces into play, so /,/,H/ // ////

it is very logical to open the position.

7.e5
Of course we cannot ignore the recapture, even
if we feel the lust to do so.
7.exd5?!
Living legend, Russian grandmaster David
Bronstein, says that positions like this one
should be evaluated by the gl-knight. Black
should not have trouble getting good play even
if he loses the right to castle.
7...8d6! 8. We2t
B %? %»M
11.263
This is the most natural and most often seen
move. But, unfortunately, we will have to get a
little dense with the theory here and look closer
at the most important alternatives.
11.8g52 W5 12.0-0-0 axb5 13.2e3 W6
14.8xd5 b4! According to A57 by Anartoly
Karpov this is a mistake. Actually, the reality
100 The Benko Gambit

is very different. (14...8xa2? 15.5xb5 fa3 is, 15. 82! is a serious improvement (15.a4?! axb5
according to the same source, much better for 16.2xb5 Wxbs! 17.axb5 Hxalt 18.80f2 &f5
Black, but after 16.bxa3! Exe2 17.2xe2 Black 19.g4 £c51 20.50g2 ©d4 21.912 Hb3 22.We2
is in serious trouble. 17...%¢7 18.8hd1 with a Ad4 2392 ©b3 24.We2 0-0 Black is much
strong attack. Or even worse: 17..%e7? 18.8c5 better, Timman — van Wely, Rotterdam 1998).
Wb 19.0d6+ ©f8 20.2c8% and White wins.) 15..%xf2+ 16.50xf2 Here it looks as if White
15.0b5 Exa2! 16.%b1 Wg6t 17.%5xa2 b3t is better. For example: 16...d4 17.bxa6 £xa6
18.5Ral Wa6t 19.9a3 £xdS and Black simply 18.8e4
wins. Thank you Fritz. 12.9xb5
After 11.9122 axb5 12.0f4? ¥xeSt 13.8d1 "This looks natural.
Of6 14.89xb5 £¢6 Black is much better. Better 12.82£42! gives Black two strong options:
is 12.8e3, bur still this cannot be the way for a) 12..Wxe5 13.Hxb5 Da6! (13..8e72
White to play. 14.8d4! (145522 Hf6 and Black is at least
11.8f4 Wc5 is a bigger tabia. Here we should slightly better) 14..Wxe2t 15.8xe2 &f8
look at: 16.8¢7 &c6 17.8b6 Ead 18.81xd5 and White
a) 12.8e3 d4 13.82 d3 14.We3 Wxe3dT is slightly better) 14.0-0-0 &e7 15.2d2 ¥b8
15.8xe3 &b4 and Black is slightly better. 16.8e3 £c6 17.8¢3 Dc7 18.g3 De6 19.82g2 g6
b) 12.b4?> ¥d4! (12..Wxb42! 13.2d2 WS and Black is better, as the sharp 20.2fxd5? does
14.2b1 d4 15.9a4 d3 16.Dxc5 dxe2 17.8xb7 not work because of 20...8xd5 21.8xd5 &xd5
and White is better) 13.We3 Wxe3t 14.&xe3 22.2xd5 Exa2 23.2d4 Wb3! and Black wins.
£xb4 and Black is much better. b) 12...%b4 Now we have two valid moves to
¢) 12.24 d4 13.bxa6 Dxa6 14.De4 b4t investigate:
15.%2 d3 16.¥e3 £.¢5 and Black is much better, b1) 13.212 He7 14.a3
Adianto - J. Polgar, Jakarta 1996. 14.8d1 d4 15.2xd4 D5 16.a3 (16.62! &c5!
11.8d2%? is an interesting, but not very (16..fxe6? 17.8xe6 dxe6 18.¥xeGt &e7
popular, move. I found only two games with it 19.2b6 with a crushing attack) 17.ex{7+ Sxf7
in my database. 18.6)xb5S Wxb5 19 Wxcdt Wxcd 20.8xc4T d5
a) 11...¥d8 is strongly met with 12.¥e3! (but 21.8xd5% £xd5 22.8xc5 £xa2 and Black has
not 12.9\f42 d4 13.5e4 axb5 14.23d6T £xd6 good winning chances.) 16..0xd4 (16... Wa5?!
15.exd6t B8 16.%eS We8 17.Wxe8T txe8 and 17.e6! with an attack) 17.8xd4 ¥a5 and Black
White has no real compensation for the pawn) is better.
12..%c7 13.0f4 De7 14.¥d4! and Whirte is 14..¥b3 15.52d1 S5 16.Dfxd5 Hxe3 17.%xe3
better. &a6! and Black is better.
b) 11..%cS! 12.b4 This is the main idea. b2) 13.0-0-0 He7 14.8xdS ¥xc3t!
12..9b6! (12..%xb4? 13.bxab £xab 14.9xd5 A standard sacrifice in this kind of position.
and White is much better, Komljenovic — Black receives a strong attack in return for the
Campos Gambuti, Cordoba 1990) 13.8f4 axb5 queen. And, just as importantly, the position is
(13...8e7 14.a4 and White has some initiative) easy for Black to play.
14.8fxd5 &xd5 15.0xd5 Wd4 16.Hc7+ &d8 15.bxc3 &©xd5 16.9xd5?
17.8xa8 Wxalt 18.50£2 He7! (18...80¢6 19.8e3! Necessary was 16.#d2! £a31 17.50b1 Hixf4
with counterplay) Here it looks as if Black is 18.&xf4 0-0 with unclear play.
better, but only practical tests will be able to 16...£xd5
show if 11.2d2? has a future. Black clearly has a strong attack. This game
11...axb5 shows exactly how dangerous Black’s initiative
11..%b4l> 12.a3 Wa5 13.2d2 Wb6 14.9f4 is.
&e7 has also been played. Now Stohls idea 17.9d2 £e6 18.82¢2 Hic6 19.f4 b4
 5.63 101

In his annotations in Chess Informant Gipslis 16.¥d22!

gives the evaluation “with counterplay” here. This leads White into serious trouble. Better
must say that, after having spent a lot of time was 16.b4, though after 16...cxb3 17.¥b2 d4!
trying to find counter chances for White, 1 18.8xd4 Hbc6 19.8xa6 £xa6 20.82 Wxe5t
have not successfully done so. 21.5e4 &5 (21.Wxb22? 22.0d61 &d8
20.£5 bxc3 21 ¥xc3 Ha3 23.4b6 mate) 22.0-0-0 Wa5 23.Hdel &d8!
White had serious problems in Czerwonski Black has a massive attack.
— Gipslis, Koszalin 1998. The best option was potentially 16.84c57
12..%b4t 13.2¢3 De7 D5 17.8xf8 Bxf8 18.0-0-0 d4 19.2xd4 Dxd4
20.%xd4 4 c6 where Black is only slightly better.
__ 16...0f5 17.8£2 £c5 18.8d1 &xf21 19.5)xf2
///
//
,,, v ///
,,,,,,
/g
A Y
=
= 0-0 20.2e2
20.70g4 BeB821.8e2 h5 22 .5\e3 Bixe3 23.Wxe3

J » //
d6 and Black is much better.
20..%c7 21.Wf4 De7 22.Dg4
7/
//////8/)
223
25.8xh6t
He8
HBxh6
23.8g4
26.5d4
Df5
Exe5t
244
27.8e2
Dho
&6
7 Z and Black wins.

,,,,,,,,,,,,
22..0\g6 23.%g3 d6
Also interesting is 23..Wb6!? 24.Hd2 h5
O 25.0f2 Be8 26.f4 d6 27.exd6 ¥xd6 and Black
72% keeps the pressure.
=
24.exd6 Bxd6 25.20d4 5 26.Df2 Dc6 27.8e2
Dxd4 28.8xd4 f4 29.¥g5 Ee8!

7 %/z///w/
14.a3?
Necessary, accordingto Khalifman, is 14.0-0-0!
when play would continue:
//w
L
14... 215 15.8£2 Was!
15..8c5 16.8xd5! &xf2 17.9xf2 &xd5
18.5xd5 b7 19.&xc4 0-0 20.52b1 and two //////

pawns for the exchange gives White much

better chances.

// /%/ ,
16.5)f4 d4 17.8xd4 HHxd4! ,,,,,,,,,,

17...£2b4?! gives Black compensation according

to Khalifman. However, after 18.Wxc4! &c6
AN

19.06d5 Dexdd 20.9c7t Pe7 21.Wxb4t

Wxb4d 22.07d5T &xdS 23.89xd5t e6
24
.9 xb4 White is better.
18.8xd4 £c5 19.8d2
19.8xc4d £a6 20.8xc5 Wxe5 21.Wed Hc6 With the obvious idea ...8de6. Black wins.
22.8xa6 Exa6 and Black is better. 30.@g4 Bde6 31.2d2 c3 32.bxc3 ¥xc3
19...0-0 20.2fd5 Dc6 0-1
with compensation for the material.
14..%a5 15.%12 a6 Game conclusions: After 6...exd5 Black has good
15..00f57 16.8b6 Wa6 17.8d1 Le7 18.2f4 play. We still need to look at the critical 12.0-0-0
and White is much better. in the next game, but all other lines were covered
102 'The Benko Gambit

in this game, and it seems as if White’s position

is unpleasant to play. So, although the position
in the critical line is evaluated as “compensation
for the marerial”, we should not forget that this Qi%l%//flfil
is an objective evaluation, and that it was White _
who did not find the correct path in our practical
example. ///A///E@ @
Game 42
Norris — B. Lalic
Aberdeen 1998

1.d4 £ 6 2.c4 c5 3.d5 b5 4.cxb5 a6 5.3 €6 6.e4

exd5 7.¢5 We7 8.We2 Dg8 9.0c3 £b7 10.2h3
c4 11.8e3 axb5 12.0-0-0 ¥b4 13.8xd5 13...%xc3+!
Another main alternative is the following; Again we have this queen sacrifice. The
13.9f4 £e7 14. 2 Da6 evaluation of it currently is that Black equalises,
14.. %2517 15.2b6 Wa6 16.2c5 (16.8xd5!? but probably has easier practical play. Taking the
8c6 17.a4 D8 18.8xbS! Re7 (18..8xb5? exchange is, of course, possible, but very much
19.5fdS! and White wins) 19.2fd5 &xd5 against the spirit of the position. White can, if
20.59xd5 Wxad 21.8c7T 2f8 22.8c5 D6 nothing else, force a draw. It seems that if he goes
23 8xe7t ©8xe7 24.5)xa8 Wxb5 25.4b6 for more, Black can then force the draw:
Hxe5 26.f4 @g/l 27 W5 ®xes 28.Dxd71 13..8xd52 14.0xd5 a5 15.&b6 Wxa2
Be8 29.8xc5 De3 30.g3 £7d5 with equality 16.8c71 she7 17.89f4 Walt 18.%c2 Ea2
Surender — Bergmann, e-mail 1999) 16...8c6 19.5fd5t d8 20.Hxb5T e8 21.¥xc4d Bxb2t
17.a3 (17.20h521 5! 18.8xf8 Lxf8 19.g4 (21..Wxb2t 22.¢bd3 Wd2+?! 23.c2e4 is probably
&e7 and Black is betrer) 17..d4 (17..%b7 better for White) 22.52d3 ¥d11 23.0ed Welt
18.8xe7 fxe7 19.90fxd5 £d8 20.%e4 0-0 24.5f4 g5t 25.8f5 Dh6t 26.8f6
21.0d6 and White has the initiative) 18.8xd4
N5 19.82d2 Exc5 20.WxcS Wby
15.56b1 Hc7!
15..065 16.0&d5 £xd5 17.5xd5 &xe3
18.%xe3 WeS 19.Wxc5 &xc5 20.0c¢3 D7
Lalic — Alterman, Pula 1997, and now 21.g3
gains a huge advantage for White.
16.23 Bxa3! 17.&c5 Wa5 18.8xa3
18.8b6 Wb4 19.8¢5 is equal.
18..b4 19.&xb4 Wixb4 20.%d2 Wb3 21.8e2
&c0!
After 21...g62 22. & e4! White is able to exploit
the fact that the d7-pawn is defenceless. 26...5g8% 27.%f5 with the oddest perpetual I
22 Bhel g6 have seen for a long time.
With an unclear position. Probably Lalic was 14.bxc3 £xd5 15.20f4 a3t 16.8b1 De7
quite impressed with Black’s tactical resources 17.8cl
and secretly wanted to sit on the other side of Here there are some alternatives:
the board. 17.8d4 £e6 18.9xe6 dxe6! 19.£4 0-0 20.g3
 5.3 103

Dbc6 21.8g2 Dxd4 22.cxd4 Bad with the chances. However, he manages to lose in just a
initiative. few moves. I think the reason this happened is
17.%d2 8e6 18.5h5 He8 19.80f4 Dbco psychology. White was shocked by his blunder
20.0)xe6 fxe6 21.f4 ©f7 22.8e2 HdS 23.4f3 and lost some of his fighting spiri.
&ce7 with unclear play, Klein — Waitzkin, New 30... 251 31.%e4 g6 32.8b12
York 1993. A better defence was 32.2b7! £b8 33.8b1 £g7
17..2bc6 18.8xa3 Bxa3 19.82b2 Ea5 34.f4 He6 35.0e3 Dd8 36.8c4 Hc8 37.8d5!
20.%d2 32..2c8 33.%2f4 Dg7 34.8b7 Ec2 35.h3 Ed2
White provokes exchanges leading to a drawn 36.8e4?
ending,. The white position has now become quite
20...b4 21.cxb4 Bxa2t 22.%xa2 3t 23.¥xd5 difficult. Yer he rerains some drawing chances
Hxb4t 24.b3 Dexd5 25.Dxd5 Dxd5 26.8c4 until this blunder. Better was 36.8b5!2.
He3 27.8xc3 e 36..0e6t 37.8g3 Dc5 38.Bb4 He2 39.0f4
You need a lot of talent to lose an endgame Bxe4t 40.Bxed Dxed 41.0xes Re6 42.f4 h6
like this. 43.h4 h5 44.e3 Bf5 45.0f3 f6
0-1

B
/ ,i, ,
Game conclusions: In this critical main line Black

///@54 can probably accept the exchange sacrifice and

still make a draw, but far more arttractive is to
sacrifice the queen for a strong attack. This does
not lead directly ro the downfall of White, but
7, % gives sufficient compensation to equalise. Again,
it is probably more demanding to play White, as
7 //////
/ defending is one of the hardest things in chess.

/ /""""/ A Chapter conclusions: The 5.f3 system has not

. / . 8
28.bd422
A horrible blunder, putting the king in the
circuir of the knight. The simplest defence was
28.8el! Dixg2 29.Hgl ©f4 30.Exg7 with instant
equality.
28...8xg2 29.Bg1 Dh4 30.2d5
After 30.0e4 g6 not too much has happened.
Winning the g2-pawn was a good step for Black
towards the full point, but just one step in a long
journey. White should still have excellent drawing
been an attractive option against the Benko since
Khalifman introduced the queen sacrifice on ¢3,
where Black gets rook and bishop for the queen.
The current evaluation is about equal chances,
bur in a position where most players would find
it easier to be Black.
The only alternative to 5...e6 and 6...exd5
that seems fully viable is 5...g6 and quick
development of the pieces, before a quick counter
strike in the centre.
In this line everything is about rapid
development of the pieces, as White’s strong
structure will otherwise give him an edge.
Chapter 7

5.9\ c3 — The Zaitsev Variation

1.d4 56 2.c4 c5 3.d5 b5 4.cxb5 a6 5.0 31 poor. Black gains active and dangerous play and
White will have to play quite carefully just to
keep the balance.

///////////////// Game 43
Arencibia — Ivanchuk

x/%//m
Cap d’'Agde 1998

7 1.d4 96 2.c4 5 3.d5 b5 4.cxb5 a6 5.2c3

axb5!
This is really the only good move for Black.
After 5..%a5? 6.8d2 axb5 7.e4 b4 8.e5 bxc3
9.%xc3 Wa4 10.Wxa4 Bxad 11.exf6 gxf6 12.82b5
White is berter.
6.e4
,,,,,,
6.50xb5 8a6 7.8)c3 transposes to the Benko
Gambit Accepted.
This developing move is an idea of Grandmaster 6...b4
Igor Zaitsev, a well-known theoretician and 6... %45 was famously refuted in the following
famous life-long second of World Champion game: 7.5 b4 (7..5e4?! 8.4d2! &xd2 9.Wxd2
Anatoly Karpov. Though not a particularly strong b4 and now 10.£e4! gives White a clear edge
grandmaster, he has invented a long list of new according to Adams back in 1991) 8.8b5 De4
moves and variation, most famously the Zaitsev 9.8c4 £a6 10.9d3 b3T 11.0e2 &xb5 12.8xb5
Variation of the Ruy Lopez, one of the most Wxa2 13.5xa2 bxa2 14.Wxed al=¥ 15.23f3 The
important openings of all, and the spectacular compensation for the exchange secures White
11.2)g5!? sacrifice in the Open Spanish, which a clear advantage. 15...%a5 16.8c4 o6 17.8d1
was first played in the Karpov — Korchnoi World fe7 18.8g5 Ha7 19.8xe7 xe7 20.9g5 Whd
Championship match in the Philippines in 21.W4 6 22.extGt gxf6 23.d6T Fe8 24 Wixf6
1978. Wxc4t 1-0 Timoshchenko - Adams, London
5.8c3 was quite popular at the end of the 1992.
1970s, but these days its reputation is rather 7.8b5 d6
106 "The Benko Gambit

All the pieces should participate! After 21.2)b5

EA oW
,,,,,,,,,,,,,,,,,, s E Black is better, but White can still fight back.
21...2e8 22.50d2

& a
., e, 7, . A, 22.hxgd Bxedt 23.¢2d1 Sxe4t 24 $e2 Bd4t
7. and Black wins.
22.. %251 23.8c2 Wxel 0-1
/%//7 // / b) 9.e5?!
/w/ / / This makes little sense here.
9...dxe5 10.Dxe5 g7 11.8c4 0-0 12.0-0 2a6!
13.8el £xb5 14.2xb5 ¥xdS
Q \

& ? % iy i
///
D&
\\\\\\

and Black is slightly better.

) 9.8f4 8g7 10.¢5
,,,,,,,,,,,,
This is also not admirable.
10...dxe5 11.&xe5 0-0 12.8c4 £a6 13.24 bxa3
8.8f4 14.8xa3 Hbd7!
8.8c4 was popular briefly, with the idea of Worse would be 14..8xb5? 15.8xa8 &xcd
replying to 8...g6 with a double pawn advance 16.8xb8 Wa5t 17.8c3 Wa6 18.8xf81 &xf8
of 9.e5! and 10.d6! with an attack. But of course with compensation, I. Rogers — Hodgson,
Black plays 8...22bd7! 9.3 &b6 10.2d3 Gereat Britain 1985.
(What can I say about 10.2e2? Well, 1 do not Black is better here. For example:
know why such a stupid move is one of the main 15.8c7
lines in A57 — Karpov’s book from Informant. 15.8g3?? &xb5 16.8xa8 Wxa8 17.2xb5 WaSt
10...Hxe4 11.0-0 g6 12.a4 bxa3 13.Hxa3 Exa3 and Black wins.
14.bxa3 £g7 and Black is much better. White 15..%¥c8 16.8a5
does not have full compensation for the pawn.) 16.d6? exd6 17.6xd6 Wxc7 18.£xa6 £1b8 and
10...g6 11.b3 27 12.£b2 0-0 13.0-0 £a6 14.a4 Black wins.
bxa3 15.6xa3 with equality, Mensch — Pinski, 16...2xb5 17.2xb5 ¥b7 18.8c6 ¥Wxb2 19.8b3
Budapest 1997. W22 and Black is much better.
8.8\f3 g6 gives us the following options: 8...g5!
a) 9.8c4 £g7 10.0-0 0-0 11.Bel Dbd7 12.a4 After some time this energetic move was
bxa3 13.Exa3 xa3 14.5xa3 found, after which Black does not have any great
This gives approximately equal play, Bacrot problems. Other moves do not deal sufficiently
— Nisipeanu, Balatonbereny 1996. Black with the tasks of development and the white
now showed an interesting way to play the threat of e4-¢5 with an attack.
position. 8..bd7? 9.3 £\b6 10.8c1 (10.8e5" dxeS
14...@g4!? 15.£f1 &deS 16.d2?! 11.8xe5 Wd7 12.%¢2 and White is much better
16.h3 with equality was better. according to grandmaster Vitaly Tseshkovsky)
16...0x12! 17.8xf2 D4t 18.%e2 6! 10..9xed 11.9e2 £5 12.8g5 €5 13.dxe6 Dixg5
This gives excellent compensation for the 14.Hxc5! dxc5 15.8¢77 ¥xc7 16.8xc7 and
piece. White was only able to defend for a few White is winning in, Nenashev — Rashkovsky,
additional moves. Tashkent 1987. Nenashev is of course known as
19.%b3 exd5 20.h3?! Alexander Graf these days, one of the strongest
20.¥xd5 was necessary, when after 20...8e6 and most original players from Germany.
Black has very strong compensation, but is not 9.§=xg5?!
yet winning. The stronger 9.8e3 will be considered in the
20...dxe4 21
.8 xe4? next game.
 The Zaitsev Variation 107

9...0xe4 10.2f4 18.%xb7 0—0 and Black is much better.

12.5f32 &bd7 13.0-0 0-0 14.¥b3 &b6
15.a4 &b7 16.8fd1 W¥d7 and Black is much
betrer, Rodewis — Richrer, Germany 1996.
But playable is 12.a4 bxa3 (12...2bd7 with the
standard plan of eating the d5-pawn) 13.&xa3
Bxa3 14.8xa3 Wb6 though Black remains
slightly better.
12...8bd7 13.2g3 b6 14.b3 0-0 15.0-0
fd7
In this line it is imporrant to keep an eye on the
f5-square. If not, bad things could easily happen
g% // /g/ A faster than you could say: [ won a piece!
HE
///// /

o 16.%4d2
16.8¢5 Ba5!? (Black avoids the complications
after 16..80fxd5 17.8cl (the compensation
10...22f6! after 17.8xd6 &xal 18.W¥xal f6 is rather poor)
In my opinion this move gives Black the 17..5xc4 18.bxcd &f6 19.¥d2 with some
advantage; which means that the whole line with compensation) 17.a4 bxa3 18.5¢3 (18.%xa3
9.82xg5 is dead. Dfxd5 19.8xd5 &xal 20.¥xal Hxd5 21.£h6
The text move is much better than the more €5 and Black is much better) 18...2g4 19.8d2
popular 10..£g7?. Now play can continue: Nxcd 20.bxcd He5 21.£477 (21.9b1 Ha8
11.%e2 D6 12.9Dxd6T £f8 13.Hxc8 Wxc8 22 Bxa3 Hxchd 23.8xa8 Wxa8 24.2c3 is much
14.d6 exd6 15.2xd67T g8 16.Df3 W15 (16..
W c6 better for Black, but White needed to try this)
17.8g3 Dbd7 18.¥c2 We6t 19.8¢2 b3 20.Wc4 21..Dxc4 22.65 Hxd2 23.¥xd2 Le5 and Black
Exa2 21.8xa2 bxa2 22.0-0 Wxc4 23.8xc4d Hed wins, Hochstrasser — Pikula, Pontresina 2000.
24.8xa2 Pxg3 25.hxg3 and White was much 16...2fxd5
berter in Timoshchenko — Golod, Slovakia This is often the consequence of the Zaitsev
1993.) 17.g4! This idea has not been tested in Variation. White loses a pawn and has to play
practice yet. (known is 17.%e5 Wxe5t 18.8xe5 very accurately to get enough compensation.
Abd7 19.8xF6 Be81 20.82d2 HHxf6 21.8b5 and 17.£h6
according to Karpov, White is slightly better)
17..%d5 18.g5! Wxd6 19.gxf6 2xf6 20.We87
W8 21.8g1t 8g7 22.We4d Ha7 23.0-0-0 and % v/@/
White is much better here. %
,,,,, Z?
11.8c4
Now White has a serious problem: how will
he get the b5-knight back into the game? In my
opinion Black is slightly better.
11.We2 Ha6! is the point behind Black’s play.
The d6-pawn is now under solid protection.
11..8g7 12.De2
Let us have a look at some plausible
alternatives:
12.a3?! ©bd7 13.axb4 Bxal 14.¥xal Qb6
15822 oxb4 16.0f3 £b7 17.Wa7 &Sxcd
108 The Benko Gambit

17...8xh6
Of course moves like 17...8xal? 18.8xal! look
more like an attempt from Black to destroy his
own position than anything else, for example:
18...e5 19.9xd6 with a very strong attack.
7 % / R
18.¥xh6 £6 19.82d3
I cannot find the ateack after 19.%h5 £5 20. %13
/4 . A 93
7
2xb5 21.8xb5S e6!. Also after 21.2xd51 &Hxd5 % 7 IZ(/////C ’?,H///c/,,,/ 7

22.Wxd5+ bh8 23 Bfel f4 24.9f5 £3! Black is

"
better, partly because 25.%xf3? does not work on
account of 25...2d7 and Black wins.
19..5f7 20.a4 bxa3 21.2xa3 W8 22.Wh4
@b4 23.8e2 d5
g
White does not have enough for the pawn.
24.9b5 Bxal 25.8xal 2h8 26.2c3 €5 27.f4 38.8xc6!
c4 Now White achieves a draw.
More natural looks 27..e4 28.Ha7 £c6 38...Wxg5 39.8a8t drg7 40.8a7t &f8 41.2a8%
29.Exf7 Wxf7 30.£h5 Wf8 31.806 We7 32.8f5 thg7 42.8a7t f8
£e8! (better than 32...e3 33.9d1 d4 34.2xh7! 151,
Wxh7 35.¥xf6t g7 36.Md8t &h7 37.Wh4t
g8 38.Wd8T &7 39.Mxb6 We4 40.Wc7t Wd7 Game conclusions: Black must be very happy with
41 ¥xd7t &xd7 42.9b2 and White has some the opening, if not with the rest of the game.
drawing chances) 33.8\ce2 .fi,gG and Black is Afrer 9.8xg5?! Black takes over the initiative and
much better: the pawns are ready to advance! should change his strategy from equalising to
28.2h5 £d3 29.h3 Bg7 30.fxe5 fxe5 31.bxcd playing for a win.
Nxch
31.. 90+ 32 %h2 W4 33 Wxf4 exfd 34.@g62 Game 44
dxc4d 35.82f3 and White has good drawing Potapov — Degraeve
chances. Cappelle la Grande 2002
32.50h2 He3 33.813 D42
Even worse was 33..e4? when White 1.d4 &6 2.c4 c5 3.d5 b5 4.cxb5 a6 5.Dc3
has 34.2cxed! dxed 35.¥xed Hxg3 36.Ha8 axb5 6.e4 b4 7.2)b5 d6 8.8f4 g5 9.8e3
(36.5bxg3?? Wd6t 37.8h4 &f5t and Black
wins) 36..82c8 37.Wxd3 Exf3 38.Wd4t Wfo
39.8xc81 kg7 40.¥d71 W7 41.9d41 with ,,,,,,,,,,,,,,,,,,

equality.
33...2c6!34.8b1 g8 with a mean exchanging

//@//
plan leaves White utterly helpless. After 35.8al
Wo7 36.8g1 We5 37 Wxg5 Hxg5 Black wins.

,/&// % /
34.%el d4 35.2ce4 Bgb 36.Hal £c6 37.2g5
We7
Better was 37...h6! 38.8)5e4 Hifxg2!? 39.8xg2
@xgz 40.@xg2 W4 41 . Wxf4 exfd 42.50h2 fxg3t

B wsang
43.9xg3 and because the bishop is “good” - h1 is 4// /7/ i e
a light-coloured square, just like the black bishop
== Y, A 7 = et
- Black has good winning chances.
 The Zaitsev Variation 109

—ige ¥
This is the critical position in the Zaitsev
variation, It is only in this line that Black can face
real dangers, and these would still mainly come
from himself.
Now he has a number of relevant moves, which 3 "oy » ,,,,,,

theoretically all look good, but a closer look brings

us back to the truth I have been promoting from
the very first pages of this book: Black should
always play the most active moves in the opening ////
when playing the Benko Gambit.
9...8g7 % @Z&//
This seems to be the best move. The g-pawn is
no more hanging than on the previous move. |
have some doubts about 9..Hxe4 10.2d3 a5
(10..9f6 11.8xg5 &g7 12.8e2 Hbd7 13.0-0
He5 14.a4 HHxd3 15.%xd3 and White has some 15.%d22!
initiative is well known) Better was accepting the challenge with 15.g4
D4 16.Dxf4 gxf4 17 8xt4 £xb2 18.20c4 and
the position is rather unclear.
15...8a6 16.8c1 &xf1 17.8xf1 ¥b8 18.2d3
De5 19.8xe5 Lxe5 20.g4!
4 After the sofc 20.g3?! ¥b3! Black has a
4,,,//,//%/ /
dangerous initiative. For example 21.Ef2 &xg3!
7z 7
7 with a better ending for Black, or 21.We2 with
the idea ©cd, which is mer with 21..8xg37T
22.hxg3 Dxg3 23.Wd2 Dxfl 24.%8xf1 £d7 and
Black is berter.
20...2f4 21.9c4
11.¥e2! A new idea in this position, which
gives White better chances. (1 1.8e2 5!
DT s

% 7 /é//,
and Black was slightly better in Cu. Hansen
— Fedorowicz, Amsterdam 1990) 11...b3} 7,
Z

12.¢bd1 ¥xa2!? 13.8cl (13.Exa2?? bxa2 and

Black wins) 13...f5 14.8)¢71 ©d8 15.5xa8 ¥xa8
and I think White is better. ///////////

10.f3 h6 11.a4 bxa3

Now this is forced; the d5-pawn is too strong.

"
12.8xa3 Exa3 13.9xa3
This position is more or less equal, but maybe

/ ’/@f/z//
a little easier for Black to play!?
13...0bd7 14.2e2 D51
Preventive, though not absolutely necessary.
Black does not want the white knight to arrive at
£5. Also OK was 14...82a6 15.9g3 &xf1 16.%xf1 21...8f6
A little better was 21...45g6 22.8xe5 (22.%¢2
Whe 17.0f5 Bg8 with unclear play.
e6 23.8f2 exd5 24.exd5 0-0 and Black is slightly
110 The Benko Gambit

better) 22...80xe5 23.¥c3 h5 24.§,Xg5 hxgd 38.h5 2b8 39.8h2 Eg8t 40.2f3 Bgl 41.h6
25.8f4 Wbs! and Black retains some small @gs 42.Bg2% Bxg2 43.@xg2 &h7 44.5063
winning chances after 26.8xe5 dxeS 27.¥xe5 &xh6 45.2g4 g6 46.b3 &3
What 28. 902 6 29.¥c7 Wxb2t 30.dg1 &f7 Va—Ya
31.Wg3 oxf3 32.5h1. Game conclusions: After 9.8e3 play remains sharp
22.Wc2 g6 23.f4 gxt4 24.8xf4 Dxf4 25.8xf4 and interesting, but certainly not to White’s
h5 26.211 hxgs 27.Wast &f8 28.8xgh advantage. With accurate play Black will retain
The white king looks unsafe, buritis impossible active and equal play.
for Black to use this to his advantage anymore.
28.. %8 29.Wc6 £d4 30.2g2 6 31.WxeBt Chapter conclusions: It is for good reasons that
thxe8 32.h4 &f7 33.%2h3 Eb8 34.Eg2 Eb4 we no longer see the Zaitsev Variation at the top
35.8c2 Bb3+ 36.52g4 Ed3 37.2d2 Eb3 level. Black has an easy and attractive game after
After 37..8xd2 38.5)xd2 £xb2 39.%f5 the 8...g5%
draw is near as well.
 Chapter 8

Five Sidelines

An old adage goes: If you want to play chess Also completely unthreatening is 4.g3 bxc4
well, you should play principled openings. In 5.8g2 d6 6.0f3 g6 7.0-0 £g7 8.5fd2 0-0
the Benko Gambit this means that you should 9.9 xc4 ©bd7 with equality.
accept the pawn sacrifice and fight for the full 4.0 e4! 5.8284
point. White can get lucky and end up better in 5.h4 WaSt 6.8d2 bxed 7.Wc2 Hixg5 8.hxg5
some sidelines, but in general Black is doing well, 26 9.%c3 Wxc3 10.bxc3 £g7 and Black is at least
which [ hope to prove with this chapter where slightly better, Murshed — Sermek, Moscow (ol)
we shall mainly concern ourselves with 4.2g5, 1994,
4992, 403, 4.89)d2. and 4.24 5..%as51 6.20d2

e
,,, @b H
Game 45
Grivas — Khalifman
Leningrad 1989 4/‘% %‘/‘4‘ ,,,,,/?y‘
/ ’/

Wakho %
1.d4 &6 2.c4 ¢5 3.d5 b5 4.8g5 o
7 /
A good move, but only in blitz. It only works

%&7 / /
when your opponent does not know the theory.
4.e4?! Another “blitz variation”, but it is not
quite as stupid as it looks: I lost a couple of games
J,/
&”
to it on the Internet. This was the punishment
for lack of knowledge. 4..Dxed! 5. W3 Wast!
(5...0£6?? This is a classic mistake in blitz. 6.d6
Hc6 7.cxb5 Hidd 8.Wxa8 N2t 9.06d1 Hixal
10.b4 exd6 11.2b2 and Black can resign with a
clear conscience. Instead after 5...0d6 6.cxb5 a6 6...85! 7.8¢3
7.8f4 White has interesting play.) 6.2¢3 (6.22d2 7.b42 ¥ixb4 8.8b1 W3 9.8b3 Hxf2! 10.8xc3
&d6 7.cxb5 Dxb5! 8.d6 £ c6 and Black is much Dxdl 11.8xb8 Dxc3 12.82e5 Bg8 13.8xc3 b4
better) 6...5xc3 7.2d2 b4 8.bxc3 b3 9.2d3 do! 14.82b2 Bg6! Black is much better. For example,
(better than 9...b2? 10.8b1 Wxa2 11.5h3 and one threat is Eg6-a6.
White holds the position) 10.23h3 ©d7 and 7.8e5 2¢8 (this is better than 7..f6 8.&c7
Black is much better. Wxd2t 9.¥xd2 ©Hxd2 10.8%xd2 and Black is
4.b3 This is not at all dangerous. 4...bxc4 slightly better) and now:
5.bxc4 g6 6.8b2 £g7 7.4 d6 8.53f3 ¢4 9.83bd2 a) 8.82¢7 Wxc7 9.9 xe4 bxed 10.e3 £a6 11.8c1
0-0 with equality. %7 and Black is probably better.
112 The Benko Gambit

b) 8.b4 Wixb4 9.9gf3 Was 10.¥b1 &Hxd2 9.d6? h6! 10.¥xc4 exd6 and Black is better.
11.50xd2 d6 12.8b2 (12.W¥xh72! Hg6! with 9. Wxcd! &b7 (9..d6 10.2d4! (10.8xg5 £ab
initiative) 12...bxc4 and Black is better 11.%c2 &xd5 and Black is slightly better)
¢) 8.9gf3 &xd2 9.Wxd2 ¥xd2+ 10.9xd2 d6 10...2a6 11.%c3 Wxc3 12.8xc3 £¢g7 and the
11.8¢3 5 12.€3 £g7 13.0-0-0 bxcd 14.8xc4 game is unclear) 10.%¥xc5 Wxc5 11.8xc5 £xd5
£d7 with an unclear game. and the position is more or less equal.
7¥2 oxf4l (7..0xf2 8.8e5! f6 9.8c3 b4 9...0xd5 10.¥xc4
10.8xf6 exf6 11.%xf2 and White is slightly 10.¢4 Db6 11.8xc4 Hixcd 12.Wxcd D6 Black
better) 8.Wxe4 bxcd 9.Wxf4 ¢3 10.bxc3 £¢7 and is at least slightly better.
Black is berter. 10...2b6 11.¥c3
7...bxc4 This looks bad, but White’s position is already
7..£5 8.Wc2 £07 9.cxb5 d6 and the position difficult. For example: 11.¥c2 &c6 12.0gf3
is unclear. 8g7 13.e3 Da4 14.8b1 Eb8 and Black is much
8.%c2 better.
8.3 &6 9.8xg5 ©xd5 10.e4 Db6!? (this is 11..¥xc3 12.bxc3 £g7 13.8cl Dad! 14.8b1
stronger than 10...hG 11.2xh6 £xh6 12.exdS Hc6 15.¢4 Bb8 16.Df3 d6 17.2fd2 Re6
¥Wh4 when Black is only a liccle better) 11.%c2 18.8c2 d7 19.8¢2 Eb6!
g7 12.8xc4d Dxcd 13 Wxcd & c6 with a huge The future FIDE World Champion shows how
advantage. to exploit such a strategic advantage: after he has
8...2f6 finished developing he just occupies the b-file.
8...20d6 9.8xg5 &g7 and now: 20.0-0 Ehb8 21.8fcl Bb2 22.a3 Ra2 23.Exb2
10.e42! Ha6 1123 Dbd 12.¥cl Qd3+ Bxb2 24.2d1 Db6 25.f3 DeS 26.8c22
13.8xd3 cxd3 14.0gf3 (14.2b1 5! and Black’s 26.8e2 Decd is also hopeless.
position looks berter) 14..8xe4 15.b4 cxb4 26...Bxb1 27.8)xb1 £xb1 28.Bcl £d3 29.a4
16.8)xe4 b3+! (16...8xd5? 17.0-0 ¥xe4 18.Eel Hbc4 30.8b3 Da5 31.8a2 c4 32.8e3 Decb
and White has a terrific actack) 17.%d2 ¥xd5 33.2b1 2xbl 34.Exbl b3 35.8Bb2 &xc3
and Black was much berter in Zysk — Hertneck, 36.2c2 &8d4
Germany 1987. 0-1
10.&f4 ‘This move is pretty safe. 10..%a6
11.£xd6 exd6 12.e3 and the position is unclear. Game conclusions: 4.8g5 could work as a surprise
weapon but Black has no worries if he knows his
theory. Of the lesser options 4.e4?! is just trying
z%.a./@
Y
// / &
, , ,, E for a cheap trick, and both 4.b3 and 4.g3 are
feeble.

/ /g8
/ / Ty
7A
Game 46
Malakhatko — Sergey Kasparov
Pavlograd 2000

// / / /
\\\\

1.d4 D6 2.c4 c5 3.d5 b5 4.2

//8%/
/// // '7///
A calm move, but unpleasant to face as Black
- an exception in the sidelines. White’s main idea
in this variation is to play e2-c¢4 immediately.
4...Dabk
An interesting way to fight for equality, because
White must play very carefully. Black wants
 Five Sidelines 113

to show that the queen’s position on ¢2 is not
perfect. Black has several alternatives:
After the natural 4..bxc4d S.e4 d6 6.8xcd
g6 7.5f3 £g7 8.0-0 0-0 9.h3!? (9.83c3"? Lg4
10.£d2! and White keeps the advantage) 9...2a6
10.9a3 Wh6 11.&e3 Dbd7 White is slightly
better.
4..d6 5.e4 gb?! After this sacrifice it is hard
to get more than a long fight for a draw. 6.cxb5
a6 7.0f3 Rg7 8.8c3 0-0 9.a4 axb5 10.2xb5
$ab 11.0-0 ©b4d 12.We2 2a6
compensation for the pawn is very questionable.
4..g6 5.00f3 2g7 6.cxb5 a6 7.e4 0-0 8.2¢3
leads to positions after 4...d6.
5.a3!
The only sensible move.
5.cxb5?! Db4 6.Wxc5 e6
6..0fxd5? 7.8d2 (7.8a3 &b7 8.¥d4 Was
9.¥d2 e5! with an attack) 7...2b7 8.%¥c1 Hc8
9.80¢3 Bxc3 10.bxc3 Zed 11.£3 D27 12.82d1
£g6 13.8b1 d5 Black is also better.
7.d6 HHfds!
Very risky is 7...8xd6? 8.¥xd6 &2t 9.82d1
Hxal 10.b3! b6 11.¥xb6 axb6 12.81c3 Hd5
13.50a4 Hxad 14.bxad Hc3t
16.82e1 D2t 17.8d1 Dxf2+ 18.8xc2 Hxhl
19.£e3 and White has good winning chances.
Hxa2 with good winning chances.
12,96 13.82d2 ¥hot 14.9¢5 Bxfl! 15.8xf1
WxgSt 1602 &xg2 17.8d1 W5t 18.ckcl
g6 19.8d2 2h6 and Black’s initiative is very
dangerous.
5.%)c3 bxcd! (better than 5...b421 6.0d 1 g6 7.e4
d6 8.8f3 £¢7 9.2d3 and the position is slightly
better for White, R. Kempinski — Alterman, Bad
Wiessee 1997) 6.e4 & b4 7.8d1 Was 8.2d2 d6
9.8xc4 £a6 and Black is already slightly better.
13.2f4 Black’s 5.e4? 2 b4 6.¥d1 bxc4 and Black is better.
5...bxc4 6.e4 D7 7.8xc4 6 8.8)c3
8.dxeG?! fxe6 9.5 &b7! 10.&f1 (better than
10.513 &xf3 11.gxf3 d5 12.exf6 dxcd 13.fxg7
fxg7 when Black has the initiative) 10...2g4!
(10...2fd5 11363 £e72! 12.8d3! wich a strong
attack) 11.%e2 Wh4 and the game is unclear.
8...exd5 9.exd5 d6 10.Hge2
In positions like this White has to be careful
about the d5-pawn. However here 10.22f3 was
an interesting choice. 10...£2¢7 (10...2g4 11.9g5
h6 12.8)ged Dxed 13.Wixed T We7 14.59d2 Wxed
15.8%e4 &d7 16.8¢3! and White is slightly
better) 11.0-0 0-0 12.b4 £e4 13.Wd3 &£d7
15.80d2 Dedt 14.EBel &6 with equality.
10...8¢7 11.0-0 0-0 12.b4!
White is slightly better, but that is all.

Ee
8.e4 £b7! 9.exd5 Bc8 10.Wxb4 Bxclt 11.cke2 12...Eb8
£xd5
According to grandmaster Ermenkov the e
T
position is unclear.

//
/&%2% )
wL/E
%%gé 2’
Z A B ///

‘%f Y0, 0 ,,,,,,,,,,, >ot

/g/ fiwfi
N //,’enE A 5
12.5f3 13.bxc52!
Better than 12.53h3 Wf6! 13.£4 &c4t 14.50e3 This releases the tension unnecessarily, so
&xf1 15.b6 axb6 16.¥xb6 £xd6 17.¥xd6 13.2b1! §g4! was betrer. This plan is worth
Wxb2 18.90d2 Wc3t 19.6f2 Hxal 20.Exfl
114 The Benko Gambit

remembering, as it is useful in some Benoni Many moves and plans are good enough for
seructures. (’'m 100% sure that some pour souls equality here, but this plan is my favourite. 4...
reach this kind of position and then have no idea d6 5.¢4 g62! (5...bxc4 leads to 4...bxc4) 6.cxb5
what to do. This is very dangerous: a bad plan is fg7 7.2 6 8.8¢3 exd5 9.exd5 0-0 10.8e2
better than no plan. 13...Ze8 14.He1 ¥d7 15.8f4 White is slightly better, Boensch — T. Pachtz,
cxb4 16.Bxb4 Hxb4 17.axbd We4 18.h3 Wha Halle 1987. For 4..%a5 5.Wc2 bxcd G.e4 see
19.%a4 and White keeps unpleasant pressure.) 5..%a5.
14.8d3 5 15.2¢3 g6 16.8el £f6 Black is doing 5.e4 ¢3!
well. 5...d6 6.8xc4 g6 7.b3 £g7 8.8b20-0 9.2gf3
13...dxc5 14.2d1 £d6 &bd7 10.0-0 Hb6 11.8el Dfd7 12. gxg7 dxe7
Now Black is at least equal. and White is slightly better, Ehlvest— areev
15.8g3 h6 16.8b2 Be8 17.2ab1 £d7 18.2d3 Polanica Zdroj 1997.
fe5 5. %25 6.Wc2 e6 7.8xcd exd5 8.exd5 do
Black can also fight for more than a draw. 9.8gf3 &e7 10.0-0 0-0 11.8el and White is
For example: 18..We7 19.8c4 2g4 with some slightly better according to Michael Adams. ‘The
initiative. English GM is one of the top experts on the
19.8) ce4 £xb2 20.Exb2 Exb2 21.¥xb2 Dcxd5 Benko Gambit.
22.9xc5 Q.g4 6.bxc3 g6 7.c4
-1 7.8 c4"? and now:
Very risky is 7..0xe4?! 8.We2 &g7 9.Wxed
Game conclusions: 4. ¥ c2 is certainly not the most &xc3t 10.2d42 &xal 11.8a5 ¥Wxa5t 12.5xa5
frightening move Black must face in the Benko, &c3t 13.80d1 2xa5 14.h4! and, according
but it is solid and sensible. to Kramnik, White has the initiative here.
(this is much better than 14.d62! &c6 15.8£b5
Game 47 0-0 16.8xc6 dxc6 17.dxe7 He8 18.¥xc6 &f5
Grachev — Tregubov 19.¥xc5 £d8! when Black has more than full
Moscow 2004 compensation for the queen) For example: 14...
h5 15.2h3
1.d4 56 2.c4 5 3.d5 b5 4.0d2 7..d6 8.8d3 £g7 9.0f3 0-0 10.0-0 Dbd7
11.2d2 &b6 12.6xb6 (12.5a52 4! 13.8c2
¥¢7 Black is better, Kramnik — Leko, Dortmund
1998) 12...axb6 13.a4 is unclear - Leko.
7...d6 8.2d3 £g7 9.Eb1 0-0 10.2gf3 5!

/Z, , , Z

4
///////x /m i
Aok 0
/7///
////////////

////@é/////
&7
“

This is a very dangerous move. .. for White.

4,..bxc4! /z,/ &
 Five Sidelines 115

11.0-0 19.8c3!
11.59b35h512.0-0 D4 13.8xf4 exfd 14.9d2 Now Black has a problem with the a5-pawn.
g5 15.g3 £g4 16.5g2 Dd7 17.gxf4 &xf3t 19...f52!
18.xf3 W6 19.f5 §eST 20.80g2 g4 21.8e2 Black should confess his sins and hope for
8h6 22.¥c3 &4 with excellent compensation, gentle punishment. 19..%d8 20.2b3 &xa4d
Korchnoi — Khalifman, Yalta (rapid) 1995. 21.8e1! White is ar least slightly better (instead
11...2bd7 12.8el Hh5 13.8c22! after 21.8xa5 @Ob6 22.2xb6 Exb6 23.Hbal
It is perhaps better to keep an eye on the Black holds the position).
c4-pawn. After 13.&2f1 2Db6 White can 20.exf5!
play the active 14.20b3 with the idea &a5. 20.8ab3 Hxad 21.8al Wd8 22.exf5 Qbo!
14...£5 15.8g5 ¥f6 with an unclear game (instead (22...gxf5? 23 Eb8 ¥Wxb8 24.8xb8 Exb8 25.8xa4
15...h6 16.9)e6 £xe6 17.dxe6 fxed 18.2d2 gives &xa4 26.%xa4 and White wins) 23.fxg6 hxgb
White the initiative). 24.8xg6 Df4 25802 a4 26.Ba3 Bf7! 27.8c3
13...9b6 14.2e3 £d7 Hg7 28.g3 Dh3t 29.89g2 &g4 and Black has
Better seems 14...20f4!? 15.a4 a5 16.g3 ©h3t compensation for the pawn.
17.%2¢2 &g4 18.Wel £5 with an initiative. 20...gxf5 21.8xa5?
15.a4 21.2ab3! Hxad 22.8a1 Qif4 (22...e4 23.59h4
Otherwise Black could have stopped this move £g5 24.5b8 ¥xb8 25.Exb8 Bxb8 26.8)xf5 £xf5
by Wes. 27.%xh5 and White wins) 23.8b7 White is
15...a5 16.Beb3 a6 much better.
Risky was 16..50xa4?! 17.8f1 5 18.83d2! 21...5xd5 22.cxd5 ExaS 23.Bb6 e4 24.8)c4
(18.exf5 e4! and the position is unclear) 18... Ba7 25.0xd6?
4 19.5f3 &6 20.2d2 and che black knight is Better was 25.2fd2 & £6 26.2xd6 Wes 27.50f1
unemployed on a4 and in long-term danger. $£¢7 28.8b3! and Black has some compensation
17.2a3 &h6! for the material, but not enough.
Black wants to exchange his “bad” bishop. 25..%d8 26.8b7 Exb7 27.2xb7 ¥c7 28.d62
18.8b2 This is unnecessarily risky. Also bad is
28.9e527 WixeS 29.%xh5 Wb2 30.8g3t 2g7

M Ed 31.2b3 c4 32.8xc4 Wbt 33.8f1

Black wins. However 28.8el!1? Qf4 29.8b3!
Wxb7 and

. %371/ £xa4 30.d6 Weo 31.8¢37 .@g7 32.9xa4 Wxb7

ST 33.d7 looks good for White.

.RLA
shs
A
ta
,,,,,,,,,,,,,,,,,, 28...Wxb7 29.2e5 D6 30.8g31 Dg4t?
30...8g7 31.8b37 Hh8 32.271 Bxf7 33.8x{f7

Y $xa4 34.8b3 £d7 and Black is just better.

2o . Now White has his glorious moment to make

this comedy of errors something to laugh about
for him. This all depends on him seeing a rather
atcractive tactical variation. However, in missing

YW &
,,,,,,,,,,,,,,,,,,,,,,,

this he gives the initiative back to his opponent,

who is quite happy to get a second (third or
fourth) chance to take the full point.
18... W82
18...5\f4 19.2¢3 5 looks better and Black has
an easy game.
116 The Benko Gambit

E ,,,,,,,,,,,,,,,,,

4 //x{? {?x
?y///flf _
/ Yy / v,
/ //x/ A //4 /? >
o 2 /é /é o/////?/
///é///& e/
/fi?é@/ /////////4
e B ,,,,,,,,,,,

31.h32 4...8b7!
The main idea of 4.8f3 is the following
This seems to be rather naive from such a good
player. The winning line was: variation:
31.8b31 g7 32.Wal Bf6 33.h3 &4 34.hxgd 4...g6?! 5.cxb5 a6 (5..8¢7 6.9¢3 a6 7.e4 and
&xg3 35.0xd7 ¥xd7 36.g5 Wxd6 37.¥xf6T Black loses a move compared to the variation
Wxf6 38.gxf0 Hxf6 39.fxg3 and White wins. 4.cxb5 a6 5.e3) G.ed! (weaker is 6.8c3 axb5
31..8f4 32.8b31 g7 33.2d5 7.d6Y &b7 8.e3 &g7 9.2xb5 0-0 10.0-0 €6
It is too late to repent. Also after 33.%al 11.24 ¥b6 when Black has some compensation,
8xg3 34. Hxd7t Hf6 35.8xc5 Wb6 36.8e6T Kuzmin — Alterman, Berlin 1991) 6..8xe4
b7 37.5d4t g6 38.fxg3 Wxd6 39.0e6 He8 7 W2 £5 8.6 bd2 Wa5 9.b4!? ¥Wxb4 10.8b1 with
40.5f47 $Bh6 there is nothing that can prevent the initiative, for example: 10...¥c3 11.¥d1!
Black from claiming the full point. 4...b4?! 5.a3 g6 6.axb4 cxb4 7.8e3 W7 8.8d4
33...%b2 34.Bxgit fxg4 35. Dxd7 €3 36.Dxf8 and White is slightly better.
2 37.Wel Wcl 38.De61 £h6 39.¥xcl &xcl 4..bxc4 5.0¢3 d6 6.e4 g6 7.8xc4 Black reaches
40.d7 e1=¥t 41.h2 ¥xf2 42.hxgs Wh4t a position from the 5.b6 variation (it does not
0-1 matter that the pawn is still on a7), but with one
move less. Here we can look at two options:
Game conclusions: | have a low opinion of 4.2d2, 7.. &gl 8.h3 2xF3 9.WxF3 Dbd7 10.854 Le7
but perhaps 7.8c4!? keeps the position unclear. 11.0-0 0—0 12.8fel and White is slightly better.
7..8g 8.5 dxe5 9.0xe5 0-0 10.0-0
Game 48 $a6 (there is an interesting tactical motif
1. Sokolov — Bareev here: 10..0fd7 11.0xf7! Bxf7 (11..8xf7?
Sarajevo 2000 12.d6+ &e8 13.0b5 and White wins)
12.d6 &b6 13.8xf7T dxf7 14¥3+ &f6
1.d4 966 2.c4 ¢5 3.d5 b5 4.9f3 15.8e4 215 16.9xf6 exf6 17.2f4 ¥d7 and the
This is another fairly unthreatening move: position is unclear) 1 1.8g5 fxc4 12.8\xc4 Dbd7
White has no real chances to get an advantage 13.%f3 White is slightly better.
here. However you should not disregard this 5.a4
move entirely. Black must be careful to choose 5.56d2 bxcd! G.c4 e6 7.dxe6 (7.8xc4? exd5
the right order of moves. 8.exd5 We7+ and Black takes a pawn) 7...dxe6!
(7..fxe6? 8.¢5 Dd5 9.Dxcd D6 10.4¢3 &Hd4
 Five Sidelines 117

11.5e4 and White is much better) 8.8c3 &c6

9.8xc4 £e7 10.0-0 0-0 11.f4 ¥d7 12.e5 &d5 T
13.8\de4 ©d4 14.2d3 Had8 The position is XA ,,,,,,,,,,,,,,,,,

/xy VY&
///

unclear, Jelen — Mencinger, Ljubljana 1987.

Z
5.0¢3? b4 6.Da4 €6 7.8g5 d6 8.4 Le7 and, -

because of the weak position of the knight on a4,

Black has at least equality.
T, ,///%7 /@/ .
5.e3?! e6! 6.dxe6 fxe6 7.cxb5 d5 According to
Paul Keres this position is already slightly better O ANA
- %y 7 /
for Black.
5% bxcd G.ed e6 7.dxe6 dxe6 with ., 7 8/’/4
D // ///
equality.
5..%as511
# ////////////////
5 7/
This is a useful check. Now White has some
problems with the coordination of his pieces.
Of course Black is OK after 5...b4 6.2bd2 7,86
g6 7.4 d6 8.82d3 £g7 9.0-0 0-0 10.8e1 Hbd7 The queen on a5 gives Black chances to castle
111 5 12.g3 a5 with equality, Vaisser — long and attack on the kingside. Therefore more
Tseshkovsky, Tashkent 1987. elastic is 7...d6 8.e4 ©Hbd7 9.2d3 h6!? 10.0-0
5...bxc4 6.9¢3 e6!? (otherwise the bishop on b7 (perhaps better is 10.b3 with the idea of castling
is not useful) 7.4 exd5 (7...5xe4?! Unfortunately long) 10...g5 with an unclear position.
for Black this does not work. 8.&Yxe4 exd5 9.4 ¢3! 8.e4 d6 9.2d3 £g7 10.0-0 0-0
d4 10.8xc4 dxc3 11.8xf7+ &xf7 12.9b31 Now Black does not have time. For example:
c4 13.8e5% @gS 14.¥xb7 &Hc6 15.Dxc6 10...8bd7 11.2f4 with pressure.
We8T 16.8e3 oxb2 17.¥xb2 dxc6 18.0-0 11.£f4 ¥c7 12.9bd2 Dbd7 13.h3 e5!
White is much berter, Rubinstein — Spielmann, You should be careful with moves like this:
Vienna 1922.) 8.exd5 d6 9.&xc4 £e7 10.0-0 experience has shown that they are either very
0-0 11.2el and Whicte is slightly better. good or very bad. Now this is good because
6.82d2 White cannot play on the queenside. This
6.5bd2? bxc4d 7.e4 Dxed 8.8xc4 and there is move prevents the breakthrough e4-e5 and also
no compensation here. shortens the front.
6.%d2 b4 should also be in Black’s interests. 14.8¢3
6...b4 7.%c2 14.dxe6? is suicidal, because the b7-bishop
7.8¢5 d6 8.80bd2 Hbd7 9.¢4 ho6 10.8xf6 exf6 begins playing! 14...fxe6 and Black is already
11.g3 g6 12.8£h3 2¢7 13.0-0 h5 and the game is better.
unclear, Razuvaev — Mainka, Prague 1992. 14...2ae8 15.2h2 28
7.g3 d6 8.2g2 Abd7 9.0-0 g6 10.%c1 ho All the pieces go to the kingside.
11.5h4 &g7 12.e4 0-0-0 The position is 16.8fel Hh5 17.¥d1
unclear, Matthias — Mainka, Lippstad 1993. This Or 17.8e2 D4 18.2g4 h5! with unclear play.
is a classic position for 5...%a5+. Black has good 17..20f4
attacking chances on kingside, whereas Whire
should play in the centre (e4-¢5).
118 The Benko Gambit

Instead 4...b4"2 5.00d2 g6 6.¢4 d6 7.2gf3 o7

8.g3 ¢6 9.2h3 exd5 10.8xc8 Wxc8 1l.cxd5?
M/g/z%77 (11.exd5 0-0 12.0-0 with equality) 11...0-0

w A
/////////// 45k 12.0-0 c4 13.¥¢2 c3 14.bxc3 bxe3 15.8b3 We4
Black is much better, W. Schmidt — Kasparov,
/my/x/ (ol) Dubai 1986.

//Mfiww%
5.5\c3 d6 6.e4 £a6
6..g6 7.8xcd fg7 8.50f3 0-0 9.0-0 £a6

@&% g % 10.6)d2 ©fd7 and the position is more or less

/// /// &

equal.

7,,,,,/,,,,,/,7 /] %,/,,} ,,,,,

/// ,,,,,, \2@/ »
,,,,,,,,,,,,,,,,,,,,,, -
% %%,//é/ /% /%
4/
18.8c2
Of course the position after 18.8xf4? exf4 is /
/
simply better for Black.
18...50f6 19.20df1 h6 20.8g3 Dh7 21.¥d2 h5
4// _
22.8ab1 h4 23.2gf1 £h6
23..We7t 24.8d1 &g5 and Black has easy
play.
24.563%d825.01h2 Hg526.b3 g7 27.%h1
Hxf3 28.0)xf3 Eh8 29.8g1 Hh5 30.Dg5 £6
In my opinion Black is still slightly better.
7 .£4!
Va1
7.0f3? is a typical mistake of club players.
7..56bd7 8.8d2 & e5 and Black is better.
Game conclusions: 4.3 is a solid, unthreatening
7...g0!
move, but Black must be accurate with his move
7..8bd72 8.20f3 Wb8 (8...g6 9.€5 and White
order to avoid being tricked into a different
is better) 9.%e2 ¥Wh7 10.g3 €5 11.dxe6 fxe6
variation a tempo down. The right answer is 12.£h3! and White is much better according to
4..2b7.
Informant.
7..e621 8.5f3 exd5 9.e5! d4 10.exf6 dxc3
Game 49
11.9d5 ¥c7 12.Wedt? (12.8xc4 £b7 13.Wd3
Ward — Kinsman
gxf6 14.0-0 Hc6 15.9f5 Lg7 16.bxc3 fe7
Wrexham 1998
17.%h5 0—0—0 18.Wxf7 and White is much
berter according to GM lzeta) 12...5d8 13.8xc4
1.d4 96 2.c4 ¢5 3.d5 b5 4.24?!
&b7 14.Wd3 White has a clear advantage.
This move is quite safe for Black, but quite
8.3
dangerous for White! The point is that if Black
8.c5 H1fd7 9.e6 fxe6 10.dxe6 Df6 11.Wf3
play quietly then White has the same structure
ds 12.8e3 £ c6! and Black is doing well (not
as in the 4.cxb5 a6 5.b6 variation, but White has
12..d62113.8d1 Wxe6 14.5xd5 Dxd5 15.8xd5
a weakness on b4. What is more, Black can play
4b7 16.&2xc4 and White is clearly better).
more actively.
8..8g7 9.5
4...bxc4
9.5d2 Was 10.Ea3 Hbd7 11.¥£3 0-0 12.8xc4
This is the most principled.
 Five Sidelines 119

Bxcd 13.9xc4 W4 14.We2 Hb6 15.9d2 e6 and This is an important tabiya position for this
Black is at least equal, Fornal — Pinski, Poland variation. Only God knows who is better.
1995. 15.2b52
9...dxe5 A very quiet and very weak move. The old rule
Also interesting is 9...2)fd7 10.%e2! (10.¢6 fxe6 says: if you can attack then you must attack. In
11.0g5 ©f8 12.dxc6 W8 and Black is better W. chess, like in war, cowardice is punished.
Greten — Zellner, Germany 1989) 10...0-0 11.e6 15.5¢57 Bh6 (15..Bf8? 16.5xh7 Hges!
with an unclear position. For example: 11...fxe6 17.0-0-0 (17.9xf8? Hd3t 18.8d1 &Hxb2t
12.¥xe61 ®h8 13.5g5 Was! (after 13..¥e8 19.¢c1 &d3+ 20.50d1 ¥b3+ and Black wins)
14.%h3 White has some initiative) 14.¥xe7 h6 17..2h8 18.8¢5 He8 with an unclear position)
15.5ge4 @6! and we have a mess. 1687t B8 17.8d1 ¥xb2 18.8d5 &b7
10.fxe5 Dg4 11.8f4 19.8xc4 and now:
11.e6 fxe6 12.80g5 exdS! 13.Wixgd £.c8 14.8d1 a) 19..9g4? 20.0-0! &f6 (20...&2d4T 21.&h1
4d4 15.8b5 Hc6 16.5xd4 Hixd4 17.2e2 0-0 2T 22.8xf2 &xf2 23.8b5 and White wins)
and Black has full compensation for the piece.
11.%e2 &Hd7 12.e6?! (for 12.84 see the game)
12..5de5 13.exf7t Dxf7 14.%e6 W8 15.8¢e2 1,/fm
Wxe6 16.dxe6 ©d6 and Black is better.
/ i
11...0d7 12.¥e2
Ty
&//g/ B/
12.e6 and now:
12...fxe6!?? ‘This is only for men with very
7 / /
strong nerves. 13.8g5! b6 14.8xe6 Le5!
15.%d2 (after 15.%xg4 Wxb2 Black is already
better) 15..8df6 (15..&xf4 16.9xf4 0-0
17.862 Dge5 18.0-0 &f6 19.8e6 and White
is at least a little better) 16.a5 ¥b8 17.g3 and
21.5xf7!1 &xe7 22.dxe7 Incredible! Black is
White has an edge.
helpless. 22...5de5 (or 22..Eg8 23.5d6T &bs
12...40de5! is stronger. 13.h3 @xf3+ 14.%xf3
24 8b1 Wd4t 25.82e3 and White wins) 23.2xh8
06 15.exf7t &xf7 16.8e2 Ef8 17.0-0 g8
&d7 24.4b5T 26 25.8b4T and Whire wins.
18.%e3 Wh6 with unclear play.
b) 19..Wxg212 20.8fl 2f5 21.¥xf7 is
12..%b8! 13.d6 exd6 14.exd6t d8
unclear.

e 7 o 19...4c3t 20.0f1 &d4 21.We2 Wxe2t

22 Bxe2 £3b6 23.5e7T 2d8 and the excitement
continues! The position is unclear.
/
15.0d582b7 16.We7+ (16.9c7 £xb2 17 2d1
Bb8 (17..2c312 18.2d2 £xd2t 19.¥xd2 EbS
20.82e2 16 21.0-0 and White is better) 18.We77
®c8 19.8e2 Wbat 20.50d2 &b7! 21.0-0!
(21.8xg4 15 22.213 &xf3 23.gx3 &f6 24 W6
€3 25.8a6 cxd2+ 26.8xd2 Wb3 27.HxcS WxeGt
28.5xe6 He8 29.Ec11 &b7 30.2c71 &a8 31.0-
0 ¥xe6 32.8xd7 8b7 33.8xb7 $xb7 and Black
is better) 21..&d4t 22.%h1 D21 23.Exf2
8xf2 24.8e5 is unclear) 16...%c8 17.8d1 Wxb2
18.8¢2 &b7 19.&c1 £c371 20.%61 W2 21.8xc3
Wxc3 22.Wxf7 Bf8 with unclear play.
120 The Benko Gambit

15...Ee8! 19..8e3 20.2d3 c4 21.8202 Dgit 22.%g3

After 15..8xb5 16.We7+ 8 17.axb5 ¥ixb5 Dge5 23.8e4
Black is probably better, but White has chances It is hard to give good advice for White.
for active play. 23..9xf3 24.8xf3 &xb2 25.8adl Le5t
16.%e7% 26.5bh3 &xb5 27.axb5 Exb5 28.2c6 Eb3t
The only move. 29.g3 Db6 30.Ehel £6 31.Ed2 a5 32.8a2 R d6
16..8xe7t 17.dxe7t txe7 18.2xb8 Exb8 33.8e8 £c3 34.8d11 £c5 35.8e2 Le5 36.5a2
19.8xc4 &b4 37.8f7 a4 38.2g8 h6 39.87 g5 40.2g4
a3 41.%2h5 Bb2 42.8xb2t £xb2 43.8d6 Ha4

E %/ // 44.8¢6 ¢3 45.2xh6 D5 46.8b6T a5 47.8c6

A A 7770

®b5 48.8xf6 c2 49.Bf1 c1-¥ 50.8xcl &xcl

i ae A4 51.%2g6 b4 52.8a2 B3
g// v 7 .t/ 0-1

//@3” // 7 Game
Black.
conclusions: 4.a4?! poses no threat to

S m Bar
Chapter conclusions: The lines considered in this
chapter do not threaten the soundness of the
Benko Gambit, but they have varying levels of
fi////ffé’///z == ///
respectability. 4.2f3 and especially 4.%c2
sensible moves, if racher quier. 4.8g5 and 4.a4
are

are less reliable, especially if Black is not caught

Black has a pawn for the exchange and a raging
attack. Meanwhile White is playing without the
h1-rook. This is more then enough compensation
for the material. Instead 19.h3 £xb2 20.5b1 ¢3
and Black wins.
by surprise. The other 4" move alternatives
mentioned in the notes all cause Black no
theoretical problems, but Benko players should
remember the trick in the 4.e4 line.
 Index of Variations 121

Index of variations

The Benko Gambit Accepted I

1.d4 D6 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.bxa6 g6 12...@g4 Game 14, 46
5...6?! Game 9, 31 12...8a6? 13.%¢c2 Game 15, 48
5...82xa6 Game 9, 37 13.e5!? Game 15, 48
6.2c3 8xa6 12..¥c7 Game 15, 48

g@% %
12..%b6 Game 16, 49
12.h3
12...%b81 Game 17, 51
12...5b61 Game 17, 51
12.. %35 Game 17, 517
12...826!1? 13. Bel Game 18, 52
,/////; 'y
> 12..%b6 Game 19, 54

AR AW r? y i
The Benko Gambit Accepted 111

1.d4 Df6 2.c4 c5 3.d5 bS 4.cxb5 a6 5.bxab g6

7.e4 £xf1 8.%0xfl d6 9.g4?! Game 9, 32 6.2¢3
742! Game 10, 33 0.g3 d6 7.b3 game 20, 57
7.3 d6 8.60d2 Game 11, 36 6...8xa6 7.3
7.g3 d6 8.8g2 £g7 9.9h3 game 25, 66
The Benko Gambit Accepted 11 7...d6 8.g3 £g7 9.8g2
9.82h3 game 21, 58
1.d4 916 2.c4 c5 3.d5 b5 4.cxb5 a6 5.bxa6 g6 9...20bd7
6.9c3 £xa6 7.4 8xf1 8.xf1 d6 9.g3
9.3 £g7 10.h3 Game 13, 44 E%/ Wes
9..£g7 10.52g2 0-0 11.2f3 Ak 27Y.
II.QgCZ!? Game
11..2bd7
12, 41
.é/% /t 'y
2w
/ 7?
/ &/
,,, |
10.Eb1
10...b6! game 23, 62
10...0-0
11.%c2 game 24, 64
11.0-0 Was game 22, 60
12.8el! 10.0-0 Db6! 11.Eel 0-0
12.. W25 Game 14, 46 12.84 game 26, 67
12...2b6!? Game 14, 46 12.e4 Game 27, 69
122 The Benko Gambit

A &4 E
. 4 V, & =7
The Positional 5.b6 78 /// =

1.d4 D6 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.b6

i L
5...e6
5...a52! Game 28, 71
6.8¢3 Hxd5
6...2b7"7 Game 29, 73
7.8 xd5 exds 8.¥xd5 D6 9.013 EbS
9...8e7" Game 29, 74
9...8b72! Game 29, 74
B Lws NE
8. £d2 &b7 Game 36, 86
8.@ge2

sim Z2
a
8...00xd5 Game 34, 84
8...2xd5 Game 35, 85

5.3
1.d4 Df6 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.f3 6!
5...d6?! Game 38, 91
,,,,,,
5...axb5 Game 38, 91
5...g6 Game 39, 94
10.g3?! Game 29, 74
6.e4 exd5
lO.igS Game 29, 74
6...2b7 Game 40, 97
10.8e5 Game 29, 74
6...c4?! Game 40, 97
10.£d2"? Game 29, 74
7.5
10.e4 Game 30, 76
7.exd5?! Game 41, 99
5..%xb6 6.2c3 d6 7.e4 gb
7..¥e7 8.We2 @gs9.2c3 £b7 10.2h3 c4
8.a4"? Game 31, 77
8.3 Game 32, 79
5...d6 Game 33, 80
‘// 1/
The Ambitious 5.¢3

1.d4 D66 2.c4 c5 3.d5 b5 4.cxb5 a6 5.e3 axb5 Wyx/ / %

5...e6?! Game 34, 83
/
5...2b7 Game 34, 83
5...g0% Game 37, 87
6.2xb5 Was5+ 7.8¢3 £b7
fi“@ ”g%fi
11.8e3
11.2d2 game 41, 100
11.5f4 game 41, 100
11.9122! game 41, 100
11...axb5
12.6xb5 game 41, 99
12.0-0-0 ¥b4 13.8xd5 game 42, 102
 Index of Variations 123

Zaitsev Variation 4.%c2 game 46, 112

4.8d2 game 47, 114
1.d4 56 2.c4 ¢5 3.d5 b5 4.cxb5 a6 5.0c3 4.Df3 game 48, 116
axb5s 4.a4 game 49, 118
5..%a5 game 43, 105
6.¢4 b4
6..Wa5?! game 43, 105
7.8b5 d6
8.2\f3 game 43, 106
8.8c4 game 43, 106
8.8f4
8...0bd7? game 43, 106

/ // //vz,x‘
a1y
A
/ f g //é

0 G
7 é’/
V/
%/V//%W'
%7///%///7

B/ WELNE
8...g5!
9.8xg5? game 43, 105
9.8e3 game 44, 106

Five Sidelines

1.d4 916 2.c4 c5 3.d5 b5

B o iveded X
7 i/xfimwzéx
7, ) &
/awg//%/
,/fi% % %
/ %/// :
%%/%%%
,,,,,,,,,,,,,,,,,

4.895 game 45, 111

4.¢3 game 45 11]
4.e4?! game 45, 111
4.b3 game 45, 111
 1 24 The Benko Gambit

Index of full games Alphabetic Index of full games

1 . Ehlvest — Fedorowicz 8 Va. Akobian — Pinski 71

2 . Kraidman — Halkias 12 Andruet — Fedorowicz 33
3 . Grischuk ~ Tregubov 17 Arencibia — Ivanchuk 104
4 . Sjddahl — Adams 19 Asanov — Rashkovsky 26
5 . V. Perez — Magem Badals 21 Azmaiparashvili — Conquest 91
6. Herraiz Lopez — lllescas Cordoba 23 Bacrot — Ponomariov 67
7 . 1. Farago — Knaak 24 Bareev — Kasparov 36
8 . Asanov — Rashkovsky 26 Bystrov — Leconte 52
9 . Radziewicz — Pinski 31 Ehlvest — Fedorowicz 8
1 0. Andruet — Fedorowicz 33 L. Farago — Knaak 24
1 1. Bareev — Kasparov 36 Gleizerov — Tregubov 76
1 2. Korchnoi — Borik 42 Grachev ~ Tregubov 114
1 3. Haba — Nun 44 Graf — Felgaer 86
1 4. Gyimesi — Sermek 46 Grischuk — Tregubov 17
1 5. Gyimesi — Stamenkov 48 Grivas — Khalifman 111
1 6. Hjartarson — Alburt 49 Gyimesi — Sermek 46
1 7. Titenko — Kaidanov 51 Gyimesi — Stamenkov 48
1 8. Bystrov — Leconte 52 Haba — Nun 44
1 9. Piket — Topalov 54 Herraiz Lopez — lllescas Cordoba 23
2 0. Sosnicki — Pinski 57 Hjartarson — Alburt 49
2 1. Oleshkevich — Kotenko 59 Hiibner — Heinemann 69
2 2. Kramnik — Topalov 60 S. Ivanov — Khalifman 99
2 3. P Varga— M. Pap 62 S. lvanov — Sznapik 94
2 4. van der Sterren — Adams 64 Kamsky — Illescas Cordoba 77
2 5. Korchnoi — Baklan 66 Korchnoi — Baklan 66
2 6. Bacrot — Ponomariov 67 Korchnoi — Borik 42
2 7. Hiibner — Heinemann 69 Kraidman — Halkias 12
2 8. Va. Akobian — Pinski 71 Kramnik — Topalov 60
2 9. Vladimirov — V. Mikhalevski 73 Malakhatko — S. Kasparov 112
3 0. Gleizerov — Tregubov 76 Norris — B. Lalic 102
3 1. Kamsky — Illescas Cordoba 77 Oleshkevich — Kotenko 59
3 2. Tukmakov — Plachetka 79 V. Perez — Magem Badals 21
3 3. Pleger — Benko 80 Petursson — Fedorowicz 87
3 4. Szczechowicz — Pinski 83 Pfeger — Benko 80
3 5. Shirov — Adams 85 Piket — Topalov 54
3 6. Graf — Felgaer 86 Potapov — Degraeve 108
3 7. Petursson — Fedorowicz 87 Radziewicz — Pinski 31
3 8. Azmaiparashvili — Conquest 91 Shirov — Adams 85
3 9. S. Ivanov — Sznapik 94 Sjédahl — Adams 19
4 0. Wojtaszek — Harikrishna 96 1. Sokolov — Bareev 116
4 1. S. Ivanov — Khalifman 99 Sosnicki — Pinski 57
4 2. Nortris — B. Lalic 102 Szczechowicz — Pinski 83
4 3. Arencibia — Ivanchuk 104 Titenko — Kaidanov 51
4 4. Potapov ~ Degraeve 108 Tukmakov — Plachetka 79
4 5. Grivas — Khalifman 111 van der Sterren — Adams 64
4 6. Malakhatko — S. Kasparov 112 Ward — Kinsman 118
4 7. Grachev — Tregubov 114 P. Varga — M. Pap 62
4 8. 1. Sokolov — Bareev 116 Vladimirov — V. Mikhalevski 73
4 9. Ward — Kinsman 118 Wojtaszek — Harikrishna 96