1.e4 e5 2.Nf3 Nc6 3.Bb5 Nf6 4.0-0 Nxe4 5.d4 Nd6 6.Bxc6 dxc6 7.dxe5 Nf5 8.Qxd8+ Kxd8 Although he has nice memories in the Najdorf from the previous game against Svidler (San Luis 2005), Topalov opts this time for the rock-solid Berlin Defence.
9.Nc3 Ne7 10.h3 Ng6 11.Bg5+ Ke8 12.Rad1 Bd7
13.Nd4 [Both players had faced this position before. 13.Rd2 Be6 14.Rfd1 Be7 15.Be3 (After 15.Ne2!? Nxe5 16.Nxe5 Bxg5 17.f4 Be7 18.Nd4 Bc5 19.Kh2 Bxd4 20.Rxd4 f6 White's initiative for the sacrificed pawn was sufficient for maintaining equality but not more in Svidler-Kramnik, Dortmund 2004) 15...a6 16.Ne2 Rd8 17.Rxd8+ Bxd8 18.b3 Be7 19.c4 h5 20.Nfd4 Bc8 21.f4 h4 22.Nc3 Rh5 and now, after the careless 23.Ne4? Black missed the opportunity to win a pawn by 23...Nxe5 in the game Polgar-Topalov, Sofia. In the present game, both players remained faithful to their favourite methods of play in this structure: Topalov activated his king's rook by means of h5-h4 and Rh5 while Svidler sacrificed his central pawn in order to open lines and endanger the enemy king.]
13...h6 14.Be3 h5N [Previously, 14...Bb4! 15.f4 Bxc3 16.bxc3 c5 had been played. Black managed to hold his own in Shirov-Almasi, Monaco (blindfold) 2002; Capturing the e5-pawn at this stage would make little sense. After 14...Nxe5 White would most likely win the pawn back with, say 15.Bf4 maintaining the better structure and pressure along the central files.]
16.f5!? With the black king stuck in the centre, such intuitive sacrifices are just natural.
16...Nxe5 17.f6 Rh5 [The position after 17...gxf6 18.Ne4 Be7 certainly looks scary for Black. However, the solution chosen by Topalov, to block the position, will leave White with a huge advantage of space and increase Black's problems of communicating his wings. In spite of succesive modifications of the position, this positional factor will persist for quite a long time.]
18.Ne4 g6 19.Bf4 c5 20.Nf3 Exchanging White's only active piece.
20...Nxf3+ 21.Rxf3 Be6 22.Rfd3 c4 23.R3d2 c6 24.Ng5 Bc5+ 25.Kh2 Bd5 26.Re2+ Kf8 Crossing the d-file would result in the loss of the f7-pawn, because the d5-bishop would be pinned.
27.Rde1 This is the type of position White was aiming for. If Black could connect his rooks his position would be simply wining, but the way it is he is reduced to passivity.
27...b5 28.c3 a5 29.a3 Rc8 30.g4! In order to increase his pressure, White needs to activate his last piece, the king. Spoiling his king side structure is not a too high price to pay.
30...hxg3+ 31.Kxg3 At first sight, this might look like suicide in a solid position. However, in case of a waiting black strategy, White would have had such a plan at his disposal: h4, Bc7-a5-b4 (please not that the black queen's rook is busy preventing make on the back rank and cannot cross the bishop's plans) exchanging the dark squared bishops in order to invade the enemy position through e7. Black cannot meet Ba5-b4 with Bd6 and c5 because this would chronically weaken the d-file.
31...Be6 32.h4 Of course, White is in no hurry to capture on e6, snce this would bring Black some relief.
32...Kg8 33.Re5 Bf8 34.Nxe6 fxe6 35.Rd1 Rh7 [Black takes his chance to recycle his rook. The exchange on e5 would be followed by Rd7 with a winning position, while 35...Kf7 36.Bg5 Rc7 would leave the king's rook out of play again. White could open a new front of action with 37.a4!? ]
36.Rxe6 Rb7 37.Re4 Kf7 38.Bg5 Re8 39.Rxe8 Kxe8 40.Kg4 In spite of the mass simplifications, White retains a huge advantage of space.
40...Rh7 [40...Rd7? would lose instantly to 41.f7+! ; Black had no time for the generally desirable blockade of the queen side with 40...a4 because of 41.Bf4 threatening Kg5. After the text move, the retreatof the bishop can be met by Rh5.]
42.a4! Now that the black rook is temporarily isolated on the king side, White puts the opposite wing under pressure. It is quite isntructive to follow how Svidler will manage to carry out his partial plans until the end of the game by just one tempo. If allowed to play a4 himself, Black would have chances to defend his fortress.
42...bxa4 [Black's coordination is too poor to maintain the tension on both wings. After 42...Bd6 43.Ra1 Bc7 44.h5!! the white king would penetrate with decisive effect, for instance 44...Rxh5 (or 44...gxh5+ 45.Kf5 ) 45.f7 Rh8 (45...Bd6?! 46.Rd1! ) 46.Bh6 Bd6 47.Kg5 and the f-pawn will decide soon the outcome of the battle.]
43.Re5 c5 44.Bf4 Rh8 45.Bg3 Bh6 46.Re7+ Kc6 47.Bf4 [Clearing the way for the king. 47.f7 looks pretty strong, too.]
47...Bxf4 48.Kxf4 Rh5 49.Re5 Rxh4+ 50.Kg5 Rh5+ 51.Kxg6 Rxe5 52.f7 Re6+ 53.Kg5 Re5+ 54.Kg4 Re4+ The theme of systematic pursue of the king is known since the game Fenton-Potter, London 1875 and was succesively developed into a nice study by Potter, Saavedra and finally Barbier in 1895.
55.Kg3 Re3+ 56.Kf2 a3
57.f8Q! [Last accuracy. 57.bxa3 Rxc3 58.f8Q Rxa3 looks quite difficult (probably impossible) to win.]
57...axb2 58.Qc8+ Kb5 [Or 58...Kd5 59.Qb7+ and Black has no way to defend his rook.]
59.Qb7+ Ka4 60.Kxe3 Ka3 61.Qb5 It's all over now. White's king is too close to allow a last moment escape.
61...a4 62.Qxc5+ Kb3 63.Qb4+ Kc2 64.Qxa4+ Kxc3 65.Qa5+ Kc2 66.Qf5+ Kc1 67.Qf1+ 1-0