1. Nf3 Nf6 2. c4 c5 3. g3 Nc6 4. Nc3 d5 5. d4
A move that Kramnik has played many times. Karpov too played it in the past. In NIC Magazine 1/2007, Albert Kapengut writes that Polugaevsky suggested this move in 1979 to Tal. The Wizzard from Riga only used it nine years later. In his match against Timman in Hilversum (1988) it was a theoretical novelty.
5. cxd5 is the most frequently played move in this position.
5... cxd4
5... e6 would lead to the Tarrasch Defence.
Sharper but more risky was 5... dxc4 6. d5 Nb4 7. e4 Bg4 8. Bxc4 Bxf3 9. Qxf3 Nc2+ 10. Kf1 with the idea e4-e5.
6. Nxd4
Now the Exchange Variation of the Grü nfeld is on the board, with an extra tempo for White. In this case it means a temporary pawn sacrifice.
6... dxc4 7. Nxc6 Qxd1+ 8. Nxd1 bxc6
Black is a pawn up but both his c-pawns are weak. Theory gives White a small edge but in practice it's quite difficult to convert it.
9... Bd7?! is passive (the bishop is better placed on a6): 10. Ne3 c3 11. bxc3 e6 12. Nc4 with a comfortable plus for White.
10. Ne3 e6 11. Nxc4 Ba6 12. Na5!?
White immediately starts putting pressure on c6. Kramnik likes this kind of positions - no queens, no spectacle, but small means, subtleties, and always 'playing for two results', as Shipov calls it.
The most frequently played move is 12. b3 which Kramnik tried a few times too. After 12... Bb4+ 13. Bd2 Ke7 14. Rc1 White keeps some pressure, as in Kramnik-Polgar, Tilburg 1997. Apparently Kramnik now thinks 12.Na5 offers more chances for an advantage.
12... Rc8
Also possible was 12... Bc5 . A nice example of how Kramnik, like a wizard, nows to make something out of nothing in this kind of positions, is the next game: 13. Bd2
12... Kd7!? is an interesting possibility, suggested by Rybka.
The first phase of White's strategy has succeeded: the c-pion is now on a black square, so that Be7 is restricted to protecting the pawn. Furthermore, Nd5 loses a stronghold. White's next task is to tie up the pawn, and then start surrounding it.
Also possible was 16... Rc7 17. Nc4 (Perhaps Kramnik was planning 17. Bf3 here, like in the game.)
17... Nb6 18. Ba5 Bxc4 19. Rxc4 Nxc4 20. Bxc7 Rc8 21. Bf4 g5! 22. Bc1 Nb6 23. Bb7 (23. Bd2!?
Rybka)
23... Rc7 24. Ba6 c4 25. bxc4 Nxc4 26. Rd1 Bf6 27. f4 1/2-1/2, Kasimdzhanov-Berelovich, Hoogeveen 1999
16... Bxe2? by the way fails to 17. Rfe1 followed by Bxd5 and Be7 drops.
17. Bf3
Überdeckung of e2. Shipov calls this, with a wink, an 'equal position, but not a drawn position'.
The first new move. Black likes to play the moves Nb4 and Bf6, without being troubled by tricks such as Rxc5 and Bxb4 (with a pin over the diagonal a3-f8).
Immediately 18... Bf6 is answered by the nasty 19. e4!? when(19. Be1 Ke7 and Black has won a tempo compared to the game)
19... Nb4?! is not so good because of 20. Rxc5! Rxc5 (20... Nxa2 21. Be3)
21. Bxb4 Rxd1+ 22. Bxd1 Be7 23. Bxc5 Bxc5 and White won a pawn, although it still ended in a draw after 24. Bc2 e5 25. a4 f6 26. Nc6 Bb7 27. Na5 Ba6 28. Nc6 Bb7 29. Na5 Ba6 30. Nc6 in the game Kopylov-Berelovich, Hamburg 2006.
19. Be1
White chooses another, quiet plan.
19. Rc2 with the idea Bc1-b2, now fails to 19... Nb4! (Shipov)
19... Rd7
On chesspro.ru, Korotylev and Landa analyse the move 19... Nb4 and they conclude that after 20. Rxd8+ Rxd8 (20... Bxd8 21. Bb7! Rc7 22. Bxa6 Nxa6 23. Kg2 and White's better bishop still guarantees a slight advantage)
21. Kg2 (21. a3 Na2!? 22. Ra1?? Bf6 23. Rxa2 Rd1=)
21... Rc8 22. a3 Nd5 , besides the rook exchange, not much has changed - White still has a tiny edge.
The waiting move 20... Rdd8!? (Landa/Korotyl ev) was certainly worth considering, followed by waiting for White to do something, but they do remark that what Navara is doing, is completely logical. He wants to keep on playing concrete.
22... g6 with the idea Bg7 and f5, was better according to Landa and Korotylev, to keep the stronghold d5. 23. Be2 Bg7 24. Kf1 f5
23. Be2!
23. e4?! seems tempting, but is premature to early becayse of 23... Bxc4 (23... Nb6? 24. e5!) 24. Rxc4 Nb6 and Black puts a bishop on d4, when all his problems are over.
23... Nb4?!
This allows a tactical finesse, with which White gains the bishop pair.
23... Ba6!? (Korotylev/Landa) was probably better. After 24. Kg2 , however, Black has to find a move.
24... Bxe2? 25. Nxc8+ Ke8 26. Rxd7 Kxd7 27. Rxc5
25. Bxb5 Nxa2 26. Rxd6 Nxc1 27. Rd7+ Kf8 28. Rxa7
A critical position. If Black would have been able to take on b3, the position would be a draw, but it's not possible. But isn't it possible to play Rb8 first, and only then take on b3?
28... Bd8
This move has been criticised widely, but it's not so clear whether it's that bad.
28... Nxb3? unfortunately, fails tactically: 29. Bd7 Rb8 30. Bc7
The alternative was 28... Rb8 29. Bc4 Nxb3 30. Bc7 Rb4 and now it would be a draw after Bxb3, but White doesn't give up his bishop pair that quickly! 31. Bd6+! (31. Bxb3? Rxb3 32. Bd6+ Ke8 33. Bxc5 Rc3=)
31... Ke8 32. Ba6! This very strong move found by Korotylov and Landa completely follows the principle that bishops will always be a 'force to reckon'. The white threat is Bc8-d7-c6! mating.(32. Be2!?)
32... Rb6 33. Bc7 Rb4 (33... Rxa6 34. Rxa6 c4 35. Rc6 c3 36. Bb8 Kd7 37. Rc7+
)
34. Bb7! Kd7 35. Bf3 Nd2 36. Ba5+ Kd6 37. Kg2 Nxf3 38. Bxb4 cxb4 39. Kxf3 b3 40. Rb7 b2 41. Ke4 and White should win (Landa and Korotylev).
29. Bc4!
Dominating the knight on c1. During the game everybody feared for Navara's life, but he stays cool.
Consistently played, but after this Navara shows his defensive abilities in full splendour.
According to the commentators, 30. Bc3! is the move, although it remains unclear how White will exactly win the knight after 30... Nxb3 31. Be5! (31. Ra3 Bf6! 32. Bxf6 Nd2!=)
31... Rb4 32. Bd3 g6 (32... Bf6 33. Bd6+ Ke8 34. Bxh7! is just possible for White)
33. Bc3 Rb6 34. Bc4 Nc1 35. Kg2 But it doesn't look good for Black.
30... Rxd8 31. Ra1 Rd1+ 32. Kg2 g5!
A beautiful, instructive move. Black sacrifices a pawn for activity.
Na 32... Ke7 33. h4! and Black is in big trouble, he has too many weaknesses:
33. Kf3
33. Ra5 Nd3! with the idea g4 and Ne1 34. Bxd3 Rxd3 35. Rxc5 Rxb3! (35... h6?! 36. Rb5! is tricky for Black, the well-known 'b-pawn problems' in this kind of endings. ) 36. Rxg5 h6 37. Rg4 Rb5 and according to Landa and Korotylev this ending cannot be won.
33. e4 g4 34. e5 h5 and the White king cannot approach.
33. h3 h6 (Not 33... h5? 34. Be2!) (but possible was 33... Ke7!? 34. Kf3 Kf6 35. Ra5 Rd7! 36. Rxc5 Rb7 37. Kg4 h6 etc.) 34. Kf3 Ke7 35. Ra5 Rd7 36. Rxc5?! Rb7 and Black wins b3 with a probable draw.
33... h5 34. h3 Ke7 35. Ra5 f5!
37. Kg2!? Shipov
Here Korotylev and Landa remark that although White has won a pawn, his bishop can never be removed from the diagonal a6-f1 because of mate in one.
39. e4!
The best try. White returns the pawn to be able to let his king rejoin the game.
39. f3 gxf3 (39... Rd2 40. e4 (40. fxg4?? Rf2#) 40... gxf3 41. e5+) 40. Kxf3 Nd3 41. Bxd3 (41. Ke2 Nxc5 42. Kxd1 Ne4=) 41... Rxd3 42. b4 Rb3 and White cannot win.
39... Rd4 40. Ke3 Rxe4+ 41. Kd2 Na2
Still the knight has not been released.
42. Rb5?!
42. Ra5! offered better chances: 42... Nb4 43. Kc3 Nc6 44. Ra6 Ne7 (44... Nd4? 45. Bd3)
45. b4 although according to the experts, eventually this position should be a draw too, at the end.
42... f4
More simple was 42... Rd4+! 43. Kc2 Rd6 and White cannot make progress, e.g. 44. b4 Rc6=
43. Ra5
Black's idea, naturally, is 43. Kc2 fxg3 44. fxg3 Re3
43... Nb4 44. Ra4 fxg3 45. fxg3
45... Nc6 46. Bb5 Nb4 47. Bc4 Nc6 48. Bb5 Nb4 49. Bc6!?
Another winning attempt. Navara is put to the test till the very end, but he just won't give in.
The best way. Black again gives a pawn to be able to hold his position.
Not 50... Nxc6? 51. Rxd4 Nxd4 52. Kxd4 and because of the distant passed pawn, the pawn ending is won.
51. Rxb4 Rxc6+ 52. Rc4 Rd6 53. b4
53. Rxg4 e5 and the passed e-pawn is enough to draw, e.g. 54. b4 Kf5 55. Rg8 e4 56. Kc4 Rd1 57. Rf8+ Ke5 58. Re8+ Kf5 59. b5 Rc1+ 60. Kd5 Rd1+ 61. Kc6 Rc1+ 62. Kb7 Rc3 63. Rf8+ Ke5 64. g4 Rg3 65. b6 Rxg4 66. Kc8 Rg6 67. b7 Rb6 68. Kc7 Rxb7+ 69. Kxb7 e3= (Landa/Korotylev).
53... e5 54. b5 Kf5 55. Rb4 e4 56. Rd4 Rh6 57. Rd1
57. Rd5+ Ke6 58. Kd4 Rh3 59. b6 Rxg3 60. Rb5 Rd3+! 61. Kxe4 Rd8=
No Black needs to play one more accurate move.
58... e2!
...and he does. Landa and Korotylev write that with this move Black "passed his endgame technique exam by academic Vladmir Kramnik".
59. Rb1
59. Kxe2 Rh2+ 60. Kd3 Rg2 61. Kc4 Rxg3 62. b6 Rh3 63. b7 Rh8 64. Rb1 Rb8 65. Kc5 Ke4=
59... Rh2 60. Kd4 Rf2 61. Re1 Rg2 62. b6 Rxg3 63. Rxe2
and drawn. A magnificent piece of endgame defense by Navara.
1/2-1/2