2024 Chess Olympiad, Part 2
Passing the ... no, ripping the torch from the still warm hands of the older generation.
I don’t blame you if you’ve already consumed all the Olympiad content you can handle. The last round wrapped up four five days ago, but the artisanal annotations here at Lit & Chess take a while to produce. The mainstream (chess) media has covered the final standings and many of the highlights. In fact, the only quibble I have with chess.com’s Olympiad wrap-up article is its lack of capital letters. Writing that India “finished a full four points ahead of five teams” doesn’t communicate their achievement as clearly as: THE FOUR POINTS THAT SEPARATED INDIA FROM SECOND PLACE WAS GREATER THAN THE DIFFERENCE BETWEEN SECOND AND FORTIETH PLACE IN THE STANDINGS!!!
Or, alternatively: GUKESH FINISHED WITH A PERFORMANCE RATING WELL OVER 3000, THE SECOND GREATEST TOURNAMENT PERFORMANCE OF THE MODERN ERA, JUST BELOW CARUANA’S PERFORMANCE AT THE 2014 SINQUEFIELD CUP!!!!
The Indian squad were, in many people’s minds, the favorites. They often made it look easy, but this doesn’t mean we should be any less flabbergasted by their performance. As a team, they lost only one game out of 44. Erigaisi, India’s Board Three, chased Gukesh for the top performance rating all tournament long, finishing with 10/11. They never seemed to be in any meaningful trouble at any point of their eleven matches. And they did this at a time when, despite the continued ban of Russia, global chess has never been more competitive. Basically, they treated the event like they were the old Soviet Union and the rest of the world … well, it’s hard to compete with the best1.
I’d like to focus on one game in particular: Gukesh vs. Caruana, Round Ten. This matchup was symbolically significant. Fabi has been second in the world for most of the last decade and just missed a chance to force a tiebreak against Gukesh for the right to play in November’s World Championship match. Any game between them feels a bit like that lost tiebreaker. At the time, I figured that Fabi would have beaten Gukesh handily, but now I’m not so sure.
In terms of the tournament, Gukesh’s victory basically assured India of first place. And the game itself was played at an extremely high level, with both players fighting hard to win for their respective teams.
Gukesh Dommaraju (IND) - Fabiano Caruana (USA), Round 10
1 d4 Nf6 2 c4 e6 3 g3 d5 4 Bg2 Nc6 5 Nf3 Rb8:
Let’s pause here to take stock. Is this weird? Yes it is: welcome to modern chess, where it seems as long as you play e5 or d5 you can get away with all kinds of strange ideas. I’ve managed to take 43 trips around the sun without having to learn anything about the Catalan, but I think black’s idea is to get white to transpose into the 1 d4 Nf6 2 c4 e6 3 g3 d5 4 Bg2 Nc6 5 Nf3 dxc4 6 00 Rb8 variation without having to deal with 6 Qa4. If white doesn’t want to allow this, he has to avoid 6 00 and instead resolve the issue of the c4 pawn: 6 cxd5 Bb4+ 7 Nc3 exd5 8 Bg5 h6 9 Bxf6 Qxf6 10 00 Bxc3 11 bxc3 00:
This kind of early middlegame is pretty typical for top-level chess. White’s advantage consists of his position being somewhat easier to play. His minor pieces are better and he going to dictate the play for the next chunk of the game with a pawn break in the center, either c3-c4 or e2-e4. Black isn’t in any serious trouble, but he could be if he’s not careful: for example, 12 Rc1 Na5?! 13 Ne5 is a problem. Gukesh played 12 Nd2, potentially playing for either break, but allowing Caruana to reroute his knight, something black often needs to accomplish in these queen pawn openings with Nc6. After 12 … Ne7 13 e3 Be6 14 Qa4 a6 Gukesh played 15 e4, challenging black in the center:
Wait a minute, something strange just happened. On move 13, Gukesh played e2-e3 when he could have moved immediately to e4. Just two moves later, he played e3-e4. Why did he throw away a tempo?
The positional choices that are being made in this part of the game are quite important. Generally, he would prefer to play c3-c4. After something like 13 e3 Bf5 14 Qb3 Qd6 15 c4, white’s position is rock solid and he will have targets on d5 and the queenside while black will have little to do. Alternatively, after e2-e4 the c3-d4 pawn structure could prove wobbly down the road. This explains Caruana’s decision to play Be6: he’s trying to shut down c4. So Gukesh returned to the e4 plan, but not before setting a tiny trap.
Caruana fell right it. He played 15 … c5?!, which is exactly the kind of move black can play if he can get away with it, far more assertive than sitting passively with something like 15 … c6. But it turns out 15 … dxe4 would have been better, given Gukesh’s response: 16 Qa3! The point is that black cannot play 16 … cxd4? because of the zwischenzug 17 e5! Qg5 18 Nf3, deflecting the black queen from its defense of the Ne72. This is awkward. Either rook to c8 is met by 17 dxc5, since the white queen now protects c3. So Caruana played the only reasonable move left, 16 … b6, and after 17 Rab1 Rfe8, Gukesh grabbed the pawn anyway, 18 dxc5:
Caruana has to be careful that he doesn’t end up seriously worse. I spent some time working on his most obvious move, 18 … bxc5. One important question is whether or not black can get enough for his pawn after 19 Rxb8 Rxb8 20 Qxc5? I spent some time on the forcing variation 20 … Rc8 21 Qb4 Qxc3 (21 … Rxc3? 22 e5! is that deflection idea again) 22 Qxe7 Qxd2 23 exd5 Bxd5? 24 Qd7 Rc5 25 Qd8+ Kh7 26 Be4+! g6 27 Bxg6+!:
The point is that either capture drops the exchange: 27 … hxg6 28 Qe7+ or 27 … Kxg6 28 Qd6+. I then wrote “long variation wrong variation” in my notes and let the computer inform tell me the reasons why: that 24 Qd6! prevents Rc5 and wins a piece immediately, that 21 … Nc6! allows black to continue playing down a pawn, and that 21 Qd6 was better than 21 Qb4 in the first place. But perhaps most important is the fact that black can play 20 … Rb2! instead of 20 … Rc8, gaining counterplay on the seventh rank and likely winning back his pawn shortly.
Instead, we got something very different: 18 … d4 19 c4:
More than any other moment, this move pair signals that both players are trying very hard to win. Fabi has to: if the US doesn’t beat India, the battle for first place is over. Gukesh, on the other hand, could have played 19 cxd4 Qxd4 20 Nb3 Bxb3 21 Rxb3 instead, with a small edge. This might have been the pragmatic choice, given India’s tournament situation and the danger that he exposed himself to in the game, but it would have seriously reduced his winning chances.
Now much of the play is forced, at least from black’s point of view: 19 … Qg5 20 Rfd1 (apparently black is OK after 20 f4 Qxc5 21 Qxa6 d3+ 22 Kh1 Nc6; I’ll just say that it’s quite messy) 20 … bxc5 (trying to maximize his chances by keeping the queens on rather than minimizing his losing chances with 20 … Qxc5) 21 f4 Qh5 22 f5:
Black can’t retreat the Be6 because he can’t afford to lose c5 for nothing. Fortunately, white’s soft underbelly is showing, ready for invasion by the black queen: 22 … Rxb1 23 Rxb1 Qe2 24 fxe6 Qxd2 25 exf7+ Kxf7 26 Qxc5 Kg8 (black doesn’t have many other good options: for example, throwing in 26 … Qe3+ 27 Kh1 Kg8 drops the all important d-pawn to 28 Rd1) 27 Rf1 Kh7:
These last two defensive king moves are incredibly impressive. Black first finds safety on g8 and then uses his king to support a future activation of his knight on g6 (the immediate 27 … Ng6 runs into 28 Qc6). The cost is that f7 is weak: what if white goes for Qe5 and Rf7?
Before we get there, Gukesh had another fascinating opportunity from the diagrammed position. What follows is a computer line; I take no responsibility for the madness: 28 Qd6!? Ng6 29 c5 d3 30 c6 Qe3+ 31 Kh1 d2 32 c7 Ne5! 33 Qxe5! Rxe5 34 c8=Q Rc5! 35 Qd7 Rc1 and white has to force a draw with 36 Qf5+:
From the previous diagram, Gukesh pinned the knight and took aim at g7 with 28 Qe5, and Caruana kept finding only moves: 29 … Qe3+ 29 Kh1 d3:
The d-pawn is here to save the day. 30 Rf7 is met by 30 … Qe1+ 31 Bf1? Rg8! 32 Qxe7 d2 and black wins. Gukesh decided to bring his long-slumbering bishop into the game, and now it’s white who has to walk the tightrope: 30 Bh3 d2 31 Bg4! This stops black’s tricks, as there’s a second defender of d1. The variations 31 Qf4 Qe1! and 31 Bd7 Rf8! show that the initiative could pass to black at any moment.
We’re getting close to the end; it’s no surprise that both players were in time-trouble. The players continued 31 … Qd3 32 Kg2 Qxc4 (finally regaining the pawn lost on move 18) 33 a4 Qc6 (33 … Qxa4 looks impossible due to 34 Rf7, but 34 … Rg8! 35 Qxe7 d1=Q regains the lost piece; this suggests that 33 a3 would have been a little better) 34 Qd4:
The final critical position. After neutralizing Gukesh’s pressure for so long, Fabi blundered with 34 … Qg6? The queen is poorly placed here, given the prospect of Bf5 in some positions, and white can quickly round up the d2-pawn. Evidently Ng6, Ng8, or Kh8 would have all kept the game going: it’s time to start using some of the other pieces. Gukesh played like a machine to finish off America’s top board: 35 Rf4! (securing the e-pawn) Qg5 36 h4 (not falling for 36 Qxd2? Ng6 37 Bf5 Kh8 38 Bxg6 Qxg6 when white’s winning chances evaporate) 36 … Qg6 37 Qxd2 Rb8 38 Qd7 (there’s no good way to defend the Ne7) 38 … h5 39 Be6 Rb6 40 Bf7 Qd6 41 Qxd6 Rxd6 42 Bxh5 (the second pawn falls) 42 … Nc6 43 Be2 a5 44 Bb5 Ne5 45 Rf5 Re6 46 Kf2 Black Resigns.
It’s a testament to the strength of both players that they could produce such a thrilling and high-quality game at the most critical moment of the Olympiad. Gukesh didn’t have to prove that he’s one of the best players in team play; after all, he won the gold medal on board one in 2022 as well3. But this year’s gold was sweeter thanks to his team’s success, and his 9/10 score made an important case for his legitimacy in November’s World Championship4.
Congrats to India well-deserved victories in the open and women’s sections; they have the chance to dominate this event for years to come. Hopefully in 2026 we’ll see one or more other teams give them a real challenge.
Speaking of which, I noticed that Vishy Anand is playing what amounts to a correspondence game against the world soon — a sequel to Kasparov’s epic struggle against the world in 1999. If we’re bringing that back, why not reach even farther into the past (1970, to be exact) and hold an India vs. the Rest of the World Match sometime in the next decade?
So much in chess depends upon these short tactical operations. My interest in this game was piqued when Robert Hess pointed out e5 on the broadcast; I’m not sure that I would have noticed it myself.
A fact that I’d forgotten until the internet reminded me of it. Also crazy: he was playing for the Indian B team in 2022, which placed ahead of India’s A team and won the bronze medal.
You could, of course, make the argument that winning the Candidate’s tournament gives him all the legitimacy he needs, but barreling through the top players in the world and finishing with a better performance against stronger opposition than Magnus Carlsen is a pretty powerful statement.