1.e4 At the press conference held before the start of the match, Topalov stated that his duel with Nisipeanu was important for him because he had practically no experience as a match player. Although a 4-game confrontation is rather short and does not allow one to display his whole arsenal, we can distinguish some basic features of Topalov's match strategy. First of all, he made use of the width of his repertoire by constantly changing the openings, thus giving the match a multi-facet character. In doing so, he seems to have aimed to increase the dynamic of play day by day. In the first two games he chose basically solid openings (1...e5 and 1.d4) while in the second half he adopted the much sharper 1...c5 and 1.e4. This could have been caused either by the necessity of getting a more clear image of Nisipeanu, whom he described as "probably the strongest player I have not played yet" (obviously, the game played between them 15 years ago is hardly relevant) or simply by a good knowledge about the general evolution of his own sportive form along competitions.
1...c5 Although Nisipeanu made his first noticeable steps in the national chess as a Sicilian player, he emerged as a first class international fighter with more solid openings such as the Caro Kann and later, the Philidor. By returning to his "first love" specially for this match, the European Champion obviously intended to surprise his mighty opponent, but we shall see soon that he will be hit by the effect of the boomerang.
2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 Nc6 In youth competitions, Nisipeanu usually played the Najdorf or the Scheweningen. In recent years, he has occasionally employed this move order, but mainly with the rather unusual idea of transposing after White's most aggressive continuations 6.Bc4 and 6.Bg5 to a... Dragon.
6.Bg5 [Here is an example from last year: 6.Bc4 Bd7 7.Bb3 g6 8.Nxc6 Bxc6 9.Nd5 Bg7 10.0-0 0-0 11.Nxf6+ Bxf6 (This whole series of exchanges have only helped Black complete his development in very comfortable way.) 12.Re1 a5 13.a4 e6 14.c3 Be5 15.Bc4 Qc7 16.g3 Kg7 17.Bb5 Rae8 18.Be3 f5 with some initiative for Black, Hamdouchi-Nisipeanu, Castelldefels 2005.]
6...e6 [Only this was meant to be the real surprise for Topalov. This position occurred in Nisipeanu's games as Black only once, 8 years ago and in the company of a lower-rated player. The European Champion's recent adoption of the four knights Sicilian was mainly based on the discovery that the move doomed by theory 6...g6 is in fact entirely playable, as can be seen from the following game: 7.Bb5 Bd7 8.Bxf6 exf6 9.Nde2 Be6 10.Nf4 Bg7 11.Qd3 0-0 12.Nxe6 fxe6 13.Rd1 f5 14.Bxc6 bxc6 15.Qxd6 Qb6 16.0-0 Qxb2 and White faced difficult problems because of the lack of stability of his knight, Nataf-Nisipeanu, Germany 2004.]
7.Qd2 Be7 8.0-0-0 0-0 9.f4 Nxd4 10.Qxd4 Qa5 11.h4!? This very rare move was played for the first time by none other than... Nisipeanu himself, in a youth championship 11 years ago! I doubt that Topalov chose it for this game as a psychological weapon, aimed to place his opponent in the uncomfortable position of playing "against himself", since he could hardly have expected the Rauzer Attack for this game. I believe it is more of a coincidence that the World Champion had chosen this line when preparing against the Rauzer time ago. The obvious aim of the advance of the h-pawn is to inhibit the thematic ...h6, which would create no real threat, since ...hxg5 would always lead to the dangerous opening of the h-file. Besides, the bishop is over-defended in view of the desirable break in the centre based on e5.
11...e5 [This counter-blow in the centre looks like the most logical answer. Nisipeanu's game continued with 11...Rd8 12.Qe3 h6 13.Be2 Bd7 and now, instead of rushing in with 14.e5 as in Nisipeanu-Tyomkin, Holon 1995, White should have first played 14.Qg3 threatening Bxh6 and forcing the enemy king to occupy an unfavourable square with 14...Kf8 and only now play 15.e5 , which would discard the capture on g5 followed by ...Nh7 as happened in the game.]
12.Qe3 exf4 [During his first matches against Karpov, where the Sicilian was frequently played, Kasparov generally refrained from this capture for as long as possible (although in some cases he overstepped the limit of admissible risk). This is probably the correct approach, but here the normal developing move 12...Be6 leads to rather unclear consequences after 13.f5 Bxa2 . It is understandable that, confronted with such an opening surprise Nisipeanu refrained from embarking in a theoretical dispute in such kind of position. Besides, after the exchange on f4 it is not entirely clear which could be White's benefit from having spent a tempo and weakened the g4-square by the advance of his h-pawn.]
13.Bxf4 Be6 [This pawn sacrifice is more or less forced, because in case of the over-cautious 13...Rd8 White would obtain a wonderful development with 14.Bc4 ]
14.Bxd6 Bxd6 15.Rxd6 Rac8 [Apart from Topalov's rapid rhythm of moving, the first technical sign that his opening play had the coherence of a home-prepared analysis consists of the fact that after 15...Bxa2 16.Rxf6! gxf6 the move h4 proves of essential usefulness by enabling the activation of White's remaining rook with 17.Rh3! when Black cannot parry the simple mating threats and keep his bishop defended at the same time. With his last move, Black puts some pressure on White's queen side, specifically, on the c3- and a2-squares.]
16.a3 Optically speaking, Black has ample compensation for the pawn. He has the comfortable e5-square at his disposal for his knight, while White's king side structure is weak. The queen side pressure looks threatening, especially that White is slightly behind in development. However, things are by far less clear if examined more concretely. First of all, White's position is much better from strategic point of view than it might look at first sight. His outpost on d5 is of higher significance than Black's e5-square, because it is situated on an open file. In case of a capture on d5, White would obtain a strong passed pawn. Although he still needs two whole tempi (Be2 and Rhd1) to obtain an optimal development, all other white pieces are placed on active squares. Even the weakness of the king side pawns is relative. In certain cases, he can simply start an attack against the enemy king by means of g4-g5, h5 and g6, although for the moment it is a bit early to think about that.
16...Rc6!? [This completely unexpected move was played after a long thought. In the press centre, we considered 16...Rc5 to be the best way of developing the initiative. Initially, our debates took such a course: 17.Kb1?! (Removing the king from the exposed c1-square in order to threaten the fork with b4) 17...Rfc8! 18.b4 Qxa3 19.Qxc5! (Forced, since 19.bxc5? would lose to 19...Ba2+! 20.Ka1 Nxe4! as found by IM Andrei Murariu) 19...h6 (For reasons that will become clear later we first tried to secure the back rank with 19...g6 but then IM Nemeth discovered that 20.Qd4 Rxc3 21.Rxe6! fxe6 22.Qxf6 would win for White.) 20.Qxc8+ (Now, 20.Qd4 Rxc3 21.Rxe6 fxe6 is inoffensive for Black.) 20...Bxc8 21.Rd8+ Kh7 22.Rxc8 Qxb4+ 23.Kc1 Nxe4 (With the king on g7 in the similar line starting with 19...g6 instead of 19...h6, this would lead to a draw by perpetual after 24.Nxe4 Qe1+, but here White has an additional possibility). 24.Bd3! f5 and in spite of the small material deficit, Black can hope to survive because of the safer position of his king.; However, after 16...Rc5 White has a more natural and in fact stronger possibility. Instead of the prophylactic king move, planning the weakening of his own position, he should simply develop with 17.Be2! Rfc8 18.Rhd1 when the planned 18...Rxc3 is refuted by 19.Qxc3! based on the weakness of the back rank, while 18...h6 allows simplifications with 19.Rd8+! reducing Black's pressure along the c-file.; The immediate sacrifice 16...Rxc3 is also not entirely satisfactory for Black after 17.Qxc3 (Only not 17.bxc3? because of 17...Qxa3+ winning the d6-rook.) 17...Qxc3 18.bxc3 Nxe4 19.Rd4 (The only safe square for the rook, avoiding the possible fork on f2) 19...Nxc3 20.g4 followed by Rh3 and Kb2. If Black could transfer his knight to c5 and centralise his king to e7 he would be out of the danger of losing, but the knight does not seem to be on the right path. Maybe the immediate 19...Nc5 followed by ...Rc8 and ...Kf8-e7 would be a better practical chance, but in that case Black would not even have a pawn for the sacrificed exchange.]
17.Rxc6 bxc6 We can evaluate now the consequences of Black's previous move. Obviously, by exchanging one pair of rooks he has given up any attacking ambitions, but has obtained certain strategic compensation. First of all, he has deprived the enemy knight from the important d5-square, while maintaining the stable control of the e5-square. This should allow Black to resist for a long time in a passive but rather solid position, inviting White to take some risks in order to make further progress.
18.Be2 Nd7 19.Rd1 Qc7 20.g3!? The first move on which Topalov spent considerable amount of time provokes a big surprise in the press centre. We were mainly worried by the massive advance of the king side pawns in view of a direct mating attack. The World Champion might have considered this to be a rather double-edged issue and preferred to consolidate his king side instead, preventing among others the enemy queen's incursion to h2.
20...Rb8 21.Qd4!? [This was definitely the most shocking move of the entire match. After I convinced myself that there was no mistake in the electronic transmission of moves and that this was the move effectively played by Topalov, I unwillingly remembered a recent comment by Kasparov about Karpov's style of play. The 13th World Champion stated that his predecessor had a considerable practical strength and a deep chess understanding, but it was practically impossible to learn from his games. personally, I have some doubts about that, but I feel that this comment would apply perfectly to Topalov. For instance, how could I explain to my 7-years old son the fact that right after consolidating his king side White completely abandons it? The teaching process would have been much simpler if White had played 21.Rd4 slowly increasing the pressure and aiming to gradually suffocate Black. In fact, I have the feeling that I would have attached to 21.Qd4 an "?!" if it had been played by anybody else, but since Topalov seems to handle such abrupt changes of plans rather well, being able to control the ensuing complications, I cannot really criticise his decision. I would rather confess my limited understanding of chess.]
21...Qxg3 With not so much time left on his clock, Nisipeanu had no reasons to refrain from this capture.
22.Qxa7 Nf6 The situation has become quite sharp. I suppose that Topalov was seduced by the idea of creating an outside passed pawn and by the temporarily hanging position of the enemy rook. However, the former aspect will be compensated by Black's similar achievement on the king side, while the latter only has temporary character.
23.Kb1 [Actually, White could have tried to justify his concept with 23.Qb6 , aiming to eliminate the main defender of the d5-square, the last remaining black queen side pawn. For instance 23...Nd5 (Or if 23...Bd5 then 24.Qd4 ) 24.Qxc6 and it is not easy to find a way for Black to take advantage of the tensioned situation.]
23...h5 Securing the back rank and preparing a dangerous candidate for promotion.
24.a4 [Since the fight will soon consist of a race between the passed pawns on opposite wings, White hurries to advance his a-pawn. On the other hand, his last move has the significant drawback of weakening the b4-square, a detail Black will fail to take advantage of. In a mor peaceflu position, such a move like 24.Qc5 , preventing the centralisation of the black queen would deserve serious attention, but here it would lead to unclear consequences after 24...Qxh4 (possible because the b8-rook is not attacked any more) 25.Qxc6 Qg3 followed by the advance of the h- and (in case of the exchange of queens) of the g-pawn. In such cases, the material balance is completely irrelevant, the decisive factor being the speed of the advancing pawns.]
24...Bg4? [This is a significant mistake already. By exchanging his bishop, Black will lose stability in the centre, leaving the e6-, d5- and f7-squares insufficiently defended. The best practical chance consisted of the centralisation of the quen with 24...Qe5 , creating the threat ...Rb4, when the c3-knight would have started feeling insecure. 25.Qd4 could be answered with 25...Qa5 followed by the transfer of the knight to e5 or, eventually, ...Kh7 and ...Rb4. In this case, White would have faced serious problems breaking Black's blockade on dark squares, while the weakness of the b4-square would have made itself felt. The attempt to take advantage of the similar weakness of the g5-square with 26.Rg1 is easily parried with 26...Ng4 . We can notice here the qualitative difference between the e4- and c6-pawns from the point of view of the key-squares.]
25.Bxg4 hxg4 26.Qd4! By returning with the queen to the centre, White clears the way to his a-pawn and questions Black's stability in the centre and on the queen side in view of the threat e5-e6.
26...Qxh4?! [Simplifying White's task. Black should have taken the e-file under control with 26...Re8 ]
27.e5 Now, there is little hope left for Black.
27...Nd5 28.Nxd5 cxd5 29.Rg1 [Topalov decides to eliminate the only source of black counterplay, the g4-pawn. In case of 29.Qxd5 Black could have put up some resistance with 29...Re8 , preventing e6, although it is likely that the connected white queen side pawns would have decided the game to White's favour.]
29...Qh2 30.Rxg4 Rc8 Apparently, Black maintains some counterplay based on the threats ...Qxc2+ and ...Rc4. However, Topalov had foreseen a tactical remedy to this problem.
31.Rxg7+! The decisive blow. White eliminates another candidate to promotion and opens the position of the enemy king.
31...Kh8?! [Still hoping to trick White. If Black had understood in this moment that the exchange of rooks is inevitable anyway, he would have probably played 31...Kxg7 32.Qg4+ Kh7 33.Qxc8 Qxe5 with better practical chances to muddy the waters by advancing the f-pawn than in the game. Since the immediate 34.a5 would lose the pawn to 34...Qe1+ , White would have had to lose an additional tempo to take the a5-square under control.]
32.Rg4! [Topalov took his time to calculate the consequences of this accurate move. The tempting 32.e6 would have led to complete equality after 32...Qxc2+ 33.Ka2 Rc4 34.Rg2+ Rxd4 35.Rxc2 Rxa4+ 36.Kb3 Ra7 ]
32...Rc4 [White's main trick consists of the fact that after 32...Qxc2+ 33.Ka2 Rc4 34.Rh4+ Kg8 he disposes over the unexpected but decisive queen retreat with check 35.Qg1+ when the only way to avoid mate would be 35...Qg6 when White could transpose to an easily winning pawn ending.]
33.Qd1 Rxg4 34.Qxg4 Qxe5 Compared to the queen ending that could have been obtained by the immediate capture on g7, Black is in a worse situation here, because of the unfortunate position of his king. Concretely, White wins an essential tempo with
35.c3! , threatening Qd4 and covering the e1-a5 diagonal at the same time, thus ensuring the further advance of the a-pawn.
35...Kh7 36.Qd4 Qe6 37.a5 f5 Black not only is a pawn down, but his only candidate to promotion is less advanced than White's.
38.b4 [White maintains the centralised position of the queen, keeping the f-pawn under observation and threatens to simply advance his connected pawns. Another simple win was suggested by IM Valentin Stoica: 38.Qa7+ Kg8 (The sixth rank is denied to the black king because of Qb6, exchanging queens) 39.Qb8+ Kg7 40.Qb7+ Kg8 41.a6 when the further advance of the a-pawn is ensured, while Black has no perpetual check, for instance 41...Qe4+ 42.Kc1 Qe1+ 43.Kc2 Qe2+ 44.Kb3 Qc4+ 45.Ka3 Qc5+ 46.Ka4 Qc4+ 47.b4 with an easy win.]
38...Qe2 39.Qxd5 By this moment, I unwillingly remembered the final part of the game Topalov-Anand from San Luis 2005. In that game, White had three connected king side pawns against Black's far advanced b-pawn and eventually missed a win. However, in our game the situation is completely hopeless for Black, because his f-pawn does not present any danger.
39...Kg6 40.Qd6+ Kg5 41.a6 f4 42.Qc5+ Kg4 43.a7 [After 43.a7 Qe4+ 44.Kb2 the threat Qc8+ followed by a8=Q would force Black place his queen on a passive square with 44...Qb7 when after 45.Qa5 Qa8 46.Qc7 followed by Qc8 the f-pawn would be two tempi too slow to generate adequate counterplay. Therefore, Black resigned.] 1-0