4...Bf5. If 4...e6 5.Bd3 Nbd7 (the 6th game saw 5...dxc4 6.Bxc4 c5, and losing a tempo each one  i..e. not losing anything  the opponents transposed into the Queen's Gambit Accepted) then White plays 6.00! (neutralizing significantly the Meran plan with 6...dxc4 and 7...b5). That's the point of the 4.e3 move order. Naturally it doesn't win by force. Last month I had an occasion to study the game of the Polish Extraliga Bartel  Jaracz where Black prevailed after 6...Bd6 7.Nc3 00 8.e4 dxc4 9.Bxc4 e5 10.Be3 b5 11.Bb3 Qe7 (Loek Van Wely plays 11...Bb7 here) 12.Bg5 h6N 13.Nh4!? exd4 14.Ng6 Qd8 15.Qxd4 Bxh2+!! 16.Kxh2 Ng4+ 17.Kg1 Qxg5 18.Qd6? (White should have gone for 18.Nxf8 Qh4!) 18...Re8 19.Qxc6 Rb8, etc. 5.Nc3 e6 6.Nh4 Bg6. This somewhat boring line is often played at the professional level. Both sides stand solidly, White has a microplus. 7.Nxg6. 7.Qb3 is an alternative, true later on White all the same takes on g6  as a rule right off. 7...hxg6. 8.a3!? Almost a novelty. Last year Kramnik played 8.Bd3 as White: 8...c5!? (8...Nbd7 Vyzmanavin  Kramnik, Paris (rapid) 1994 and other games) 9.Qb3 Qd7 10.cxd5 exd5 11.dxc5 Bxc5 12.Qb5 Qxb5 13.Nxb5 Kd7 14.00 Nc6 15.Rd1 Ke7 16.Bd2 Bb4 17.Nc3 Rac8 18.Rac1 Rhd8 19.a3 Bd6 20.Ne2 Ne5 draw agreed, Kramnik  Gelfand, European Club Cup 2005. More topical is 8.g3 Nbd7 followed by for example, 9.Bg2 (or 9.Bd2 Be7, and now 10.Rc1 Topalov  Vallejo, Monaco (rapid) 2005 or 10.b3 00 11.Bg2 Kramnik  Gelfand, Monaco (blindfold) 2005) 9...dxc4 (accepting the pawn sac) 10.Qe2 Nb6!? (10...Be7 Крамник Kramnik  Gelfand, Monaco (blindfold) 2003) 11.00, when 11...Bb4!? (instead of 11...Be7  Kramnik  Anand, Frankfut (rapid) 2000, Sasikiran  Shirov, Corsica (rapid) 2005) 12.a3 Ba5N allowed Black to solve all his problems in Ponomariov  Shirov, Foros 2006. Frequently played is also 8.Bd2 Nbd7 and here, a la Kasparov, 9.Rc1 (9.Qb3 Qc8!? 10.g3 Be7 11.Rc1 Nb6 12.cxd5 exd5 13.Bg2 Nc4 14.Qc2 Qd7=, draw agreed, Drozdovsky  S.Savchenko, Odessa 2006; 9.g3  8.g3), where Black scores well after 9...Bd6!, as in Ivanchuk  Cheparinov, KhantyMansyisk 2005. 8.Qb3 Qc7 (or 8...Qb6) is yet another trend. Well it can't be denied that even 8.h3 (Karpov  Akopian, Dagomys 2006 and others) is more popular than Topalov's choice. 8...Nbd7 9.g3N. A new move. The previous game on the subject saw 9.h3. Somebody may ask  why didn't Topalov opt for another move order: 8.g3 Nbd7 (after all Black rarely plays anything else) and now 9.a3? Frankly I don't know. A bizarre, theoretically less precise move order is often the strongest from the practical standpoint robbing the opponent of a couple of minutes. In the case at hand Topalov could also cut off the possibility of 8.g3 Bb4. 9...Be7 10.f4!? Here is the idea of Francisco Vallejo, one of Topalov's seconds. Playing almost exclusively by pawns White seizes more space. Commonly even a good move has its drawbacks  in this case it could be a weakening of some squares (e3, e4, g3) and a lag in development. But since Black's arrangement is rather passive it's difficult for him to exploit these minuses. Soon we will see Topalov left with pluses solely. 10...dxc4. By this exchange Kramnik closes an issue of the blockading move 11.c5, which could have been a reply to 10...Qc7 or 10...00 (in the latter line White had ideas with g3g4 as well). Attractive is a desperate jump 10...Nh5 with an attack on the g3 pawn (suggested by Andrey Deviatkin). True I myself lost quite a number of games by such jumps. The least probable course of the events is 11.cxd5 (11.Qf3 looks normal; 11.Bg2 dxc4!? 12.Qe2) 11...Nxg3 (this move isn't necessary) 12.hxg3 Rxh1 13.dxe6 with compensation for the exchange. There is something antipositional in the continuation 10...Ne4 11.Nxe4 dxe4. But if you have a strong wish it's playable too. 11.Bxc4. 11...00?! As you understand, my duty is to point some mistakes of the one who lost the game. The castling raises doubt however strange it may seem. Certainly, 11...c5?! 12.d5! exd5 (12...Nb6 13.Bb5+) 13.Nxd5 doesn't lead to equality. 11...a6 could be associated with a piece sacrifice: 12.e4 b5 13.Be2 c5 14.e5 cxd4!? 15.exf6 Nxf6. But if someone says  White is simply winning here what would I respond? Normal is 11...Qc7!? (Shipov), preparing the long castling as required. In general it's unpleasant when your opponent may castle both sides. In case of 12.e4 Black may think about 12...Rd8!? 13.e5 (13.Be3 c5!?) 13...Nd5 14.Nxd5 cxd5. It seems to me that another decent option is 11...Nb6!? 12.Be2 c5 13.dxc5 (here 13.Bb5+ Kf8! isn't dangerous for Black) 13...Qxd1+ 14.Nxd1 Bxc5 15.b4 Bd6 16.Bf3, and to deliver himself from positional subtleties Black may sacrifice a pawn  16...Rc8!? 17.Bxb7 Rc2. The compensation is vague, but the pieces are active. 12.e4! Black's problem is that there's little he can do with his opponent's pawn centre. 12...b5!? White is clearly better after 12...c5?! 13.e5! Ne8 (13...cxd4 14.exf6 Nxf6 15.Ne2!) 14.d5 Nb6 15.Ba2 (there's no need in a more exotic 15.b3) 15...exd5 16.Nxd5. Also suspect is 12...Qc7 13.e5! Nd5, to which White may reply 14.Nxd5 cxd5 15.Bd3 (threatening to checkmate after h2h4h5; the knight d7 is quite miserable), and after 15...Qb6  to the extent of 16.Be3 Qxb2 17.h4 Bxa3 (or 17...Rfc8 18.h5 g5 19.Rb1) 18.Kf1! (but not 18.h5?? Qg2! 19.Rg1 Bb4+!  it transpires that the a3 pawn was of certain importance). The was a choice however: 12...Rc8!? 13.Be2!? (13.00 c5!, and bad is 14.e5? cxd4 15.exf6 Nxf6!; the strongest 14.d5 Nb6 15.Ba2 exd5 16.e5! Ne4 17.Nxd5 c4! 18.Nxe7+ Qxe7 shouldn't frighten Black) 13...c5 14.e5 cxd4 15.Qxd4!? (15.exf6 Nxf6 gives Black a compensation for the piece though I don't dare say it's sufficient after 16.Na2 Qb6!?, intending Rfd8, Nd5) 15...Bc5 16.Qd3 Nd5 17.Nxd5 exd5, and here again due to the restless knight d7, White's chances should be better after the exemplary 18.b4 (18.Qxd5 Qb6, and bad is 19.Qxd7?? Bg1!  Monokroussos) 18...Bb6 19.Bb2. 12...Nb6 13.Be2 (or perhaps 13.Ba2) 13...c5 14.Be3 is possible as well where White may claim a small but durable advantage as they said in good old times. 13.Be2! b4! Consequently played. 14.axb4 Bxb4 15.Bf3. 15...Qb6? Not without a shadow of doubt I declare this move to be a fatal mistake. Most likely after the principled 15...c5! White's advantage isn't decisive. Let's consider some possibilities: a) 16.Be3 Nb6 17.Bf2!? (17.Qd3 cxd4, when both 18.Qxd4 Qe7! and a somewhat stronger 18.Bxd4 Bc5 19.Ne2 e5! 20.fxe5 Nfd7 aren't dangerous for Black) 17...Nc4! (17...cxd4 18.Qxd4+/=, where White has the upper hand after both 18...Qxd4 19.Bxd4 Rad8 20.000!, and 18...Qe7 19.e5! Nfd5 20.00 Bxc3 21.bxc3 Rac8 22.Bxd5!?) 18.Qe2 (18.Qb3 Na5 19.Qa4 Nc4 20.000 looks doubleedged) 18...cxd4 19.Qxc4 dxc3 20.bxc3 (20.Qxb4? Qd2+ 21.Kf1 cxb2!, and White will have to give up the rook for this pawn) 20...Ba5!=; b) 16.Kf1!? probably retains some plus (and in principle it's a pleasure to play this), but I suppose it's not what Kramnik was afraid of; с) 16.e5 cxd4 17.Qxd4 Nd5 18.Bxd5 Bc5+/= (it's not a great finesse: 18...exd5?? is simply bad on account of 19.Qxb4) 19.Qd3 (if 19.Qe4 exd5 20.Nxd5 Nf6!!, and White is forced to play 21.Nxf6+ gxf6=, where he has nothing but equality: 22.Bd2 fxe5 23.f5 Qd4!) 19...exd5. Maybe White can resolve to capture a pawn here 20.Qxd5!? (20.Kf1 g5!? Sakaev, Yemelin) with a possible continuation 20...Qe7 21.Kf1!?. The lag in development is obvious but so far White controls all the key squares in his camp and to all appearance he shouldn't be checkmated… Of course in any case Black has drawing chances. 16.00!+/. White's advantage is determined by the evidently better pawn formation, control of the centre and the pair of bishops. The black knights have no steady posts and only embarrass each other. 16...e5. Or 16...c5 17.e5 Nd5 (17...Bxc3 18.bxc3 Nd5 19.c4 Nc3 20.Qd3 cxd4 21.Bxa8 Rxa8 22.Be3! dxe3 23.Qxc3 e2+ 24.Rf2  Monokroussos, Black unavoidably cedes the epawn and even after losing his cpawn White's potential will suffice to win; 17...cxd4? 18.Na4 is completely bad) 18.Nxd5 exd5 19.Be3, and Black's position is difficult as 19...cxd4 20.Bxd4 Bc5 21.Bf2! doesn't bring relief. Perhaps he should have already refrained from the pawn advances. 17.Be3! Rad8. Or 17...exd4 18.Na4! (but not 18.Bxd4? Bc5, trading the darksquared bishop). In the event of 17...exf4 White would face a difficult choice between 18.gxf4, 18.Bxf4 и 18.Na4 as all these moves are good enough. 18.Na4!? The following play presented a systematic (or an unhurried to be more precise) massacre. To go into nuances wouldn't make much sense. 18...Qb8 19.Qc2! exf4. After 19...exd4 20.Bxd4 Nb6 Black is stably bad as well. 20.Bxf4 Qb7 21.Rad1!? Rfe8 22.Bg5 Be7 23.Kh1!? A blunder 23.Rf2? would have given Black good drawing chances after 23...Nxe4! 24.Bxe4 Bxg5 25.Bxc6 Qb4 because the sacrifice 26.Rxf7 is hardly correct. 23...Nh7 24.Be3!? Many people couldn't help playing 24.Bxe7 Rxe7 25.e5, but Topalov decided not to change anything  probably due to his big advantage on the clock. 24...Bg5 25.Bg1! Nhf8 26.h4!? Be7 27.e5! Nb8 28.Nc3!? The least forced among all the reasonable continuations. 28...Bb4 29.Qg2!? Not a bad move and also a trap (White anticipates his opponent's idea). To a more natural 29.Ne4 Black would probably answer 29...Ne6. 29...Qc8. In case of 29...Ne6? White achieves an overwhelming position right away: 30.d5 cxd5 31.Nxd5+. 30.Rc1!? Bxc3. Black's defence doesn't become easier after this exchange. Once again 30...Ne6? 31.d5 was bad, but it was worth taking the queen away from c8 to some other place. 31.bxc3! Ne6 32.Bg4. Here and on the next moves Topalov is obstinately unwilling to proceed to an assault (h4h5). 32...Qc7 33.Rcd1 Nd7 34.Qa2. White decides not to touch the c3 pawn either. 34...Nb6 35.Rf3. 35...Nf8? It was possible to conduct a total defence by 35...Rd7 36.Rdf1 (if 36.Rxf7?! Rxf7 37.Bxe6 Nd5 38.Bxf7+ Qxf7, and all the white pawns one of which is an extra pawn are placed on the disadvantageous dark squares; 36.c4!?) 36...Qc8! (Shipov), or to go for complications  35...c5!? 36.d5 Nf8 37.d6 Qb7 (Monokroussos) with some practical chances to survive. 36.Rdf1! Now Black's position collapses. 36...Re7 37.Be3! Nh7 38.Rxf7! Nd5. If 38...Rxf7 39.Rxf7 Qxf7, then, of course 40.Be6+. 39.R7f3. This way Topalov could have won this position for a dozen of moves. Mathematically, better was 39.Be6!, but actually Kramnik resigned at once. In the 10th game Vladimir Kramnik scored his first really convincing win in the match and thus leveled the official score. At the point when Veselin Topalov blundered his position was probably defensible. But Kramnik had an initiative throughout the game and it was him who introduced a sound opening novelty. Welldeserved victory. Recovering a bit from an inexpressible terror of the "toilet scandal" and listening to people who doesn't take a keen interest in chess I can draw a conclusion that for the general public a match with the "toilet scandal" isn't that boring and disgusting as series of draws from another tournaments in which people don't see neither rhyme nor reason. Catalan Opening Е08

8.Bf4! In principle White is planning to develop the knight at d2 (at c3  only when it doesn't imply a pawn sacrifice). It's logical to free this square immediately. 8.Qc2 is also played where Black recently scored well replying with Topalov's 8...b6 9.Bf4 Ba6!..
8...Nbd7. Let's remember the line 8...b6 9.Nc3! Ba6 (9...dxc4?! is already very risky; of course Black may play 9...Bb7 or 9...Nbd7, but it's rather passive) 10.cxd5 cxd5 11.Rc1 Nc6 12.Nxd5 Qxd5 13.Ne5 Nxd4 14.Bxd5 Nxe2+ 15.Qxe2 Bxe2 16.Bxa8 Rxa8 17.Rfe1 Bb5 18.Rc2 Nd5 19.Rec1. Last year Topalov lost twice in this position  to Ponomariov in Sofia and later on to Kramnik in Dortmund.
9.Qc2. An identical position may arise via 4...Be7 move order. But there it made less sense to send the bishop to f4 (from c1) than in the case in question (from d2).
9...a5!? A very rare variation. 9...b6 is seen more frequently (for example Karpov  Kramnik, Monaco (blindfold) 2001). Or 9...Nh5 10.Bc1 (backwards!) 10...Nhf6 (also backwards!). [There is a strategic risk in both 10...f5 and 10...b5 11.Ne5! Dizdarevic  Short, Bosna  Solingen 1988  here of course 11...Nxe5? 12.dxe5 bxc4 is bad due to 13.g4].
But in so doing Black accepts exactly what he was trying to avoid (a position from the line 4...Be7 5.Nf3 00 6.00 c6 7.Qc2 Nbd7 arises with a loss of three tempi by both sides). Commonly White chooses between 11.Nbd2, as in Gelfand  Adams, Wijk aan Zee 2006 and about thousand other games, and 11.b3 with approximately the same amount of practice. From time to time Black tested 9...Ne4.
10.Rd1!? To 10.Nbd2 Black plays 10...Nh5!? and what a pity, the bishop can't retreat to c1 already while on e3 it isn't placed suitably. If 10.Nc3, then 10...dxc4! is possible. 10.b3 has been tried a couple of times but this gives Black a hook on the queenside.
10...Nh5. 10...a4 11.Ne5 (11.Nbd2!?; 11.c5!?) 11...Qa5 12.Nc3 (12.Nd2 Nxe5 13.Bxe5 Ng4!= Andersson  Spassky, Turin 1982, trading the bishop as to 14.Bf4 Black has 14...g5) 12...Qa6! 13.Rac1!? dxc4 14.Ne4 a3 15.Nxc4 axb2 16.Rb1 with a very slim edge for White, Beliavsky  Eingorn, USSR (ch) 1990.
11.Bc1 b5. Total chess. The game Marin  Pogorelov, Barcelona 1993/4 saw 11...Nhf6 12.Nbd2 b5 13.c5!, and White's prospects on the kingside are better. Commenting on that game Marin suggested 11...f5 having in mind 12.b3 Nhf6 13.Ba3 Bb4, nevertheless this idea is ambiguous.
12.cxd5N. This is already a novelty. After 12.c5 f5! Black can't be inferior. The game Rahman  Ghaem Maghami, Doha 2003 went 13.Nc3 g5 14.a3 Bf6 15.Qd2 h6 16.h4 g4 17.Qxh6 gxf3 18.Qxh5 fxg2 19.Qg6+ with a draw, which Black shouldn't turn down: 19...Bg7?! 20.Qxe6+. To 12.e4 Black plays 12...bxc4! 13.e5 f5 (or even 13...f6, since 14.g4?! fxe5 15.gxh5 e4 looks somewhat frightful for White), and now 14.exf6. I will not try to prove White's compensation after all possible captures on f6.
12...cxd5.
13.e4! Seemingly from now on it is possible to talk about a small but tangible White's advantage. This assessment (of course this is just an assessment not an ultimate truth) will be kept up to Topalov's blunder on move 24. After 13.Ne5 Bb7 14.e4, probably the most precise is 14...dxe4! (less clear is 14...Rc8 15.Qe2 Nhf6 16.Nxd7!, but not 16.Qxb5 dxe4!; about 14...Nhf6 see below) 15.Nc6 Bxc6 16.Qxc6 b4!? 17.Qxe4 Rc8  Black controls the d5 square and his position is hardly worse.
13...dxe4. It is important to find out how White would answer to 13...Nhf6. But it's not that simple. Maybe 14.Ne5!? (14.Nc3?! b4! is unable to pose problems for Black; or 14.Nbd2 dxe4! 15.Nxe4 Bb7=; after 14.exd5 Nxd5 15.Nc3 Nxc3 16.bxc3 Bb7 White most likely doesn't obtain an edge; to 14.e5!? Black replies 14...Ne4, and his position is rather healthy than the contrary  15.Ne1 f5!?) 14...Bb7 15.Nxd7! Qxd7 (15...Nxd7 16.exd5! with an initiative) 16.e5, and here Black at least can't establish the knight on e4: 16...Ne4? 17.f3 Rac8 18.Qe2 Ng5 19.h4+.
14.Qxe4 Rb8. The rest is weaker: 14...Qb6? 15.Ng5! (but not 15.Qxa8? Bb7) 15...Bxg5 16.Qxa8 with a sound extra exchange; 14...Nb6?! 15.Ne5 (or 15.Qe2) 15...Nf6 16.Qe2, winning a pawn without a visible compensation; 14...Ra6?! 15.Qe2!?, and the rook is obviously misplaced.
15.Qe2! Stronger than 15.Ne5 Nhf6!? (an attempt to sacrifice a pawn by 15...Bb7!? 16.Qe2 Nhf6 17.Bxb7 Rxb7 18.Nc6 Qe8 is also possible, most likely Black's position is tolerable; 15...Nxe5? is bad on account of 16.Qxe5+ with the fork on b8 and h5) 16.Qe2 Nxe5 17.dxe5 Nd5, and I believe Black shouldn't be in trouble.
15...Nhf6. Equally principled is 15...Bb7!? 16.d5!? (16.Ne5  15.Ne5 Bb7 16.Qe2) 16...Bxd5!? (after 16...exd5 White has an evident initiative for the pawn although an advantage is yet to be proved) 17.Nd4!  here White wins a pawn or an exchange but I'm not absolutely sure of his advantage: 17...Nhf6 (or maybe 17...Bxg2!? 18.Kxg2 Nhf6 19.Nc6 Qe8 20.Nxb8 Qxb8) 18.Bxd5 Nxd5 19.Nxe6!? (19.Nc6 Qc7! 20.Nxb8 Rxb8) 19...fxe6 20.Qxe6+ Rf7 21.Qxd5 Qb6 (Notkin).
16.Bf4! Rb6 17.Ne5! 17.Nc3 runs into 17...Ba6!
17...Nd5! To all appearances the best choice.
17...Bb7?! 18.Bxb7 Rxb7 19.Nc6 Qe8 20.Nc3! (20.Nxa5);
17...Nxe5?! 18.dxe5 Nd5 19.Be3! Rb7 (19...Rb8 20.Bxd5 exd5 21.Nc3+/) 20.Nc3! Nxc3 21.bxc3 Qe8 (21...Rd7 22.Bc6!+/; 21...Qc7 22.Bxb7 Bxb7?! 23.Qxb5), and here instead of taking the exchange White may play 22.Rdb1!?+/ ;
17...Ba6?! 18.Nc6! Qe8, and White has a promising choice. Interesting is for example 19.b4!? (19.d5 b4!; 19.Bc7 Rxc6! 20.Bxc6 Qc8 21.Bxd7 Qxd7 22.Bxa5 Bb7) 19...axb4 20.d5 with an initiative.
A more serious alternative was probably 17...Nb8!? 18.Nc3 (18.d5 Nxd5 19.Bxd5 exd5 isn't dangerous for Black) 18...Ba6. But if White refrains from 19.d5?! b4 20.dxe6 Bxe2 21.exf7+ (this isn't clear), and just looks for a good square for his queen instead, he would have good chances to develop his initiative.
18.Bxd5! Dennis Monokroussos whom I had praised so much yesterday, came up with an amazing idea 18.Nc3?! Nxf4 19.gxf4. Even the young Tal couldn't be struck with a thought to do such thing to his own pawn structure.
18...exd5. To 18...Nxe5 19.Bg2!? looks good (19.dxe5 exd5 isn't quite convincing, when 20.Be3 could be met with both 20...Rb7 and 20...Re6; in the event of 20.Nc3 Black may play 20...d4!? 21.Be3, and here even 21...dxe3!? 22.Rxd8 exf2+, Zagrebelny. In the game Black might have transposed into this line by playing 19...Nxe5!? after 18...exd5 19.Nc3). Now if 19...Ng6 (or else White will carry out d4d5 fast), then 20.Be3, and the black pieces are poorly placed.
19.Nc3.
Black loses a pawn but he automatically obtains compensation for it in the form of the scope opening for the black pieces after it's gone.
19...Nf6!? Interesting was 19...Bb7!? 20.Nxb5 a4 (Shipov) 21.Nc3 Nxe5!? 22.Bxe5 Qd7, but having for an object the fight for equality I better like 19...Nb8!? 20.Nxb5 f6! (driving the white knight to an inferior position) 21.Nf3 (21.Ng6? hxg6 22.Bc7 Re6!; to grab the pawn by 21.Nc6 Rxc6! 22.Bxb8 looks dangerous for White because of 22...Bf5!?) 21...Ba6 22.a4 Bb4 (apparently this is more precise than 22...Qd7 23.Rac1 Bxb5 24.axb5, when Black should play 24...Rb7!).
20.Nxb5 Ba6. 20...a4 21.Nc3! a3!? deserved some attention but there's no real reason to disapprove of Topalov's choice.
21.a4.
21...Ne4. After 21...Bb4 22.Rdc1!? White (as in many lines that we will see below) gives the pawn back leaving Black with certain problems: 22...Qe8 23.Rc2 (23.Rc7 Bxb5 24.axb5 Rxb5 25.Qc2) 23...Bxb5 24.axb5 Qxb5 25.Qxb5 Rxb5 26.Nc6 (lines by Maxim Notkin).
Topalov might discard 21...Qe8 in view of 22.Bd2 (it seems that after 22.Nd3 Bxb5!? 23.axb5 Rxb5 24.Re1 Bd8! Black holds) 22...Bxb5 (or 22...Bb4 23.Bxb4 axb4 24.Rdc1!) 23.axb5 Qxb5 24.Qxb5 Rxb5 25.Nc6!?. This one looks unpleasant. There is a point in 21...Re6 (in order to regain the pawn slowly) 22.Rac1 Qb6, but after 23.Qd3 or 23.Rc2 again Black has no clear equality.
22.Rdc1! Exactly with this rook. Another one is waiting for the afile to open.
22...Qe8. Protecting the c6 square c6 and exerting additional pressure on b5. If 22...Bg5!? 23.Bxg5 (23.Nc6!? is also possible) 23...Nxg5, then White retains some advantage by 24.Ra3!? (unclear is 24.Rc6 Rxc6 25.Nxc6 Re8! with idea 26.Qg4 Bc8!  Shipov; or 24.Nc6 Re8 25.Qg4 Qf6 26.Ne5 h5!).
23.Rc7!? The activation of the rook is a good decision at least from the practical standpoint. Black's choice is confined now. Of course White had alternatives.
23.f3!? Ng5! (23...Nd6 24.Qe1!?; 24.Qd2!?) 24.Bd2 Bxb5! (24...Bf6 25.Bxa5 Bxb5 26.axb5 Qxb5 doesn't convince in view of 27.Qd1!, and if 27...Bxe5 28.dxe5 Rh6, then 29.b3!) 25.axb5 Qxb5 26.Qxb5 Rxb5 27.Rxa5 Rxb2 28.Bxg5 Bxg5 29.f4 Bf6 30.Rxd5 h6, and this ending doesn't look winning for White.
Interesting is 23.Nc4!? Re6 (23...dxc4 24.Qxe4; 23...Rxb5 24.axb5 Bxb5 25.Rxa5 Bb4 26.Rxb5 Qxb5 would hardly secure Black the wellbeing) 24.Ne3!. But this is sophisticated and it's not completely clear whether it's really good. Objectively the humble 23.Rc2!? wasn't bad.
23...Bd8! 24.Ra7! Eccentric is 24.Rd7?! Re6! (but not 24...f6 25.Qg4! g6 26.Qh3 h5 27.Rxd8 Qxd8 28.Nxg6 as pointed by Deviatkin; 24...Qe6 is strongly met with 25.Ra7) with unclear play: 25.Qc2 (or: 25.Ra7 Bxb5! 26.Qxb5 Bb6; 25.Qf3 Bxb5!? 26.axb5 Bb6 27.Rxd5 Nf6; 25...Bb6!?) 25...Bb6 (Black may win the exchange for two pawns by 25...Bxb5 26.axb5 Bb6 27.Rxd5 Nf6 28.Rc5 Bxc5 29.Qxc5; 25...Rxe5?! 26.Rxd8 is hardly good) 26.Rxd5 Qa8! (Shipov) 27.Qxe4 f5. Again White may opt for the modest 24.Rc2!?.
24...f6? A blunder. Topalov saw that Black would have had some problems after the natural 24...Bxb5 25.axb5 Qxb5 (25...Rxb5? 26.Nd7!+) 26.Qxb5 Rxb5. With material balance and symmetrical pawn structure White's initiative is unquestionable. Anyway I can't call Black's position lost. Let's consider two important continuations:
а) 27.Nd7 Re8 28.Re1 (in case of 28.Ra8, 28...Rxb2! should be sufficient  29.Re1 Rb7 30.Nc5 Rbe7 31.f3 Nd6!=) 28...g5! (weaker is 28...f5?! 29.f3 Bf6! 30.Nxf6+ Nxf6 31.Rc1 Rxb2 32.Be5! with good winning chances) 29.Bc1, and as far as I can see Black is able to hold by 29...g4! 30.Ra8 Kg7 31.h3 h5;
b) 27.Ra2
b1) After 27...g5!? 28.Be3 the quiet moves like 28...h6 or 28...Kg7 don't solve the problems as sooner or later such moves will all be used up. Searching for salvation Black should resort to the drastic measures: 28...f5!? with a possible continuation 29.Ra8 Bc7!? 30.Rxf8+ Kxf8 31.f3 Rb3!?;
b2) 27...Bb6 28.Rb7 Nd6 29.Rxb6 Rxb6 30.Nd7 Ra6 31.Nxf8 Kxf8 32.Bxd6+ Rxd6 33.Rxa5 Rb6 34.Rxd5 Rxb2 leads straight to the Mark Dvoretzky's endgame files. Black is a pawn down in a tangled rook ending;
b3) 27...f6!? 28.Nc6 Bb6! 29.Rb7 Rf7! 30.Rb8+ Rf8 31.Ne7+!? Kf7. Here White can win the exchange in two different ways but the draw is a more probable outcome than his win: 32.Nc8 (or 32.Rb7 Ke6 33.Nc6 Rc8 34.Na7 Bxa7 35.Rxb5 Bxd4 36.Raxa5) 32...Rxc8 33.Rxc8 Bxd4 34.Rc2 Nc5!?
The ideas 24...Re6?! 25.Rxa6!? Rxa6 26.Nc7! Bxc7 27.Qxa6 g5 28.Be3!? Bxe5 29.dxe5 Qxe5 30.Rd1! Qxb2 31.Rxd5 Qa1+ 32.Kg2 Qxa4 33.Rxa5+/ and even more so 24...Nf6?! 25.Rxa6 Rxa6 26.Nc7 Bxc7 27.Qxa6 Nh5 28.Bd2 Bxe5 29.dxe5 Qxe5 30.Bc3+/ should be considered among the unlucky ones.
Topalov could have delayed the resolution (and it was probably what he wanted to do) by 24...g5!?, but after 25.Bd2!? (or 25.Be3!? Bxb5 26.axb5 Qxb5 27.Qxb5 Rxb5, which leads to the position we have already considered; 25.Nd7?! gxf4! isn't trustworthy) Black all the same can't play for a win: 25...Re6?! (he would better try to make a draw after 25...Bxb5!? 26.axb5 Qxb5 27.Qxb5 Rxb5 28.Bxa5) 26.Be3!? f6 27.Nd3! Bxb5 28.axb5 with White's advantage after 28...Bb6 (28...Qxb5 29.R1xa5!! Bxa5 30.Qh5+) 29.Ra6! Qxb5 30.R1xa5 Bxa5 31.Rxe6. The rook can't be caught: 31...Kf7?? 32.Ne5+.
25.Nd7!
What had been prepared in reply to 25.Qg4?! may never be revealed.
25...Rf7 26.Nxb6 Rxa7.
27.Nxd5!+. Perhaps this capture was overlooked by Topalov. In another lines Black is fine but now he's simply two pawns down. Just like his opponent in the previous game, Kramnik converted his advantage with an intentional deliberation.
27...Rd7 28.Ndc3. But not 28.f3?! Rxd5 29.Qxe4 Qd7 30.Nd6 Rxd6 31.Bxd6 due to 31...Bb7!  Yemelin (the point was 31...Qxd6? 32.Qe8+ Qf8 33.Qe6+ и 34.Qxa6).
28...Rxd4?! Trying desperately to lose a piece. Better was 28...Re7 with extremely illusive chances for a positive result.
29.Re1. Kramnik refuses to take Topalov's piece but it doesn't have an influence upon the outcome (in contrast to their game in Sofia 2005).. 29.f3 Bb6 30.Kg2!+.
29...f5 30.Qc2 Rb4 31.Nd5?! Dragging out though not spoiling anything. Again f2f3 won on the spot.
31...Rxb5 32.axb5 Qxb5 33.Nc7 Qc4. Black is completely lost after 33...Bxc7 34.Qxc7! Qxb2 35.Qd8+ Kf7 36.Qd7+ Kg6 (or 36...Kf6 37.Be3!) 37.Qe8+! Kf6 38.Rxe4 fxe4 39.Qc6+.
34.Qd1!? 34.Qxc4+ Bxc4 35.Ra1+, and in case of 35...Bf6 (35...g5 36.Rxa5! gxf4 37.Ra8) 36.Rxa5 Bxb2 Black loses one of his kingside pawns and all his hopes with it: 37.Rxf5.
34...Bxc7 35.Qd7! h6 (35...Bxf4 36.Qe8#) 36.Qxc7 Qb4. It is assumed that in a hopeless position one should try to preserve the queens: 36...Qd4!
37.Qb8+ Qxb8 38.Bxb8 Nd2. The agony would have lasted a bit longer after 38...a4.
39.Ra1. Now the a5 pawn is lost: 39...Nc4 40.b4!.
39...g5 40.f4! After all Vladimir decided to touch the f2 pawn (it began to cause damage to the king's safety).
40...Nb3 41.Ra3 Bc4 42.Bc7 g4 43.Bxa5. Black resigned.