Professional Documents
Culture Documents
The Kaufman Repertoire For Black Kaufman PDF
The Kaufman Repertoire For Black Kaufman PDF
Eight years after his acclaimed The Cl!ess Advantage in Black and White,
grandmaster Larry Kaufman is back with a completely new repertoire
book, covering the entire scope of chess openings for both White and
Black, in one tome .
Using the latest versions of top engines like Komodo and Houdini, the
former Senior World Champion and computer expert has refined his
analysis of ready-to-go and easy-to-digest lines almost to perfection.
His main new co"nviction is that l.d4 gives White better chances of an
advantage than l.e4, and he h�s changed his recommendations accord
ingly.
Larry Kaufman has based his repertoire on sound, practical lines that do
not outdate rapidly and are suitable for masters while accessibl-e for
amateurs. He regularly and successfully uses the openings he recom
mends himsel£
The Kaufman Repertoire for Black and Wl1ite contains many improvements
on existing opening theory and offers a good balance between narra
tive and variations.
"The author has done a very good job at presenting a playable and in
teresting repertoire for both Black and White."
Carsten Hansen, ChessCafe
"Simply the best comprehensive repertoire book that I have ever seen."
John Watson, The Week in Chess
Larry Kaufman
Black Introduction . • • • • s
Magnus Carlsen's defenses!
Chapter 1 • • • • • . • • . • . 7
Unusual Opening Moves
Chapter 2 • • • • • • 17
English Opening
Chapter 3 • • • . . • 25
Queen's Indian versus Reti
Chapter 4 • . . . 37
Anti-Griinfeld
Chapter S • • • • • • • • • • 53
Queen's Pawn Openings
Chapter 6 . • . • 65
Neo-Griinfeld
Chapter 7 .................... . 75
Griinfeld Defense- Non-Exchange lines
Chapter 8 . • • • . • • 99
Griinfeld Exchange
Chapter 9 . • . • . • . 127
Center Game and Ponziani
Chapter 10 • • . . • • • • . . • 133
Bishop's Opening and Vienna
Chapter 11 141
Gambits
Chapter 13 169
Italian Game
3
Chapter 14 • . . • • • 179
Spanish Offshoots
Chapter IS 199
Breyer Defense
4
Black Introduction
Against I.d4, I switched to the Griinfeld, both in my own play and for this book. The
Semi-Slav is still a good alternative, but there are some problems in the 5.�g5 lines and
also a big problem of reaching the Semi-Slav without allowing unpleasant options like
the Catalan or the Slav Exchange.The Griinfeld is in much better shape than it was eight
years ago, and it seems that finding an advantage against it is an extremely challenging
task. My own experience is that since I have been aiming for the Griinfeld and for the
Breyer, my results have improved noticeably. Here too we are following the recent pref
erence of Carlsen.
I was pleasantly surprised to learn that the English Opening. l.c4, is not much of a
problem for tl1e Griinfeld player, contrary to my opinion eight years ago. I show how
the move l...g6! eitl1er transposes to the Griinfeld or leads to near-equality in all cases.
As for I.tLlf3 , we can play the Griinfeld anyway, covered in the Anti-Griinfeld chapter.
In the Reti chapter I give some alternatives for Black. So my overall conclusion is that
contrary to my belief eight years ago. the Griinfeld does not have major move-order
problems.
The section on l.e4 e5 is the only major part of this book that has much in common
with my previous work. Even here, most of my recommendations have changed. I kept
the same main lines against the King's Gambit and Goring Gambit, but against almost
all other white tries I have made major changes. Many players are reluctant to meet l. e4
5
The Kaufman Repertoire for Black
with J ..eS because there are so many ways White can vary before we get to play our
.
own line ( in this case the Breyer, on move 9 of the Spanish). This is true, but almost all
of them are inferior. In fact I would say that only the Italian, the Spanish with 6.d3, and
the Spanish with 9.d4 lead to positions ( with best play) where I would rather play
White than Black, and just marginally so. Quite a few of the white options that I actu
ally face in tournaments fail even to equalize the game. When people try to take me out
of book early, I am usually quite content! In this book I don't take the attitude that Black
is always happy with a draw; once White makes one or two second-rate moves I start to
look for a black advantage.
6
Chapter 1
K�. .t if • .t � K
•• A •·••••
Black can easily equalize in many ways, such as l . ..dS 2.i.b2 i.g4, but I recommend in
Game 1.1 playing for the advantage with the pawn exchange l. .es 2-ltb2 hb4 3be5
.
f!J£6. Black ends up two tempi ahead. which easily trumps the slight profit White made from
the pawn exchange. Simply put, if three tempi equal a pawn, surely two are worth more than
the modest difference in value between these pawns. The important point to remember is not
to play .. .lt:)c6 until after ...0-0, ... d7 -dS, and ... c7 -cS. Black's ultimate aim is ...d5-d4.
Next we come to the similar but far more respectable Larsen's Opening, I .b3.
K � .tif• .t � tK
••••• , ••
7
The Kaufman Repertoire for Black
Aside from the late great Danish grandmaster, this was also played by Bobby Fischer
and recently by the American star Hikaru Nakamura, who especially likes to play it in
blitz chess. Its main drawback is that the knight on b I becomes a problem piece, as its
natural development to c3 blocks the bishop. For this reason, l.b3 is an excellent first
move when giving knight odds! In normal chess, Black can equalize by playing a re
verse Queen's Indian with l ...dS 2..�b2 c5 3.e3 a6 4.t:Df3 t:Dc6, but I prefer to try for a
small advantage by l . ..e S 2.�b2 tD c6 3.e 3 tD f6 4.�b5 �d6 , as recommended in Chess
Advantage, but I now recommend answering the usual (but strange-looking) 5.lDa 3 by
the equally strange-looking S ...lDa S . Black's main plan is ... 0-0, ...l:re8, and .....if8, or in
the 5 ...t:DaS line ...c7-c6, .....ic7, and ...d7-dS. See Game I.2.
Finally we look at Bird's Opening, l .f4, which could also be called a reversed Dutch.
E .t"ir� .t. � E
' .t. ' .t. .... .t. ' '
If White aims for a Leningrad by l tDf6 2 .tDf3 dS 3.g3 I recommend the rare 3 ...c6
..•
4.�g2 'i¥b6! to provoke the undesirable move e2-e3, which permits .....ig4. Black aims
to clear all the minor pieces off except his knight and one white bishop, on the grounds
that knights are better than unpaired bishops on a crowded board. If White goes for a
normal Dutch with 3.e 3 we pin immediately, with ideas of ...t:Dd7, .....txD, and ...e7-eS.
All this is covered in Game 1.3. All in all, I prefer Black's chances by a smidgeon after I.f4.
8
Chapter I - Unusual Opening Moves
lf 3.f4 d6 4.fxe5 dxeS S .heSlLJf6 6.lbf3 wins a pawn) 3S.i..fs lld8 36.a4 g6
0-0 7.e3 lbc6 8.i..b2 l:Ie8 9.i..e2 l:Ixe3N 37.i..e4 hd3 38.�d3 J:Ixd3 39.l:Icc l aS
1 O.c3 l:Ixf3 l l.hf3 i..d6 I2.0-0 lbeS 40.llabI l:I8d6 4 1 .Wh2 l:If6 42.l:Ib2 gS
13.d4 lbeg4, White must play 14.g3, 43.g4 'it>g7 44.llc4 J:Ifd6 4S.'it>g3 lld2
when the knight fork will leave Black a 46..llcc2 .llxc2 47..llxc2 e4 48.llc4 .lle6
pawn up with the better position as well. 49..llc3 'it>f6 SO.f4 gxf4+ S I.'it>xf4 'it>e7
52.g5 hxgS+ 53.Wxg5 'it>d6 54.'it>f5 .lleS+
3. ... lbg8-f6
5Hti>f6 .lldS 56..llc2 .lld3 57.lle2 c4 58.g4
4. c2-c4
c3 59.g5 .lld2 and White resigned.
4.lbf3 0-0 S.e3 dS 6.i..e2 cS 7.0-0 lbc6
1 2. lla 1 -a2 �a5-c7
8.i..b2 i..aS! N, planning ...d5-d4, is better
1 3. �g3xc7 lLld5xc7
than the immediate 8...d4 given in Chess
14. lLlb1 -c3 �c8-f5
Advantage, which is answered by 9.c3.
1 5. l:ta2-a1 lla8-d8
4. ... 0-0 1 6. l:ta1 -c1 lbc6-e5
5. lLlg1 -f3 d7-d5 1 7. lLlf3xe5 'iff6xe5
6. e2-e3 l:l.f8-e8
7. a2-a3 �b4-a5
8. �f1-e2 c7-c5
9. c4xd5 lLlf6xd5
1 0. 0-0 lLlb8-c6
11. �e5-g3
9
The Kaufman Repenoire for Black
Or 3.c4 tlJf6 4.e3 d5 5.cxd5 tlJxd5 6.a3 I really like this move here. It answers the
�d6. threat against e5 , and envisions castling,
....lle8. and ...�f8. In case of �xc6 at any
time, after ... dxc6 the bishop will no lon-
.i .t'if� .i
ger be blocking a pawn and will be well
''' '''
placed on d6.
�.t
�'J 5. tt:lb1 ·a3
3 . ... tt:lg8-f6
4. .tf1 ·b5 .tf8-d6
10
Chapter I - Unusual Opening Moves
eels the threat of tL:\c4 and prepares 14.d3 llad8, Black is already for choice
...c7-c6. The fact that Nakamura against here, with more space and ideas like
Ponomariov refrained from S .tt:la3 in fa ...eS-e4.
vor of S .�xc6 followed by 6.d3 and
9. ... g7-g6
7.tt:ld2 suggests that he feared S.tL:\a3
1 0. 'fVd 1 -f3 c7-c6
�aS.
11. 'fVf3-g3 ttJf6-h5
6. ttJg1 -f3 1 2. �g3-h3 ttJh5-g7
13 . g2-g4?1 �d6-a3
ln case of 6.tL:\c4 tt:lxc4 7.�xc4 0-0
1 4. �b2xa3 'fVe7xa3
8.tLle2 c6 9.tL:\g3 �c7 I 0.0-0 dS l l.�e2
1 5. e3-e4 ruB-dB
l::te8 12.c4 aS 13.a3 i.e6 Black has a nice
pawn center without suffering too much Black already has a likely winning advan
pressure against it. tage. He has the center, while White's
Or 6.�e2 �e7N 7.tt:lf3 e4 8.tt:leS 0-0 pieces are randomly scattered.
9.0-0 a6 IO.d3 l:te8=.
1 6. �h3-f3 'fVa3-e7
6. ... 'fVdB-e7 1 7. ttJh4-g2 d7-d5
7. ttJa3-c4 1 B. e4xd5 e5-e4
1 9. 'fVf3-g3 c6xd5
Or 7.'fVcI 0-0 8.�e2 cS 9.d3 tL:\c6
20. �c4-e2
IO.tt:lbS �b8 ll.c4 dS 12.cxdSN tt:lxdS
13.0-0 �e6 - Black is playing the
Maroczy Bind against the Scheveningen
Sicilian, with colors reversed. In general
the Bind is effective in this situation,
more so than against the Dragon set-up. .t.
7. ... ttJa5xc4 .t. �
B. �b5xc4 0-0 '� 'iY
� ���!:::.Ci:J£3:.
� � .ti
21 . 0-0 b7-b6
22. .llf1 -e1 �cB-b7
23 . ..Q.e2-c4
II
The Kaufman Repertoire for Black
�
l i ?tJ�
�
�- � �
1:[ w
Analysis diagram
31 . tl:lf4-h5??
White resigned after making this blun A) S.d4 e6 6.0-0 cS 7 .lbc3 �d6 8.a4
der, but his position is probably lost any 0-0 9.h3 �xf3 IO.ihf3 a6 - although
way. White has the bishops, his set-up calls for
lbeS , which is impossible. Black has the
initiative on the queenside and the right
YO 8.8 (AOI) Game 1.3 bishop;
D Danielsen,Henrik B) S .b3 hf3 6.�xf3 eS 7.fxe5 lbxeS
• Lafuente,Pablo 8.�b2 �d6 9.0-0 0-0 1 o.lbc3 c6
Odense, 20 I I (7) l l.lbe2 "iVe7 12.lbd4 g6 - White's king
is weakened, e4 may be a home for a
1. f2-f4
knight, and Black will regain the bishop
Bird's Opening, a reversed Dutch once pair since 13.�e2 is too passive;
Black plays ...d7 -dS. In my opinion it's C) 5 .0-0 � 6.�xf3 eS 7.d4 e4
strictly 'for the birds' . 8.�e2 .id6. In such a very closed posi
tion Black's good bishop plus knight may
1 . ... tl:lg8-f6
be nearly equal to White's two bishops,
2. tl:lg1 -f3 d7-d5
and Black has more space.
12
Chapter I - Unusual Opening Moves
like the game move because Black plans 14oo.aSN I S .a3 fS is also good. Either way
to give up the bishop pair with oo.�g4 Black is already for choice.
and ..ixf3, so he will want to exchange
00 •
1 5. d3-d4 tLld7-f6
bishops on the long diagonal to kill
White's bishop pair. I5 oo.fxg4N 16.hxg4 eS 17.dxeS lLlxeS
I 8.�g3 lLlef7 I 9.llaeI llae8 favors Black
6. �c1 -b2 �f8-g7
due to the pressure on the backward e3
7. e2-e3 �c8-g4
pawn.
More accurate is 7 00.0-0 8.0-0 �g4 9.h3
1 6. c2-c3
(9.lLlc3 lLle4 I O.lLla4 �aS favors Black
13
The Kaufman Repertoire for Black
I � 23 . ... tLle4-g5
j. j. j. ·� j. 24. 'i'Vf3-g3 J:%f8xf4
'iV j. � � j. 24...tLle6! is winning.
j. j.
�. �� 25. llc2-f2 lle8-f8
�� �� � 26. llf2xf4 l:tf8xf4
27. Wg1 -g2 l:tf4-f7
� t2J �
� �w 27...liJe6! 28.l:te! llf6 29.�xc7+ ltJxc7
30.c4 l:te6 Black is a clean pawn up in the
1 6 . ... f5xg4?1
ending.
16...c5!N 17.llac l cxd4 18.exd4 e6 19.g5
28. l:tf1 xf7+ 'flic7xf7
tLlfe4 and Black's powerful knight outpost
29. 't\Yg3-e5+ 'il'f7-f6
plus queenside play with ... aS and pres
30. 'il'e5-c7+ Wg7-g8
sure on c3 give him a clear advantage.
31 . Wg2-g3 b7-b5
1 7. h3xg4 tLld6-e4
1 8. tLld2xe4 tLlf6xe4
1 9. J:%f1 -c1 ?I
1 9. ... e7-e5
20. l:tc1-c2 e5xf4
21 . e3xf4 �b6-c7
22. l:ta1 -f1 .l:la8-e8 32. c3-c4?
14
Chapter I - Unusual Opening Moves
IS
Chapter 2
English Opening
I used to regard the English Opening ( l .c4) as a major problem for a would-be
Griinfeld player. The best move is supposed to be l. ..e S, but if I'm not keen to take on
me Sicilian as White with l.e4 how happy can I be playing against it a tempo down?
The Symmetrical Defense l. ..cS is a serious option but probably doesn't give full equal
ity. Simply playing Griinfeld moves doesn't work as after l.c4 lbf6 2.lbc3 g6 3.e4 we
are tricked into playing the King's Indian, while 2...d5 3.cxd5 lLlxdS 4.g3 g6 S.�g2 is
probably a bit favorable to White.
Over the past few years I have come to appreciate that l .c4 is very well met by l . ..g6!
if y ou hope to reach the Griinfeld.
The idea is this: Black meets lbc3 by ...cS , meets e4 by ...eS, and meets d4 by ...lbf6.
Let's look at some possible move-orders after l.c4 g6:
17
The Kaufman Repertoire for Black
This line is not thought to be good for Black, but extensive computer analysis seems to
indicate that Black is fully equal. See Game 2.2.
For the Reti ( I .tLlf3 tLl f6 2.g3 dS 3.�g2 g6 4.c4 dxc4) and the King's Indian Reversed
(4.0-0 �g7 5.d3 0-0 6.tLlbd2 d4!?) see Game 2.3.
Unfortunately this simple solution to the English problem does not work against
I. tbf3 , because after I ...g6 White can play 2.e4!. However, after I.tbf3 the anti-Griinfeld
is OK for Black. For alternative solutions to the I .tbf3 move order, see the Reti chapter.
5 . ... d 7-d6
after tbc3 or g3 has been played.
I have played S ...tLlh6 here, heading to
2. tt:lg1 -f3 g7·g6
wards d4, but I don't recommend it as
3. g2-g3 it.f8·g7
6.h4! is rather strong.
4. �f1 -g2 tLlb8-c6
6. tt:lb1 -c3 it.c8-f5
:i A'ii'* ,..i This system, played repeatedly by Vladi
.l i i .l i ..t i mir Malakhov, makes a lot of sense. Black
.. .l aims to exchange the important bishop
.l on g2.
!J:, 6 ... tbh6 7.d4 cxd4 8...Q.xh6 ..Q.xh6
9. tbxd4 seems to be a bit better for
White, so I have given up·on the ... tbh6
idea.
18
Chapter 2 - English Opening
:i i.�
•• � .. . .a •
.. .. i
•
'ff��
� �Cjjtj�
j
� �� 8
� n�
7. d2-d3 1 5. . .. �g7xb2
In the event of 7.b3 �d7 8.�b2 tLlh6 IS ... b6 directly may have been more pre-
9.d3 �h3 IO.�d2 ..Q.xg2 II.'i.t>xg2 0-0 cise, as Black's king will be perfectly
12.e3 tLlfS 13.tLldS tLleS 14.tLlxeS dxeS= happy on g7. Anyway, chances are even.
the weak white pawn on d3 offsets his su
1 6. .Ub1 xb2 b7-b6
perior bishop.
1 7. d3-d4 c5xd4
Should White play 7 .h3 �d7 8.'i.t>h2
1 8. tt:lf3xd4 llf8-d8
t:L)f6 9.d3 0-0 I 0..2.e3 a6 I I.'+Wd2 llab8
1 9. tt:ld4xc6 'iWd7xc6
12.g4 �e6 13.tLlgS hS 14.tLlxe6 �xe6
20. 'ifa4xc6 tt:le7xc6
IS .gS tLle8 16.llab l '*'kd7=, then
21. llf1 -d 1 \t>g8-f8
White's weakened king offsets the
22. f2-f4 \t>f8-e7
bishop pair.
23. \t>g1 -f2 lld8-d7
7. ... 'W'd8-d7 24. lld 1 -d3 a7-a5
8. lla1 -b1 �f5-h3 25. b4xa5 tt:lc6xa5
9. �g2xh3 1'Vd7xh3 26. llb2xb6 tt:la5xc4
1 0. tt:lc3-d5 1'Vh3-d7 27. tt:le3xc4 llc8xc4=
11. 'iWd 1 -a4
19
The Kaufman Repertoire for Black
Stalemate.
20
Chapter 2 - English Opening
3. ... �g8-f61
4. �g1 -f3
4. ... e5xd4
5. e4-e5 �f6-e4
1 1 . 0-0 lUB-eS
1 2. .tf3xc6 d7xc6
1 3. h2-h3
13.tbf3 .ig4=.
7.tbc3 tbxc3 8.bxc3 tbc6 9.'ird3 (9.'ii'e3 Black's bishop pair more than offsets hjs
.ie7=) 9....te7 1 o . .th6 .tfs 1 1..igs crippled pawn majority.
.ie7 leads to a draw by repetition. Black
can avoid this by 9 ... .iaS!? IO..ih6 'ii'e7
1 1 . .ie2 tbxeS with balanced chances if RE 2 3 .4 (A07) Game 2.3
he wants a real fight. 0 Markowski, Tomasz
7 .�d 1 fs 8.exf6 tbxf6 9..tgs 0-0=. • Romanishin,Oleg
Moscow, 2004 (6)
7. ... �e4-g5
1 . g2-g3 d7-d5
7 ... tbxd2 8..txd2 tbc6 9.'ii'e3 .txd2+ is
2. �g1 -f3 �g8-f6
also about equal.
We would be more likely to reach this po
8. �f1 -e2 �b8-c6
sition by l.tbf3 tbf6 2.g3 dS.
9. 'ii'd 4-e3
3. .tf1 -g2 g7-g6
9.'ti'f4 tbe6 IO.'ti'g3 tbg7 =.
21
The Kaufman Repertoire for Black
22
Chapter 2 - English Opening
1 8. �d2-c3 �g7-h6
.i'i¥ .i�
1 9. J:tc1 -d 1 �d7-e6
i iAi
IJl A i
i IJl
t2J ��CD� i �
��'V/jj ���� i
� n� IJl
t2J�� CD�
1 4. �e3-c1 �� ��Si�
Else Black regains the pawn favorably. � 'if n �
1 4. ... a7-a6
This was just a sample continuation of the
1 5. �c2-b1 tbd5-b4
gambit. White's shut-in rook and offside
1 6. �c1 -d2 �e6-g4
knight are good value for the pawn.
1 7. llf1 -c1 �d8-d7
23
Chapter 3
The idea of J.lbf3 to avoid the Griinfeld is this: after l .lLlf3lLlf6 2.c4 g6 3.lLlc3
and now 3 ...d5 , White does not have to play 4.d4, transposing to the Griinfeld. He can
instead play 4.cxd5 lLlxdS and then 5.'ii'a 4+ or 5.'ii'b 3 or 5.e4 or 5.g3 or 5.�c2, all of
which pose some problems for Black. I think that Black should not fear these lines. and
I give adequate responses in the anti-Griinfeld chapter. In this chapter I cover some al
ternate solutions to this move-order, in case Black is not happy with one of the anti
Griinfeld lines. Feel free to skip this chapter if you are following my primary recom
mendation, although the first game can also be used as an alternative to playing the
Neo-Griinfeld against an early g3.
So what can Black do if he fears the anti-Griinfeld? One solution is to answer 3.lbc3
by 3 �g7 then when White plays 4.e4 (4.d4 dS is the Griinfeld), play 4 c5 (other
..• , .•.
wise you must play the King's Indian) . Now White's only try for advantage is 5.d4,
hoping to get a variation of the Maroczy Bind vs. Accelerated Dragon after Black takes.
But Black may surprise him with 5 'ii'a 5 , as I played in a World Senior Championship
•..
game against IM Rukavina in 2009, getting a winning game but only drawing it. This
should lead to positions similar to the Accerated Dragon Maroczy Bind if White plays
accurately; perhaps Black is a tad better off here than in the line with S...cxd4. Still, un
less you are an Accelerated Dragon fan, these options leave something to be desired,
and don't fit with the Breyer Defense to J.e4 given in this repertoire.
25
The Kaufman Repertoire for Black
.i � .t'iV� .t � K
j. j. j., j. j. j. j.
hoping for 2.c4 ..ig7 3.e4 eSI as analyzed in the previous (English) chapter. Of course,
if 2.d 4 we play 2 ...ttJf6 and we are back on track for the Griinfeld. This is a solution fa
vored by Peter Svidler, perhaps the strongest consistent Griinfeld player in recent years.
The only problem with l...g6 is 2.e4!, after which we either have to play a Pirc or Mod
ern Defense with 2...d 6 or 2 .....ig7, or else play 2 ...cS when we are playing the Hyper
Accelerated Dragon. So the choice of tltis line or l... t2Jf6 may come down to whether
you prefer 2...t2Jc6 or 2..g6 in the Sicilian! Again, this won't work for our repertoire,
only for those who don't fear the Maroczy Bind. However the option to play the Pirc or
Modern makes some sense, as White has already played tt:lf3 so he cannot play the dan
gerous lines with an early f4 or �e3 with f3 or �gS. This is often Svidler's preference.
.i � .t 'iV � .t � .i
l. j. j. j. j. j. j.
Here the idea is to meet 2.c4 by 2 ...ttJc6, and then answer 3.ttJc3 by 3 ... eS, as Grischuk
played repeatedly against Gelfand in their 20II Candidates' Match. Tltis would be my main
reconunendation, except that Wltite can play 2.e4 and we must play the Sicilian. At least
this time Black can choose any Sicilian, not just the Accelerated Dragon. As a practical mat
ter, few players open with l.t2Jf3 and then transpose to the Sicilian, since if they wanted to
26
Chapter 3 - Queens Indian versus Reti
play against the Sicilian they would probably invite it by opening l.e4. So this can be
your solution if you at least have the Sicilian as a second defense. But I can't make it my
choice for this book as I can't assume a knowledge of the Sicilian by the reader.
The solution I came up with is original, to my knowledge. I don't mean that the moves
are original, just the concept of using the Queen's Indian as a companion to the
Griinfeld. The idea is to meet I .t!Jf3 t!Jf6 2 .c4 by 2 ... b6.
E � i.'iV� i. E
.l .l.l.l.l.l.l
.lj �
Now if 3.d4 �b7 4.t!Jc3 e6 we are in the Queen's Indian, but not White's most recom
mended option with 4.g3. So you might say why not 4.g3 e6, the main line Queen's In
dian? My answer is that if White plays 4.g3 (or 3.g3 �b7 4.d4 or 4.�g2) we don't play
4...e6 but instead choose 4... g6. This double fianchetto against the white fianchetto was
recently recommended in a Yearbook article by the very strong grandmaster Sergey
Tiviakov, and I used his analysis as the starting point for my own in this book. I must ad
mit that the computers are not too fond of this Tiviakov Variation, but as the positions are
closed it is reasonable to question their judgment. I feel that this choice is the most in the
spirit of the Griinfeld, as we do fianchetto the king's bishop in both lines, and it is not un
usual in some Griinfeld lines to fianchetto the queen's bishop as well.
So in this chapter you will find both the Tiviakov Variation and the regular Queen's
Indian lines without g3. I hope you find this solution (or one of the above-mentioned
ones) acceptable. But my preferred option is the Anti-Grunfeld.
If you believe as I do (and also I think world # I Magnus Carlsen) that the Breyer and
the Grlinfeld are the best answers to l.e4 and l.d4, then it follows from the above that
White's optimum opening play, if he prefers the Anti-Griinfeld to the real one, might
be l.tLlf3 dS 2.d4 or l...tl:Jf6 2.c4 or l ... g6 2.e4 or l ...cS 2.e4. But very few players are
equally at home on the White side of the Sicilian, the Queen's Gambit, the Queen's In
dian, and the Symmetrical English! If you are one of them, see the l.tL'lf3 chapter in the
White side of this book!
Now for the games. First we look at the Tiviakov Variation I .t!Jf3 t!Jf6 2.c4 b6 3.g3 �b7
4-.�g2 g6 5.d 4 �g7. Note that this position can also be reached by I .d4 t!Jf6 2 .c4 g6
27
The Kaufman Repertoire for Black
3.tLlf3 .2.g7 4.g3 b6 S..2.g2 .2.b7 (as actually played in Game 3. 1) if Black doesn't like
the Neo-Griinfeld 4...d5 . However this is not so important because White can avoid it
by playing 3.g3 first (before tLlf3), after which 3 ... b6? doesn't work. The resultant po
sitions are similar to the King's Indian. The computer doesn't like Black either here or
in the King's Indian, but it seems that the Tiviakov line is better than a normal King's
Indian so it may be fme for Black even if the computer doesn't agree.
Next we come to I.tl:lf3 tLl f6 2.c4 b6 3 .d4 �b7 4.tLlc 3 e6 5.�g5 h 6 6.�4 �e7
(Game 3.2). Black equalizes without much difficulty.
The remainder of this chapter deals with the Petrosian/Kasparov Variation 5.a3 in
the above sequence. It is the main line of the Queen's Indian excluding 4.g3 (which we
meet with the Tiviakov line in Game 3.1 ). After 5 ...d5 6.cxd5 tl:lxd5 White has three se
rious choices. The move 7.�d2 is met by 7 ...tl:ld7 in Game 3.3, when White can isolate
the Black dS pawn but only at the cost of the bishop pair. In Game 3.4 we examine the
traditional 7.e3, when I deviate from the recommendation of most books (7...g6) pre
ferring 7 �e7, which appears to equalize comfortably thanks to a 20I0 novelty. Fi
•..
nally we cover the most popular 7.'ii'c 2 , which we meet by trading knights, playing
8 ...h6 next to prevent 9..2.gS if White takes with the queen, or 8 ...c5 if he takes with
the pawn. See Game 3.5. In all cases it seems that Black can equalize the game.
28
Chapter 3 - Queen's Indian versus Reti
I O.ltJc2 aS l l.e4 c6 1 2.lleI cxdS Black has a pleasant King's Indian posi
13.exdS ( 1 3.cxdS �a6�) 13...ttJe8 - tion, and IS.f4? can be well-met by
though the computer says this is equal, I IS...�a6.
prefer Black, who has ideas of ...ltJd6,
...�a6, and ...f7-fS.
1 0. ... d7-d6
QI 2 . 1 6 (Ei2) Game 3.2
1 1 . e2-e4 �b7-c8
0 Carlsen,Magnus
1 2. h2-h3
• Leko,Peter
Miskolc match, 2008 (8)
1. d2-d4
1. ... tt:lg8-f6
2. c2-c4 e7-e6
3. tt:lg1 -f3 b7-b6
4. tt:Jb1 -c3 �c8-b7
1 2. ... tt:Jf6-d7 5. �c1 -g5
The actual game went 12...ltJhS 13.�h2 S.a3 transposes to 4.a3 games.
�d7 14.�f3 �h8 IS.�e3 tlJf6 16.�g2
5. ... h7-h6
ltJg8 1 7.'it'd2 fS 1 8.exfS gxfS 1 9.f4 e4
29
The Kaufman Repertoire for Black
9 . ... c5xd4
6 . ..Q.g5-h4
6. ... ..Q.f8-e7
I8.llacI �b7=. Black plans ...<it>f8-e7.
This gives comfortable equality, unlike
1 0. ... tl:lb8-c6
the risky 6...�b4!?.
1 1 . ..Q.d3-c2
7. e2-e3
I I.Ilci dS I2.cxdS tbxdS 13.�g3 llc8=;
For 7.�c2 cS 8.dxcS bxcS 9.e3 0-0 see I l.�g3 dS 12.cxdS tbxdS=.
the next note.
11. ... d7-d5
7. ... 0-0 1 2. ..Q.h4xf6 ..Q.e7xf6
8. ..Q.f1 -d3 1 3. c4xd5 ttJc6-b4
14. d5xe6 ..Q.b7xf3
8.�c2 cS 9.d.xcS bxcS IO.�e2 tbc6
1 5. g2xf3 ttJb4xc2
II.IldI d6 12.0-0 tbhS 13.he7 't!Hxe7=.
1 6. �d1 xc2 'fHd8xd4
White's pressure down the d-file just off
1 7. l:ra 1 -d 1
sets Black's central pawn dominance.
8. ... c7-c5
9. 0-0
30
Chapter 3 - Queen's Indian versus Reti
White has little compensation for his bro- 9... hd5 I O.e4 �b7 l l.dS exdS 12.exd5
ken kingside. �d6 13.0-0-0 0-0 I4.�b5 scores too
well for White though the computer rates
it as even.
QI 4. 1 9 (EI2) Game 3.3
1 0. d4xc5 ..Q.f8xc5
0 Jobava,Baadur
• Almasi,Zoltan I O ... bxcS Il .e3 �d6 1 2. b4 0-0 is also
Rij eka Ech, 20 I 0 (8) okay for Black, though in this case Black's
compensation for his isolated queen
1. d2-d4 tLlg8-f6
pawn is superior development rather than
2. c2·c4 e7-e6
the bishop pair as in the main line.
3. tt:lg1 -f3 b7·b6
4. a2-a3 �c8-b7 1 1 . e2·e3 0·0
5. tLlb1 ·c3 d7-d5 1 2. �f1 -d3 tLld7-f6
6. c4xd5 tLlf6xd5 1 3. 0·0
7. �c1 ·d2 tLlb8·d7
8.lLlxd5 exdS 9.g3 (if 9.b4 �d6 I O.�gS The text is a computer improvement over
f6 II.�h4 cS 12.�g3 �xg3 13.hxg3 c4 the actual game, which continued
I4.e3 0-0 the protected passed pawn off I3 ...lLle4? ! 14.�b4 'ti'e7 IS.hcs bxcS
sets the slightly weakened black kingside) 16.b4! and with a great knight to be
9 ... �d6 IO.�g5 lLlf6 J J. ..Q.g2 0-0 12.0-0 posted on d4 vs. a bad bishop White had
lle8 - chances are balanced. full compensation for the pawn and won.
31
The Kaufman Repertoire for Black
1 5. ... 'ifd8-f6
1 6. l:f.a1 -c1 a7-a5
1 7. �b4-c3
1 0. ... c6-c5
1. d2-d4 tbg8-f6
2. tbg1-f3 e7-e6
3. c2-c4 b7-b6
4. a2-a3 �c8-b7
5. tbb1 -c3 d7-d5
6. c4xd5 tbf6xd5
7. e2-e3 �f8-e7 1 1 . e3-e4
32
Chapter 3 - Queen's Indian versus Reti
21 . �c2xe4 lle8xe4
&ii� 22. l:tb1 ·b8+ l:tf7-f8
� � 23. l:tb8xf8+ �g8xf8
� ttJ 24. �e3xc5+ �f8·g8
8��
Black's advanced passer gives him a slight
l:l � 'if n �.
edge.
33
The Kaufman Repertoire for Black
34
Chapter 3 - Queen's Indian versus Reti
I O.�d3 '\fl'c7 I 1 .'\fl'bI �e7 I 2.0-0 0-0 In the actual game, Black played 1 2...�e7
13.�e3 llac8 14.lldI llfd8 1 5.a4 h6=. 1 3.�d3 lt:Jf6 14.'Wb5+ 'Wd7 IS.tL:leS
Black plans ...tL:lf6 and perhaps ...lt:Jh5-f4. �xbS 16.�xb5 + �f8 1 7.f3 lt:Je8
1 8.�d7 lld8 19.�c6 - White has a slight
1 0. ... c5xd4
edge here, but Black held the draw
1 1 . c3xd4 lla8-c8
· eventually.
1 2. '\fl'c2-b3
1 3. �f1 -d3 g7-g5
12.'\fl'b l �e7 13.�d3 0-0 14.0-0 llc3
14. �f4-e3 �f8-g7
IS.llc l llxc l + 16.�xc l 't\Va8 1 7.lla2 fS
1 5. 0-0 0-0
18.�c4 �xe4 19.he6+ �h8= - it's not
1 6. lla1 -c1 4Jd7-f6=
obvious who should be better here but
the computer says it is equal.
K 'ii � .t �
l .t � l i l
' l
35
Chapter 4
Anti-Grunfeld
This name is applied to the use of the opening move I .tt:lf3 to aim for the usual
Queen's Gambit lines without allowing the Grunfeld Defense. White plays I .tLlf3 tLlf6
2.c4 g6 3.tLlc3 and if 3 . . . �g7 , 4.e4 d6 S .d4 is the King's Indian. So a Grunfeld player
will likely play 3 ... d5 ,
after which 4.d4 would be a Grunfeld. However White can try other fourth moves in
stead, hoping to benefit by saving the tempo d2-d4 or in some lines by recapturing on
c3 with the d-pawn instead of the b-pawn. Actually this was the hardest chapter of the
book for me to write, because I believed that White had more than one path to a
non-trivial advantage in the Anti-Grunfeld. However, when I really got into the analysis
deeply, I found satisfactory lines for Black in each case, and my present opinion is that
with best play the Anti-Grunfeld confers no larger advantage on White than he can ob
tain against the normal Grunfeld, which is to say very little. In several lines though, it
seems that Black has hardly any choice if he wishes to stay close to equality, so you had
best study this chapter rather thoroughly, unless you want to play altogether differently
against i .tt:lf3 as discussed in the Reti chapter.
In Game 4. 1 we look at 4.�a4+ �d7 5.�3 dxc4 6.'i¥xc4 a61 7.d4 (and other 7th moves
for White) . Now if 7 . . . �g7 8 .e4 0-0 we would transpose to the Hungarian Variation of
the Russian System against the Griinfeld (which I recommend in that chapter for Black) ,
but with the difference that Black's bishop is on d7 rather than c8. This seems to be in
White's favor, as he no longer can play . . . �b7 (after . . . bS) , and also because he needs d7
as a retreat square for his knight in case of eS. So we don't play this way. Instead we play
7 ... b5 8.�3 cS (the novelty 8 . . .�g7 9 .e4 tt:lc6 ! ? is also discussed and doesn't seem bad)
9.dxc5 �g7 I O.e4 �c6 and Black doesn't seem to have any theoretical problems.
37
The Kaufman Repertoire for Black
In Game 4.2 White provokes an endgame by S.e4 ttJxc3 6.dxc3 �xd l + 7.�xdl . The
novice might say that White has lost the right to castle for nothing, but with queens
and a pair of knights off the board the white king will be quite happy on c2 or (after
.ic4) on e2. Black's equalizing strategy is to play . . . f7 -f6 and . . . e7-e 5 , together with
. . . ttJd7 , so as to meet the natural i.e3 by . . . .icS , exchanging off White's better bishop.
It seems that this works fine, as Black has a nice plus score in grandmaster play from
here in the current century.
Next we look at two queen moves, S.�b3 and S.�c2. The first can transpose into a
Griinfeld line which I recommend for White after S .... ttJb6 6.d4 �g7 7.e4 0-0 (recom
mended by Delchev) 8.�e3 �g4 9 .l:::tdI, so I don't recommend playing this way. Instead
we play 7 ...�g4!. Other seventh white moves can be met by 7 . . . �e6 hitting the queen.
The move S.�c2 was unveiled by Topalov in his 2 0 1 1 Candidates' match with
Kamsky: Topalov got an advantage and should have won the game but didn't. For a while
I considered this novelty to be the refutation of 3 ... d5 in the anti-Griinfeld, but as shown
in my notes to Game 4.3 I no longer consider 5 .'iYc2 to be any problem for Black, if he
reacts in the sharp manner shown, namely S ... ttJc6 6.d4 ttJdb4 7.'iYa4 �d7 S.�dl eS!.
Game 4 . 4 features the move S.�a4+. The usual reply historically was 5 . . . �d7 , but
White seems to keep a slight edge. This line bothered me quite a bit. Delchev and
Agrest recommend 5 . . . c6, and although the computers hate it I think it is playable but
not quite equal. The current fad is for the computers' recommended S ... ttJc6, which
leads to an endgame in which Black has an awful pawn structure-. However the com
puters have shown that Black's superior development and piece play are enough to
equalize here, and recent grandmaster games confirm this. No one would have played
this way before computers, but today the motto is 'whatever works' .
Finally we come to perhaps the best o f the fifth moves, namely S.g3 . After the natu
ral moves S ...�g7 6.�g2 0-0 7.0-0 I favor 7 ...c5. Now after White trades knights he can
play either 9 .d3 or 9.d4. The move 9 .d4 is rather dangerous after 9 ... cxd4 I O.�e 3 ! but
the reply 1 0 . . . d3 ! seems to lead to near-equality for Black. Instead 9 .d3 ttJc6 I 0. .2.e3 is
a dangerous gambit which gives White a pull if declined. So we take on b2 and retreat
38
Chapter 4- - Anti -Griinfeld
to f6 when attacked. Game 4-.5 shows how Black can steer the game to the safety of an
even-material bishops-of-opposite-color draw, or can play an original line I give which
leads to the interesting balance of two bishops, rook, and extra pawn versus two rooks
and a knight. So it seems to me that although S . g 3 should keep at least some edge. it is
quite tiny with best play.
I conclude that the Anti-Griinfeld does not appear to give White any more of an edge
than he gets in the real Griinfeld, but White does have a variety of interesting options
here, so there is at least a reasonable case to be made for starting the game with 1 .tLlf3 if
you expect to see the Griinfeld.
EO 6 1 .8 (A 1 5 ) Game 4.1
7. d2-d4
D Jankovic,Alojzije
• Grandelius,Niels 7 .e4- bS 8 .\We2 cS 9 .e5 tLlhS 1 0.tLle4- tLlc6
Khanty-Mansiysk o1, 20 I 0 (4-) 1 1 .tLlxc5 �fS 1 2 .d3 �c8 1 3 .�e3 (so far
Ponkratov-Kurnosov, Taganrog 2 0 1 1 )
1. tt:Jg1 -f3 tt:Jg8-f6
1 3 . . . e6N 1 4-.llc 1 �g4- 1 5 .�c3 b4- 1 6.�c4-
2. c2-c4 g7-g6
�xf3 1 7 .'iVxf3 b 3 ! - the idea is to block
3. tt:Jb1 ·c3 d7-d5
the White queen's access to a4- in case of
4. �d1 -a4+ �c8-d7
. . . li:JxeS �d 1 . White is in trouble.
5. �a4·b3 d5xc4
7. g 3 bS 8 .\WM (after 8 .�b3 cS 9 .�g2
6. �b3xc4
ti:Jc6 I O.d3 i.g7 1 1 . 0-0 �c8 I 2 .\Wd I 0-0
If 6 .'ihb7 ? li:Jc6 7 .ti:Jb5 tLldS 8 .ti:Jbd4- Black's advantage in development and
tLlcb4- 9.a3 �b8 1 0.�xa7 �a8 1 1 .'il'b7 space easily offsets White's extra center
i.g7 1 2 .l:Ib 1 cS 1 3 .axb4- ti:Jxb4- 1 4-.e3 pawn) 8 . . . c5 9 .�g2 �c6 I O .d3 h6
�b8 1 5 .\Wa7 cxd4- Black wins material. 1 1 .0-0 i.g7N and Black is better as
White must respond to the threat of ... g S ;
6. ... a7·a61
7 . d 3 b S 8 .\Wf4- c S 9 .g3 �c6 I O.�g2 �g7
1 1 . 0 -0 ti:Jbd7N 1 2 .a3 h6 - this permits
i. � "if� .t i. Black to answer \Wh4- by . . . g6-g5 . I al
i iAi i . i ready prefer Black's game.
i � .i 7. ... b7-b5
'if
ttJ
ttJ_
� � ti � � � �
� ii£, � ii£, �
Very similar to the Hungarian Variation of
the Griinfeld (which I recommend here)
against the Russian System.
39
The Kaufman Repertoire for Black
8. '\Wc4-b3
40
Chapter 4 - Anti-Griinfeld
1 4. ... �d8-c7
1 5. l:!f1 -e1
EO 5 2. 8 (A I S ) Game 4.2
0 Suba,Mihai
• Tseshkovsky,Vitaly
Arco Wch-sen, 20 I 0 (9)
Black actually played 1 5 . . . .!Ifc8 1 6 . .Q.gs 6 .bxc3 �g7 7 .�e2 (for 7 .d4 or 8 .d4, see
.tbS 1 7 .'i¥a2 e6 1 8 . .l:Iac l .l:Iab8 1 9 .tt::lx b5 Chapter 7 on the Griinfeld) 7 . . . cS 8 . 0-0
axbS 2 0.c6 tt::l e S 2 I ..txf6? �xf6 2 2 .b3 tt::l c 6 9 . .l:Ib l 0-0 I O.'i¥c2 b6= or l O .'iVa4
axb3 2 3 .'iYxb3 'i¥b6 24.c7 !Ixc7 tt::l e S l l .tLlxeS �xeS 1 2 .f4 �g7 1 3 .'i¥b3
25 . .l:Ixc7 'i¥xc7 2 6 ..txb5 'iYaS 27 . .ll b l b6=.
tt::l xf3 + 2 8 . gxf3 .Q.es 2 9 .h3 'i¥d8 3 0 .'i¥e3
6. ... �d8xd1 +
"i¥h4 3 I .<.t>g2 .l:!.d8 3 2 .�fl .tf4 3 3 .'i¥c5 ?
7. <ot>e1 xd 1
.l:Id2 34.'i¥c8+ Wg7 3 5 .'iYc3 + �h6 and
White resigned. Although this endgame was a t one time
considered to favor White slightly, it has
1 6. �c4-b4 'ir'c7xc5
been over ten years since White has won
1 7. �b4xc5 lt:Jd7xc5
a single grandmaster level game from
1 8. e4-e5 lt:Jf6-g4
here, while Black has won several with
1 9. �c1 -f4 lt:Jc5-e6
dozens of draws.
41
The Kaufman Repertoire for Black
7. ... f7-f6
Analysis diagram
8. ... e7-e5
9. '.t>d1 -e2
9 . .te3 tt::l d 7 1 0 .We2 �cS 1 1 Jlhd 1 �xe3 Black is slightly better as White's kingside
1 2 .<iii x e3 We7 1 3 .tt:ld2 aS 1 4.f3 tt:lb6 pawns are inferior and his bishop is
1 S . .I1l.e2 �e6 1 6.tt:lb3 (so far Ivanchuk- slightly bad since three white pawns are
Nepomniachtchi, Havana 20 1 0) stuck on light squares.
42
Chapter 4 - Anti-Griinfeld
43
The Kaufman Repertoire for Blask
Analysis diagram
A) 7 .�f4 �e6 :
A I ) 8 .�c2 tLlc6 9 . l:i. d i (in case of
9.e3 tLlb4 I O .'iYd i 0-0 I l ...Q.e2 cS
1 2 .0-0 l:i.c8 I 3 .�eS �xeS I 4.tLlxeS
cxd4 I S .exd4 ltJ4dSN White has no at
tack to offset his isolani. I would prefer Analysis diagram
44
Chapter 4 - Anti-Griinfeld
9. a2-a3?1
5. ... tt:lb8-c6
6. d2-d4 tt:Jd5-b4
7. 'ifc2-a4 �c8-d7 1 0. ... a7-a6
8. 'ifa4-d1 e7-e51 1 1 . tt:Jb5xd4 tt:lc6xd4
45
The Kaufman Repertoire for Black
EO S 2 . 1 1 (A I S) Game 4.4
0 Markos,Jan
• Li Chao B
Shenzhen, 20 1 1 (4)
1. tt:Jg1-f3 tt:Jg8-f6
2. c2-c4 g7-g6
3. tt:Jb1 -c3 d7-d5
4. c4xd5 tt:Jf6xd5
1 3. ... 'tl¥d8-f6!N 5. 'tl¥d1 -a4+
1 4. e2-e3
46
Chapter 4 - Anti-Griinfeld
i:
9. g2-g3
8 .'ltfd4 tt:lxc3 9 .dxc3 (in case of 9 .'ltfxh8? Or l l .a3 1;lb8 1 2.e3 0-0 1 3 .tt:la4 �c8
tt:lxa2 I 0. b 3 tt:lxc I I I . 'ltfb 2 tt:lxe 2 1 4.tLlcS (so far Vitiugov-Areschenko,
1 2.<;t>xe2 �e6 Black has way too much Olginka tt 2 0 1 1 ) 1 4 .. JlbSN 1 S .d4 eS
for the exchange, namely two pawns, 1 6 .a4 l:!.b8 1 7 .dxeS tt:lb4 1 8 .0-0 .txeS
better development, the bishop pair, the 1 9 .1;ld ! tt:lc2 2 0 .llb l tt:la3 2 l..l:i.a l tt:lc2
better pawn structure, and a safer king! ) with a draw by repetition.
9 ... f6 I O .e4 �e6N I I ..te3 l:Id8
1 1 . ... 0-0-0
1 2 .'ltfxa7 'lifxe4=.
47
The Kaufman Repertoire for Black
1 3. kb2xc3
48
Chapter 4 - Anti-Griinfeld
EO 2 7. 1 6 (A 1 S) Game 4.5
D Meier,Georg
• So,Wesley
Lubbock, 20 I 0 (9)
9. ... tLlb8-c6
1 0. �c1 -e3
7. ... c7-c5
8. tLlc3xd5
49
The Kaufman Repertoire for Black
�xf8 1 6 .'ifb l .tg4 1 7 . .txc6 llc8 1 8 .e3 At first I wanted to recommend this, be
�eS 19 . .te4N c4 20.dxc4 .llxc4 Black cause two bishops, a rook, and an extra
has the better of a probable draw. pawn are virtually equal in strength to
two rooks and a knight. However White
1 1 . ... �b2-f6
seems to have the initiative here, so now I
1 2. 'iYd1 -a4
would make this only a second choice in
After 1 2 . ttJd4 'i¥xa2 1 3 . tt:Jxc6 bxc.6 case you don't like the game continua
1 4.£.xc5 .tg4 I S .f3 £.e6 1 6 .d4 .ll f b8 tion.
1 7 . ll a l 'irc4N White doesn't have 1 9 .e3 .tg7 2 0 . l:f.fb 1 l::!.c8 2 1 . lla3 .l::t c 7
enough for the pawn. 2 2 .f4 eS 2 3 .ttJc4 exf4 24.gxf4 £.f8
2 5 . �f2 f6 2 6 .e4 �f7 - Black should be
1 2 . ... 'iYd5-d7
okay here, although he must be careful.
1 3. 'iYa4-b5
White's active rooks give him the sunny
1 3 .Lc5 b6 1 4.hb6 ttJd4 1 5 .'iVd l axb6 side of a probable draw.
1 6.tt:Jxd4 l!1xa2 1 7 .'i¥b3 .lld 2 1 8.'i¥b4 lla2
1 4. 'iYb5xc5 'iYe6xa2
1 9 .'iVb3 .i':!.d2 with a draw by repetition.
Black is under pressure on the queenside,
but his extra pawn offsets this.
1 5. .l:i.f1 -e1
Analysis diagram
50
Chapter 4 - Anti-Griinfeld
51
Chapter 5
which we meet by 3 h6 , welcoming the bishop for knight exchange. White has to make
...
the trade or play a bad French Defense. I think that the exchange results in a very pleasant
game for Brack, who can use his unopposed bishop to pin the c3 knight. See Game 5 . 1 .
53
The Kaufman Repertoire for Black
i: � .t tv � .t 1:
.t. & & & & &
�· &
aims at �d2 and 0-0-0. The remedy seems to be a timely . . . tLle4. See Game 5 . 3 .
After l .d4 tLlf6 2.tLlf3 g6 White has three popular options, all covered in Game 5 .4. He
can choose the Colle 3 .e 3 , which is considered ineffective against the king's fianchetto,
the Torre 3 .�g5 �g7 4.tLlbd2 and now 4. . . 0-0, ready to meet 5 .e4 or 5 .c3 by 5 . . . d 5 , or
the London System 3 .�f4 �g7 4.e3 d6 S .h3 0-0 6.�e2 when I like 6 ...c5 7.c3 'ilr¥b6.
54
Chapter 5 - Queen's Pawn Openings
This move is rare but good. It seems cause the bishop pair is a bigger factor
wrong to spend a tempo to force the ex than the crippled majority) 6.exd5 "tifxdS
change, but the point is that White's 7 .tt:lf3 0-0 8.ile2 "tWaS 9."tifd2 tt:ld7
knight on c3 is then misplaced, because 1 0 .a3 tt:'lb6 1 1 ..ld.b 1 �xc3 1 2 ."tifxc3
White would want to play c3 -c4 after the 'ifxc3 + 1 3 .bxc3 Ue8N.
trade. More usual is 3 . . . tt:lbd 7 , but why
guard against a bad trade?
4. �g5xf6
Or 4.ilh4 e6 5 .e4.
Analysis diagram
4. ... e7xf6
5. e2-e3
55
The Kaufman Repertoire for Black
56
Chapter 5 - Queen's Pawn Openings
1. d2-d4 t2Jg8-f6
2. .fi.c1 -g5
57
The Kaufman Repertoire for Black
5. ... e7-e6
6. tLlg1 -f3 �f8xc5
7. �f1 -d3
1 5. ... h6xg5
1 6. l:rh1 xh8 �d8xh8
1 7. �b5-f1 �c8-d7
1 8. 0-0-0 J:[a8-c8
7. ... h7-h6
8. �g5-h4
8. ... g7-g5
1 9. tt:le5xd7?!
Also after 8 . . . 0-0 9 . 0-0 �e7Black is at
least equal. In the event of 1 9 .g3 .fl.cS 20 .�d2 �h2
2 1 .tLlxd7 '>t>xd7 22 .a3 Wc7 23 .Wb 1 Wb8
9. �h4-g3 ttJf6-h5
Black is better with an extra center pawn
1 0. �g3-e5 ttJc6xe5
and better development.
11. tt:lf3xe5 tt:lh5-f6
1 2. h2-h4 �c5-d6 1 9. ... tt:lf6xd7
1 3. �d3-b5+ \t>e8-e7 20. �d4xh8 .l:!.c8xh8
14. �d1 -d4 a7-a6 21 . e3-e4 d5xe4
58
Chapter 5 - Queens Pawn Openings
:1
59
The Kaufman Repertoire for Black
60
Chapter 5 - Queen's Pawn Openings
more than enough compensation for a Although White has two pawns for the
pawn. exchange, he can't keep them and is
clearly losing.
1 5 . l:!.d4·d3 tt:Jb8·a6
26 . ... l:!.d7·c7
1 6. g2-g3?1
1 6. ... tt:Ja6·b4
1 7. .:t.d3·b3 c7·c5
1 8. a2·a3 tt:Jb4·d5
1 9. e3·e4 c5·c4
20. .:t.b3·b5 tt:Jd5·e3 33 . ... f7·f5
21. .:t.b5·c5? .:t.e8·d8
3 3 . . . c3 ! .
22. tt:lf3·d2 �c8·e6
23. b2·b3 l:!.d8·d7 34. tt:Je4·c5 l:!.d7·d 1 +
24. b3xc4 l:!.a8·d8 35. l:!.g1 xd1 l:!.d8xd1 +
25. l:!.c5·d5 �e6xd5 36. 'i¥i>c1 ·b2 l:!.d1·d2
26. e4xd5 37. 'i¥i>b2·c3 l:!.d2xc2+
38. 'i¥i>c3·d4 tt:Je3·f1
39. e2·e4 f5xe4
40. �f3xe4 l:!.c2·a2
41 . tt:Jc5·e6+ 'i¥i>g7·f6
42. tt:Je6·f4 tt:lf1 xg3
43. �e4xg6 tt:lg3·e2+
44. tt:Jf4xe2 'i¥i>f6xg6
45. tt:Je2·f4+ 'i¥i>g6-h6
46. 'i¥i>d4xc4 l:!.a2xa3
47. 'i¥i>c4·b5 l:!.a3-g3
White resigned.
61
The Kaufman Repertoire for Black
Analysis diagram
62
Chapter 5 - Queen's Pawn Openings
1 0. c3xd4?
1 0. ... tt:Jf6-d5
1 1 . �f4-g3 tt:Jd5-b4
7. ... �d8-b6 1 2. we1 -d2?!
63
The Kaufman Repertoire for Black
1 2. �e6-f5
1 3 . .!lh1 -c1 e7-e5
14. .!lc1 -c4?
64
Chapter 6
Neo-Griinfeld
This name refers to lines in which Black plays the typical Griinfeld moves even though
White has not played tLlc3 . Usually this means that White has chosen an early kingside
fianchetto, although first in Game 6 . 1 we consider (after l .d4 tt:lf6 2.c4 g6) 3.f3 .
White aims for the Samisch Attack against the King's Indian or to exchange on dS (after
3 . . . d5) and then kick the knight when it has no opposing knight to capture. So I avoid
the problem by playing 3 ... li:Jc6 (as also recommended by Delchev and Agrest) , which
although it looks a bit cheeky has the virtue of making f2-f3 look silly. Its point was to
cover e4, but if Black focuses on d4 that becomes irrelevant, even harmful as li:Jf3 is
now illegal.
The remainder of this .chapter deals with 3.g3 �g7 (3 . . . c6, as recommended by
Avrukh, is much safer but basically it's playing for a draw) 4.�g2 dS .
65
The Kaufman Repertoire for Black
The obvious try for an advantage is now S.cxdS lt:lxdS 6.e4 lt:lb6 7.lt:le2. We hit back at
the center with 7 ...c5 S.dS 0-0 9.0-0 e6. Black should defer the exchange of pawns
(which clears c4) until White has played either lt:lbc3 or a4. Normally Black develops
his knight to a6. Computers tend to like White due to the advanced passed pawn but
human experience suggests that Black has adequate piece play. See Game 6. 2.
I n Game 6.3 White simply ignores . . . dS and just plays lt:lf3 and castles. Black should
take on c4, and then when White plays tt:la3 to regain the pawn, give it back by . . . c3 !
followed by . . . cS , giving White the same weak pawns he gets in the Griinfeld Exchange
variation but without the big center to compensate. This line is not very promising for
White and not so popular any more.
In Game 6.4 White exchanges on dS but refrains from e4, just developing his knight to
f3 and castling. This is the main line now of the Neo-Griinfeld. Black should voluntarily
retreat his knight to b6 before castling, so as to be able to play . . . tt::l c 6 next with attack
on d4 to force the rather tame defense move e3. Then after both sides castle Black
should refrain from . . . e7 -eS due to d4-d5 , instead playing the 'improving move'
9 . . . l:le8 and if 1 0 . .!le 1 , another improving move: 1 0 . . . aS . Black needs to play sharply
and accurately in this position; one sloppy move can leave him with an awful game. As
of now it seems that Black is fine, but this is cutting-edge theory so I would keep
abreast of developments in this line. One of the key games in the notes was played just
days before my deadline.
Analysis diagram
66
Chapter 6 - Neo-Griinfeld
A 1 ) In case of S .dS tt:ld4 6 .ttle2 cS I fa S .e4 dxe4 6.dS tt:leS 7 .fxe4 �g7 8 .tt:lf3
vor Black due to the monster knight; tt:lfd 7 9 .�e 2 0-0 I O .tLlxeS �xeS
A2) S .tLle2 exd4 6.tt:lxd4 �g7 7.tt:lxc6 ( ! O . . . tLlxeS) I I ...th6 lle8 1 2 . 0-0 c6
bxc6 and I like Black due to the possibil 1 3 .�d2 tLlf6 1 4.h3 ..td7 I S . .!lad l
ity of getting in . . . d7 -dS in one go, which �c7 =. Black's powerful outpost bishop
would be impossible if this arose from a compensates for White's space edge.
King's Indian Defense;
5 . ... tt:Jf6xd5
A3) S .dxeS tt:lxeS 6.�e3 �g7 7 .tt:lc3
6. e2-e4 tt:Jd5xc3
0-0 8 .�d2 d6 9.b3 lle8 1 O .tLlge2 a6
7. b2xc3 e7-e5
1 l..l:i.d 1 ( 1 I .ttlf4 c6 1 2 .�e2 bS is simi
lar) 1 l . . .bS 1 2.cxbS axbS 1 3 .tt:lxbS dS
and White is in trouble; .i ..t 'i¥ � 1: .i
B) 4.dS ttJeS S .e4 d6 and now: i i i i i
� i "
.i ..t 'if � ..t .i i
ii i ,i i i i � � "
ri � i � . � '
.�� -
� � -�
-� � J:t 1t 'li' w � tD J:t
�
� 15,: � � 8. d4-d5
J:t; tLJ �;'�i' w � l2S l:t
8 . ..tbS �d7 9.tt:le2 .2.g7 I O . ..te3 ( I O . .ti:b 1
Analysis diagram
tt:laS 1 1 .0-0 c6 1 2 .�d3 0-0 1 3 .�e3
B 1 ) 6.f4 tt:led7 7 .ttlc3 �g7 transposes �e6=) I 0 . . . 0-0 l l .dS tt:las 1 2 .�d3 ..th6 !
to next note; 1 3 . ..tf2 ( 1 3 . .2.xh6 �h4+) 1 3 . . . b6=.
B2) 6.tLlc3 �g7 7 .f4 tLled7 8 .ttlh3 Black plans . . .�e7 and . . . tt:laS-b7 -cS .
(8.tt:lf3 0-0 9.�e2 tt:lcS =) 8 . . . 0-0 9 . .ie2
8. ... tt:Jc6-a5
tt:lcS 1 O .tt:lf2 e6 1 1 .0-0 exdS 1 2.cxdS
9. h2-h4
.Ue8N 1 3 .�f3 hS 1 4.h3 aS I S Jie l tLlfd7
1 6 .�e3 b6=. It is hard to see any good 9 .�e3 c6=.
plan for White here;
9. ... .if8-c5
B3) 6.tt:le2 �g7 7 .tt:lbc3 0-0 8 .tt:ld4 c6
9.f4 tt:led7 ! O .dxc6 tt:lcS l l .�f3 .!le8 and 9 . . . �e7 1 0 .g3 0-0=. Black intends to
due to the threat of .. . e7 -eS , White cannot meet h4-hS by . . . g6-g S .
keep his extra pawn so he stands worse.
1 0. .ic1 -a3
4. ... d7-d5
I O .hS c6 1 1 .hxg6 fxg6=.
Black only plays this Griinfeld move when
1 0. ... �d8-d6
White has a knight on c3 to exchange.
1 1 . .ia3-b4 .ic5xb4
5. c4xd5 12. �d1 -a4+ �c8-d7
67
The Kaufman Repertoire for Black
1. d2-d4 ttJg8-f6
1 8. f3-f4?1 2. c2-c4 g7-g6
3. g2-g3 �f8-g7
1 8 .h5 jld7=.
4. �f1 -g2 d7-d5
1 8. ... �c8-g4 5. c4xd5 ttJf6xd5
1 9. ltJg1 -f3 e5xf4 6. e2-e4 ltJd5-b6
20. e4-e5?1 'ird6xb4 7. ltJg1 -e2 c7-c5
21. c3xb4 ttJa5-c4 8. d4-d5 0-0
22. l:rd1 -d4 ..ig4xf3 9. 0-0 e7-e6
23. d5-d6+ c7xd6
24. e5xd6+ ttJc4xd6
25. g2xf3 l:rb8-c8+
26. 'lt>c1 -b1 ttJd6-f5
27. l:rd4xf4 l:rh8-d8
1 0. ttJe2-c3
68
Chapter 6 - Neo-Griinfeld
1 3. tt:Jb1 -d2
22 . ... �f5-d3
1 3 .ltJa3 ltJeS 1 4.h3 tt:lb4 1 5 .�e3 ?
( 1 5 .ltJe4 b6=) 1 5 . . . b6 1 6.'ti'e2 �fs 22 . . . h5 ! 2 3 .h4 tt:ld3 24.'ti'a 1 lleS favors
1 7 ..l:Ifd 1 'lli'd 7 was Nakamura-Topalov, Black as now f2 -f4 would leave g3 too
Monaco (blindfold) 2 0 1 1 . Black is clearly weak.
better, though White won.
23. l:tf1 -d1 'tWd7-f5
1 3 . ... tt:lc4xd2 24. h2-h3 h7-h5
1 4. �c1 xd2 �c8-f5 25. tt:Jc3-a2 �d3-c2
1 5. �d2-e3 'f¥d8-d7 26. lld1 -f1 tt:lb4xa2
27. l:rd2xc2 tt:la2-b4
1 5 . . . c4 was probably better and equal.
69
The Kaufman Repertoire for Black
It is headed for d4, where the knight will GI 2 . 1 4 (D7 7) Game 6.3
obviously be better than the White D Maiorov,Nikita
bishop. • Gharamian,Tigran
Lille, 20 I I (6)
32. Wk'b2-c3 a7-a5
33. �d2-d3 tt:lc6-d4 1. d2-d4 tt:lg8-f6
34. l:rd3-e3 tt:ld4-e2 2. c2-c4 g7-g6
35. .l:te3xe8 I:td8xe8 3. tt:lg1 -f3 �f8-g7
36. Wk'c3-d3 tt:le2-d4 4. g2-g3 0-0
37. �h1 -g1 f6-f5
I would play 4 . . . d5 first as explained in
38. g4xf5 g6xf5
the notes to Carlsen-Giri. After 5 . .ig2
39. �g1 -h 1 �g7-h7
dxc4 6 .LLla3 c3 ! 7.bxc3 cS 8.0-0 0-0 we
40. Wk'd3-c4 wh7-h6
are back to the current game.
41 . d5-d6 �f4xd6
42. l:rf1 -g1 �d6-f6 5. �f1-g2 d7-d5
43. �g2-f3 l:!.e8-d8 6. 0-0 d5xc4
44. �c4-c1 + f5-f4 7. tt:lb1 -a3 c4-c31
45. Wk'c1 -c3 l:!.d8-d7
It is wise to damage White's pawn struc
46. l:rg1 -g4 l:!.d7-g7
ture this way.
With an extra pawn and a super knight vs.
8. b2xc3 c7-c5
a bishop, Black is winning.
70
Chapter 6 - Neo-Griinfeld
>
Ci5 't{ � . tb ��
�' � �' � . �
.s: �"� l:! FJt
After 9 .tLle5 tLlc6 l O .tLlxc6 bxc6 1 L �.xc6 1 6.tLlb3 ! �xb3 1 7 .axb3 'i¥xb3 1 8 .f4
�h3 1 2 . .!le i .ll c 8 1 3 .�f3 i:Yas 1 4.'i¥d2 'iiHe6 and White has the bishop pair and a
.l:lfd8 Black's compensation for the pawn pawn center for the pawn. I would prefer
is obvious; Black's chances, but it's a game.
9 .tLlc4 tt:lc6 1 O.tLlfeS tLldS 1 1 .�b2 llb8 -
1 6. ... �b6-b2
Black is doing fine here.
1 7. tt:Ja3-c4 �e6xc4
9. ... tt:Jb8-c6 1 8. �e2xc4 c5xd4
1 0. �d1 -e2 1 9. c3xd4 tt:Jc6xd4
20. �c4-a4
1 0 .�b2 'i¥b6 1 1 .'i¥e2 ( 1 1 .tLlc4 �a6
1 2 .tLlce5 �fS =) 1 1 . . . �f5 and now:
A) 1 2 . .l:!.fc 1 .l:lfd8 1 3 .tLlc4 (in case of
1 3 .�fl tt:laS 1 4. tLleS .l:l:ac8 1 5 . tt:lac4
tt:lxc4 1 6 .tLlxc4 'i¥e6 Black's more active
pieces more than offset White's extra cen
ter pawn) 1 3 . . . 'i¥a6 1 4.�fl 'i¥a4=;
B) 1 2 . .!lfd 1 l::rfd 8 1 3 . tLlc4 (after
1 3 .�fl tt:laS 1 4. tLleS .!lac8 Black has the
initiative; White can do little) 1 3 . . . �a6
1 4.�fl 'i¥a4=.
71
The Kaufman Repertoire for Black
8. e2-e3
Black has a pawn plus the dominating If both sides had already castled White
rook on the seventh rank. He should win. could play 9 .d5 now and answer
9 . . . .ixc3 by I O.dxc6, but as it is 8 .d5 ?
just loses a pawn to 8 . . . .ixc3 +.
GI 1 . 1 0 (D7 1 ) Game 6.4 8 .0-0 tt:Jxd4 9.tLlxd4 'f!ixd4 I O .tLlbS and
D Carlsen,Magnus now:
• Giri,Anish A) I O . . . 'i¥e5 I l .a4 0-0 1 2 . .if4 'f!ixb2
Wijk aan Zee, 20 I I (3) 1 3 .tt:Jxc7 ( 1 3 . .l:!.b l 'flia2 I 4.tt:Jxc7 l:Ib8
1 5 .tLld5 e5 1 6.tt:lxb6 axb6 I 7 . .id5 'i¥a3
1. d2-d4 tt:lg8-f6
I 8 . .ic i 'iYe7 I 9 .'f!ib3 l:Id8) I 3 .. Jib8
2. c2-c4 g7-g6
( 1 3 . . . 'f!ixa I 1 4.'f!ixa 1 .ixa 1 1 5 . .l:!.xa I
3. g2-g3 .tf8-g7
.l:!.b8 1 6 .tLld5 .l:!.a8 1 7 .tt:lc7 .l:!.b8 with a
4. �f1 -g2
draw by repetition) I 4.a5 ;
4.tt:lc3 d5 5 .cxd5 tLlxdS 6 . .ig2 tLlb6 7 .e3 B) I O . . . ifc4 I I .'f!ib3 0-0 I 2 .'flixc4
0-0 8 .tLlge2 (8 .tLlf3 transposes to the tt:Jxc4 I 3 . tt:lxc7 .l:!.b8 I 4.tLld5 .ie6
game) 8 ... e5 9 .0-0 exd4 I o.tt:Jxd4 'file?= I 5 .tt:Jxe 7 + �h8 I 6.tt:ld5 .
72
Chapter 6 - Neo-Griinfeld
73
The Kaufman Repertoire for Black
B2) 1 2.b3N a4 1 3 .l:tb l axb3 1 4.axb3 1 9.c4 lbxc4 20 .�b4 tt:'ld6 2 l .�e3 �f8
..ifs 1 S .l:tb2 eS 1 6.dxe5 �e7 = ; White has some but not enough compen
B 3 ) 1 2 .l:td l �c4 1 3 .�c2 lbb4 sation for the pawn.
1 4.'iWb l eS 1 S .a3 exd4 1 6.axb4 dxc3
1 9. ... e5-e4
1 7 . llxd8 .ll a xd8 1 8 .�c2 axb4 and
although Black has only rook and two
pawns for the queen, the threat of . . . b3 is
hard to meet.
11. ... e7-e5
1 2. d4-d5 tt:lc6-b4
1 3. e3-e4 c7-c6
1 4. a2-a3 c6xd5
1 5. a3xb4 a5xb4
1 6. .l:Ia1 xa8 b4xc3
1 7. b2xc3 tt:lb6xa8 20. tt:lf3-g5?
Probably Carlsen was trying to avoid a It's not often that Carlsen loses in 2 2
draw against a lower-rated opponent: moves with White, but a poor opening
1 8.'iWxdS tt:'lb6=. followed by three consecutive bad moves
will do the trick. I guess if you are going to
1 8. ... tt:la8-b6
make this many errors in a tournament, it's
1 9. .l:Ie1 -d1 ?
wise to get them all in in one game!
74
Chapter 7
This chapter covers all the lines after l .d4 tt:lf6 2.c4 g6 3.tt:lc3 d5 in which White does
not exchange pawns immediately on dS .
.i � .i. iV � .i. .i
' ' ' ' ' '
�'
'
� �
t2J
� � � � � �
.t[ .iil 'iY � .iil lLJ .tl
This includes lines with an early e2-e3 , lines with kgS on move 4 or S , lines with �f4
on move 4 or S , the Russian System, based on �b3 to recapture on c4 with the queen,
and miscellaneous white tries. The one common theme in all these lines is that Black
almost never should play passively. That means no early ... c7 -c6 to defend dS. Usually
Black will take on c4 (unless White can recapture with his fl bishop) and/or play an
early . . . c7 -cS . This normally forces White to play concretely, meaning he cannot just
make moves by general principles, he must calculate variations.
In Game 7 . I we look at unusual white fourth moves, such as h2-h4, g2-g4, and
�a4+. This last move intends to play the Russian System with Black's bishop on d7
rather than c8. This does spoil some of Black's options, but it creates a new one, namely
a surprising . . . b7 -bS ! on move 8. The other moves aren't very good, as 4.h4 is met by
fighting in the center with 4 . . . cS while 4.g4 is met by 4 . . . dxc4, clearing a square for
the knight to move to if attacked.
In Game 7 . 2 we look at lines with 4.e3. Lately White has been playing this with the
idea of trading twice on dS and then playing ttJg l -e2-c3 to hit the queen, which we
meet by . . . c7 -cS d4-dS e7 -e6 and isolating the queen's pawn. If White just develops
and castles without trading on dS, Black plays a quick . . . c7 -cS and finds himself playing
the Tarrasch Defense, where White's extra tempo is only enough to equalize.
Against 4.�g5 tt:le4 5 .�h4, we trade knights and take on c4, meeting 7 .e3 by
7 . . . ke6 to guard the pawn. If we want equality we soon play . . . c7 -cS (Game 7 . 3 ) . If we
want to go for the win, we defend the pawn by a quick . . . ttJd7 -b6 as in Game 7 .4.
75
The Kaufman Repertoire for Black
When .tg5 is played on the fifth move (after 4.tLlf3 �g7) it is usually with the idea of
meeting 5 ...tbe4 either by 6.cxd5 , sacrificing the bishop pair (Game 7 . 5 ) or by 6.�f4
(Game 7 . 6 ) . In the former case Black has a pleasant choice between trying to demon
strate that the bishops make up for a possible white minority attack or offering a gam
bit pawn, for which he receives rich compensation. I give both options. In case of
6.�f4 we trade knights and play . . . c7 -c5 , which usually leads to an early queen ex
change. White retains central dominance, but in the endgame Black's queenside major
ity should play a role so chances are close to even.
Now we come to the lines involving 4.�f4 (or 4.tLlf3 .tg7 5.�f4, which may trans
pose) . White can play either 5 .tbf3 or 5.e3 (we meet the rare 5 Jlc l by 5 ... tt:lh5) but in
both cases I recommend castling on move 5 . Now if White plays whichever move he
did not play on move 5 ,
we play 6 ...c5 7 .dxc5 tbe4 which equalizes; see Game 7 . 7 . I f White fails to defend his
pawn on move 6, we take it (5.tbf3 0-0 6 . .l:!:cl dxc4) and meet 7 .e4 by ... �g4, . . . tbh5 ,
and ... �xf3 which sacrifices the bishop pair to inflict doubled backward pawns and an
isolated pawn on White, a roughly even deal; see Game 7 . 8 .
I f White accepts the gambit after 5.e3 0-0 by taking twice o n d 5 and then o n c 7 , the
move 8 . . . tt:la6 ! regains the pawn and leaves Black with the bishop pair to offset his infe
rior pawn structure and inferior center. A few accurate moves give Black full equality
(Game 7 . 9 ) .
Finally we come to 5.e3 0-0 6 ..l:!:cl , which I believed until recently to favor White. I t
now seems that 6 ...�e6! 7.�3 c5 ! leaves White with n o advantage a t all, while 7.tLlf3
dxc4 is no better (Game 7 . 1 0) . I have often played �f4 lines as White, as they are quite
tricky and promise an edge in most variations, but apparently they yield nothing
against precise play.
Finally we come to the Russian System (4.tbf3 �g7 5 .�3 dxc4 6.'i¥xc4 0-0 7.e4) ,
which is my recommended line for White.
76
Chapter 7 - Griinfeld Defense - Non-Exchange lines
.! � j_ if ,.! • �
� � � � � j_ �
.- � .._ ,
I conclude in the White portion of the book that only the Hungarian Variation (7 . . . a6)
gives Black near-equality, so I must recommend it here. White has two main options
against this. He can play the aggressive 8.e5 bS 9.�b3 tt:Jfd7 1 0.e6 ! fxe6 l l .�e3 , after
which White attacks the queenside after the recommended l l tt:Jf6 or the kingside af
..•
ter the dubious I I .. .lLJb6. Black should return the pawn by 1 2.a4 b4! , after which his
better development and piece activity compensate fully for his inferior pawn structure
(Game 7 . 1 1 ) . Or White can play the positional 8.�e2 which we meet by 8 . .b5 9.'in>3 .
cS ! l O.dxcS �e6 l l .�c2 tt:Jbd7!. This leads to positions where White has won the
bishop pair at the price of an isolated d-pawn and a slight inferiority in piece place
ment. I won't claim that Black has 1 00% equality here (if I did I couldn't recommend
this for White) , but I think he is closer to equality than in other major defenses to the
Queen's Gambit. Moreover, the positions are rich enough that either side can reason
ably play for a win. See Game 7. 1 2 .
1. d2-d4 tt:Jg8-f6
2. c2-c4 g7-g6
3. tt:Jb1 -c3 d7-d5
77
The Kaufman Repertoire for Black
B 1 ) 7 .�f4 tt:la6 8.e4 0-0 9 .tt:lf3 cS Three minor pieces are a match for queen
1 0 .e5 (in case of 1 0 .d5 bS 1 1 .tLlxb5 and pawn, but here it is queen and two
tt:Jxe4 1 2 .ihe4 �xb S , Black's safer king pawns so Black is better;
and better development give him the B22) 8 .ti:Jf3 bS 9.ti:Jxb5 tt:lxe4
edge) 1 O . . . �e6 1 1 .exf6 �xc4 1 2.fxg7 B 2 2 1 ) I O .tt:lxc7 tbc6 1 l .ti:Jxa8 'IiVaS +
�xf1 1 3 .gxf8'iV+ 'iVxf8 1 4.Wxfl cxd4 1 2.�d2 ti:Jxd2 1 3 .ti:Jxd2 tt:lxd4 1 4.tt:lc7
1 5 .tLlxd4 ttJb4N 1 6 . .l:i.d 1 l'ld8 1 7 . a3 tt:ldS �c6)
1 8 . tt:lde2 tt:lxf4 1 9 .ti:Jxf4 .!:i:xd l +
2 0 .ti:Jxd 1 'i:Vd8 .
Analysis diagram
Analysis diagram
5. d4xc5
78
Chapter 7 - Griinfeld Defense - Non-Exchange lines
7. ... �f8xc5
8. e3xd4 e5xd4
9. lt:Jg1 -f3 0-0
So Black 'settled' for winning two pieces
1 0. �f1 -e2 tt:Jb8-c6
for a rook, with a clearly winning posi
tion.
K A "iV
1 7. b2-b4 .l:i.f8-e8+
i i
1 8. 'it>e1 -f1 �c5-f8
. �� 1 9. l:ta1 -d1 a6-a5
tt:'J A 20. b4-b5 tt:Jc6-e5
� i 21 . tt:Jf3xe5 �c7xe5
22. .l:i.h1 -h3 �f8xa3
23. �d3xd4 �e5-e2+
24. 'it>f1 -g1 �a3-c5 0-1
79
The Kaufman Repertoire for Black
6. ... �d8xd5
7. tt:Jg1 -e2 0-0
8. tt:Je2-c3 �d5-d8
9. jLf1 -e2 c7-c5
1 0. d4-d5 e7-e6
11. e3-e4 e6xd5
1 2. e4xd5 b7-b6
80
Chapter 7 - Griinfeld Defense - Non-Exchange lines
1. d2-d4 tLlg8-f6
2. c2-c4 g7-g6
3. tLlb1 -c3 d 7-d5
4. �c1 -g5 tLlf6-e4
5. �g5-h4 tLle4xc3
6. b2xc3 d5xc4
7. e2-e3 ]Lc8-e6
8. tLlg1 -f3 c7-c5
1 4. �d1 -b3
If Black wants to play for a win he can try
1 4.�f3 lt:Jd7 1 5 J�e 1 (in case of 1 5 .d6 to hold the pawn by 8 . . . gg7 9 .ge2
�xf3 1 6.'i¥xf3 'i¥f6 1 7 .'i¥d5 'i¥e5 = the tt::l b 6, transposing to Avrukh-Popilski
advanced passer is a bit weak) 1 5 . . . tt::l e 5 below.
1 6.�e2 'i¥d6= and although the queen is
9. �f1 -e2 ]l.f8-g7
not the best blockader, Black can expand
1 0. 0-0 0-0
on either wing.
1 1 . l::t a 1 -b1 c5xd4!
1 4. ... �d8-d6
In an earlier game in the match Kamsky
This is recommended by grandmaster played 1 1 . . . �d5 ? and should have lost.
Delchev. The game move 1 4 . . . lt:Jd7 wasn't
bad, but it seems better to blockade the
.� � 'if .� �
passer and to prevent �f4.
J.' J. J. J. A J.
The actual game went : 1 4 . . . tt::l d 7
A J.
1 5 . � e 3 ( 1 5 .�f4 a6 1 6.a4 J;l e 8 1 7 .nfe I
�d4=) 1 5 . . . �d4 1 6 . J;lad 1 �xe3
1 7 .fxe3 'i¥g5 1 8 .e4 'i¥e3 + 1 9 .�h 1
lt:Je5 2 0 . a4 �ab8 2 I .J;lde 1 �g7 2 2..�.a6
'i¥g5 2 3 .�xb7 llxb7 2 4 .tt::l b 5 J;ld7
2 S .'i¥c3 �g8 2 6 .lt:Ja3 J;le8 2 7 .a5 bxaS
2 8 .'i¥xa5 tt::l d 3 2 9 . J;l d 1 J;lxd5 3 Q .J;ld2
1 2. tLlf3xd4
llxe4 3 1 .'i¥xa7 J;Ifs 3 2 . J;ldd 1 nxfl +
3 3 . llxfl 'i¥f4 3 4.na 1 tt:Jc 1 and White 1 2 .cxd4 'i¥d7 1 3 .tt::l g 5 �dS 1 4.e4 �c6
resigned. 1 5 ..ixc4 'i¥xd4N 1 6 .'i¥b3 e6 1 7 . J;lbd 1
'i¥b2 1 8 .�xe6 'i¥xb3 1 9 .�xb3 tt::l d 7=i=. I
1 5. J:rf1 -d1 tLlb8-d7=
would rather play Black here due to the
Black will centralize his rooks. The pres pressure on the e4 pawn and the excel
sure on the dS pawn restricts White's ac lent coordination of the black minor
tivity. pieces.
81
The Kaufman Repertoire for Black
1 2 . ... �e6-d5
1 3. 'iYd1 -c2 'iYd8-d7
1 4. l:If1 -d1
1 4. ... �g7xd4
1 5. c3xd4
Draw agreed here. A likely continuation
1 S .e4 �eS 1 6..i'hd5 'iYc7 1 7 .g3 f6 would have been 2 3 . . . exf6 24 . .!:tbxc3
1 8.'ij'a4 �g7=. .llx c3 2 S ..!hc3 l:rf7 26.l!c6 .ll e 7 and
White's more active rook is balanced by
1 5. ... 'iYd7-e6
his inferior pawn structure. A draw is the
1 6. �e2-f3 �d5xf3
proper outcome.
1 7. g2xf3 tt:Jb8-d7
1. d2-d4 tt:Jg8-f6
2. c2-c4 g7-g6
3. tt:Jb1 -c3 d 7-d5
4. �c1 -g5 tt:Jf6-e4
5. �g5-h4
1 8. 'iYc2-e4
S .�f4 ti:Jxc3 6.bxc3 jLg7 7 .e3 cS 8 .ti:Jf3
1 8 . llxb7 llfb8 1 9 .d5 'iYh3 2 0 . llxb8+ transposes to the game Elianov-Kovchan .
.!:i.xb8 2 l .'iYe4 Ilb2 22 ..Q.xe7 c3 23 . .Q.d6
5 . ... tt:Je4xc3
hS=. There are many possible lines from
6. b2xc3 d5xc4
here, but most likely either White or
7. e2-e3
Black will end up giving perpetual
check. If 7.'iYa4+ c6 (this is a rare but strong
move) 8.'iYxc4 �aS 9 .ttJf3 �e6 l Q .'ij'dJ
1 8. ... b7-b6
.Q.g7 1 1 .e3 ttJd7 1 2 .'ik'c2 tt:Jcs +, Black
1 9. J:td1 -c1 .!:!.a8-c8
plans . . .�fS with the initiative;
20. l:rb1 -b4 c4-c3
7 .e4 bS 8 .a4 c6 9.tbf3 �g7 1 0 . .Q.e2
21 . J:tb4-b3 tt:Jd7-f6
ti:Jd7 = . White may have enough compen
22. 'iYe4xe6 f7xe6
sation for the pawn, but not enough to
23. �h4xf6
claim any advantage.
82
Chapter 7 - Griinfeld Defense - Non-Exchange lines
9. ... tt:Jd7-b6
1 0. 0-0 �f8-g7
1 1 . a2-a4
83
The Kaufman Repertoire for Black
1 2. ... 0-0
I 'i8 !J *
' &· ' ' .t '
� ..t '
'
1'3:, ' ' 1'3:,
1'3:, [}:, Cjj
'iY � 1'3:,. 1'3:, 1'3:,
l:I �
1 7. ... �a4-c61N
1 3. .!:!.f1 -b1
The actual game continued 1 7 . . . b 5 ? !
1 3 .e4 �d7 1 4.�g3 e6=. White's better 1 8 .�f3 �c6 ? 1 9 .tLlb2 ? ( 1 9 .tbd2 ! was
center and Black's doubled pawn give winning) 1 9 . . . .ic2 2 0 .c4 b4 2 1 .�a2
White enough for the pawn, but no ad tbc3 22 . .ixc6 tbxa2 and Black eventually
vantage considering the weakness on a4; won.
1 3 .tbd2 cS.
1 8. �h4-g3 e7-e6
1 3 . ... 'fk'd8-e8 1 9. tbc4-e5 �g7xe5
20. �g3xe5 f7-f6
Or 1 3 . . . ..id7 1 4.tLld2 �e8 , transposing.
21 . �e5-g3 �e8-d7
1 4. tbf3-d2 .ie6-d7 22. e3-e4 tbd5-e7
1 5. tbd2xc4 .id7xa4 23. f2-f3 f6-f5
1 6. 'fk'c2-a2
1 6 . ... tbb6-d5
1 7. 'fk'a2-a3
84
Chapter 7 - Griinfeld Defense - Non-Exchange lines
9. ... e6xd5
8. 'il¥d 1 -d2
8S
The Kaufman Repertoire for Black
1 O . . . a5 . For 1 O.b4 0-0 1 l .e3 see the next Black having the bishop pair I would pre-
note. fer his chances.
1 4.llfc l l:rc8 1 5 .lLla4 b6 1 6 .llab I tt:lf6=. mally I would stop the score here as the
The bishop pair offsets the weak pawn on remaining moves are not thematic, but as
c6. I've analyzed the game thoroughly I give
my analysis.
1 1 . ... a7-a5
1 6. �e2-d3 �f8-d6?!
1 1 . . . c6 1 2 .0-0 'li'd6= is also okay, but I
1 7. f2-f4 tt:Jb8-d7
wanted to rule out 1 2 .b4 and the minor
1 8. �d2-c2 tt:Jd7-f8
ity attack.
1 9. .!:rf1 -f3 c6-c5?
1 2. b2-b3
should admit my earlier mistake by
Now the minority attack is less effective, playing 1 9 .. .'.tig8 , but White is already
but it's hard to suggest a better plan for clearly better.
White.
20. tt:Jc3-b5
1 2 . ... c7-c6
2 0 . .ld.g3 + .
-
1 3. 0-0 llf8-e8
14. a2-a3 20 ..
. . c5xd4
21 . e3xd4 �d6-b8
22. .l:!.a1 -c1 ?!
2 2 . llafl ! ± .
22 . ... �c8-d7
23. �c2-c5?? tt:Jf8-e6!-+
24. '\Wc5-c3
1 4. ... �g7-f8?!
86
Chapter 7 - Griinfeld Defense - Non-Exchange lines
31 . f4-f5? f7-f6
32. .l:i.f3-g3 g6-g5 . .. ..t 'IV • •
33. 'fk'e2-h5 .!:!:e7xe5 i i i i i ..t i
34. 'fk'h5-g6+ �g7-f8 .. i
35. 'fk'g6xh6+ �f8-e7 i
36. .l:i.g3xg5 tt:Jc7xb5 � � .. � .
87
The Kaufman Repertoire for Black
1 0. c5xd4
1 1 . c3xd4 'ii'd 5-a5+
1 2. 'ifd 1 -d2 0-0
1 7. ... �e6-d7N
1 5 . ... .!la8-c8
88
Chapter 7 - Griinfeld Defense - Non-Exchange lines
E • ..t tt E•
.l' i i ..L
, ..t .L
i,
21 . ... l::tf8-d8
89
The Kaufman Repertoire for Black
E . :E
�· � :�
7. e2-e4
1 3. 0-0
7 .e3 i.e6 8.tt:lgS �dS 9.e4 h6 l O.exdS
hxgS l l .i.xgS tt:lxdS 1 2.�xc4 tt:lb6 1 3 .tt:le2 tt:lc6 1 4.0-0 �xb2N I S Jlb l
1 3 .i.b3 tt:lc6 1 4.dS ( 1 4.tt:le2 aS 1 S .a3 a4 lUeS 1 6 . llxb2 tt:lxc4 1 7 . llxb7 tt:lxe3
1 6 .i.a2 tt:Jxd4 1 7 .tt:lxd4 'iVxd4 1 8.'iVxd4 1 8.fxe3 .!ld2 1 9.tt:lc3 c6 20 . .!lc7 .!lad8
i.xd4 1 9 . .!lxc7 l:Iac8 with equality) 2 J ..!:hc6 .l:!.c2 2 2 .tLldS Iha2=.
1 4 . . . tt:ld4 l S .0-0 'iVd7 1 6 . .!le l .!lfe8 is
1 3 . ... tLlb8-d7
equal. White cannot keep the bishop pair.
He has more space but an isolated 1 3 . . . tt:lc6 ! 1 4.tLldS ( 1 4.tt:le2 i.xb2
d-pawn. I S . .!lb I tt:leS transposes to the game)
90
Chapter 7 - Griinfeld Defense - Non-Exchange lines
.i .. j_
il l �
91
The Kaufman Repertoire for Black
This looks odd but it keeps the knight out White must return the pawn, after which
of e S , keeps open the options of . . . �b7 the bishop pair fully offsets Black's infe
or . . . �a6 or . . . �g4, and prepares . . . .ll f7 rior pawn structure.
and . . . e7-e6 or . . . e7-e5 . 1 6 . . . fxe5 1 7 .tt:lgS .ll e 7 1 8 . .!lc5 h6
19 .tt:lf3 .ll c 8 20.llgc 1 .!lf7 is also equal.
1 3. J:lh1 -g1
1. d2-d4 tt:lg8-f6
2. c2-c4 g7-g6
3. tt:lb1 -c3 d7-d5
4. �c1 -f4 �f8-g7
5. e2-e3
92
Chapter 7 - Griinfeld Defense - Non-Exchange lines
?:;;
.i tD
8 '1'3::, !J.' B !':::,
� 'if w
14. d4-d5
93
The Kaufman Repertoire for Black
1 0. e5-e6
White has only the bishop pair for the
pawn, so he is half a pawn down without I O .fLe2 c5 l l .e6 cxd4!N (grandmaster Pe
.
94
Chapter 7 - Griinfeld Defense - Non-Exchange Jines
95
The Kaufman Repertoire for Block
.i 6 ..t .� .� �
. ··��.� i .. ., i, .1.. i
1 9. ... �g7-h6
i ii; ;m ... .
After 1 9 .. Jlab8! Black's bishop pair and � o/ii .� �
better placed pieces easily offset his poor liS I lZJ
pawn structure. � -� ���
20. tt.:lc3-d1 tt.:lb4-d5 J:ll it w � a:
2 0 . . . a5=.
8. �f1 -e2
21 . 'iVa3-b3 'iVd6-g3??
8 .a4? ! b S ! 9 .1\Vb3 cS ! I O.dxcS i.e6:f
2 I . . . 'i!Vb6=. I I .'i!Va3 ?! b4! 1 2 .'ii'xb4 tt:'lc6+ I 3 .'i!Va3 ?
llb8-+ ; the threat of . . . l:!.b3 forces White
22. 'iVb3xb7 �h6xe3+
to part with material.
23. tt.:ld1 xe3 tt.:Jd5xe3
The recent fad 8 .'ii'a4 is a serious try here,
24. llf1 -f2 tt.:le3-d5
but is well met by 8 . . . c5 ! 9 .dxc5 .td7 = .
25. 'ifb7-c6 tt.:ld5-f4
Now I O .'ifb4 is met by I O . . . aS ! , I O.'ifa3
26. «t>g1 -f1 l:!.a8-b8
by I O ... i.c6 , and I O.'i!Vb3 or I O.'i!Vc2 by
27. �e2-c4 1-0
I 0 . . . 1\VaS ! , in all cases with fine play for
Black.
GI 1 0.5 (D9 7) Game 7.12 8 .'i!Vb3 bS 9 .Yl.e2 transposes to the game;
D Riazantsev,Alexander 8 .�f4 bS 9 .'i!Vxc7 (stem game Euwe
• Rodshtein,Maxim Alekhine, Wch m- 1 2 Netherlands I 9 3 5 )
Moscow, 20 1 1 (5) 9 . . . 1\Vxc? I O .�xc7 Yl.b7 ! .
96
Chapter 7 - Griinfeld Defense - Non-Exchange lines
1 0. d4xc5 �c8-e6
1 1 . �b3-c2 tt:lb8-d7
,�K - ·,.!
.l 1 6. ... �d8-d6!N
.l '�
Previously 1 6 . . . a5 was played.
97
The Kaufman Repertoire for Black
98
Chapter 8
Grunfeld Exchange
In this chapter we look at lines where White exchanges pawns immediately ( 4.cxd5
tbxd5 ) , although 4.tbf3 kg? S . cxdS tbxdS will transpose to that chapter.
X � ..t � • .t X
� � � � � �
White plays ttJf3) to play 8 ... e5 ! , and to meet the bizarre 8 .kb5 by . . . �e6. We also
look at the equally bizarre 5 .tba4, which is better than it looks. I recommend S ...�g7
6.e4 lbb6. The basic rule is that it is okay for the knight to be driven to b6 if White
has paid a price for this.
Now we come to the real Exchange Griinfeld, 5 .e4 tbxc3 6.bxc3 �g7.
99
The Kaufman Repertoire for Black
In Game 8.2 we consider all the infrequent seventh moves. The queen check is met by
7 ...tLld7 intending . . . 0-0, . c7-c5 , . . . a7-a6, . . . b 7 -b6, and . . . i.b7 . Although Black does
. .
not put heavy pressure on White's center, White pays a price for this with his wander
ing queen. The move 7 .�a3 is similarly met by 7 ... tt:Jd7, intending . . . c7 -cS . Instead,
7.�gS is met by the immediate 7 ...cS, trying to provoke d4-d5 . The main move in the
game, 7.�bS+, is met by 7 ... c6 8.�a4 0-0 9.tLle2 bS 1 0.�b3 aS with good counterplay:
In all these sidelines Black is fine.
In Game 8 . 3 we consider unusual eighth moves after 7.tt:Jf3 cS . Against 8.�e2 we
pressure the center by 8 ...tt:Jc6 9.�e3 �g4. Against 8.h 3 , pressure on d4 won't work,
but after 8 ... 0-0 9.�e2 tLlc6 1 0.�e3 we switch to an attack on e4 by 1 0 ...cxd4 1 1 .cxd4
fS ! . As for the game move 8.�bS + , we block with the knight on c6, castle and play
. . . .ig4, leaving out the pawn exchange on d4. Again all the lines are harmless.
Now we come to the important lines with i.e3 , meaning either 7.tLlf3 cS 8.�e3
�as 9.�d2 0-0 1 o..!:tc 1 or Kramnik's preferred order 7.�e3 cS 8J:k 1 �aS 9.�d2 0-0
1 o.tt:Jf3.
I favor putting direct pressure on the center with 1 O .. l:!.d8 . If White just develops we
.
force dS by 1 1 .�e2 �g4, and then play . . . c5-c4 before White can do so. If White ad
vances l l .dS we play . . . e7 -e6. These lines are a bit complicated and must be studied
1 00
Chapter 8 - Griinfeld Exchange
carefully. Points to remember: . . . lt:Ja6 is often a good move, . . . c5-c4 as a pawn sacrifice
is sometimes viable, and often the a2 pawn becomes a target. See Game 8 .4. I didn't
fmd any advantage for White, but this is a critical line and White might yet find some
thing.
The next four games all feature the so-called 'Modern Exchange' variation, meaning
7.lt:Jf3 cS s.nb 1 0-0 9.�e2 . Games 8 . 5 , 8 . 6 , and 8 . 7 all feature the popular sequence
9 ...cxd4 1 O.cxd4 �aS+ 1 l .�d2 ( 1 1 .�d2 leads to an endgame which is covered in
Game 5) 1 l ...�xa2 1 2.0-0. This is a rather dangerous (for Black) gambit by White,
whose compensation for the pawn is rather obvious (several tempi and an extra pawn
in the center) . The usual move here (recommended by Avrukh) is 1 2 . . .�g4, after
which White generally regains his pawn and Black has to play precisely to reach an
endgame that he can draw. Instead I go for the second-most popular move, 1 2 ...b6 ,
which intends to hold the pawn, at least for a while. After 1 3.�cl �b7 White can force
an immediate repetition, but few players will choose a gambit as White just to force a
quick draw at the first opportunity. In Games 8 . 5 and 8 . 6 I give two reasonable contin
uations for Black which appear to equalize, though they are complex and tricky. In
Game 8 . 7 I give the greedy 1 3 ...�e6 in case you can't allow the repetition draw. But my
preferred option, given in Game 8 . 8 , is the same line as recommended by Delchev and
Agrest, namely 9 ...ll:Jc6 1 O.dS ll:JeS 1 1 .ll:JxeS hes 1 2.�d2 e6 1 3.f4 �c7 ! , which pre
vents the normally desirable 1 4.c4. There are some scary lines here, but it seems that
Black can hold the balance.
Now we come to the old way of playing the Exchange Variation, which has again
become the main line. This is 7.�c4 cS 8.ll:Je2 lLlc6 9.�e3 0-0 1 0.0-0. In Game 8 . 9 I
give my second-string defense, 1 O ... e6, which aims to restrain the White d-pawn,
and also gives the queen the square e7 where it is less subject to attack than on c 7 .
Still, I prefer 1 0 ... �c7 1 1 .l::t c 1 b 6 (Game 8 . 1 0) , primarily because we may want to
play a later . . . e7 -e5 without the loss of a tempo. Some of the resultant positions (after
. . . e 7 -e5 and d4-d5) are evaluated as good for White by the computers, but I don't
trust computer evaluations much in highly blocked positions such as these. Comput
ers love protected passed pawns, but when they are securely blockaded by a knight
they aren't much of an asset.
At the last moment I decided to include coverage of the 1 O ...b6 gambit line as a third
black option. That was Griinfeld guru Peter Svidler's choice against World Number 1
Magnus Carlsen just days before finishing the book. At this writing, my analysis indi
cates that it may be the best choice of all for Black, since the gambit seems quite sound
if accepted, while if it is declined Black has the choice transposing to I 0 .. .'� c7 or aim
ing instead for . . . e7 -e6 and . . .'�i:Ye7 , which is probably a bit better. See Game 8 . 1 1 .
101
The Kaufman Repertoire for Black
5. ilc1 -d2
Analysis diagram
1 02
Chapter 8 - Griinfeld Exchange
7. �d2-e3 0-0
.i � .t 'i¥ E �
.t. .t. .t. .t. .t. .t .t.
� .t.
Analysis diagram
88
ttJ it. I 2 .i.xc4 ( I 2 .�b3 lt:lxb2 ! = ; note that
88 888 this would be bad with Black's a-pawn on
a6. This shows the point of the 8 .�b5
l:t 'iY w it. tt:J l:t
move) 1 2 . . . tbxc4 I 3 .�b3 lt:lxb2
1 4.'ifxb2 flic7 I S .i.b4 a5 1 6.tbge2 axb4
8. h2-h3
1 7 .'ifxb4 fS ! N ( 1 7 . . . �d6=; 1 7 . . . 1i..g 4=;
A) 8 . .tbs (this bizarre but strong move 1 7 . . . b6!N also favors Black) I 8.0-0 fxe4
aims to provoke 8 . . . a6 9.�e2 as in some 1 9 . .ll ac i i.fs 20.a4 'ifes +. With two
variations the loss of protection for the bishops for two knights in a very open
knight by the a7 pawn is important) position, Black can afford to lose his b7 or
8 . . .�e6 9 . tbge2 (9.tLlf3 �g4 I O .�e2 e4 pawn and still have enough compensa
transposes to the 8 .tLlf3 line, with each tion. White is unlikely to hang on to both
side having lost a tempo) 9 . . . c6 I O .�d3 of his isolated pawns for long;
lt:lc4 I I .�xc4 1i..x c4 1 2 . 0-0 lt:ld7 D) 8 . f4 lt:lc6 9.d5 lt:laS I O.�d4 e5
1 3 .'iYd2 �aS 1 4. �fd 1 .!:!.fd8=. This is a I I .i.xe5 i.xe5 l 2.fxe5 �h4+ 1 3 .g3
computer improvement over I 4 . . . �ad8, �e7 I 4.'i¥d4 .!:td 8 I S .b4 tbac4 with
which was equal in Wang Yue-Carlsen, equality.
Medias 2 0 1 0, won by Black. Presumably
8. ... e7-e5!
the idea is to retain the bishop pair if
White plays �h6 now or later. Either way 8 . . .f5 is often played (I played it once my
Black is fme; self) but inferior.
B) 8.lt:lf3 1i..g 4 9.1i..e 2 tbc6 1 O.dS i.xf3
1 1 .gxf3 ( 1 Ltxf3 tLleS 1 2 .�e2 tbec4
1 3 .1i..c i c6 1 4.dxc6 bxc6 I 5 .�c2 'ifd4
I 6 .0-0 �fd8 1 7 .a4 �cS+) 1 I . . .lt:la5
I 2 .1i..d4 'ifd6 1 3 .1i..x g7 Wxg7 I 4.�d2 .t.
'iff6 1 5 .b3 l:tad8 =; 88
C) 8 .�e2 lt:lc6 9 .d5 (9 .lt:lf3 1i..g 4 trans ttJ it. 8
poses to the 8 .tbf3 line) 9 . . . tbeS I O.�d4 8 8
8 '8
c5 ! I I .1i..x c5 ( I I .dxc6 tbxc6 l 2 .�xg7 ' 'i¥ <;t> it. ttJ �
:
Wxg 7 +) 1 I . . . lt:lec4
1 03
The Kaufman Repertoire for Black
1 5. a2-a3
I S . .!lei tt:lc6 1 6 .lLlb5 llfc8 1 7 .i.cs 'i¥f6 Black's bishop is superior to White's
1 8 .tt:lbd4 i.d7 1 9 . .!ld l lLla4=. knights.
1 04
Chapter 8 - Griinfeld Exchange
I also like 9 . . . eS .
1 0. �a4-b3 a7-a5
1 05
The Kaufman Repertoire for Black
1 9. ... c6xd5
20. h5xg6 f7xg6
21 . 'irc1 -h6+?!
11. ..
. tt:Jb8-d7
Here a draw was agreed, although Black is
1 2. h2-h4 a5-a4
clearly better after 22 . . . �fS . White's at
1 3. �b3-c2 tt:Jd7-b6
tack has failed, his king is in the center,
14. 'ird1 -c1 tt:Jb6-c4
and his queenside pawns are effectively
1 5. �e3-h6 e7-e5
all three isolated from each other. Curi
1 6. �h6xg7 Wg8xg7
ously, the two players in this game and I
1 7. h4-h5 'ifd8-e7
ended up tied for 1 st place (along with
1 8. �c2-d3 �c8-e6
one other grandmaster) , with Vaisser tak
ing first and Jansa second on tiebreak
points. Perhaps Jansa would have won this
World Senior Championship if he had not
agreed to a draw here.
1 06
Chapter 8 - Griinfeld Exchange
1 07
The Kaufman Repertoire for Black
1 8. �c2xe4 J:!.f7xf3 . .. .t tv � :1
1 9. �e4xe7 J:!.f3-f7 .l' .l .l .l .l '.t .l
.
·
20. �e7-h4 , .l
-�
K .t .l
.l
'i¥
!'::,, � 7. tt.:lg1 -f3
,!'::, !'::,
7 . .te3 cS 8J:tc 1 (Krarnnik 's preferred
l:L w
move-order) 8 . . . �a5 9.�d2 0-0 1 0.t2Jf3
transposes.
20 . ... �c6-e6
7. ... c7-c5
There was nothing wrong with regaining
8. .2.c1 -e3
the pawn by 20 . . . il.xc3=.
This variation is very popular recently
21 . c3-c4 il.g7-f6
with strong grandmasters.
Better and equal was 2 l . . .�e5 , prevent
8. ... �d8-a5
ing Ild6 while keeping an eye on h2.
9. �d 1 -d2
22. �h4-g3 �e6-e4
9 .t2Jd2 and now:
In the game Black played to regain his A) 9 . . . cxd4 is a reasonable alternative to
pawn by 22 . . . �a6, but after 2 3 .�e4 castling. It avoids the draw given below,
�xa2 24.�d7 he had to fight for the draw, but I consider castling to be the more
successfully. The text equalizes, as the iso correct move. 1 O.t2Jc4 ( 1 O.cxd4 t2Jc6
lation of all five white pawns surely offsets 1 l .d5 tt.:ld4 1 2 .l:!.c1 .td7 1 3 .�d3 �a4
White's one pawn lead in material. 1 4. 'ti'g4 il.d7 1 5 .'ti'd1 draw by repeti
tion)
1 08
Chapter 8 - Griinfeld Exchange
I O . . . 'i¥d8 ( I O . . . dxe 3 ?! I I .tlJxaS il.xc3+ tt:le2 + 2 l..� xe2 il.xe2+. With the bishop
1 2 .�e2 ..ixaS I 3 .Wxe3 was Aronian pair in a wide open position and no bad
Sutovsky, Ningbo Wch-tt 2 0 I I . Black has pawns or pieces Black is better.
two bishops, a pawn, and some threats
1 2. ... f7-f6
against the exposed white king for his
queen, but it wasn't enough and Black
lost. The computers consider this sacrifice
unsound) I l .cxd4 fS 1 2 . .!lb i fxe4
I 3 .�e2 0-0 I 4. 0-0 il.e6 I S . .!lxb7 �ds
I 6 . .!lbS ttJc6 I 7 .'i¥d2 e6=. Although
Black has one more isolated pawn than
White, his pieces are active and central
ized;
B) 9 . . . Qd 7 ? ! and now:
B I ) I O.tt:lb3 'i¥xc3 + I I .Qd2 'i¥b2
1 2 . ..ic i 'i¥c3 + 1 3 .Qd2 was a quick draw
1 3. �g5-f4
by repetition played in Shulman
Dominguez Perez, Lubbock 2 0 I I . 1 3 .c4 �xd2+ 1 4.il.xd2 .l::!.e 8 1 S .�e2
B 2 ) I O . ..ie2 'i¥xc3 I I. . i::!: c l 'i¥b2 tt:ld7 1 6 .il.e3 exdS 1 7 .exdS fS =.
1 2 . .!lc2 'i¥b4 I 3 . .!lxcS tt:lc6 I4 . .ll c4 'i¥b2 1 3 .il.e3 tt:lc6 1 4 . .2.d3 ( 14 ...ie2 exdS
I S .dS tt:ld4 is also equal. I S .exdS c4 1 6 .lld 1 ttJe7 1 7 .�xc4 �e6
B 3 ) but I 0 Jlc I ! is a correct and strong transposes) 1 4 . . . exdS 1 S . exdS c4!
pawn sacrifice according to Komodo. 1 6.il.xc4 �e6 1 7 . � d 1 tt:le7 I 8 .dxe6 ! ?
( 1 8 . 0- 0 �xdS I 9 .'i¥e2 Wh8=)
9. 0-0
1 8 ... .!lxd2 I 9 ..lhd2 'iYxc3 ( 1 9 ... tt:lfS ! ?
1 0. l:ta 1 -c1 l:tf8-d8
2 0 . .ll d 3 .l:le8=, if you wish to avoid an
1 1 . d4-d5
immediate draw) 2 0 . .2.b 3 �c 1 +
I I . ..ie2 �g4 I 2.dS c4 I 3 . 0-0 tt:ld7 2 I .lld ! 'i¥c3 + 2 2 . W e 2 . This was
I 4.h3 il.xf3 I S .�xf3 bS I 6 . .ll fd l 'i¥a3 = . Potkin-Svidler, Moscow ch-RUS 2 0 1 0
Black's mobile queenside majority and and Pashikian-Cornette, Aix-les-Bains
pressure on White's two weak pawns off Ech 20 1 1 . White avoided the drawing
set the bishop pair. 2 2 .lld2 in both games but both were
drawn anyway. With a rook and bishop
1 1 . ... e7·e6
and bishop pair for the queen White is
1 2. �e3-g5
the equivalent of a pawn down, but his
I 2..�e 2 exdS I 3 .exdS b S ! I 4 ...ixcS �b7 strong passer evens the chances. Svidler
I S .c4 'i¥xd2 + I 6.tt:lxd2 bxc4 1 7 .il.xc4 played 2 2 . . . aS , Cornette chose
�xdS =; 22 . . . 'i¥b4, and the computer likes
1 2 .c4 'i¥xd2+ 1 3 .tt:lxd2 b6 I 4.�e2 ttJa6 22 . . . 'i¥as 2 3 .lld3 fS = .
I S .0-0 tt:lb4 I 6.a3 tt:la2 1 7 .llc2 ttJc3
1 3 . ... 'ifa5·a4
1 8 . ..if3 exdS 1 9 .cxdS ?! ( 1 9 . exdS �fs
1 4. c3-c4
2 0 . .!lcc 1 .!lacS=) 1 9 . . . il.a6 2 0 Jifc 1
1 09
The Kaufman Repertoire for Black
1 4. ... e6xd5
1 5. e4xd5 tt:Jb8-a6
1 6. .!tf1 -e2 tt:Ja6-b4
White seems to have enough compensa
1 7. 0-0 �a4xa2
tion for the piece with his threats against
1 8. �d2-e3 .1i.c8-f5
the queen and the b7 pawn to get a draw
1 9. �e3xc5
based on the following analysis, though
this is far from certain.
22 . ... �a2-a3
23. �b5xb7 tt:Ja6-b4
24. Iie1 -e7 .!tg7-f8
25. .!:Ie7xh7 �f5-e6
26. .iic 1 -e1 .!::!:a 8-b8
27. �b7-e4 �a3-d3
28. W4'e4xe6+ <Jtg8xh7
29. �e6xf6 �d3-f5
30. �f6-h4+ �f5-h5
1 9. ... tt:lb4-a6! 31 . �h4-f6 �h5-f5
V2-1h
This is a computer improvement over
the actual game, which continued
I 9 .. .lt:Jxd5 ? ! 2 0 . cxd5 'i¥xe2 2 I..!He l ? ! GI 4. 1 4 (D8 5 ) Game 8.5
( 2 1 .�c 7 ! lld7 2 2 . d 6 ± ) 2 1 . . .'i¥a2 D Wells,Peter
2 2 .�b5 'iYa6 ? ! (22 . . Jhd5 2 3 .'i¥xb 7 • Michaelsen,Nils
.!ladS+) 2 3 . 'i¥b 3 = 'iYb6 24.'ifa2 'ifa6 Israel, 1 9 8 8
2 5 .'iVb3 'il'b6 with a draw by repeti
1. d2-d4 tt:lg8-f6
tion.
2. tt:lg1 -f3 g7-g6
20. �c5-b5 3. c2-c4 .1i.f8-g7
4. tt:lb1 -c3 d7-d5
2 0 .'ife7 'i¥b2 2 I ..llce i .lle 8 2 2 .�c l
5. c4xd5 tt:lf6xd5
'ifxc i +.
6. e2-e4 tt:Jd5xc3
20 . ... 'fka2xe2 7. b2xc3 c7-c5
21 . .!::!:f 1-e1 �e2-a2 8. lla1 -b 1 0-0
22. d5-d6 9. �f1 -e2 c5xd4
I I0
Chapter 8 - Griinfeld Exchange
1 4. �e2-c4 �a2-a4
1 5. �c4-b5 �a4-a2
1 6. l::i:f1 -e1
1 6 . ... l::U8-c8
1 7. �c1 -d1
1 1 . �c1 -d2
I I�
j_ J.
i 17. ... e7-e6!
1 8. �b5-a4
1 1 . ..
. �a5xa2 1 9. ... e6xd5
1 2. 0-0 b7-b6 20. �a4-b3 �a2-a3
1 3. �d1 -c1 21 . e4xd5
1 3 . .l:!e l i.a6 1 4.i.g5 �e8=. After 2 l .i.xd5 'i/e7 2 2 .i.g5 'i¥f8 White
has compensation for the pawn, but per
1 3. ... i.c8-b7
haps not quite enough.
Ill
The Kaufman Repertoire for Black
23 . ... �d4-c3!N
1. d2-d4 tbg8-f6
2. tbg1 -f3 g 7-g6
3. c2-c4 �f8-g7
4. tbb1 -c3 d7-d5
5. c4xd5 ttJf6xd5
1 12
Chapter 8 - Griinfeld Exchange
1 13
The Kaufman Repertoire for Black
w 9. ... c5xd4
1 3. 'li'd1 -c1
1 14
Chapter 8 - Griinfeld Exchange
1 4. �e2-c4 1 7. . .
. f7xe6
1 8. tt:Jf3-g5
In case of 1 4 . .ll e 1 i1l.a6 1 5 .�d 1 .ll c 8
1 6 .�a3 �d7N 1 7 .dS �c4 White has full
compensation for the pawn but perhaps
no more than that.
1 4. �e6xe4
1 5. .l:!.f1 -e1
22 . ... e7-e6
23. �a3xf8+ l:Ia8xf8
24. l:Ib1 -c1 �g7-g8
25. �c4-b3 l:If8-e8
26. 'ife1 -e5 t2Jc7-d5
This line, also recommended by Delchev
27. l:Ic1 -e1 �g8-f7
and Agrest, is my first choice. There is not
28. h2-h4 h7-h5=
so much you need to learn here, and the
line seems to be sound.
1 0. d4-d5
1 16
Chapter 8 - Griinfeld Exchange
ous advanced passed pawn and much 1 4 . ..tc4 a6 1 5 .a4 bS 1 6.axb5 axbS
better coordinated and developed pieces, 1 7 . ..txbS exdS 1 8 . ..tc6 lla6 1 9 .�xdS
so full compensation. lld6 2 0.0-0 �fS 2 I .'ifc2 llxdS 22.exf5
After 1 2.'ifc2 e6 1 3 .f4 �g7 1 4.dxe6 llxfS 2 3 .c4 'ifd4+=;
..ixe6 1 5 J lxb7 'iVaS 1 6 .e5 Itab8 1 4.dxe6 1i'xd2+ 1 S . .ixd2
1 7 .llxb8 llxb8 1 8 .�d3 ? ( 1 8 . .ie3N �fs
1 9 . .id3 �xd3 2 0.�xd3 llb2=)
1 8 ... ..ixa2 1 9. 0-0 .ib 1 2 0 .'ife2 'ifxc3
2 1 .�c4 .tfs Black is just a pawn up.
1 2. ... e7-e6
1 3. f2-f4
1 3 . ... �e5-c7
1 17
The Kaufman Repertoire for Black
1 6.fS .ixfS 1 7 .llxb7 'tiYd6 1 8 . .ic4 l:tfb8 In case of 1 9 . .l::rb 3 li'c8N 20Jle 1 bS
1 9 .l:tb3 (or 1 9 .l:txb8+ l:txb8 2 0 .'iib' g S Black clearly has the initiative a t n o cost.
.ixc3 2 1 JhfS .id4+ 2 2 .'1t>h l l:tb4 White's queenside pieces are just targets.
2 3 .�d3 f6 24.'iib'g 4 �eS 2 S . �c4 �d4
1 9. ... �a5-b6
2 6 .�d3 with a draw by repetition)
20. Wg1 -h1 �b6xc5
1 9 . . . l:txb3 2 0.axb3 l:te8 2 1 .l:te 1 l:teS
21 . �a3xc5 �d8-d7
2 2 .l:te3 �c7 2 3 .g3 hS 24.l:txeS 'iib'x eS
2 S .'iib'e 3 M=.
1 6 . ... lla8-b8
1 7. �c1 -a3
1 18
Chapter 8 - Griinfeld Exchange
1 1 . �d1 -d2
Black has superior pieces, superior pawn 1 1 .dxc5 'iiaSN 1 L�.b3 b6 1 3 .cxb6 axb6
structure, and a safer king. 1 4.'i¥c2 .ia6 1 5 . llfd l l:lfc8 - Black plans
to bring his knight to c4 which should
give him good compensation for the pawn
GI 5 . 7 (D8 6) G ame 8.9 as his pieces will be more active and White
D Svetushkin,Dmitry has more weak pawns than Black.
• Zhou Jianchao 1 1 .llb 1 ( 1 1 .llc l can be met by 1 1 . . .'VIie7)
Khanty-Mansiysk ol, 20 I 0 (3) 1 1 . . .cxd4 (this exchange tends to favor
White, but here it is played to avoid a later
1. d2-d4 <1Jg8-f6
dxcS and is j ustified by the loss of tempo
2. c2-c4 g7-g6
that llc l would imply) 1 2 .cxd4 b6
1 19
The Kaufman Repertoire for Black
1 3 .�a4 ..Q.b7 1 4. .l:tfd 1 (else 1 4 ... a6 will 2 9 .�xg5 gxh5 3 0.�xh 5 + Wg8
be a good move) 1 4 .. .toa5 1 5 .�d3 .ll c 8=. 3 l .�g 5 + Wf8 3 2 .�h6+ would have led
Black plans ... a7 -a6 and ... b7 -b5 . to a draw by perpetual check.
1 6. .. . h7·h5
1 7. �e3·g5 �d8·d7
1 8. e4·e5?1 J:!.c8xc1
1 9. J:!.d1 xc1 t2la5·c6
20. �d3·b5 J:!.f8·c8
21 . �g5·f6 it.g7-f8 27. Ild6·d7??
22. �d2·g5 a7·a6
27 .d5 ! Ire 1 2 8 .tt:le2 .l:tc8 29 .tt:lf4 was a
23. .!::rc 1 xc6
draw by repetition.
1 20
Chapter 8 - Griinfeld Exchange
4. c4xd5 tt:lf6xd5
5. e2-e4 tt:Jd5xc3
.
6. b2xc3 �f8-g7
7. �f1 -c4
1 1 . J::l:a 1 -c1
12I
The Kaufman Repertoire for Black
rook is needed on f8 .
1 8 .f4 exd4 1 9 .cxd4 .!lfe8 ? ! ( 1 9 . . . cxd4!
1 4. .\:&h6xg7
2 0.lLlxd4 .!:!.fe8 2 1 .e5 b S ! =) 2 0 .e5 ? !
1 4 . .ll fd 1 eS 1 5 .�xg7 �xg7 1 6 .d5 ttJaS ( 2 0 .d5t) 2 0 . . . cxd4 2 1 .lLlxd4 1tVdS ? !
1 7 .�d3 c4 1 8 .i.c2 i.c8 1 9 .lLlg3 lLlb7 ( 2 1 . . .b5= ) 2 2 .tLlf3 ? ! (22. . �fd 1 t)
2 0 .ttJfl ttJd6= is similar to the game. 2 2 . . . lLlc4= and Black eventually won in
Here too the computer likes White due to Korobov-Nepomniachtchi, Rogaska
the protected passed pawn, but I think Slatina tt 2 0 1 1 .
Black is fine thanks to his ideal blockad 1 8 .d5 lLlc4 1 9 .�d3 bS 20.a4 a6 2 1 .ld.b 1
ing knight. 1tVd7 2 2 .lLl g3 was Zilka-Heimann,
1 22
Chapter 8 - Griinfeld Exchange
1 7. tt:le2-g3
26 . ... h7-h6
1 7 . f3 fS ( 1 7 . . . tLlc8=) 1 8 . I:tce 1 f4
1 9 .1Wd3 tLlc8 2 0 .�b3 tLld6 2 l .c4 bS 2 6 . . . tLlbS 2 7 .1lxf8 .l:!.xf8 2 8.�a4 tLld6
2 2 .tLlc3 a6 - the computer favors White 2 9 .�c2 tLlbS would draw by repetition.
here because of the protected passed Probably Black wanted to win.
pawn, but I think Black is fine with his
27. l:lf2xf8 �d8xf8
wonderful blockading knight and much
28. tt:lh2-f3 :cas-es?!
superior bishop, which will drop back to
c8; 28 . . . l:.f4 2 9 .lle l tLlf7=.
1 7 .a4 f5 1 8 .f3 tLlc8 1 9 .I:tcd l tLld6=.
29. ..t>g1 -f2 �c8-g4
1 7. ... f7-f6 30. tt:lf3-h2 �g4-d7
31 . ..t>f2-e3 g6- g5
1 7 . . . f5 1 8 .exf5 gxfS 1 9.1l¥g 5 + �h8
32. h4xg5 h6xg5
20 .l:i.cd 1 �xdS=.
33. tLlh2-f3 g5-g4
1 8. h2-h4 tt:le7-c8 34. tLlf3-h4 l:Ie8-f8
35. �c1 -h1 .!:!:f8-f4
1 8 . . . f5 =.
36. g2-g3 �f4-f6
1 9. �c4-d3 tt:lc8-d6 37. a2-a4 J:rf6-h6
38. .!:!.h1-b1 .l:Ih6-h8
1 9 . . . c4! 2 0 .�c2 tLld6 2 l . f4 exf4
39. .i:tb1 -h1 tt:ld6-e8
22.�xf4 1l¥e7 = (Avrukh) .
40. l:th1 -f1 tt:le8-d6
20. f2-f4 41 . .!:H1 -h1 �h8-h6
42. l:th1 -b1 ..t>g7-f7
20.c4t.
1 23
The Kaufman Repertoire for Black
1. d2-d4 t2Jg8-f6
2. c2-c4 g7-g6
3. t2Jb1 -c3 d7-d5
4. c4xd5 t2Jf6xd5
5. e2-e4 t2Jd5xc3 Analysis diagram
6. b2xc3 �f8-g7
was better according to Komodo, to dis
7. �f1 -c4 c7-c5
courage d4-d5 and to provide a good
8. t2Jg1 -e2 t2Jb8-c6
square for the queen on e7) 1 4.cxd4
9. �c1 -e3 0-0
"ild6 1 5 .d5 tbaS 1 6.�b5 .l:I.xc l 1 7 . .l:I.xc l
1 0. 0-0 b7-b6
llc8 and was eventually drawn. Now
1 8 Jl:xc8 + i.xc8 1 9 . ti:Jd4 would have
given White some edge.
1 1 . ..
. 'ifd8-c7
1 2. t2Je2-d4
The alternatives:
1 2 .cxb6 axb6 1 3 . .l:lb 1 li:JeS 14 . ..id5
llb8 .
1 24
Chapter 8 - Griinfeld Exchange
Analysis diagram
1 25
The Kaufman Repertoire for Black
25 . ... ttJe5-c4!
1 26
Chapter 9
The Center Game starts with l .e4 e5 2.d4 exd4 3.�xd4 tLlc6 .
Now the normal retreat for the queen is to e 3 , although the alternate choice of a4 has
long been championed by veteran IM Walter Shipman. It has the drawback of failing to
defend the e4 pawn as securely though. After 4.�e3 t2Jf6 5.t2Jc3 �b4 6.�d2 0-0
7.0-0-0 l;Ies
.I ..t � .l •
l i ... i : i i; i
.. . . .. f�ili
127
The Kaufman Repertoire for Black
White must choose between the 'safe' 8.�c4 and the gambit 8 .�g3 . In the former case
Black's threats against the e4 pawn and the c4 bishop tend to put him a bit ahead in the
attacking race. The gambit move 8.�g3 can be accepted either by 8 .. Jhe4 (usually
recommended) or by 8 ... tt:lxe4, my choice due to the given game, which may lead to
Black getting four pawns for a piece. See Game 9 . 1 .
after which Black can equalize either by 3 ... tLlf6 or by 3 ...d5 . I choose the second move
because it offers more winning chances, and also because my computer analysis actu
ally shows a tiny edge for Black. White's best move is 4.'il¥a4, after which 4 . . . �d7 is an
exciting gambit recommended in my earlier book. However my current computer
analysis makes it look Jess appealing than the safe 4. . f6, which seems to give Black a
.
good position for free. If White responds S.d3 as in the given game, play resembles the
Philidor/Hanham Defense but with colors reversed. The Hanham ( l .e4 d6 2 .d4 tt:lf6
3 .tt:lc3 eS 4.tLlf3 tt:lbd7) is not such a bad defense to l .e4, but it does not give full
equality, and even with the extra tempo of White's first move it is hardly something to
avoid as Black, especially since the move �a4 may turn out to be inappropriate. See
Game 9.2.
1 28
Chapter 9 - Center Game and Pomiani
1 3. wc1 ·b1 ?I
1 29
The Kaufman Repertoire for Black
1 30
Chapter 9 - Center Game and Pomiani
131
The Kaufman Repertoire for Black
tbeS 2 7 .tbxe5 .llx eS 28 .�fl g6 29.tLlb3 Or l l .cxd4 tt::l g 6 1 2.g3 dxe4 1 3 .tt::lxe4
£t.xb3 ?! (Black plays for a draw from a i.h3 , and White's castling problems give
better position, probably due to the huge Black the edge.
rating difference) 3 0 .axb3 .ll e 7 3 1. .�e2
11. �e6-f7
llc7 3 2 .Ildl .llc 3 3 3 Jhd5 .llx b3 34.f4.
1 2. 0-0 ttJb8-d7
Somehow White won this completely
equal and drawish position. White's pieces and pawns seem to be dis
tributed rather randomly, while Black
1 1 . ttJf3xd4
dominates the center.
132
Chapter 10
and now 3.tLlc3 or 3.d3 , I recommend avoiding the move 3 . . . tLlc6 , instead planning to
put a pawn on that square. There are two reasons for this. First of all, I don't favor plac
ing the knight on c6 if White fails to attack the e5 pawn, because it's generally not ad
visable to obstruct pawns which are not part of the castle. After all, I am recommending
the Breyer Defense to the Spanish, in which Black actually retreats the knight volun
tarily from c6 to b8 and then moves it to d7. Why go to c6 in the first place if you don't
need to do so? The second reason is that the moves ..tc4 and tLlc3 both invite a black
pawn to c6. In the former case it is because of the possibility of ... d7 -dS to chase the
bishop, while after tbc3 the pawn move takes away the two most advanced squares to
which the knight might otherwise later go. In general, it is advisable to place your
pawns in front of the enemy knight with two squares in between for precisely this rea
son.
When White plays 3.d3 in the Bishop's Opening, I advocate the immediate 3 . c6, to ..
achieve a quick ... d7 -dS. White has nothing better than 4.tt:Jf3 dS , when after 5.�b3 I fa
vor the bishop check on b4. The idea is that after the obvious 6.c3 we can retreat to d6,
no longer having to fear tLlc3 with pressure on our dS pawn. Assuming a later tt:Jbd2, we
end up with a position somewhat similar to the Breyer, but this time we are playing
against it! The Breyer is an excellent defense to the Spanish, but it is still not quite equal,
133
The Kaufman Repertoire for Black
so if we can play against it when we are Black we should have no complaint! White can
instead invite a bishop trade by 6.�d2, but the trade is fine for Black, as a recapture by
the knight again forfeits the chance to pressure dS by lbc3 , while the queen recapture
allows Black a very good pawn sacrifice option. See Game 1 0 . 1 for all of this.
If White chooses 3.tbc3 (or reverses his second and third moves) , I favor 3 ...�c5 .
With the knight on c3 the plan with . . . d7 -dS is not so inviting, as the pawn comes un-
der heavy attack, so the move ... c7 -c6 is not urgent. After 4.d3 c6 S.lbf3 we typically
play . . . d7 -d6, ... lbbd7, . . . �b6, ... h7 -h6, and . . . tbd7 -f8-g6, taking advantage of the de-
lay in castling. This way we may be able to dispense with . . . . .ld.e8 later. This plan seems
to give full equality. See Game 1 0 . 2 .
I fWhite does not play �c4 i n the Vienna (after 2.tbc3 tbf6) ,
he can choose 3.f4, which is a kind of 'King's Gambit' which Black should not accept
due to 4.e5 . Fortunately, the reply 3 ...d5 ! seems to solve all problems. After the usual
4.fxe5 tLlxe4 I don't even know how White can equalize. See Game 1 0 . 3 .
1. e2-e4 e7-e5
2. jLf1 -c4 tLlg8-f6
3. d2-d3
1 34
Chapter 1 0 - Bishop 's Opening and Vienna
We can compare this to our recom 6.c3 �d6 (if Black plays S . . . �d6 , White
mended Breyer. White has saved a tempo has the better prospects after 6.tLlc3)
in getting his bishop to b 3 , while Black 7 .tt:Jbd2 (7 .£.gs dxe4 8.dxe4 h6 9.£.h4
has saved two tempi in getting his knight 't/Ve7 I O .tiJbd2 tiJbd7 1 l .'t/Ve2 �c7
to d7 . White has not yet played h2-h3 or 1 2.0-0 0-0 1 3 .h3 tLlcS and with ideas
c2-c3 , but probably will. Black has not like ... b7 -b6 and . . . �a6 or . . . a7 -aS or
played ... a7 -a6 or . . . b7 -bS but may not . . . ttJxb3 or ... ttJc5-e6-f4 Black is for
need to do so. He has played the useful choice) 7 . . . 0-0 8 .0-0 tt:Jbd7 9.lle 1 lle8
. . . c7 -c6 and has not committed his 1 o.ttJfl h6 1 t .t2lg3 tt:Jf8 1 2.h3 tt:Jg6
bishop to b 7 , which is often not a very
good square against the d3 set-up. In
i. .i. 'i¥ 1: �
short, Black is playing a much improved
i i i i
Breyer Defense. 8.c3 ttJcS 9 . ..ic2 ..ig4
(9 .. .'tWc7 is also okay. Black is playing the
i .i. ,. ,. ,
Hanham Defense with White having
i i
played the passive d3 instead of d4) !'!,
1 O .h3 ..ihS 1 1 .tLlbd2 tLle6 1 2 .tLlfl tiJd7 .; � !'!,, !'!, /iJ Cfj !'!,
1 3 .tLlg3 ..ixf3 1 4.'ti'xf3 g6 1 S . ..th6 lle8 � :!'!, ' !'!, !'!,
1 6.llad 1 �gS 1 7 .�xgS 't!VxgS 1 8.d4 :s � 'if � �
tLlb6 1 9 .�b3 llad8=. With all the pawns Analysis diagram
1 35
The Kaufman Repertoire for Black
i. � .t . • i.
.� � � � �. �
'ill �
�� 'iV
1 4. ... a7-a51
136
Chapter I 0 - Bishops Opening and Vienna
1 9. ttJc3-b5 �d6-b6
20. c2-c4 e5-e4
21 . d3xe4 d5xe4
22. ttJf3-d4 .l:!.f8-d8
5. t2Jg1 -f3
137
The Kaufman Repertoire for Black
138
Chapter I 0 - Bishop's Opening and Vienna
bxc6 1 1 .tt:le2 �cS Black has the bishop After 8 .�b S + tt:lc6 9.0-0 0-0 1 0.a3
pair while his isolated pawns are all on ii.xc3N 1 l .�xc3 �b6 1 2 .1txc6 bxc6
closed files and not easily attacked. 1 3 .dxcS �xeS+ 1 4.�d4 �e7 Black is
again better due to the strong knight and
5 . ... �f8-c5
the option of . . . c6-cS .
6. d2-d4
8. ... d5xe4
6."ife2 �f2+ 7 .Wd 1 tt:lxc3+ 8 .dxc3
9. �d2xb4 c5xb4
(8 .bxc3 �cS 9.d4 �e7 1 0."iff2 0-0
1 0. tt:Jf3-g1 0-0
I I .ii.d3 cSN Black has better pawn struc
11. �f1 -c4 tt:Jb8-c6
ture and a safer king) 8 ... �b6 9.�gS
1 2. c2-c3 b4xc3
"ifd7 1 O.�d2 0-0 1 1 .l::!.d 1 �a4 1 2 .a3
1 3. b2xc3
�fS Black has better development and
better pawn structure.
6. ... �c5-b4
7. �c1 -d2 c7-c5
�
!JL � .l
�
� ��
� .l � :a: - 'i¥ � tt:J Z1
.i. � .. 1 3 . ... �d8-c71
t2J t2J
� � � !JL �� Kramnik actually played 1 3 . . . � g S ?
1 4.�e2 ..tfs 1 S .tt:lh3 ..txh3 1 6.gxh3 .
:a: 'iV � !JL :g:
Now 1 6 . . . tt:laS would have kept the ad
vantage. He played 1 6 ... e 3 ? and went on
8. tt:lc3xe4
to lose.
In case of 8.a3 1txc3 9 .1txc3 tt:lc6
1 4. �c4-b3 tt:Jc6xe5
I O .�bS cxd4 l l .�xd4 0-0 1 2.1txc6
15. d4xe5 �c7xc3+
bxc6 1 3 .�b4 l::!. e 8 1 4. 0-0 �b6 I S .1td6
1 6. �e1 -f2 �c3xe5
�g4N Black is better thanks to his strong
1 7. l:Ia1 -c1 �e5-f6+
outposted knight;
1 8. �f2-e1 �c8-e6
If 8 .1td3 tt:lxd2 9 .�xd2 cxd4 1 0.tt:\xd4
tt:lc6 I l .tt:lxc6 bxc6 1 2 .0-0 0-0 1 3 .�hl Black has three pawns for a knight, better
�aS Black has the bishop pair, a pin, and development, and a much safer king. He
White's eS pawn is weak; should win.
1 39
Chapter 11
Gambits
fn this chapter we consider all the lines in which White starts with 1 .e4 and subse
quently sacrifices a pawn, usually for development. It is said that 'the only way to refute
a gambit is to accept it', and this is in most cases my advice. It is also said that having ac
cepted a gambit, you should look for a way to return the material to achieve a good po
sition, and this too is often my preference. But it's not good to generalize too much;
each gambit should be studied independently. These gambits are mostly pretty rare in
rop level chess, but amateurs will often encounter them, and will be likely to fall into
..uious traps if they have not studied the gambit in question.
Black can accept (2 . . . exf4) and then return the pawn by 3 .tt:J f3 dS or try to hold on
w it by 3 . . . g5 or by Bobby Fischer's recommended 3 . . . d6. Since I believe that Black
has the better chances after returning the pawn, I see no real point in learning the
lines where he tries to hang on to it. However I recommend the move-order l .e4 eS
2.f4 dS (!) 3 .exd5 exf4 4.tbf3 , which transposes to 2 . . . exf4 3 .lbf3 dS 4.exd 5 . The
point is that with the normal move order White can play the Bishop's Gambit
3 .i.c4 when after 3 . . . dS he can choose either capture, with the bishop capture be
ing generally considered the better one. However by playing . . . d7 -dS on move 2 we
cut out that option; White can still choose 4 . .tc4 but he has already taken on dS
with the pawn.
Back to 4.tbf3 , Black should develop by 4 tbf6 rather than expose his queen by
...
4 . �xd5 . White has a couple ways to try to come out a pawn ahead, but they don't end
. .
well for him. If White plays normal moves he often has to surrender the bishop pair to
�vin back his pawn on f4, which is the main reason I like this line for Black.
141
The Kaufman Repertoire for Black
In Game 1 1 . 1 we look at 5 .�c4, the most popular move. In the game White manages
to regain the pawn without losing the bishop pair, but he ends up with much worse
pawn structure. In Game I 1 . 2 we consider all other fifth moves, where White's attempl
to stay a pawn ahead is refuted by a novelty in the note to move 1 2.
is a rather dangerous one for Black to accept by 4 . . . <1Jxe4, although the computers pre
fer Black. We cannot play 4 . . . tbc6 as I'm not recommending the Two Knight's Defense
to the Italian. Fortunately for us the move 4. .�b4 seems to solve all of Black's problems
.
is well met by 3 ...'fk'e7!. Technically this is 'declining' the gambit as White could pta,
4.'ifxd4, but that clearly favors Black, so White should renew his offer by 4.cxd4 'ifxe4+,
when White struggles to prove compensation for the pawn. This line was popular around
1 900 and was shown to me as a kid by an elderly Norman Whitaker, a top player from t:hf
1 920s and now known to have had a long criminal career as a swindler. See Game 1 1 .4.
1 42
Chapter I I - Gambits
The Goring Gambit ( I .e4 eS 2.tt:Jf3 lt:Jc6 3 .d4 exd4 4.c3) had some popularity in my
youth, but is fairly uncommon now. Still, it pays to know what to do.
Declining by 4 . . . d5 gives fairly equal chances, but I think Black should try to refute this
gambit by accepting 4 ...dxc3 . Then play branches, depending on whether White sacri
fices a second pawn by S .�c4 or just recaptures S.lt:Jxc3 . In the latter case it seems to
me that White lacks sufficient play for the pawn unless he chooses to head for an even
endgame. See Game 1 I .S for this and for the Scotch Gambit 4.�c4, which is met by
4 ... �cS 5 .0-0 d6 6.c3 �g4 with only White having problems.
Perhaps the most respectable of the gambits in this chapter is Evans' Gambit. It is attrib
uted to a Captain Evans around the year I 8 3 0 , not to the late grandmaster Larry Evans.
For the rest of the 1 800s it was practically the main line of chess, but it almost died out
around 1 900 due in part to 'Lasker's Defense' . Kasparov brought it back from obscurity
with a few wins in the mid- I 990s, including one over his then main rival Anand. It
was again revived by Nigel Short, especially with a game against Nielsen which cast
doubt on the line I recommended for Black in my previous book (the game was played
just weeks after I completed that book) . Things are still not completely clear, but I be
lieve I have found a way to get the better chances for the black pieces. I also give a safe
line in case I'm wrong. The gambit goes: l .e4 eS 2.lt:Jf3 lt:Jc6 3.�c4 �cS 4.b4!?.
I 43
The Kaufman Repertoire for Black
The idea is that White normally wants to play 4.c3 and S .d4 here, but this has draw
backs, so White hopes to achieve this with a gain of a tempo at the price of the b-pawn.
Normally you need at least two tempi and more typically three to justify a gambit, but
the tempo is quite valuable in this position due to threats on f7 . Black should accept by
4. .. �xb4, when after 5.c3 �aS is the best winning try (S . . . �e7 is the safest move, for
which see Game 1 1 .6) and after 6.d4 I now recommend 6 ...d6 rather than 6 . . . exd4 due
to the above-mentioned Short game. When White attacks f7 by 7.�3 we defend with
7 .'t!Yd7. If White then takes on eS we return the pawn by 8 . . . it..b 6 with the idea of
••
. . . tlJaS , when Black should end up slightly better. If White does not take on move 8
Black should be able to retain the pawn without conceding too much compensation..
See Game 1 I . 7.
KG 2.8 (C3 6) Game 1.1..1. l O .tlJeS tlJxeS l l . llxf4 'ifg6 , but with a
D Mista,Aleksander losing position;
• Roussel Roozmon,Thomas After 6 .CiJc3 tlJxc3 7 .dxc3 �xd l +
Brno, 2008 ( 1 0) 8 . �xd l kd6 9.tle l + �f8 White has lit
tle for the pawn.
1 . e2-e4 e7-e5
2. f2-f4 d7-d5 6. ... �c8-e6
3. e4xd5 e5xf4
6 . . . it..e 7 7 .�xd5 (7 .d4 gS 8 .tlJc3 is a
better try for White) 7 . . . �xd5 8.tlJc3
�d8 9 .d4 gS also favors Black, but not
9 . . . 0-0? 1 0 .�xf4 as in Carlsen-Wang Yue,
Medias 2 0 1 0, won by White.
4. ttJg1 -f3
1 44
Chapter 1 1 - Gambits
king. The white knight blocking his c KG 2.9 (C3 6) Game 11.2
pawn makes any queenside play difficult D Sulskis,Sarunas
for White. • Melkumyan,Hrant
Benasque, 2009 (8)
7. ... �f8-e7
8. d2-d4 1. e2-e4 e7-e5
2. f2-f4 e5xf4
After 8 .lt:Jc3 c6 9 .d4 g5 1 0 .lt:Je4 lt:Jd7
3. tt:lg1 -f3 d 7-d5
1 1 .�b3 h6 1 2.c4 lt:J5f6 White has only a
little compensation for the pawn. I prefer the move order 2 . . . d5 3 .exd5
exf4 to rule out taking on d5 with the
8. ... 0-0
bishop in case of �c4 instead of lt:Jf3 .
9. tt:lb1 -c3 tt:ld5xc3
1 0. b2xc3 �e6xc4 4. e4xd5 tt:lg8-f6
11. 'iWe2xc4 �e7-d6
1 2. 'iWc4-b5 b7-b6
1 3. tt:lf3-g5
5. �f1 -b5+
1 45
The Kaufman Repertoire for Black
regained his pawn, but the price of hav A greedy computer suggestion is
ing to advance his king was too high. 1 2 . . . .lle 8N 1 3 .tt:lc3 (in case of 1 3 .kxf4_
�xf4 1 4.l::!.xf4 �b8 1 S .�fl �xb2 Black
5. ... c7-c6
will win a pawn) 1 3 . . . eS 1 4.�c4+ �h
6. d5xc6 tt:lb8xc6
1 S .dxeS .Yl.xeS 1 6.�xf4 �b6 + 1 7.�h l
7. d2-d4 �f8-d6
�xb2 1 8 . .Yl.xeS :!lxeS 1 9 .�xc6 llae8
8. 'ifd1 -e2+
2 0.h3 �xc2 . Here Black is for choice as
8 . 0-0 0-0 9 . tt:lbd2 �g4 1 O .c3 (or his pieces are slightly better placed.
1 0.tt:lc4 .Yl.c7 l l ..Yl.xc6 bxc6 1 2.lLlceS
1 3. tt:Jb1 -d2
.Yl.xf3 1 3 .tt:lxf3 l:!.e8 1 4.�d3 aS 1 S .�d2 .
tt:le4 1 6Jlae 1 fS and despite his inferior 1 3 .�xe6+ �h8 1 4.�h3 ( 1 4.tt:'ld2 llae8
pawn structure, I prefer Black due to the 1 S .�h3 cS 1 6.tt:lc4 f3N 1 7 .tt:lxd6 �xd6
powerful outposted knight) 1 O . . J:te8 1 8 .�h 1 cxd4 1 9.�xf3 tt:Jds 20.�dr
1 1 .tLlc4 .Yl.c7 1 2 .�d2 �ds 1 3 .tt:la3 tt:le4 l:rxfl + 2 1 .'iix fl l:rf8 2 2 .�g 1 tt:'lb4
and again the powerful e4 knight out 2 3 _jt,d2 tt:'lxc2 Black regains the pawn
weighs the doubled pawns. with advantage due to the advanced
passer) 1 4 . . . .lla e8 1 S .tt:lc3 ttJdS 1 6 .kd2
8. ... �c8-e6
tt:le3 1 7 .�xe3 llxe3 1 8 . .llf3 �b6
9. tt:Jf3-g5 0-0
1 9.llxe3 fxe3 2 0.�xe3 �xb2 2 1 .l:lfl
1 0. tt:Jg5xe6
�xc2=. A draw is to be expected.
Not 1 0 . .Yl.xc6? �g4.
1 3 . ... J:!.a8-e8
1 0. ... f7xe6 1 4. tt:Jd2-c4
1 1 . �b5xc6 b7xc6
1 4. ... g7-g5
1 46
Chapter 1 1 - Gambits
1. e2-e4 e7-e5
2. �f1 -c4 tt:Jg8-f6
3. d2-d4 e5xd4
4. tt:Jg1 -f3
1 47
The Kaufman Repertoire for Black
1 48
Chapter I I - Gambits
1 . e2-e4 e7-e5
2. d2-d4 e5xd4
3. c2-c3
1 49
The Kaufman Repertoire for Black
1 7 Jhfl =.
1 7. ... �e7xe5?!
Analysis diagram
1 7 . . . tt:ld7 ! .
1 4 ... �a6 ! I S .'iYxa6 tt:lb4 1 6.�e4 'iYxe4
1 8. .!:ta1 -e1 �e5-c5+
1 7 .tt:lxe4 tt:lxa6 White may have compen
1 9. Wg1 -h1 lLlb8-c6
sation for a pawn, but he is down two!
20. .l:!.f3-f5 �c5-d6
5. ... tt:lg8-f6 21 . .!:tf5-d5 �d6-h6
6. lLlb1 -c3 �f8-b4 22. �d2-e2?! �h6-f6
7. tLlg1 -f3 lLlf6-d5 23. l::!.d5-f5
8. �d 1 -d2 �e4-e7
9. �f1 -c4
1 50
Chapter 1 1 - Gambits
5. ... �f8-b4
Analysis diagram
6. �f1 -c4 �b4xc3+
151
The Kaufman Repertoire for Black
8. ... lt:Jg8-f6
13 . ... jLc8-d7
After 9.e5 tt:JxeS 10.ti:Jxe5 dxeS With 15.f4 0-0 16.fxe5 tt:Je4 17.l:!.fe1
1l.�xd8+ (11.'t�Hb3 'i¥e7 transposes to �c6 18.�c4 tt:Jgs 19.�xc5 ld.fc8
the game) 1l...Wxd8 12.jlxf7 We7 20."ihc6 ti.xc6 White regains his pawn,
13.i.b3 ..ie6 14..lle1 �xb3 15.axb3 but now he has a clearly worse endgame
tbd7 16.f4 aS 17...ie3 .llhf8 18.fxe5 We6 with three isolated pawns.
Black is for choice, as the king is a great
15. ... tt:Jf6-e4
blockader and the white bishop is of the
16. �b3-b4
wrong color to disturb it. Also Black has a
potential outside passed pawn. After 16.�a3 tt:Jd2 17.�b4 0-0-0
18.l:!.fd1 l::the8 19.c4 .ld.e6 White has
9. ... �d8-e7
nothing for the pawn.
152
Chapter 11 -Gambits
6. d2-d4 tt:lc6-a5
!53
The Kaufman Repertoire for Black
1. e2·e4 e7·e5
1 3 . ... 'ifa5·a6! 2. lLlg1 ·f3 lLlb8·c6
3. �f1 ·c4 �f8·c5
Black actually played 1 3 . . . ..tb4 (in
4. b2·b4
Robson-Perelshteyn, Lubbock 20 1 0, Black
played 1 3 . . . �h5 ?? and had to resign after Nigel Short is the top practitioner of the
1 4.ltJd5 !) 1 4 . .txb4 �xb4 1 5 .�d3 b6 Evans Gambit.
1 6.ltJc3 ..td7 1 7 . .!lfd 1 l:!.ad8 1 8.a4. White
4. ... �c5xb4
was slightly better due to his central con
5. c2·c3 �b4-a5
trol, though Black drew. The text move is a
computer improvement over both games,
preserving the bishop pair.
1 4. 'ifd1 ·f3
1 4. ... c7·c6
1 5. .!:!.f1 ·e1 l:!.f8-e8
1 54
Chapter 1 1 - Gambits
think this is better than 6 . . . exd4 as 9.i.bS a6 1 0.�a4 (after 1 0.�xc6 'i¥xc6
recommended in Chess Advantage, due 1 1 .<1Jbd2 f6N White is down a pawn and
partly to Short's 2 0 0 3 game with Nielsen the bishop pair with no clear plan of at
with 7 .�b3 . tack) 1 0 . . . i.a7 and the threat of . . . b7-b5
ensures a good game for Black.
7. �d 1 -b3
9. ..
. <1Jc6-a5
In the event of 7 . 0-0 i.g4 8 .�b3 �xf3
1 0. 'irb3-c2 f7-f6
9.�xf7 + rt!;f8 1 0.gxf3 (after 1 0.1lxg8
1 1 . a2-a4
lhg8 1 1 .gxf3 exd4 1 2.�xb7 �e8 Black
is much better) 1 0 . . . 1lb6 1 1 .1lxg8 I:rxg8
1 2 .d5 tt:las 1 3 .�b5 �h4 White's awful
pawn �.tructure gives Black the edge.
7. ... 'ird8-d7
1 1 . ... tt:Jg8-e7!N
1 55
The Kaufman Repertoire for Black
Next Black can play . . . �e6 with an extra White has some but not full compensa-
pawn. tion for the pawn.
1 56
Chapter 12
In the Scotch, which was resurrected two decades ago by World Champion Garry Kasparov,
after 3 .d4 exd4 4.tbxd4 (4.c3 is the GOring Gambit in the Gambit chapter, while the
Scotch Gambit 4.�c4 �c5 5 .c3 tbf6 is simply a transposition to the 4.c3 lLJf6 5.d4 exd4
line of the Italian Game) I recommend 4...tbf6 rather than 4 ...�c5, mainly because after
the latter move, 5.lLJxc6 'ii'f6 6.'ii'f3 ! seems to offer White a slight edge in most if not all
lines. Now after 4. . . tt::lf6, 5.tbxc6 bxc6 6.e5 �e7 7.�e2 tbd5 8.c4 my earlier book recom
mended" s ....lta6, which often leads to an endgame where White has rook, bishop, and
knight vs. two rooks and two extra but doubled pawns. This endgame is about equal but
rather hard to win as Black, and since 8 ... tt::lb 6 has been revitalized recently I now recom
mend that move. If White responds with 9.tbd2 or 9.b3 we play 9 ... a5, aiming to exchange
off this pawn getting rid of an isolated pawn while giving White a backward one on b3
(see Game 1 2 . 1 ) . I don't even see how White fully equalizes in this line. After the more ag
gressive 9.tbc3 the rare 9 ...�b7 was played successfully by Kramnik this year and this
seems to solve all ofBlack's opening problems (see Game 1 2.2).
The Scotch Four Knights Opening is not very promising for White, but there are
some tricks so you need to study the line. Black has to be willing (in the lines I rec
ommend) to put up with a bad pawn structure in the endgame in return for the two
bishops, but with queens off the bishops are a major plus so this is a good enough
deal for Black. If White avoids this ending he may achieve a slightly better pawn
1 57
The Kaufman Repertoire for Black
structure but at the cost of conceding central domination to Black. See Game 1 2 . 3 for
this as well as the Belgrade Gambit S.tt:ldS ! ? , which we meet by S ...tt:lb4 when I don't
see how White equalizes.
Unusual fourth moves after the four knights come out, including 4.g3 and 4.a3 , should
lead to equality with correct play. In the former case simple development with 4 ...�c5
and 5 . . . d6 suffices, while after 4.a3 Black should respond with 4 ... d5 when White may
find himself just playing the black side of the equal Scotch Four Knights with the use
less move a3 included.
The sacrificial 4.tt:lxe5? is unsound but tricky. Keeping the extra piece is probably
theoretically best, but for practical play just retreat the knight to c6 when it is attacked
and give back the piece by answering d5 with . . . �b4, after which Black is just better.
See Game 1 2 .4 for these three lines.
Most novices play 4.�c4 after the Four Knights Opening, until someone teaches
them not to do so due to the 'fork trick' 4 . . . tt:Jxe4. However, computer analysis fails to
show full equality for Black in the traditional lines after that move. Fortunately I discov
ered a major novelty here (S.tlJxe4 dS 6.�d3 tt::lb 4!) which appears to give Black abso
lutely equal chances.
The main line of the Four Knights is 4..ltb5 , but recently 4...�c5 . formerly thought
bad due to the 'fork trick' (again!) , now appears to be a complete solution to the prob
lem of that opening, thanks to the discovery of the strength of 9 . .�e7 ! . For both of
.
1 58
Chapter 12 - Scotch and Four Knights Opening
�
9. b2-b3 .i �
9.tt:ld2 aS and now: ���
A) 1 o.'iWe4 g6 1 l .i.d3 i.g7 1 2. 0-0 :g:
0-0 1 3 .tt:lf3 �a6 1 4.i.g5 f6 1 5 .�h4
9. ... a7-a5
llfe8 1 6.c5 �xd3 1 7 .�xd3 tt:ldS 1 8 .�c4
1 0. �c1 -b2 a5-a4
�f7 1 9 .exf6 tt:lxf6 20.�xf7+ \t?x£'7
11. ttJb1 -d2 a4xb3
E E" 1 2. a2xb3 .!:ta8xa1 +
.t. .t. � :t .t. 1 3. �b2xa1 �e7-a3
.t. -� .t. 1 4. �e2-d1 �f8-b4
�
Analysis diagram
1 59
The Kaufman Repertoire for Black
9. ... ..tc8-b7
1 60
Chapter 1 2 - Scotch and Four Knights Opening
161
The Kaufman Repertoire for Black
32 . ... tt:Jb2-d1
33. '1Wf6-h8+ Wd8-d 7
34. �f1 -b5+ c7-c6
35. �b5xc6+ wd7-c7
36. d5-d6+ '1Wd2xd6
.! A if * ..t .!
-� � i 'i i " i i i
'·�. �
.
1 62
Chapter 1 2 - Scotch and Four Knights Opening
This position is known as the 'Scotch After 1 1 .lL'la4 h6 1 2 . ..ih4 l:Ie8 1 3 .c4 ..ig4
Four Knights' as it can arise from either 1 4.�xg4 lt:lxg4 1 5 .ii.xd8 Ilaxd8
the Scotch or the Four Knights Opening. 1 6.cxd5 l:Ixd5 = Black's better placed
5 .lL'ld5 is the Belgrade Gambit. I recom pieces offset his inferior pawn structure.
mend 5 . . . lt:lb4 and now: The bishops of opposite color make a
A) 6.lL'lxf6+ 1Wxf6 7 . ..ic4 ..ic5 8.0-0 draw likely.
d6 9.e5 dxe5 l O.lL'lxe5 0-0 l l .lL'ld3 1 1 .tLle2 h6 1 2 . ..ih4 j;_d6 1 3 .lL'ld4 c5
lt:lxd3 1 2.j;_xd3 j;_d6 and Black is just a 1 4.tLlf5 ii.xf5 1 5 .�xf5 .llb 8 1 6.b3 ..ie5
pawn up; 1 7 .l:Ic 1 ii.f4 1 8 .llb 1 �d6 1 9 .g3 g5
B) 6.lL'lxd4 lt:lxe4 7 .lL'lb5 lt:lxd5 8.1Wxd5 20.gxf4 gxh4 2 1 .1Wf3 llb4 22 .c4 <;t>h8=.
'ii' e 7 !N (also suggested by Andrey White and Black have equally bad pawns.
Obodchuk in his book The Four Knights
1 1 . ... ..ib4-d6
Game) 9 .1Wd4 d5 1 Q.j;_f4 tLld6+ 1 1 .<ot>d 1
lt:lxb5 1 2. .�xb5 + c6 1 3 .lle 1 �e6 l l . . . h6, provoking 1 2 . ..ixf6 'i¥xf6
1 4.ii.d3 �c5 - White has some develop 1 3 .'lj'xf6 gxf6, is fine if you want to draw,
ment for the pawn but with his king stuck while 1 1 . . .ge7 is best if you must avoid a
in the center that should not be enough; draw. The text is a good compromise,
C) 6 . ..ic4 lL'lbxd5 7 .exd5 j;_b4+ 8 . ..id2 since White will have to lose a tempo to
�e7 + 9 .�e2 j;_xd2+ I O.Wxd2 �xe2+ reach the above endgame and so rarely
1 1 .Wxe2 0-0N 1 2.lL'lxd4 a6 1 3 .a4 .lle 8+ plays this way here.
14 .<.t>d3 d6 1 5 . .llhe 1 ..id7 - Black has the
better bishop, the safer king, and the op
tion to play . . . c7 -c6 or . . . c7 -c5 at any mo
ment.
5. ..
. ..if8-b4
6. lt:ld4xc6 b7xc6
7. jLf1 -d3 d7-d5
8. e4xd5 c6xd5
9; 0-0 0-0
1 0. jLc1 -g5 c7-c6
1 2. .ttf1 -e1
,j_ � ·�
1 2 Jhe 1 .l:!.b8 1 3 .lL'ld 1 h6 1 4 . ..ixf6 1Wxf6
.l .l .l
1 5 .�xf6 gxf6 1 6.b3 ..id7 1 7 .lt:le3 ..ib4
� 1 8 . .l:!.d l ii.c5 = . Black's bishop pair offsets
� his bad pawns;
1 2 .h3 llb8 1 3 . .llab l h6 1 4 . ..if4 l:le8
1 5 .b3 I!b7 1 6.�xd6 �xd6 1 7 .llfe l
.llb e7 and Black's control of the open file
offsets his slightly inferior bishop.
1 63
The Kaufman Repertoire for Black
1 64
Chapter 1 2 - Scotch and Four Knights Opening
Analysis diagram
7. ... ttJc6-d4
1 65
The Kaufman Repertoire for Black
_,
1
ii
[!J [!J l!J Analysis diagram
l3J l3J ;�
A) Everyone plays 6 . . . dxe4?! 7 .�xe4
b(
�d6 8 .0-0 0-0 here, but after 9 .lle l
(9 .c3 is also good) we reach the position
1 4. ... f7-f5=
discussed in the note to move 8 below,
1 4 . . . �g4 ! ? is the computer's preference: but with colors reversed. I call it slightly
1 5.h4 �h6 1 6.llf2 fS=. better for Black there, which means
1 66
Chapter 12 - Scotch and Four Knights Opening
167
The Kaufman Repertoire for Black
1 0. i.b5-e2
1 68
Chapter 13
Italian Game
The Italian Game, or Giuoco Piano, is the favorite opening of novice players, but it is
also used by grandmasters-and even World Champions on occasion. In my opinion, it is
the most promising alternative to the Spanish for White, and should keep at least a tiny
edge.
After l .e4 eS 2.lbf3 lbc6 3 ...Q.c4, I recommend the safe 3 ...�c5 rather than the Two
Knights 3 . . . ltJf6, which is a risky gambit due to 4.ltJg5 . In general, computer analysis
does not show full compensation for the pawn in that line, though Black certainly has
decent practical chances.
In Game 1 3 . 1 we examine the gambit line 4.0-0 lbf6 S.d4 ..Q.xd4 6.lbxd4 lbxd4 7.f4.
The recommended response (after 7 ...d6 8.fxe5 dxeS 9 ...Q.gs 'fte7 I O.lba3) is an aston
ishing computer suggestion, the ridiculous looking I O .. J:tg8 ! . I will do my best to ex
plain it. Note that I don't cover 4.d4, because 4 . . . exd4 transposes to the Scotch Gambit,
although 4 . . . �xd4 is also a good move.
Now after 3 . . . ..tc5 4.c3 (for 4.b4 see the Evans Gambit in the Gambit chapter, while
4.ltJc3 tt:lf6 5 .d3 h6 is a safe line for Black) 4... lbf6 White usually plays S.d3 in master
chess, because the obvious S .d4 has supposedly been analyzed to equality. However as I
show in Game 1 3 . 2 , White has two paths to a tiny edge after this move, so it deserves a
higher status and more people should try it. Still, it's nothing for Black to be scared of,
as White's edge is probably less than he can obtain in the main line Spanish. As the
positions after 5 .d4 exd4 6.cxd4 �b4+ 7...Q.d2 (Greco's ancient 7 .ltJc3 is a dubious
gambit) are rather op�n. tactics prevail and there's not much point in talking about
strategy, except that when White accepts an isolated d-pawn in return for piece play,
Black should generally seek equal exchanges.
1 69
The Kaufman Repertoire for Black
After S.d3 , I recommended S . . . a6 in my earlier book, but recently 5 ... 0-0 has been
shown to be a good move (Carlsen chose it not long ago) , since the pin 6.�g5 is well
met by 6 ...h6 7.�h4 �e7 ! , 6.b4 is met by 6...�e7! , and 6.�b3 is met by 6 .. d5 . Induc .
ing White to castle cuts out some options like tt:lb 1 -d2-fl -g3 (without the need for
lle 1 ) and also takes the sting out of the pin 7 .�gS after 6 . . . d6, since White can no lon
ger attack an early . . . gS by h4 with a rook behind it. After 6.0-0 the move 6 . . . d5 is
sometimes played, but here it is rather risky so I advise 6 ...d6. Then play can take on a
symmetrical nature if both players bring their queenside knights to the kingside, White
by .l::i: e 1 and tt:lb 1 -d2-fl -g3 , Black by . . . tt:'lc6-e7-g6. This results in p ;�itions where
White's edge due to having the move is really tiny. Moreover, with all the pieces on the
board and a fluid pawn structure Black has every right to play for a win if he is the
stronger player. See Game 1 3 . 3 .
1 70
Chapter 1 3 - Italian Game
1 0. ... l:th8-g8!
171
The Kaufman Repertoire for Black
Analysis diagram
1 72
Chapter I 3 - Italian Game
Analysis diagram
I 73
The Kaufman Repertoire for Black
1 6 . ... tt:Jb6xc4N
1 74
Chapter 1 3 - Italian Game
6. ... d7-d6
C 1 ) S .d4 i.b6 9.dxe5N lt:Jxe4 1 0.'i¥e2
tLlcS 1 l .exd6 ( 1 I .i.c2 dxeS 1 2 .l';Id1
:i ..t tv
�eS 1 3 .'ihe5 li:Jg6=) 1 l . . .cxd6 1 2 .i.c2
fl & &
.lle S 1 3 .l::re 1 i.fS and Black's lead in de
�& �
velopment offsets the isolated pawn .
..t1
& Other moves are similar to the game:
�8• C2) S . h3 tt:lg6 9 .lle 1 h6 1 0.d4 i.b6
'� 8 8 tb _ 1 l .li:Jbd2 lieS 1 2 .i.c2 c6 1 3 .tLlfl dS !N
8 ?8; .8 8 !::, (White keeps a pull against all other
.tr tt:J � if �� moves) 1 4.exd5 (if 1 4.dxe 5 ? ! lt:Jxe4
1 S .i.xe4 dxe4 1 6.�xdS l::rx dS 1 7 .l::rxe4
Our plan as Black is to transfer this knight l::r d 1 1 S.l::r e 1 �xe 1 1 9.tt:lxe 1 ..te6 Black
quickly to g 6 , its optimum square. will easily regain the pawn on eS with a
clear bishop-pair advantage. I saw this
7. h2-h3
whole sequence at move 1 3 in a game,
A) 7 .tLlbd2 tt:le7 S .d4 exd4 9 .cxd4 i.b6 but failed to see that the pawn on eS was
1 0.e5 ( 1 0 .h3 dS=) 1 0 . . . li:Jfd5N 1 1 .tLle4 doomed, so I wrongly rejected 1 3 . . . d5 !)
h6 1 2 .exd6 cxd6 1 3 .'i¥b3 lt:Jc7 1 4.d5 14 ... exd4 1 5 .dxc6 dxc3 1 6.'i¥xdS �xdS
..tfS = 1 5 .tt:lg3 i.h7 1 6.a4 tLla6 1 7 .i.d2 1 7 .bxc3 bxc6=.
1 75
The Kaufman Repertoire for Black
1 3. jLc4-b3
Or:
A) 1 3 . ttJhS �c7 1 4.�xdS cxdS
I S .ttJxeS �xf2+ 1 6.�xf2 ttJxeS 1 7.d4
1 0. ... d6-d5
ttlg6=. Black's safer king nearly offsets his
A good alternative was I O . . . h6 1 1 .tLlg3 isolated pawn. The bishops of opposite
(after 1 1 .�b3 �b6 1 2 .ttlg3 !le8 1 3 .d4 color make the isolated pawn safe here;
.ie6 White's edge is miniscule) 1 1 . . . dS B) 1 3 .d4 exd4 1 4.tt:Jxd4 �b6 I S . ..ixdS
( 1 I . . J:te8 - perfect symmetry! Usually if cxdS and Black's bishop pair roughly bal
this is the case after ten or more moves it ances the isolated d-pawn;
means that White has lost most of his ini C) 1 3 .'iifc 2 .l:!.e8 1 4.�d2 �e6 1 5 J:tad i
tial advantage. 1 2.d4 �b6 gives White �c7=.
just a tiny edge) 1 2.�b3 dxe4 1 3 .dxe4
13 . ... l:!.f8-e8
'iifx d 1 1 4 . .ld.xd 1 aS=.
1 4. jLc1 -d2
11. e4xd5 ttJf6xd5
In case of 1 4.d4 exd4 I S Jhe8+ �xe8
l l . . . cxdS 1 2 .�b3 �d6 1 3 .d4 exd4 1 6.�xdS cxdS 1 7 .tt:lxd4 b6 the bishop
1 4.tt:Jxd4 �d7 1 S .ttlg3 l:lfe8 1 6.�e3 pair balances the isolated pawn.
l:teS 1 7 . .ic2 .l:!.ae8 1 8.�d2 aS. The iso
1 4. ... .ic5-b6
lated d-pawn gives White a tiny edge,
though with all pieces on the board Black 1 4 . . . �e6 I S .�c2 �b6 was a more accu
can play this way. I prefer the game con rate move-order, transposing to the game
tinuation though. while making I S .d4 less effective.
1 2. ttJf1 -g3 h7-h6 1 5 . �d1 -c2
1 2 . . . �c7 is perhaps objectively the best 1 S .d4 exd4 1 6. lhe8+ 'ifxe8 1 7 .tt:Jxd4
move: 1 3 . .ixdS ( 1 3 . .ib3 �b6 1 4.�e2 �xd4 1 8. cxd4 �e6=.
1 76
Chapter 1 3 - Italian Game
1 77
Chapter 14
Spanish Offshoots
In this chapter we consider all the ways White can vary from the main line of the
Spanish (Ruy Lopez) Opening from moves four through nine. These are sidelines for a
reason; generally speaking Black can achieve approximate equality if he follows my rec
ommendations. Moreover Black has fair winning chances in most of these lines; my
own record as Black in these sidelines is extremely good. But if you are unprepared,
they can be rather dangerous.
The idea of the Spanish move 3 .�bS is to exert indirect pressure on the eS pawn,
hoping to wring a concession from Black. It turns out that the ' threat' to win a pawn is
an illusion, because after 4.�xc6 dxc6 S .tLlxe S ? 'ifd4 6.tt:lf3 'ifxe4 Black will just be up
the bishop pair in an open position with queens exchanged, which favors the bishop
pair. But once e4 is guarded Black usually responds to the threat by . . . b7-b5, which at
least gives White the option of a queenside attack starting with a2-a4. In general I rec
ommend meeting a2-a4 with . . . �d7 , assuming it is legal. Black should not play
. . . b7 -bS too early, because of potential problems on f7 if he is not yet ready to castle.
Let's look at the moves one by one now.
l .e4 eS (beginners are taught that this is the best move, for good reason) 2.tbf3 tbc6
and now:
3.�bS
·.t'
8
:· tLJ
8 8 8 8 8 8
� t2J j, 'if ��
This, the Spanish or Ruy Lopez, is the main white weapon in master play. There is no
easy route to equality for Black.
3 ...a6 For the Berlin 3 . . . tbf6 see Chess Advantage. The text gives Black the option of
chasing off the bishop later.
4.�a4 A very important alternative is 4.�xc6 dxc6 5 .0-0, which I recommended for
White in my last book. It was a favorite of Bobby Fischer and is still moderately popular
1 79
The Kaufman Repertoire for Black
in grandmaster play. The defense 5 ... f6 6.d4 i.g4 comes close to equality (Game 1 4. 1 ).
In my opinion there is only one path to full equality for Black, namely 5 ... �g4, which I
give in Game 1 4. 2 . The idea is to offer the bishop after 6.h3 h5 . White can't accept im
mediately but may do so later, or may choose to provoke an early endgame.
4...tiJf6 5.0-0 White invites Black to play the 'Open' defense by taking on e4. White will
win back the pawn, with a slight edge in general. I'm not recommending the Open Spanish
for Black. Instead White can defend the pawn by 5 .d3 , which we meet by 5 ... d6 followed by
... g7-g6 and ...�g7 (Game 1 4.3), which should give us something like an improved version
of the Breyer. The Center Attack 5 .d4 (Game 1 4.4) peters out to at least equality for Black.
The Wormald Attack 5.'�e2 (or its brother the Worrall Attack 6.'iVe2) is well met by playing
as in the Marshall Gambit, which is totally sound in this case. See Game 1 4.5.
5 ....ii.e 7 This is the conservative move; I don't believe in early aggression for Black.
6Jiel White hopes for 6 . . . 0-0 7 .�xc6 dxc6 8 .tt:lxe 5 , winning a pawn. Instead 6.d3
is met by 6 . . . b5 7 .�b3 d6! 8 .a4 �d7 (see Game 1 4. 6 ) , while 6.d4 and 6 .'tWe2 trans
pose to lines mentioned in the previous note. The Deferred Exchange Variation 6.�xc6
dxc6 (Game 1 4. 7 ) requires some study, as each white seventh move requires different
treatment, but Black has equality if he learns the lines.
6 ...b5 7 .ii.b 3 d6 Black can also choose the move order 7 . . . 0-0 to 'bluff' White into
.
avoiding 8 .c3 for fear of the Marshall Gambit 8 . . . d5 . However I don't recommend this,
because if White avoids the Marshall by 8 .a4 we no longer have the reply 8 ... �d7 , and
if White avoids the Marshall by 8 .h3 it is not clear that Black benefits from avoiding
8 . . . d6 9 .c3 transposing to the main lines.
8.c3 Instead, 8.h3 or 8 .d3 loses the bishop pair to 8 ... tt:la5 , while 8.d4 could land
White in the famous Noah's Ark trap after Black exchanges twice on d4 and plays
. . . c7 -c5 and . . . c5-c4. So White just prepares d2-d4 and makes a retreat for his bishop.
8 0-0 9.h3 Here 9.d3 is met by ... tt:la5 , ... c7-c5 , ... tbc6, ... .l'le8, ...h7-h6, ...�e6, and
.•.
...� with equality (see Game 1 4.8), while 9.d4 is met by 9 ...�g4. Then 1 0.d5 lDaS
1 1 .�c2 c6 1 2.h3 �c8! 1 3 .dxc6 'iVc7 (to recapture on c6 with the queen) equalizes (Game
1 4.9) . Instead, 1 0.�e3 may be the most dangerous of White's Spanish offshoots. I meet it
with 1 O ... exd4 1 l .cxd4 d5 1 2.e5 tt:le4, which is at least pretty close to equal (Game 1 4. 1 0) .
1 80
Chapter 1 4 -Spanish Offshoots
Now we reach the most important ninth move position in chess. Black usually plays either
9 ... ltJa5 (Chigorin) or 9 ... �b7 (Zaitsev) or 9 . ..t2Jb8 (Breyer) , which I recommend.
181
The Kaufman Repertoire for Black
..
� � .� �:
� .l .t .� �
.l ' -� ttJ � IS} � �
� � ·:s ttJ �� 'if
�
�
I believe that this line is the only full
ttJ
equalizer to the Spanish Exchange, which
Black's bishop pair offsets White's better I recommended for White in my previous
pawn structure. It is hard for either side to book. I have also played the gambit line
initiate action here, though Black can try with 6 . . . �h5 which offers good practical
. . . a7-a5 and . . . a5 -a4. chances for a pawn, but White can return
the pawn and keep a slight edge.
7. d2-d3 �d8-f6
RL 8.8 (C69) Game 14.2
0 Vachier-Lagrave,Maxime This line was considered the 'refutation'
• Shirov,Alexey of 5 .0-0 until Bobby Fischer took up the
Germany Bundesliga, 2009/ 1 0 ( 1 4) White side and showed that the existing
analysis was wrong. However now things
1. e2-e4 e7-e5
have gone full circle and this line once
2. ttJg1 -f3 ttJb8-c6
again looks best and rather pleasant for
3. �f1 -b5 a7-a6
Black.
4. �b5xc6 d7xc6
5. 0-0 8. tt:lb1 -d2
After 5 .ttJc3 f6 6.d4 exd4 7 .tLlxd4 c5 8 .�e3 �xf3 9.l\Vxf3 l\Vxf3 1 0.gxf3 tt:le7
8 .tLlde2 'iYxd l + 9.tbxd l ..te6 1 0.�f4 l l .tLld2 (after l l .f4 exf4 1 2.�xf4 0-0-0
0-0-0 l l .tLle3 tbe7 1 2.lld l llxd l + 1 3 .tLld2 g 5 1 4.�e3 .!lg8 1 5 .ttJf3 g4
1 3 .Wxd l g 5 1 4 . ..tg3 f5 1 5 .�e5 llg8 1 6.hxg4 �xg4+ 1 7 .�h l ..tg7 1 8. llabl
1 82
Chapter 1 4 -Spanish Offshoots
1 83
The Kaufman Repertoire for Black
1. e2-e4 e7-e5
2. tt:lg1 -f3 tt:lb8-c6
3. �f1 -b5 a7-a6
24. ... J:!.d8-d 1 +
4. �b5-a4 tt:lg8-f6
2 4 . . We7 , planning . . . c6-c5 , looks fme.
. 5. d2-d3 d7-d6
6. c2-c3 g7-g6!
25. �e2-e1 l:!.d 1 -d3
26. tt::lf3-e5 lid3-d8
27. wg1 -g2 we8-e7
28. f2-f4 l:!.h8-h5
29. i:!c2-f2 tt::le 6-c5
30. l:[e1 -e3 I:!.d8-d1
31 . f4-f5 we7-f6
32. tt:Je5-g4+ 'it>f6-g5
33. tt::lg 4-e5 'it>g5-f6
34. tt:Je5-g4+ 'it>f6-g5
1 84
Chapter 1 4 -Spanish Offshoots
I O .il.c2 il.b7 l l .lt:Jfl ( l l .a4 b4 1 2.a5 The game actually continued 2 1 . . . .l:tc8
J::f.b 8 1 3 .lt:Jc4 dS 1 4.lt:Jcd2 il.a8 =) 2 2.iLc l Uab8= 2 3 .dxc4 bxc4 24.lt:Jd5
l l . . .lt:Jb8 1 2.tt:lg3 tt:lbd7 1 3 .d4 .!:le8=. lt:Jd7 2 5 .lt:Jb4 lt:Jc6 2 6 .lt:Jd5 lt:Je7
Black has reached a main line position of 27 .lt:Jxe 7 + "ifxe7 2 8 .il.e3 tt:lf6 29.'iYc2
our Breyer Defense two tempi up (he has .l:!.c6 3 0 . lle 2 ? ! "ifb7 3 L�c l "i¥b3
already played . . .il.g 7 and White lacks h3) . 3 2 .'iVxb3 cxb3 3 3 .lt:J e l lt:Je8 34.lt:Jd3 fS
185
The Kaufman Repertoire for Black
3 5 .lLlb4 ktcc8 3 6.f4 �fl 3 7 .fxe5 dxeS 7 .eS tt:le4 8 . lLlxd4 0-0 9 .tt:lfS dS and
3 8 .exf5 gxfS 3 9 . .l:rfi �g6 40.lLld5 .!lb7 now:
4 I .tt:le3 tt:ld6 42 . .l:rdi llc6 43 .tt:lfl .l:ra6 A) I O . exd6 �xfS I I .dxe7 �xe7
44.tt:ld2 kf6 45.tt:lf3 e4 46.tt:ld4 .l:ra i 1 2.kxc6 bxc6 I 3 .�f3 ;
47 .tt:le6 tt:lc4 48.lLlf4+ �f7 49 . .l:rfl �gS B) I O.tt:lxe7+ tt:lxe7 I l .c3 tt:lcS I 2.�c2
S O J:i.ee 1 tt:lxb2 S I .g4 lLlc4 5 2 .gxf5 �xf4 kfS = . No more bishop pair here means
and White resigned. no advantage, as White's remaining
bishop is not a good one;
22. �e2xd3 tt:la5-b3
C) I O.i.xc6 bxc6 l i .lLlxe7 + �xe7
23. �d3-c2 tt:lb3-c5
1 2 . .lle I .!le8 1 3 .f3
24. tt:Je3-d5 tt:lf6xd5
25. e4xd5 �e6-c8
1. e2-e4 e7-e5
Analysis diagram
2. tt:lg1 -f3 tt:lb8-c6
3. �f1 -b5 tt:lg8-f6
1 3 ... lLld6 (New York master Paul Brandts
4. d2-d4 e5xd4
surprised me with this move back in the
5. 0-0 a7-a6
I 960s) 1 4.b3 ( 1 4.kf4 tt:lfS I S .�d2 aSN
6. �b5-a4 �f8-e7
is pleasant for Black) 1 4 .. .f6 1 5 .�b2 lLlf7
We would reach this position by 3 .. a6 1 6.f4 fxeS 1 7 .fxe5 kfs I 8 .tt:ld2 lLlg5 -
4.ka4 lLlf6 S .d4 exd4 6.0-0 �e7 or Black will be better once his knight
5 .0-0 i.e? 6.d4 exd4. This line is called blockades on e6, as his bishop is clearly
the Center Attack. more active than White's.
7. ... b7-b5
8. �a4-b3
1 86
Chapter 1 4 -Spanish Offshoots
Draw agreed.
1 87
The Kaufman Repertoire for Black
1. e2-e4 e7-e5
2. t2Jg1 -f3 t2Jb8-c6
3. �f1 -b5 a7-a6
4. �b5-a4 t2Jg8-f6
5. 0-0
9. d2-d3
S .'�e2 (the Wormald Attack) S . . . bS
6 . .tb3 �e7 7 .c3 (for 7 .0-0 0-0 see the 9.exd5 tbxdS 1 0.lt:Jxe5 (after 1 0.d3 �b7
game) 7 . . . d5 8 .d3 (for 8.exd5 lt:JxdS 1 1 .lLlxe5 lt:JxeS 1 2.�xe5 �d7 1 3 .lLld2
9.t2Jxe5 tbxeS 1 0.�xe5 lt:Jf6 1 1 .0-0 0-0 �d6 1 4.�e4 c5 1 S .i.xd5 i.xdS Black has
see the note to move 9) 8 ... 0-0 9 . 0-0 - the bishops and a big lead in development
see the game. for the pawn, more than enough. His
pieces point menacingly towards White's
5. ... �f8-e7
king) 1 O . . . tt:JxeS 1 1 .�xe5 t2Jf6 1 2.d4
6. �d1 -e2
i.d6 (in the real Marshall White would re
treat his rook to the first rank. But here a
queen retreat will lose more time after
1 3 ... .!:!.e8) 1 3 .'iVg5 .!:!.e8 1 4.t2Ja3 i.b7N
1 5 .f3 cS 1 6.'iVh4 cxd4 1 7.cxd4 i.e?
1 8.i.d2 tL:ldS 1 9 .'iVf2 �h4 20.g3 �f6=.
Black has full compensation for the pawn,
which is isolated and under attack. Black
has a safer king and a much better knight.
9. ... �c8-b7
1 0. t2Jb1 -d2
This is the Worrall Attack, favored by I O.l:Id 1 .!:!.e8 1 1 .tLlbd2 transposes to the
among others grandmaster Sergey next note;
Tiviakov. 1 0 .i.g5 dxe4 1 1 .dxe4 tL:lxe4! 1 2.�xe4
i.xgS 1 3 . .tdS i.e7N 1 4.c4 (if 1 4.i.xc6
6. ... b7-b5
fS 1 5 .�xe5 i.xc6 1 6.tL:ld4 i.d7 1 7.a4
7. �a4-b3 0-0
i.d6 1 8.'�'d5+ Wh8 Black is better with
8. c2-c3 d7-d5
the two bishops in an open position)
This is the Marshall Gambit when White 1 4 . . . �d6=. White should eventually re
has played 6.!te 1 rather than 6 .'�'e2 , but gain his lost pawn and equalize.
in this position few players accept the
1 0. ... .I:!.f8-e8
gambit, as Black gets a better version of
1 1 . .U.f1 -e1
the Marshall.
1 88
Chapter 14 -Spanish Offshoots
••
t i'
1 2. e4xd5
Says the computer - I even prefer Black. 24.lLJc5 aS 2 5 .l:tc l WihS favors Black but
White still has chances to survive.
1 3. tt:Jd2-e4 h7-h6
1 4. �c1 -d2 l:ta8-d8 24. ... e5xd4
1 5. :aa1 -d1 'Wd7-g4 25. jLd2xf4 tt:Je7-f5
26. �f4xc7 l:td8-c8
27. jLc7-b6?
1 89
The Kaufman Repertoire for Black
5 . ... �f8-e7
6. d2-d3
8. ... �c8-d7!
This is probably White's best option in After 9 .!fi.d2 b4 1 0 .c3 0-0 1 1 .h3 .!:!.b8
the Spanish if he wants to avoid the main 1 2. .�c4 'i!Yc8 Black is doing well because
lines. It makes more sense here than after the bishop on d2 impedes the develop
6.l:i:e 1 , as with d2-d3 played that move is ment of the knight;
not urgent. 9 .tt:lc3 tt:las 1 o . .ia2 b4 1 1 .tLle2 0-0
1 2.< 2lg3 cS=. Black may sacrifice a pawn
6. ... b7-b5
by ... b4-b3 next against routine moves, or
Of course 6 . . . d6 is playable, but then after may simply bring his knight back to c6.
7 .c3 White can answer a later . . . b7-b5 by The game Erenburg-Kaufman, US Chess
!fi.c2 instead of �b3 . League (internet game) 2 0 I 1 , continued
1 3 .ltJh4 b 3 ! (N) 1 4.cb3 lDc6 I S .b4 ltJb4
7. �a4-b3 d7-d6
1 6 . .ic4 and now 1 6 . . . lDe4 would have
1 90
Chapter 1 4 -Spanish Offshoots
given me the advantage after 1 7 .de4 £i.h4 £i.e6= . With ideas of ... 'i:Wd7 or . . . tt:JaS or
1 8 .'ii'd 6 tLlc2 followed by . . .£i.e6. . .. £i.e? Black has adequate counterplay.
9 . ... 0-0 1 5. ... �e7-f8
1 0. tt:lb1 -d2 1 6. d3-d4 c5xd4
1 7. c3xd4 tt:lc6-b4
1 O .lle 1 tt:JaS 1 1 .£i.c2 cS 1 2.tt:Jbd2 trans
1 8. �c2-b3
poses to the game.
1 0. ... tt:lc6-a5
1 1 . �b3-c2
191
The Kaufman Repertoire for Black
1 1 . ... tt:Jc5-e6
1 2. �c1 -e3
1 92
Chapter 1 4 -Spanish Offshoots
Analysis diagram
1 93
The Kaufman Repertoire for Black
1 5. d3-d4
1 94
Chapter 14 -Spanish Offshoots
1 7. lLlf3-h4
1 95
The Kaufman Repertoire for Black
1 96
Chapter 14 -Spanish Offshoots
1 3 .h3 (it's not clear that including this After 1 8 .h3 �fS 1 9 .'li'e2 'li'd7 Black is
move is helpful, as the black bishop gets close to equal, since the exchange of
the option to go to g 6 . Sometimes Black light-squared bishops favors him.
even plays I O . . . �hS earlier, planning to
1 8. ... a6-a5
meet l l .h3 as in this game, but White
may choose something other than For 1 8 . . . 'li'd7 ! 1 9 .h3 .ifs see the previ
l l .h3) 1 3 . . . �h5 1 4.ltJc3 ltJxc3 I S .bxc3 ous note.
ltJaS 1 6 .�c2 ltJc4 1 7 .g4 �g6 1 8.�f5
1 9. h2-h3 �g4-e6
transposes to the note to move 1 6.
20. tL'lf3-h2
1 3 . ... tL'le4xc3
With 20 .ltJd2 ! ? llc8 White has a tiny
1 4. b2xc3 tL'lc6·a5
edge.
1 97
The Kaufman Repertoire for Black
20 . . . c5=.
21. 4Jh2-g4
1 98
Chapter 15
Breyer Defense
The Breyer is in my view the best defense to 1 .e4. When I wrote my last book eight
years ago it was just one of many defenses to the Spanish, but now it is the most popu
lar such defense in high-level play. It is the main defense of the world's top rated player,
Magnus Carlsen, and has also been played more than once in the past three years by
World Champion Anand and title contenders Kamsky, Mamedyarov, Svidler, and
Adams. It was also a favorite of World Champion Boris Spassky. It does not allow any
quick forced draw by repetition (unlike the Zaitsev Defense, which is also a good line if
a draw is fine) , nor any quick queen trade. In fact it usually leaves all the pieces on the
board for the first 2 0 moves or so, making it pretty much ideal if you must win on the
Black side of the Lopez. My own results with it have been excellent. I recommended the
Berlin in my last book, which still looks pretty decent, but defending a slightly inferior
endgame from the opening is not everyone's cup of tea. Another factor in my choice is
that sidelines such as an early d2-d3 are more annoying in the Berlin because Black
lacks the option to chase away the bishop by ... b7-b5. One final factor in favor of the
Breyer for amateur players is that if your opponent is not familiar with it, when you
play 9 ... tt:lb8 (the Breyer move) he will think you are a hopeless patzer and will under
estimate you! Even if your opponent does know of the Breyer, as of this writing there
are no recent books on it, so unless he reads this book he may not be well prepared.
The defense is credited to Gyula Breyer, a Hungarian master who was among the top
ten players in the world in 1 9 1 8 (per Chessmetrics website) but who died in 1 9 2 1 at
only age 2 7 from a heart attack. To come up with the move 9 . . . tt:lb8 at that time took a
real genius, as the belief in 'development' was very strong since the time of Morphy.
The basic idea behind the move is that the knight has fulfilled its function on c6 and
now is just in the way of the c-pawn and the bishop when it goes to b7. Previously this
problem was solved by 9 . . . tt:la5, chasing the bishop back to c2, and then . . . c7-c5 , but
the knight normally had to return to c6. The beauty of the Breyer is that White stil� ,
must retreat his bishop to c2 in the main line in order to complete the maneuver
tLld2-fl -g3 , but Black's knight will be on d7 rather than c6, which makes the bishop
fianchetto to b7 attractive. Black will usually fianchetto both bishops, will usually
answer a2-a4 by . . . c5-c4 and . . . tt:lcS , will aim for . . . c7-c6 against an early d4-d5, and
will sometimes get in the shot ... d6-d5 himself.
Now let's look at the moves of the Breyer Defense. For comments on the first nine
moves see the Spanish offshoots chapter.
l .e4 eS 2.tt:Jf3 tt:lc6 3 .�bS a6 4.�a4 tt:Jf6 5 .0-0 �e7 6 ..Se l bS 7.�b3 d6 8.c3 0-0 9.h3
tt:lb8 (!)
1 99
The Kaufman Repertoire for Black
This makes it a Breyer. It only makes sense now as a response to h2-h3 , because h2-h3
rules out the strategy of answering d2-d4 by . . . �g4, so pressure on d4 is now point-
·
less.
1 O.d4 The quiet move d3 is a serious alternative, designed to avoid the need for an
unprovoked �c2 to defend the e4 pawn. It is likely to lead to the same position as
1 O .d4 but with the extra move b2-b4 included for White, which may be either good or
bad. See Game 1 5 . 1 .
1 O ...ttJbd7 This comes just in time to defend the e5 pawn.
1 I .tbbd2 See Game 1 5 . 2 for 1 l .tbh4 .l':Ie8 1 2.tbf5 �f8 . See Game 1 5 . 3 for 1 1 .c4 c6.
1 1 ...�b7 1 2.�c2 White doesn't really want to play this, as the bishop is currently ac-
tive on b3. He does this for three reasons: he wants to play ttJfl , which would lose a
pawn if played now. Also, he may want to play a2-a4 and then attack the b5 pawn by
�d3 . Finally, he may want to advance his b-pawn. For the alternative 1 2.a3 c5 see
Game 1 5 .4.
1 2 ...J::!. e8 Now that the bishop doesn't attack f7, there is no drawback to this move,
which prepares to put indirect pressure on e4 by retreating . . . �f8.
1 3.tbfl The alternative 1 3 .a4 �f8 1 4.�d3 is very popular lately, to force Black to
block his bishop by playing . . . c7 -c6. On the plus side, . . . c7 -c6 lets out the queen and
controls the center. See Game 1 5 . 5 .
1 3 .. �f8 1 4.tbg3 (in time to defend e4) 1 4...g6 This keeps out the white knight and
.
enables a black fianchetto to defend the king strongly. For the alternative pin 1 4.jlg5
h6 1 5 . .th4 g6 see Game 1 5 . 6 .
1 5 .a4 For 1 5 .�g5 h 6 1 6.�d2 � g 7 see Game 1 5 . 7 . For 1 5 .b3 � g 7 1 6.d5 llc8 see
Game 1 5. 8 .
.1 5 . c 5 This move i s triggered b y a2-a4 because now after 1 6.d5 c4 the black knight
.
will attack the a-pawn, restrain the b-pawn and can sometimes land on the b3-square.
1 7 .�g5 Experience has shown that if Black chases the bishop back to e3 by playing
. . . h6, that move gives White a target, while if he doesn't, the bishop is more active on
gS than on e3 . Most players do chase the bishop back, but I vote with the minority in
this instance, choosing 1 7 ... �e7. See Game 1 5 . 9 .
200
Chapter I 5 -Breyer Defense
20 1
The Kaufman Repertoire for Black
202
Chapter 15 -Breyer Defense
203
The Kau!man Repertoire lor Black
2 04
Chapter I S -Breyer Defense
1 4. ... l:tf8-e8
1 5. �b3-c2 tt:Jd7-f8
1 6. a2-a3 a6-a5
1 7. b2-b3 tt:Jf8-g6
1 8. tt:Ja4-b2 �c8-d7
1 9. tt:Jf3-h2 h7-h6
20. tt:Jh2-f1 tt:Jf6-h7
205
The Kaufman Repertoire for Black
1 4. b2-b4 e5xd4
1 5. c3xd4 c5xd4
1 6. tt:lf3xd4
1 8. ... g7-g6
1 9. �b1 -a2?1
206
Chapter 1 S -Breyer Defense
1 6. . . . J:!:a8-c8
1 7. �c1 -a3
1 7. ... ttJf6-h5!
207
The Kaufman Repertoire for Black
1 8. g2-g3 tt:lh5·g7
1 9. �d3-f1 tt:lg7-e6=
In an earlier game with Navara (in their 3 0.bxc5 .ta8 3 1 .llxe8+ .ll xe8 3 2 .l:td6 h6
Khanty-Mansiysk, rapid playoff, 2009) 3 3 .tbxf7 iYxf7 34.iYxf7+ <;t>xf7 3 5 .l:txa6
Shabalov played 2 0 .h4 but lost after jLds 3 6.c6 gS 3 7 .i.e3 and White's two
2 0 . . . iYb6 2 1 .�h3 l:tcd8 , though far advanced passers offer full compensa
20 . . . exd4 2 l .cxd4 �g7 might have been tion for the knight.
better.
30. ..
. �e4-a8-+
20 . ... iYd8-b6 31 . b4xc5 iYa7·b7
21 . iYc2-a2?! 32. c5-c6 iYb7xc6
33. iYa2xf7+ wg8-h8
2 l .dxe5 ttJxeS 2 2 .lLlxe5 dxeS 2 3 .�xf8
34. f2-f3
a'.xf8 = .
208
Chapter 1 5 -Breyer Defense
1. e2-e4 e7-e5
2. tbg1 -f3 lbb8-c6
3. �f1 -b5 a7-a6
4. �b5-a4 tbg8-f6
5. 0-0 �f8-e7
6. l::.f1 -e1 b7-b5
7. �a4-b3 d7-d6
8. c2-c3 0-0
9. h2-h3 tbc6-b8
1 0. d2-d4 lbb8-d7 1 7. d4xe5
11. lbb1 -d2 �c8-b7
1 7 .tbg4 exd4 1 8 .cxd4 'ti'c8 1 9 .tt:lxf6+
1 2. �b3-c2 l::.f8-e8
tt:lxf6= 20 .e5 dxeS 2 l .dxe5 tbhS 2 2 .JI..e4
1 3. tbd2-f1 �e7-f8
cS and now Black is a bit better due to the
weakness of the eS pawn.
1 7. ... d6xe5
1 8. tbh2-g4 g6-g5
1 9. �h4-g3 tt:lf6xg4
20. h3xg4 �d8-e7
21 . a2-a4 tbd7-c5
209
The Kaufman Repertoire for Black
1 5. ... h7-h6
210
Chapter 15 -Breyer Defense
21 1
The Kaufman Repertoire for Black
2 1 . . . .ll a 8=.
1 8. ... a6-a5
1 9. iLd2-e3
28. "ti'f3xe3?1
212
Chapter 1 5 -Breyer Defense
Black actually played 3 3 . . . tLle4 and White Usual is 1 7 ...h6 1 8 .�e3 tbcS 1 9.'t!Vd2 h5,
eventually reached a draw. The text keeps but lately 2 0.hc5 has been annoying.
a serious advantage for Black, though of Normally giving up a good paired bishop
course White retains decent drawing like this for a knight is dubious, but the
chances. protected passer and the crippled black ma
jority justify it. So I prefer Postny's move.
1 8. �g5-e3
RL 2 5 . 1 6 (C9 5 ) Game 1.5.9
D David,Alberto White reasons that he has provoked a
• Postny,Evgeny worsening of Black's bishop location, but
Belgium tt, 2008 /09 (9) it's a minor point here.
213
The Kaufman Repertoire for Black
White aims for f2-f4 or tt:lg4. 2 6.�f2 .ic8 2 7 .tt:ld4 b4 is perhaps a tad
better for White .
1 9. ... .l:re8-b8
26 . ... �f8-g7
This prepares to reactivate the bishop by
27. cot>g1 -h1 �b7-c8
. . . .ic8 , while preparing . . . b5 -b4. How
28. tt:le2-d4 b5-b4
ever I think that 1 9 . . . tLlcS was a better
29. tt:ld4-c6 b4xc3
way to activate the bishop by
30. b2xc3 ld.b8-b2
. . . .ib7-c8-d 7 :
31 . �f4xe5 d6xe5
1 9 . . . tt:lc5 ! 2 0 . f4 ( 2 0 .tt:lg4 tt:lxg4
2 1 .'iHxg4 kc8 2 2 .�e2 i.d7 =) 2 0 . . . exf4
2 I ..ixf4 .if8 2 2 .tLlf3 tt:lfd7 and Black is
doing well, with ideas of . . . tLlb6 or if : "1'1
2 3 .tLld4 tLleS .
20. f2-f4
32. �c8-b7
33. tt:lc6-d4 �c7-a5
34. tt:ld4-c6 �a5-c7
35. tt:lc6-d4 �c7-c5
36. tt:ld4-c6 �b7xc6
37. d5xc6 �c5xc6
38. lla1 -d1 .ila8-f8
39. �g3-h4 �c6-b6
40. .l:!.f2-e2 �g7-f6
41 . �h4-e1 �f6-g5
22 . ... tt:le5xf3+ 42. l:td1 -d5 �g5-f4
43. a4-a5 �b6-f6
2 2 . . . tt:lfd7 was probably better, and Black
44. �c2-d3 l:!.f8-b8
is doing well.
45. �d3xc4 �b2-b1
23. �d1 xf3 tt:lf6-d7 46. ttd5-d1 l:tb1 xd1
24. .l:le1 -f1 �e7-f8 47. �e1 xd1 �f6-c6
25. tt:lg3-e2 tt:ld7-e5 48. �c4-d5 �c6xc3-+
214
Chapter 1 S -Breyer Defense
This regains the pawn but makes Black's And White resigned. If he moves the
attack much stronger. queen forward, he will get mated.
215
The Repertoire in Practice
I played in the 2 0 1 1 U.S. Open just as I was nearing completion of this book, and so I
decided to play strictly according to the repertoire in every game. While my fmal result
was nothing special, the openings were a smashing success. I got an advantage in every
opening but one in which I 'only' equalized as Black against grandmaster Zapata. In
most games I was nearly winning by move 1 2 . Let's look at the openings round by
round. For comparison, my own (USCF) rating was 246 5 .
I n the first round as White against an 1 8 7 8 rated player, the game went l .d4 f5
2.�g5 h6 3 .�h4 g5 4.�g3 �g7 5.e3 ttJf6 6.h4 d6?! 7.hxg5 hxg5 SJhhS+ �xhS
9.1Llc3 with advantage, though 9 .tLlh3 ! g4 1 O.tt:Jf4 tbe4 1 1 .�h4 would have given a big
advantage.
In the second round as Black against a 2 0 1 2 rated player, I played the N eo-Grlinfeld
and he mixed up his move-order and just blundered a pawn on move 8 .
I n the third round as White against a 2 2 0 0 rated player, play went l .d4 d5 2.c4 e6
3 .1Llc3 1le7 4.cxd5 exd5 5.�f4 1Llf6 6.e3 0-0 7.�d3 c6 8.�c2 1Llbd7 9.1Llf3 l:i:e8 1 0.0-0
lt:lfS l l .h3
216
The Repertoire in Practice
This position usually arises with White's bishop on g 5 . Even that position is a bit better
for White, but the bishop is clearly better placed on f4.
1 1 . .. lLJg6 1 2 ...1:i.h2 .te6 1 3.l:l:fc 1 .l:lc8 1 4.l:l:ab 1 cS 1 5 .dxc5 iLxcS
Now 1 6 .b4! was best, with a clear advantage as Black has no compensation for his iso
lated d-pawn here.
In the fourth round as Black against a 2 2 7 2 rated opponent, the game went l .c4 g6
2.lLJc3 cS 3.t2Jf3 iLg7 4.e3 lLJf6 S.d4 0-0 6 ..te2 dS (transposing to an e3 Griinfeld)
7.0-0 cxd4 8.exd4 lLJc6 9.cxdS ttJxdS 1 0.h3 ..l:i.e6 .
This position is an exact transposition to the main line of the Tarrasch Defense t o the
Queen's Gambit, but with colors reversed. However, although here I am Black, I am not
a tempo down as in the Tarrasch line White plays �g5 and later �e3 , but here I played
. . . il.e6 directly! So as Black I have achieved a position from the White repertoire against
a second-rate defense, so of course this is a success for Black. Play continued 1 1 .l:l:e1
l:rc8 1 2.�gS h6 1 3.�e3 'iVaS 1 4.'iVd2 Wh7 (taking the bishop on e3 was also good now
or on the last move) 1 S .a3 t2Jxc3 16.bxc3 .l:lfd8 1 7 .l:l:ab 1 and now I missed I 7 . . .ii.d 5 !
.
217
The Kaufman Repertoire for Black
In round 5 as White against a 2 5 3 9 rated grandmaster the game went l .d4 dS 2.c4 c6
3.tt:Jf3 tt::lf6 4.tt::lc 3 dxc4 S.a4 �fs 6.tt::le 5 tt::lb d7 7.tt::lxc4 tt:lb6 s.tt:Jes aS 9.g3 e6 l O.jLg2
�b4 1 1 .0-0 0-0 1 2.e3 h6 1 33We2 �7 1 4J:tdl 'flfe7 1 5 .e4 l:tfd8 1 6.�f4 tt::lfd7 1 7.tt::ld 3
�d6 1 8.jLxd6 'flixd6
And now I missed 1 9.'i¥e 3 ! tt::lc4 2 0 .'i¥c l with advantage, as Black can neither take on
d4 nor achieve the needed e5 break.
In round 6 as Black against a 2 5 3 9 rated grandmaster play went l .e4 eS 2.tt::l f3 tt::lc6
3.�b5 a6 4.�a4 tt::lf6 5.0-0 �e7 6.l:re l bS 7.jLbJ d6 8.c3 0-0 9.a4 �d7 1 0.d4 h6 l l .h3
l:Ie8 1 2.tt::lbd2 �f8 l J.jLc2 exd4 1 4.tt::lxd4
And now i should have played 1 4 . ..lbe5 ! 1 5 .tt:lfl c5 1 6.tt:le2 �c6 with full equality.
In round 7 as White against a 2 2 3 9 rated FIDE master it went l .d4 tt::lf6 2.c4 g6 3.tt::lc 3
dS 4.tt::lf3 jLg7 5.11Vb3 dxc4 6.'flfxc4 0-0 7.e4 tt::lc6 8.�e2 jLg4 9.d5 �xf3 1 0.gxf3 tt::le S
l l .'flib3 c6 1 2.f4 tt::le d7 1 3.dxc6 bxc6
218
The Repertoire in Practice
;8
.:
14.�e3?, which allowed Black the equalizing combination 14 .'�i'a5 1 5 .0-0 �ab8
•.
1 6.'�'c2 �xb2 ! . Instead either 1 4.0-0 or 1 4.e5 gives a clear advantage; castling leaves
White up the bishop pair in an open position, while 1 4.e5 provokes a dubious pawn
sacrifice with 1 4 . ..ttJ dS .
In round 8 as Black against a 2 2 0 3 rated opponent, play went l .e4 eS 2.f4 dS 3.exd5
exf4 4.tt:lf3 t2Jf6 S ..tc4 tLlxdS 6.0-0 .te6 7.Wi/e2 �e7 8.d4 0-0 9 .tb3 �e8 1 0.c4 tt:le3
•
(the computer says 1 O tt:lf6 was a much bigger advantage for Black) l l ..txe3 fxe3
...
1 2.tt:lc3
In round 9 as White against a 2 3 3 1 rated opponent the game went l .d4 e6 2.c4 fS
3 .tt:lf3 tt:lf6 4.l2Jc3 dS S.�gS (S . .tf4 is also strong) S c6 6.e3 .td6 7 . .td3 0-0 8.0-0
•..
219
The Kaufman Repertoire for Black
White is obviously better as Black has traded his good bishop for White's weaker one,
and Black also lags in development.
Three weeks later I played in the Atlantic Open and again stuck strictly to the repertoire.
I got five out of five favorable openings. All opponents were rated above 2 1 0 0 , one was
IM and one was FM. In my three White games I won one in 1 7 moves and in the other
two I was probably winning by about move 2 0 . In the Black games I was better by
moves 7 and 1 4 respectively.
Round 1 , White vs. 2 1 7 4: l .d4 dS 2.c4 c6 3.t2Jf3 t2Jf6 4.lLJc3 e6 s . .igs h6 6 ..itx.f6 �x.f6
7.e3 g6 8.�d3 i.g7 9.0-0 0-0 1 0.e4 dxc4 l l .eS �e7 1 2.�xc4 bS?! 1 3 ..id3 (computer
says 1 3 .ile2 was even better) 1 3 ... l2Jd7 1 4.l:rc1 .itb7 1 S.lLJe4 l:!:ab8
Now 1 6. �e 1 keeps Black bottled up and his bishop pair is ineffective and does not
compensate for his many problems. I actually played 1 6.lLJd6 , when 1 6 . . . c5 ! would
have equalized. Black missed this and I won easily.
220
The Repertoire in Practice
Round 2 , Black vs. 2 2 6 3 (tournament co-winner) : l .e4 eS 2.�c4 tt:lf6 3.d3 c6 4.tt:lf3 dS
S .exdS?! cxdS 6.i.b5+ �d7 7.a4 i.d6 (computer likes 7 . . . a6 even more) 8.d4 e4 9.tt:le5
0-0 1 0.0-0
And now 1 O . �e6 ! was best, with advantage to Black. I played 10 ...�c7 and still kept
..
Round 3 , White vs. 2 1 7 2 : l .d4 ctJf6 2.c4 g6 3.ltJc3 �g7 4.e4 d6 S.ttJf3 0-0 6.ilLe2 eS
7.0-0 tt:lbd7 8.�e3 exd4?! 9.tt:lxd4 tt:lcS 1 0.f3 tt:le6 l l .ctJc2 a6 12.�d2 J:!.bS 1 3.l::la dl
�d7
Now best was 1 4.c5 ! tt:le8 1 5 f4 ! with a probably winning advantage. I played 14.f4 di
.
Round 4 White vs. FM, 2 246: l .d4 f5 2.i.gs tt:lf6?! 3.�xf6 exf6 4.e3 dS S.c4 i.b4+
,
6.tt:lc3 �xc3+?! 7.bxc3 c6 8.�c2 �e6 9.cxd5 cxdS?! 1 0.�d3 �d7 l l .ltJe2 tt:lc6 1 2.J:!.b l
.!lb8?! 1 3.tt:lf4 g6
221
The Kaufman Repertoire for Black
1 4.c4 (computer says 1 4.h4 was objectively best and close to winning) 14 ... dxc4
1 5.�xc4 kxc4 1 6.'ifxc4 'iff7?? (Black should play 1 6 . . . llc8 with a bad but maybe not
lost position) 1 7.'ifxc6 +! Resigned.
Round 5 , Black vs. IM, 2 5 2 1 : l .e4 eS 2.�c4 tt:lf6 3.d3 c6 4/t:Jf3 dS 5.�b3 �b4+ 6.c3
�d6 7.tt:lbd2 0-0 8.0-0 tt:lbd7 9.llel l:!.e8 I O.tt:lfl h6 l l .tLlg3 ttJfS 1 2.h3 tt:lg6=
1 3.tt:lh2?! �cS (maybe 1 3 . . ..te6 was even better) 1 4.tLlh5
And now I should have traded knights and then played . . . ke6 , with advantage. I actu
ally played 14 ...�e6 directly, which allowed 1 5.tt:lxf6+ 'ifxf6 1 6.d4, but after 1 6 ...�b6
I retained an edge.
So as you can see, the repertoire worked very well, even against two grandmasters. In
fourteen games, only one player, a GM playing White, managed to reach an equal ope
ning against my repertoire! Now if only I could play the whole game as well as the ope
ning. . !
.
222
Index of Variations ( Black)
U n usual O pen ing M oves
I .b4 (Polish) l ...e5 2 ...1lb2 bb4 3 ..be5 lLlf6 (Game 1 . 1 ) . .8
:i .. .t if � .t .. .i l .b3 (Larsen) l . . . e5 2 . .2.b2 lLlc6 3 .e3 lLlf6 4 ..2.b5 .2.d6
.t. .t. .t. .t. .t. .t. .t. .t. 5 .lLla3 lLla5 (Game 1 .2) . . . . . . . . . . . . . . . . . . . . . . . . . . .9
l .f4 (Bird) 1 ... lLlf6 2.lLlf3 d5
- 3 .e3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
- 3.g3 c6 4 ...\lg2 'iWb6 (Game 1 . 3) . . . . . . . . . . . . . . . 12
223
The Kaufman Repertoire for Black
2 24
Index of Variations (Black)
225
The Kaufman Repertoire for Black
1 .e4 e5 2.t2Jf3 t2Jc6 - Other Gam bits and Scotch and Fou r Knights
Open i ng
3 .d4 exd4
- 4.�c4 (Scotch Gambit) 4...�c5 5.0-0 d6 6.c3 �g4 . . 1 5 1
- 4.c3 (Goring Gambit) 4 ... dxc3
- S .�c4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 5 1
- S .tllx c3 (Game l l . S) . . . . . . . . . . . . . . . . . . . . l S I
3 .�c4 .2-cS 4.b4 (Evans Gambit) 4 ...�xb4 S .c3
- S ...�e7 (Game 1 1 . 6) . . . . . . . . . . . . . . . . . . . . . . . . I S 3
- S ...�aS 6.d4 d6 7 .�b3 �d7 (Game 1 1 . 7) . . . . . . 1 S4
3 .d4 (Scotch) 3 ... exd4 4.ttlxd4 ttlf6 S .tll xc6 bxc6 6.eS �e7
7 .�e2 tlld S 8.c4 tll b6
- 9 .tll d 2 aS
- 9.b3 aS (Game 1 2. 1 ) . . . . . . . . . . . . . . . . . . . . . . . . 1 S 8
- 9.tll c3 �b7 (Game 1 2.2) . . . . . . . . . . . . . . . . . . . . 1 60
3 .tllc 3 ttlf6 (Four Knights)
- 4.d4 (Scotch Four Knights) 4 ... exd4
- S .ttldS (Belgrade Gambit) . . . . . . . . . . . . . . . . . 1 62
- S.ttlxd4 (Game 1 2.3) . . . . . . . . . . . . . . . . . . . . 1 62
- 4.a3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 64
- 4.tll xeS (Halloween) . . . . . . . . . . . . . . . . . . . . . . . . . 1 6S
- 4.g3 (Glek) (Game 1 2.4) . . . . . . . . . . . . . . . . . . . . 1 64
- 4.�c4 (Italian) 4 ... tllxe4 S .ttlxe4 dS 6.�d3 . . . . . . 1 66
- 4.�bS (Spanish) 4 ... .2-cS (Game 1 2 .S) . . . . . . . . . 1 66
226
Index of Variations (Black)
1 .e4 e5 2.4Jf3 4Jc6 3.�b5 a6 4.�a4 4Jf6 5.0-0 �e7 6 .�e1 b5 7.�b3
d6 8.c3 0-0 9.h3 4Jb8 - Breyer Defense
I O.d3 (Game 1 5 . 1 ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 0 1
I O.d4 tt:lbd7
- l i .tt:lh4 l:i:e8 1 2.tt:lf5 �f8 (Game I 5.2) . . . . . . . . . 202
- l l .c4 c6 (Game I 5.3) . . . . . . . . . . . . . . . . . . . . . . . 204
- l l .tt:lbd2 �b7
- 1 2.a3 c5 (Game 1 5 .4) . . . . . . . . . . . . . . . . . . . . 205
- 1 2.�c2 l:i:e8
- 1 3 .a4 .af8 1 4.�d3 c6 (Game 1 5 .5) . . . . . 207
- 1 3 .tt:lfl �f8
- 1 4 . .ag5 h6 1 5 . .ah4 g6 (Game l 5 .6) . 209
- 1 4.tt:lg3 g6
- I 5.�g5 h6 1 6.�d2 _ag7 (Game I 5.7) 2 1 0
- I 5 .b3 �g7 1 6 .d5 l:i:c8 (Game I 5.8) . 2 1 1
- I 5.a4 c5 1 6.d5 c4 (Game I 5.9) . . . . . 2 1 3
227
The Kaufman Repertoire for Black
Index of Players
Numbers refer to pages.
228
Index of Players
229