1.d4 d5 2.c4 c6 3.Nf3 Nf6 4.e3 Bf5 5.Nc3 e6 6.Nh4 Bg6 7.Nxg6 hxg6 From the point of view of general match strategy, a courageous decision. Kramnik has nothing against repeating the opening variation that had led him to a rather convincing defeat in the 9th game.
8.Rb1!?N [Another opening surprise prepared by Topalov. The idea behind his last move is similar to that of 8.a3 , as played in the 9th game, namely to prepare c4-c5.]
8...Nbd7 9.c5 [This might look a bit too straightforward to offer White an advantage, but is not without poison. The pair of bishops and the considerable advantage of space on the queen side offer White interesting possibilities. If compared with the previous game where this variation was played, White cannot prevent the central break ...e5 by playing f4 before c5 because Black would take over the control of the e4-square by means of ...Bb4 (even if with loss of time after 9.g3 Be7 10.f4). From this point of view, 9.a3 deserves some attention and if 9...a5 then 10.f4 but apparently Topalov intended to give the game an independent course in order to avoid the opponent's preparation.]
9...a5! [A well timed move. Before starting his counterplay in the centre, Black makes sure the a-file will be open for his rook. In case of 9...e5 10.b4 it would be too late for 10...a5 already, because of 11.b5 with a promising queen side position for White.]
10.a3 Practically forced in view of the threat ...b6.
10...e5 11.b4 axb4 12.axb4 Qc7 13.f4 Probably still following the plan prepared at home. White prevents Black's counterplay from taking threatening forms and forces him take a decision at once.
13...exf4 [In case of 13...e4 White could gradually start active operations on both wings. The advance of the b7-pawn to b5 is not a reliable solution against White's intended queen side attack because of the permanent danger of a piece sacrifice for two pawns on b5.]
14.exf4 Be7 [Topalov probably relied on the fact that the weakness of the e4-square is not so easy to take advantage of. If 14...Ne4 (now or on the next few moves), then simply 15.Nxe4 dxe4 16.d5 with initiative in the centre. However, Kramnik's accurate regrouping will eventually enable the generally desirable knight jump.]
15.Be2 Nf8 16.0-0 Ne6 17.g3 [White has to be careful. He has no time for 17.Qd3 because of 17...g5 , solving all Black's problems.]
17...Qd7 18.Qd3 Ne4 It's now or never. White threatened to consolidate his advantage of space by means of Be3 and Bf3, when Black would have been reduced to passivity.
19.Nxe4 dxe4 20.Qxe4 Qxd4+ 21.Qxd4 Nxd4 22.Bc4 In spite of simplifications, White retained his advantage of space and the pair of bishops. However, the rook's activity along the a-file and his lack of weaknesses offer Black reasonable counter-chances.
22...0-0 23.Kg2 Ra4 24.Rd1 Rd8 25.Be3 Bf6 26.g4 [A questionable move. By weakening the f4-square without previous preparation, White practically commits himself to the further advance of pawns, which will offer Black stable squares for his minor pieces and excellent prospects for counterplay. A more restrained alternative was 26.h4 , leaving Black with long-term problems regarding the defence of the f7- or (if ...f6 is played) the g6-pawn.]
26...Kf8 27.Bf2 Enabling the threat g5, which would have led nowhere if played right away because of the intermediate ...Nf5.
27...Ne6 28.Rxd8+ Under certain circumstances, g5 would have consolidated White's advantage of space. This would be impossible here because of ...Nxf4 and ...Bxg5, which reveals one of the positive aspects of the recommended 26.h4.
28...Bxd8 29.f5 [With so many pawns placed on dark squares, White's advantage after 29.Bxe6 fxe6 would probably have a purely symbolic character.]
29...gxf5 30.gxf5 Nf4+ 31.Kf3 Nh5 32.Rb3 Bc7 33.h4 Nf6 Black has regrouped in optimal way. The f5-pawn is safely blocked, preventing an eventual attack of the b7-pawn by means of Bc8, while the white king's activity is restricted by a barrier along the fourth rank. We can also notice that the advantage of space has a relative character already. Since the white pieces cannot easily sustain the pawn's offensive, the whole space left behind can be used mainly by Black for counterplay.
34.Bd3 The pressure against the f7-square did not promise White much and Topalov decides to remove the bishop from the active but slightly exposed position in order to enable b5.
34...Nd7 35.Be4 [But this looks over-ambitious already. The immediate 35.Bb1 , preventing the enemy rook's activation aong the second rank, looks like a safer decision.]
35...Ne5+ 36.Kg2 Ra2 37.Bb1 Rd2 38.Kf1 Ng4 39.Bg1 The white pieces have been pushed even further back from the own pawns, which leaves the whole king side in some danger.
39...Bh2 Intending the elimination of the main defender of the dark squares.
40.Ke1 Rd5 41.Bf2 Ke7 42.h5 Nxf2 43.Kxf2 Kf6 44.Kf3 Rd4 45.b5 Finally, White has managed to carry out this thematic advance, undermining the c6-square and opening new perspectives for his bishop. In spite of the inevitable loss of the h5-pawn, this will more or less guarantee a draw.
45...Rc4 46.bxc6 bxc6 47.Rb6 Rxc5 48.Be4 Kg5 49.Rxc6 Ra5 50.Rb6 Ra3+ 51.Kg2 Bc7 52.Rb7 Rc3 53.Kf2 Kxh5 54.Bd5 f6 55.Ke2 Kg4 56.Be4 Black has managed to maintain his minimal material advantage and avoid the exchange of rooks at the same time. However, his extra-pawn is hardly relevant and his last remaining chance consists of creating some threats against the enemy king, possibly combined with the relative lack of stability of the enemy bishop along the b1-e4 diagonal. We can see that White's advantage of space finally does him a good service. If we lift the whole king side structure with, say, 2 ranks, Black would get reasonable winning chances already, one of the possible ideas being the exchange sacrifice in order to remove White's last pawn.
56...Kf4 57.Bd3 Rc5 58.Rb4+ Kg3 59.Rc4 Re5+ 60.Re4 Ra5 61.Re3+ Kg2 62.Be4+ Kh2 63.Rb3 Ra2+ 64.Kd3 In order to avoid Black's playing on in the conditions of limited time, the white king has to attack the g7-pawn, when a draw would become obvious. However, the shorter path could prove dangerous here. Once the king would reach h5, there would be such a mating construction available: ...Rg5+ Kh4 Bg3 (or e1) mate.
64...Bf4 65.Kc4 Re2 66.Kd5 1/2-1/2