Fabiano Caruana – Alexander Donchenko

Saint Louis Masters op Saint Louis (9), 07.12.2024

Alexander Donchenko comments. The game is also included in the “Lucky bag” of this issue, you can replay it here

Before the final round, Caruana led the tournament by one point ahead of six players. An undivided victory was more or less a guarantee of reaching the Candidates Tournament via the FIDE Circuit. I was lucky and unlucky to play against him out of the six chasers. On the one hand, he is of course the strongest possible opponent, but on the other, there is much more at stake for him, but he still doesn't have to win. Even if he knows how to handle it as a very experienced top player, it's a psychological pressure on him and I had to see my advantages somewhere.

1.e4 c5! Of course 1...c5 is no better than many other first moves. However, it's not just a challenge to him, but also to myself. I don't necessarily want to win this game, but I want to give myself a chance to play for it. In almost any solid opening like the Berlin or Marshall Gambit, Caruana will be able to freely determine the risk and steer the game more easily into the type of position he wants.

2.Nf3 d6 3.Bb5+ A bit of a surprise, but I was prepared to be surprised. I hardly prepared anything due to lack of time and after the games so far in the tournament I was pretty sure that Caruana wouldn't play a particularly critical line, possibly to save preparation time in double rounds like I did.

3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 In the Najdorf main position White can of course play countless variations, but the risk of running into very special preparation is also higher.

3...Nd7 3...Bd7 4.Bxd7+ Qxd7 is the other main variation, which leads to more simplified positions. 5.c4 (5.0–0 Nc6 6.Re1 Nf6 7.c3 e6 8.d4 cxd4 9.cxd4 d5 10.e5 Ne4 both Caruana and I (and many other players) have played with White) 5...Nc6 6.Nc3 g6 7.d4 cxd4 8.Nxd4 Bg7 9.Be3 Nf6 10.f3 0–0 11.0–0 Black is very solid and White has a slight space advantage. It's hard to believe that the number two in the world can easily get out of his comfort zone in this standard position.

4.0–0 a6 5.Be2!?

A rare move that many strong players have tried in order to get a quiet, playable position without theory. I wasn't prepared for this particular variation.

If Caruana had 'only' wanted to secure his tournament win, he could easily do it from theory, for example 5.Bd3 Ngf6 6.c3 b5 7.a4 Bb7 8.axb5 axb5 9.Rxa8 Qxa8 10.Bxb5 Bxe4 11.d4 Bxb1 12.dxc5 Be4 13.Re1 e6 14.Rxe4 and the game will soon end in a draw. This is widely known and I already had a short draw against Patrick Zelbel in the Bundesliga from this position.

5...Ngf6 6.d3 That doesn't look very ambitious. At the same time, White claims that his structure is still very flexible, so that he can react to any black setup, while Black has at least already committed his knight to the suboptimal square d7.

6...g6 Not the most precise move. This is both easy and difficult in such positions. I have a lot of sensible moves, but hardly any possibility of reaching an end by calculating variations and finding a 'best' move.

Most of the top players played 6...e6! 7.a4 b6 and here you can see how White can aim for a completely different structure. 8.Nfd2!? Be7 9.f4 0–0 10.Nc3 Bb7 The structure is more reminiscent of closed Sicilian or even color-swapped Dutch. Black has no objective problems and both sides still have many interesting possibilities.

7.c4!

The best move, aiming for a Maroczy build-up, even with a loss of tempo through d3–d4. That would be a structure I want to avoid at all costs in this game.

7...Bg7 More flexible and somewhat more precise is 7...b6! 8.Nc3 Bb7 but I had respect for 9.d4?!

This is objectively not good, but you are instinctively more careful with sharp attempts when your opponent is still in their preparation. You can also play a somewhat passive position for two or three moves against a computer. But if d4 is on the board in a second, you have to calculate these next moves against a computer and that almost never ends well.(9.Rb1 Bg7 10.b4 Rb8 and the black position is very healthy.) 9...Bxe4 (9...Nxe4 10.Nxe4 Bxe4 11.Ng5 Bb7 12.Bf3 Qc8! 13.Re1 Bg7 14.dxc5 Nxc5 15.Qxd6 0–0 16.Rxe7 Bxf3 17.Nxf3 Ne6= is also a good variation for Black, but finding it would at least take a lot of time.) 10.Nxe4 (10.d5 Bxf3 11.Bxf3 Bg7 12.Bg5 0–0) 10...Nxe4 11.Re1 Bg7 12.Bf1 Nef6

8.Nc3 Now White is ready for d3–d4.

8...e5!? A very radical solution to the problem. I thought abstractly that although I have a somewhat passive knight on d7, my bishop on g7 has more potential than the g2 bishop.

8...0–0 9.d4 cxd4 10.Nxd4 would be White's dream. A Maroczy structure with all the pieces on the board is usually very unpleasant to play for Black.

9.Ne1?! Surprisingly, Caruana doesn't find the best plan. Maybe he simply underestimated my next move.

9.a3! Nf8 10.b4 b6 11.Rb1 Ne6 12.bxc5 bxc5 13.Qb3 leads to risk-free pressure for White, although of course the position still retains a lot of potential.

9...Nf8!

The knight immediately moves to its best square and 'saves' time without castling.

9...0–0 10.f4 exf4 11.Bxf4 Ne5 12.Qd2 The knight is not as good as on e6 in the game and White has a free hand to improve his position.

10.f4 10.b4!? b6 (10...cxb4?! 11.Qa4+ N6d7 12.Qxb4 Nc5 13.Be3) 11.bxc5 bxc5 12.Rb1 Ne6 13.Nf3 0–0 14.Qb3 was still possible without any major problems.

10...exf4 11.Bxf4 Ne6

The knight is perfectly positioned and also arrives with tempo.

12.Be3?! Often this distance between bishop and knight is good for the bishop because it restricts the knight, but here only its potentially better squares become inaccessible.

Better is 12.Bg3! when I planned 12...h5!? (12...0–0 13.Kh1 Rb8 14.a4 is the choice of computer.) 13.h3 Nd7 14.Kh1 Ne5 15.Nc2 White objectively has some advantage, but I think this type of position will always be double-edged and I'd be happy with that.

12...0–0 One of the rare moments when I take a risk for psychological reasons. I saw the pawn sacrifice with ...b5 very early, because I'd missed it in another game recently. And I intuitively had the feeling that Caruana didn't have it on his mind, so I put castling first to hopefully play it with more effect on the next move.

12...b5!? 13.cxb5 axb5 14.Nxb5 0–0 is objectively at least as good, but here White can play a better move than Dd2. 15.a4 Ba6 16.Rb1 Qd7 17.b4=

13.Qd2 13.a4! would have thwarted my plan, but even here my position is quite comfortable. 13...Bd7 14.Nc2 Bc6 15.Qd2 Nd7 16.Bh6

13...b5!

I felt I could play the structure more easily than Caruana and for the first time I thought I was actually fighting for an advantage. I also saw how his time consumption increased and he no longer radiated the usual absolute self-confidence.

14.Rd1 A 'neutral' move, but it was better to simplify the position immediately and defend by exchanging.

14.cxb5 axb5 15.Nxb5 d5 was my idea, but it's not the only move. (15...Rb8!? 16.a4 Ba6 17.Nc2 Bxb5 18.axb5 Rxb5 19.b4) 16.e5 (16.exd5? Nxd5 17.Bh6 Bxh6 18.Qxh6 Rb8 19.a4 g5!) 16...Nd7 17.d4 cxd4 18.Nxd4 Nxe5 19.Nxe6 Bxe6 20.Nf3; 14.Bh6 b4 15.Nd1 Bxh6 16.Qxh6 Nd4 17.Qd2 Bg4 18.Bf3! Be6 These two variations are somewhat passive for White, but my advantage remains small and it's a long way to real winning chances.

14...Rb8 14...Nd4! was stronger, but like Caruana I'm a bit undecided on how to change the structure. 15.Bxd4 cxd4 16.Nd5 Nxd5 17.cxd5 wasn't entirely clear to me over the board, but I have a pretty clear advantage here too. 17...Qe7 18.Nf3 Bg4; 14...b4 15.Nd5 Nxd5 16.cxd5 Nd4 17.Nf3 didn't appeal to me because White's play is too straightforward. He has almost no choice but to exchange off my active pieces.

15.b3?! I speculated a bit on this natural move, because now I can exploit the square c3.

15.h3! Bb7 (15...b4 16.Nd5 does little for me here.) 16.Bf3 bxc4 17.dxc4 Nd4 18.Bh6 and again Caruana would have most of the weaknesses under control and the game continues as normal.

15...b4! 16.Na4 This move looks like a sad necessity.

16.Nd5 Nxd5 17.cxd5 Bc3! 18.Qc1 Nd4 Now everything would come with tempo, which makes a big difference compared to 14...b4. 19.Rf2 Nxe2+ 20.Rxe2 Bg4 21.Nf3 f5!

16...Bd7?! It's very recklessly played and I'm lucky that the punishment is hard to see.

16...Nd4! was a good move that does not allow any counterplay. 17.Bxd4 cxd4 18.Nc2 Ng4 19.Bxg4 Bxg4; 16...h5!? is also very strong, but shows either incredible intuition or cheating. 17.Bh6 Bh8!! 18.Bxf8 Qxf8 Hardly any strong player would decide to do this without need, although in hindsight you can clearly see the advantages.

17.e5! I only realized this move while I was walking around the tournament hall after Bd7 and had to fight a lot against the panic that I was ruining what I could already feel was a very good game.

17.Nb2? Qa5 18.a4 Nd4–+ is strategically untenable for White.

17...Ne8!

That still works best, but here Caruana had the best and last chance to escape the strategic pressure.

17...dxe5 18.Nxc5 Nxc5 19.Bxc5 Re8 20.Bd6 Ng4! 21.Nc2 (21.Bxb8? Bh6!–+) 21...Rc8 and I have lost all control of the position.

18.exd6? Now all my ideas fit together perfectly. That was the moment to change the flow of the game.

The objectively and practically only move was the impressive 18.Bg4!! In a sense, White uses the e5–pawn to break through the dark-squared blockade by force.18...Bxa4 (18...Bxe5 19.Bxe6! Bxe6 20.Nf3 Bg7 21.d4; 18...f5 19.exf6 Nxf6 20.Bxe6+ Bxe6 21.Nf3=) 19.Bxe6! Bc6 (19...fxe6 20.Rxf8+ Kxf8 21.bxa4 Bxe5 22.Qe2) 20.Bg4 Bxe5 21.d4 cxd4 22.Bxd4 In all these variations I lose control of d4 and even a pawn win can't make up for that.

18...Bxa4 19.bxa4 Bc3 20.Qc1 Nxd6

I've exchanged off my worst piece, permanently ruined the white pawn structure and have more active pieces. In a higher sense, the position is won, but of course you still must not offer any counter chances.

21.Bg4 f5?! I thought this obligatory move was necessary and underestimated how good my position already was.

21...Nd4! held on to a clear winning position, but I saw ghosts after 22.Bh6 (22.Kh1 h5 23.Bh3 Re8–+) 22...Re8 23.Bg5 For the computer, however, there are no special threats here. 23...Qa5 24.Qf4 N6f5–+

22.Bf3 f4! My game is based on keeping the Bf3 on the board and letting it shoot into the void.

23.Bd2! 23.Bd5 fxe3 24.Bxe6+ Kg7–+

23...Bxd2

23...Nd4! 24.Bxf4 Qe7! is another computer solution, but it's difficult to estimate that a pawn sacrifice keeps the white pieces completely passive here. 25.Kh1 Kh8 26.Bg5 Qe6

24.Qxd2?! Very natural and expected by me, but it doesn't coordinate the pieces well enough for the defense.

24.Rxd2! Nd4 25.Bd5+ Kg7 26.Nc2 Qf6 27.Qa1! is the hidden idea here, when Caruana has serious chances of holding the game. (27.Nxd4 Qxd4+ 28.Rdf2 a5 would be very similar to the game and practically untenable.) 27...Nxc2 28.Qxf6+ Kxf6 29.Rxc2 g5 30.g3!

24...Nd4 25.Bd5+ The bishop is stably positioned in the centre and yet has no real function, a rare sight.

25.Nc2!? Qf6 26.Nxd4 Qxd4+ 27.Qf2 Nf5 28.Qxd4 Nxd4 would be objectively tougher, but an endgame without any perspective is always the last option.

25...Kg7 26.Nf3 Qf6

The difference to 24.Rxd2 doesn't look dramatic, but here I can improve my position very directly and exploit my advantage.

27.Kh1 Rbe8! 27...a5?! 28.Ng1 g5 29.g3 and there is no direct plan to break through with Black.

28.Nxd4 28.Rde1 a5 29.Ng1 no longer functions in the same way as on the previous move: 29...g5 30.g3 Rxe1 31.Qxe1 Nc2! 32.Qd2 Ne3–+

28...Qxd4 29.a5 The only attempt to gain some form of counterplay, but the options are now too limited to surprise me.

29.Rf3 Re3 30.Qf2 a5–+ will soon be impossible to hold, also because no potential exchange of heavy pieces will improve White's position.

29...g5

By now I understood that the position must be won, but I tried not to think about it so as not to get nervous. Fortunately, I have the position completely under control and can only trip myself up.

30.Qc2!? 30.Rde1 Nf5 31.Rxe8 Rxe8 32.Qd1 Qf6–+

30...Nf5 31.Qa4 One last somewhat tricky attempt. Objectively, many moves win, but where I have to calculate, a mistake can always happen.

31...Rf6!

Threatens mate in two moves and effectively ends the game.

31...Ne3 32.Qd7+ Kh8 33.Rde1 Qxd3 34.Rg1 is also very won, but Caruana would at least have hope here thanks to his active queen.

32.h3 32.Qxe8 Ng3+! 33.hxg3 Rh6+ 34.Qh5 Rxh5#

32...Ng3+ 33.Kh2 Nxf1+ 34.Rxf1 Re7

I won an exchange and kept many advantages of my position. The rest is no longer difficult.

35.Qd1 35.Be4 Qe5 36.Kh1 g4–+

35...Qe5 36.Kh1 36.Qf3 Qe2–+ swaps the queens.

36...Qe2 37.Qa1 37.Qb1 Re3–+

37...Qxd3 38.Bf3 Qxc4 Of course Caruana could have resigned already, but in a last round it's understandable that he continues to play. From personal experience, it's also easier to calm down while a game is still technically in progress.

39.Rc1 Qd4 40.Qb1 Rfe6 41.Rf1 41.Kh2 Qf2 42.Rf1 Re1–+

41...Qc3 42.Qf5 Re5 43.Qg4 Qd3 44.Rd1 Qf5 45.Qxf5 Rxf5 46.Rd6 c4 Caruana had finally seen enough and resigned the game. Without a doubt that's the biggest win in a single game I've ever played. Against a player of his calibre, it always takes a bit of luck to get such a good fighting position and Caruana probably missed the moment when he could have pulled the emergency brake. Nevertheless, I'm very proud that I didn't squander the advantage and kept my nerve throughout. I crowned an exhausting yet successful tournament with this win and shared the tournament victory with Caruana with 6.5/9 points each.

0–1