Lessons from Kramnik

We have entered an era where it is not always obvious what the best players are doing. They are better than previous generations, they play all positions well, and they are fighting against players who also do everything well, and what makes the difference is not apparent to me.

But while Kramnik's play is subtle and deep, there are games which makes it look as though what he is doing is as simple as it looks.


Kramnik brought to several apparently settled opening systems a new clarity in pursuing White's main plans. In the Grunfeld, it was White's central majority; in the King's Indian, the Queen's-side attack.


This is a fine multi-layered game: a stubborn insistence on retaining a small initiative in the opening, an unexpected pawn sacrifice, and a switch to attack.

[Event "Horgen"]
[Site "Horgen SUI"]
[Date "1995.10.31"]
[Round "10"]
[White "Vladimir Kramnik"]
[Black "Rafael Vaganian"]
[Result "1-0"]
[ECO "E12"]
[PlyCount "69"]
[EventDate "1995.10.21"]

1. Nf3 Nf6 2. d4 e6 3. c4 b6 4. a3 Bb7 5. Nc3 Ne4 {Unusual.} 6. Nxe4 Bxe4 7. e3
Be7 8. Bd3 d5 9. Bxe4 dxe4 10. Nd2 f5 11. f3 Bd6 12. Qa4+ c6 13. O-O Qh4 14. f4
{Now the position is blocked, it will be hard for White to show much.} O-O 15.
c5 $1 bxc5 16. Nc4 Qe7 17. dxc5 Bxc5 18. b4 Bd6 19. Bb2 {If Black's extra pawn
is the one on e6, it does not count for much!} Bc7 20. Rfd1 c5 21. bxc5 Qxc5
22. Rac1 Qe7 23. Qb5 {Pressing.} Bb6 24. a4 Bc5 25. Bd4 Bxd4 26. Rxd4 {The
black Queen's-side is stuck.} a6 27. Qb6 Ra7 28. Nd6 Rd7 {Now what?} 29. Rc8 $1
{Exchanges ease the defence, we are told, but White needs a way in.} Rxc8 30.
Nxc8 {Black's whole position seems en prise.} Qa3 31. Qxe6+ Kf8 32. Qxf5+ Ke8
33. Qe6+ Kd8 34. Qb6+ Ke8 35. Nd6+ (35. Nd6+ Ke7 36. Rxe4+ Kf6 37. Nc4+) 1-0

An immensely revealing comment: some positions cannot be treated lightly. It was always the case that Kasparov should never be allowed to build up an attacking force on the King's-side, yet here he falls foul of the same strategy.

[Event "Dos Hermanas"]
[Site "Dos Hermanas"]
[Date "1996.??.??"]
[Round "?"]
[White "Kasparov, Garry"]
[Black "Kramnik, Vladimir"]
[Result "0-1"]
[ECO "D48"]
[PlyCount "70"]
[EventDate "1996.??.??"]
[Source "ChessPublishing"]
[SourceDate "2010.01.22"]

1. d4 d5 2. c4 c6 3. Nc3 Nf6 4. Nf3 e6 5. e3 Nbd7 6. Bd3 dxc4 7. Bxc4 b5 8. Bd3
Bb7 9. O-O a6 10. e4 c5 11. d5 c4 12. Bc2 Qc7 13. Nd4 Nc5 14. b4 cxb3 15. axb3
b4 16. Na4 Ncxe4 {All highly theoretical.} 17. Bxe4 $146 Nxe4 18. dxe6 Bd6 19.
exf7+ Qxf7 $1 {White's pieces have been drifting West, so Black takes a chance
to move towards the enemy King. A pawn goes, but Kramnik is not thinking about
an endgame...} 20. f3 Qh5 21. g3 O-O 22. fxe4 {Now it's pawn and piece...} Qh3
{Kramnik made an amusing comment here, criticising his illustrious opponent
for moving too fast; the position required perhaps an hour's thought, in his
opinion.} 23. Nf3 $2 Bxg3 24. Nc5 $2 Rxf3 {Raising the stakes to a Rook! But
if White enjoys the consolation of his Rook on a1, it is not a piece that has
much to say for the moment.} 25. Rxf3 Qxh2+ 26. Kf1 Bc6 27. Bg5 Bb5+ 28. Nd3
Re8 29. Ra2 {The Rook pokes his nose in.} Qh1+ 30. Ke2 Rxe4+ 31. Kd2 Qg2+ 32.
Kc1 Qxa2 33. Rxg3 {Is Black running out of pieces to attack with?} Qa1+ 34. Kc2
Qc3+ 35. Kb1 Rd4 0-1


Kramnik's clarity showed itself to good effect in positions with material inequality. His win against Leko was a model of how to coordinate a pair of Rooks, and against Kasparov sheer persistence paid off.

[Event "Kramnik - Leko Classical World Champion"]
[Site "Brissago SUI"]
[Date "2004.09.25"]
[Round "1"]
[White "Peter Leko"]
[Black "Vladimir Kramnik"]
[Result "0-1"]
[ECO "C42"]
[PlyCount "130"]
[EventDate "2004.??.??"]

1. e4 {Notes by Raymond Keene *** 1. e4 No surprise; Leko plays little else. I
felt a pang of sympathy for those commenting live on this game. After the
combinational flurry ending on move 23, it was too easy to reach for the
script that was titled "And White converts his material advantage". However,
that account had to be shelved hastily as Kramnik emphatically showed the
virtues of Black's position, rapidly attaining at least equality, then having
some pressure, and finally seeing Leko go astray in a much simplified ending
where the best White could hope for was a draw.} e5 2. Nf3 Nf6 {The Petroff
Defence was originally designed as an equaliser, but it also gives Black
dynamic counterchances.} 3. Nxe5 d6 4. Nf3 Nxe4 5. d4 d5 6. Bd3 Nc6 {This move
order has supplanted the ancient debate between 6...Bd6 and 6...Be7.} 7. O-O
Be7 8. c4 Nb4 {As we shall see, the Petroff is a defence where Kramnik appears
to be equally at home with both White and Black.} 9. Be2 O-O 10. Nc3 Bf5 11. a3
Nxc3 12. bxc3 Nc6 13. Re1 Re8 14. cxd5 Qxd5 15. Bf4 Rac8 16. h3 {So far, we
are still well inside 'theory', the current habitual move order of modern
masters. 16.h3 is an interesting psychological ploy by Leko, as it was
Kramnik's own choice when he played against Anand in Corus 2003.} Be4 {Kramnik
- Anand, Wijk aan Zee 2003 saw 16...Bf6 but, according to Huzman, after 17 Nh2
Qa5 18 Bd2 Rcd8 19 Bf3 h6 20 Ng4 Bxg4 21 hxg4 Bg5 22 Bxg5 hxg5 White could
have got a clear advantage by 23 Qc1 Rxe1+ 24 Qxe1 Kf8 25 Rb1 Re8 26 Qc1.} 17.
Be3 Na5 {"Provocative," said Leko afterwards, but also "the most thematic in
this position, It reminds me of the Grunfeld, which I love to play with Black.
". He had this position as White against Bologan at Dortmund earlier this year,
but his opponent that day played now 17...Rcd8; the game was drawn in 26 moves.
} 18. c4 {Leko is provoked! This forcing sequence leads ineluctably to a
situation where Black will have rook and bishop for the white queen. Also
possible was 18 Nd2 as 18...Bxg2 fails to 19 c4 Qc6 20 d5 Qg6 21 Bh5 Bf3+ 22
Bxg6 Bxd1 23 Bf5.} Nxc4 19. Bxc4 Qxc4 20. Nd2 Qd5 21. Nxe4 Qxe4 22. Bg5 Qxe1+
23. Qxe1 Bxg5 24. Qa5 Bf6 {!? A tremendous decision. Black could probably draw
with 24...Bf4 25Qxa7 b6 followed by ...Bd6 when it is doubtful that White can
break through. The text is considerably more ambitious. Black creates a passed
pawn for himself, but he must also take account of the fact that White's
passed apawn could become exceedingly dangerous.} 25. Qxa7 c5 {! This is a key
move, creating a situation with rival racing pawns. One might expect this
still to favour the side with the material advantage, but Leko's Queen and
Rook are out-numbered and out-manoeuvred by Kramnik's pieces.} 26. Qxb7 Bxd4
27. Ra2 c4 28. Re2 Red8 29. a4 c3 30. Qe4 Bb6 31. Qc2 g6 32. Qb3 {In the press
conference afterwards, Leko with cheerful self-deprecation commented that
"Once I got my queen to b3, I then spent time trying to get it to e4!"} Rd6 33.
Rc2 Ba5 34. g4 Rd2 35. Kg2 Rcd8 36. Rxc3 {Leko, with only a few minutes left,
tires of the bind that Black is exerting and chooses to go into and endgame
where only Black has winning chances. It might have been better to stay
passive and trust in his blockade.} Bxc3 37. Qxc3 R2d5 {Susan Polgar,
commenting live online, pointed out that an ending with the same material and
the same distribution of pawns on the king's-side occurred in
Gurgenidze-Averbakh, Baku 1961, and it saw a win for the rooks. However,
Leko's pawns stand better placed for defence here than did Averbakh's.} 38. Qc6
Ra5 39. Kg3 Rda8 40. h4 R5a6 {When Leko played h4, Kramnik should have quickly
replied ...h7h6, so as to meet h4h5 with ...g5. It's important to retain pawns
here, as explained below. After the game, Leko mentioned that he was thinking
about the idea Qe1!? but rejected it as too ugly for such a game(!).} 41. Qc1
Ra5 42. Qh6 {White's last two moves inhibit this ...h6 manoeuvre by Black.}
Rxa4 43. h5 R4a5 44. Qf4 {? This is a blunder, which allows Black to fix the
pawn formation and ultimately gang up on the weak white fpawn. White must play
44 hxg6 as they say in the beginner's books, swap pawns in the ending to
reduce the opponent's winning chances. Even after the superior 44 hxg6 hxg6 it
is not obvious that White can reach the safe haven of a draw, as the basic
black strategy of piling up with his rooks against the white f-pawn still
seems valid. However, with only two pawns each on the board, White can place
his g-pawn on g5, so that even if black trades both rooks for queen and f-pawn,
the resulting king and pawn ending is a draw.} g5 45. Qf6 h6 {This is a very
clever move from Kramnik, clearly overlooked by Leko, whose last move could
have no other purpose than to threaten h6. If now 46Qxh6, then ...R8a6 traps
the white queen.} 46. f3 R5a6 47. Qc3 Ra4 48. Qc6 R8a6 49. Qe8+ Kg7 50. Qb5
R4a5 51. Qb4 Rd5 52. Qb3 Rad6 53. Qc4 Rd3 54. Kf2 Ra3 55. Qc5 Ra2+ 56. Kg3 Rf6
57. Qb4 Raa6 58. Kg2 Rf4 {Black weaves a net inexorably around White's sorry
f-pawn.} 59. Qb2+ Raf6 60. Qe5 Rxf3 {Kramnik's manoeuvres, though lengthy,
have finally triumphed. If White does not trade his queen for Black's rooks,
the white king will soon be hunted down. If White does exchange, then Black
has an easy win in the king and pawn endgame.} 61. Qa1 Rf1 62. Qc3 R1f2+ 63.
Kg3 R2f3+ 64. Qxf3 Rxf3+ 65. Kxf3 Kf6 {White resigns. The pawn endgame is a
trivial win, e.g. 66 Ke4 Ke6 67 Kd4 f5 68 gxf5+ Kxf5 69 Ke3 g4 70 Kf2 Kg5 71
Kg3 Kxh5 72 Kg2 Kg5 73 Kg3 h5 and the black pawns march down to promote. A
marvellously dramatic encounter and a fine riposte to critics who thought the
match would be dull. In the press conference after the game, Kramnik looked
drained, Leko cheerful and phlegmatic. Kramnik insisted throughout that the
position objectively is drawn, and pulled many sceptical faces when describing
his win. The following day, both players remarked how hard it is to move on
from such a battle, Kramnik musing "I tried to get the two rooks out of my
mind, and to stop trying to coordinate them in different attacking patterns. I
think it was four o'clock in the morning when I was able to sleep." Leko added,
"I needed until five a.m. to get rid of those two black rooks."} 0-1

A sharp modern game with many phases, which also exemplifies the ancient idea of transformation of advantages.

[Event "Corus Group A"]
[Site "Wijk aan Zee NED"]
[Date "2004.01.17"]
[Round "6"]
[White "Vladimir Kramnik"]
[Black "Zhong Zhang"]
[Result "1-0"]
[ECO "A17"]
[WhiteElo "2777"]
[BlackElo "2639"]
[PlyCount "155"]
[EventDate "2004.01.10"]

1. Nf3 Nf6 2. c4 e6 3. Nc3 Bb4 4. Qc2 O-O 5. a3 Bxc3 6. Qxc3 b6 7. e3 Bb7 8.
Be2 d6 9. O-O Nbd7 10. b4 c5 11. Bb2 Re8 12. d4 Ne4 13. Qb3 Rb8 14. Rad1 Qc7
15. Nd2 Nef6 16. Rc1 a5 17. Rfd1 e5 18. Bf1 axb4 19. axb4 h6 20. Ba1 Rec8 21.
Bb2 Re8 22. dxc5 dxc5 23. Nb1 cxb4 24. Qxb4 Nc5 25. Nc3 Rbd8 26. Rb1 Bc6 27.
Ba1 Rxd1 28. Rxd1 Rd8 29. Nb5 Bxb5 30. Bxe5 Qxe5 31. Rxd8+ Be8 32. Qxb6 Qe7 33.
Rc8 Nfd7 34. Qc7 Kf8 35. g3 Ne6 36. Qc6 g6 37. Bh3 f5 38. Bg2 Kg7 39. Bd5 Nec5
40. Qa8 Bf7 41. Qa1+ Qf6 42. Qa7 Qa6 43. Qxa6 Nxa6 {White has one more pawn
than he deserves for the two pieces, but he lacks a lever on the dark squares.}
44. f3 Nac5 45. Kf2 Kf6 46. Rc6+ Be6 47. Ke2 Ke5 48. f4+ Kf6 49. Kd2 Ke7 50.
Bf3 Nb3+ 51. Kc3 Na5 52. Ra6 Nxc4 53. Rxe6+ Kxe6 54. Kxc4 {Transforming the
game into a pawn-up B vs N position. With all the pawns on one side of the
board, the Knight may be seen at its best. White must create a passed pawn,
while Black will seek to exchange the remainder.} g5 55. Bd5+ Kf6 56. Bc6 Nf8
57. Kd5 Ng6 58. Kd6 Ne7 59. Bb7 Ng6 60. Bf3 Nf8 61. Bd5 Ng6 62. Bc6 Nf8 63. Bd7
Ng6 64. h3 h5 {Provoking a crisis. White gives up the extra pawn to get the
King in.} 65. fxg5+ Kxg5 66. h4+ Kg4 67. e4 Kxg3 68. exf5 Nh8 {Corners are not
kind to Knights; anything would have been better.} 69. Ke7 Kxh4 70. Kf6 Kg4 71.
Kg7 Kg5 72. f6 h4 73. Bc8 Kh5 74. Bf5 Kg5 75. Be6 {Zugzwang, if I may be so
bold.} Ng6 76. f7 Nf4 77. Bc8 Nh5+ 78. Kh7 1-0