Norway Chess in 10 Diagrams
In which I try (and probably fail) to avoid being too cranky about chess commentary, endgame blunders, opening choices, and the tournament format.
The 2024 edition of Norway Chess is in the books, and much of what happened was more predictable than surprising. Magnus Carlsen remains the best in the world at all time controls and wins the tournament? Check. The Ding Liren we used to know is still missing in action? Check. The “three points for a win” gimmick fails to reduce the number of drawn games? Check and mate. Nonetheless, there was a lot of interesting chess, so let’s take a look back at one key moment from each round.1
Round One: Caruana-Nakamura
We’re just out of the opening, but Nakamura had already visited the confessional at least once, possible twice. He was using his prep from the Candidates Tournament, and was, by his own admission, a little bored. So perhaps he wasn’t completely focused when he played 17… exf3, a small inaccuracy that shifted the computer evaluation about a half pawn in Caruana’s favor. If I was annotating I might give this move a “?!” or take a short look at alternatives like 17… Ne7. Unfortunately, the evaluation system running on chess dot com knows nothing about subtlety, and so a dull yellow “?” appeared on the screen. This prompted the commentary team to overemphasize that a mistake had been made (the first of the tournament, oh my!) rather than accepting the move on its own terms and trying to figure out why Nakamura preferred it to the alternatives.
I’m not trying to bash Danny Rensch, Anna Rudolph, and David Howell. They’re doing an excellent job trying to balance their commentary for an audience that possesses a wide range of chess expertise. But we need to stop treating elite players as if they are part of some reverse Turing test, where the more they mimic the machine the more impressed we are with their play.2 Sometimes to beat another human you have to play some less than perfect moves or steer the position towards something so complex that computer-like accuracy is impossible. Making a big deal out of a half pawn swing also obscures the fact that the drawing margin in chess is quite wide, as it was in this game: Nakamura had no problem holding a draw.
Round 2: Firouzja-Caruana
This was most exciting game of the tournament to this point (and one of the most exciting overall). The opening was a weird London (more about that later), leading to a situation where both kings are exposed and under attack at the same time. What’s black to do? Retreating is no good: 28… Ne7? 29 Nd6+ Kg7 30 Qd7 is devastating. So Caruana correctly increased the pressure with 28… Rhg8!, eerily foreshadowing the queen sacrifice on the h-file that appeared in the game Carlsen-Ding. Had I been playing I might have been so concerned about Qxf3 or Rh6 that I would have missed the real threat, Qxh3+! But Firouzja was prepared and played the only saving idea, 29 Nd8+!, more or less immediately. One black piece is now deflected, either the knight after 29… Nxd8 30 Qd7+ or the rook, as in the game: 29… Rxd8 30 Qxc6 (now there’s time for this) 30… Qxf3 31 Qc7+ Ke8 32 Qxh7 and well-earned peace was declared.
Round 3: Praggnanandhaa-Carlsen
If you were writing a book about the notable absence of certain openings (such as the Kan Sicilian) from the repertoires of top GMs, you could put this position on the cover. Carlsen obviously wanted to mix things up against the youngest player in the field. It didn’t work out in this game, but it woke him up enough that he won his next three. This kind of strategy is unique to Carlsen: he can win games so consistently against other elite players that taking some risks and losing a game or two doesn’t really affect him in the way it would for any of the other top guys.
Anyway, back to the game. Pragg’s execution of the former champ was a model of strategy for these Maroczy Bind positions. He plays f5 to invite e5. Carlsen does not want give up d5, defending on e6 instead with the queen, so Pragg has to find a way to make the invitation compulsory. He does so by playing 15 Rad1 Rc8 16 Qe2. 17 e5 is coming next, ripping open the center, so black has to play 17… e5. The rest of the game was nice, but this shift of heavy pieces on the d-file is what I hope to remember in my own games.
Round 4: Firouzja-Ding
More than anyone else in the event, Ding seemed happy to adopt the openings played by his peers, running out the same line against the London that Fabi deployed against Firouzja in round two and, against Pragg, the same variation on the black side of the Giuoco Piano as Naka played against Fabi in round one. That’s fine if you’re comfortable with the line, but that’s not what happened here. In round two, Caruana played the accurate 12… c4, forcing the white queen back to d1. The point is that if white plays 13 Qa4?! instead, he runs into 13… b5! 14 Qxb5? Bc2! and the queen is trapped: a6 and Nd7 are coming. Ding played 12… Nh5?! instead and got a much worse version: 13 gxh5 c4 14 Qa4! Bxh4+ 15 Kd1! Bd3 16 Bxd3 cxd3 17 Qb5.
Objectively (at least according to the computer) Ding’s position on move 17 was no worse than Nakamura’s after 12… exf3 in round one. In my opinion, there are two reasons why this isn’t so: first, it’s much easier to go seriously wrong playing Ding’s position; the pressure on black’s kingside is oppressive. Second, Ding was already in pretty terrible form, which made it all the more essential that he get out of the opening unscathed. He didn’t here, the second in a four game losing streak, from a guy who once went 100 serious games with a loss.
Round 5: Carlsen-Firouzja
The middle rounds featured three theoretically drawn endgames that were lost at the eleventh hour. Fabi lost two of these, but each one was fairly high on the difficulty scale. Firouzja lost the other, from the position above. Black has to activate his rook to harass the pawns and king from behind, something along the lines of 77… Rb1. You’re giving up a second pawn here, and it’s not likely to be comfortable the rest of the way, but at least these split f- and h-pawn rook endings are theoretically drawn. Firouzja played 77… Rxd6? instead, a crazy kind of mistake to make at this level. The point is that after 78 Kxd6 Kf7 79 Ke5 Ke7 80 f6+ Kf8 we’ve reached a well-known3 corresponding squares position.
Black needs to be able to meet Ke6 with Ke8, so those squares correspond. He needs to be able to play Kf7 after Kf5, so those squares correspond as well. When the white king is on e5 it threatens to go to either f5 or e6, so the black king needs to be able to move to f8 in response, the only available square touching f7 and e8. The problem that white can do an extra little shuffle: 81 Kf4 Ke8 82 Ke4. Black is in zugzwang and resigned, as 82… Kf7 is met by 83 Kf5! and 82… Kf8 by 83 Ke5! Ke8 84 Ke6.
Round 6: Carlsen-Ding
The infamous “queen sac, mate in two” game that made it really, really clear that something was wrong with Ding and perhaps shook him up enough that he managed to draw the rest of his games to avoid adding to the unfolding disaster. Despite the menacing nature of white’s attack, the computer thinks that black’s actually doing fine, although he has to remain accurate and active: 28… Rb2! 29 Re4 h6 30 30 Bxd4!? Bxd4 31 Re8+ Rxe8 32 Qxe8+ Qf8! (32… Nf8? 33 Qf7 with mate to follow) 33 Qxd7 Be5! White has a pawn (for the moment) but black has taken the initiative, the most important feature of positions with opposite-colored bishops and heavy pieces remaining. What happened instead is memorable, but it goes without saying that the world champ shouldn’t fall for 28… Qf8 29 Re4 Rb2? (right idea, wrong order) 30 Qxh7+! and mate next move.
Round 7: Caruana-Firouzja
This was the best two game mini-match of the tournament. I don’t know exactly what inspired Caruana to sacrifice a piece, given his lukewarm follow-up, but the position was hard for Firouzja to defend. Here he’s got some options: 22… Bg4 as in the game, or 22… Ng4 23 Qe2 Rh7 or even 22… Bxh3 23 fxe5 Qxe3+ 24 Rxe3 Ng4. The one thing it seems black can’t do is slowly unwind and improve his position. The game continued 23 R3f2 (apparently a mistake, the computer wants 23 fxe5) 23… Nh5. Black could have played 23… Rg8! instead, with the idea that 24 hxg4 Nxg4 hits the queen while threatening mate, whereas any “normal” move, such as 24 Re1 is met by 24… Bxh3! Instead white won another pawn and survived in the ending, but you can’t blame the players for the draw, they gave it their all.
Round 8: Firouzja-Nakamura
This was Nakamura’s best chance to take the lead and potentially win the tournament, but it was not to be. White’s position is hanging by a thread: check out 34 Bf1? Bxg2 35 Bxg2 Nc4 and wins. But unfortunately for our confession booth addicted hero, Firouzja had everything under control: 34 Qxf4! Qxf4 35 Nxf4 Nxc4. Things are still looking up for Naka, who is up a piece and attacking white’s rook. 36 Rc2? Rxd4 37 Nxh3 Ne5 is quite bad, but white calmly played 36 Re2! (activating the rook; the bishop is trapped) 36… Rxd4 37 Nxh3 Kh6 38 Re7 and the game was drawn on move 56. A miraculous escape that must have been extremely frustrating for Nakamura, who has been playing inspired chess of late, only to finish runner-up in his last two tournaments.
Round 9: Firouzja-Carlsen
This was not a very exciting round for the classical games, and the position above reflects that sad truth. However, it does provide time to talk about the format. I wish we could conclude that awarding three points for a win in classical chess and forcing players who draw to play a rapid tiebreak makes for more decisive chess. It does not. I suspect that some of this is psychological. In a normal tournament format, you might feel a lot of pressure to win a game to make up ground on the tournament leader; here you might be a little more willing to inch forward in the standings by telling yourself that you’ll just win the rapid game instead.
What does lead to more decisive games is this crazy time control where the players don’t get more time at move 40, just the addition of the increment.4 This helps explain some of the bad endgame play midway through the event, and somewhat justifies the inclusion of the position above. Carlsen’s in no trouble, of course—even if he loses the d-pawn for nothing more than some trades the four versus three ending is almost certainly going to be drawn. But it’s annoying to be slowly tortured when you’re never going to get more time, so let’s give a cheer for active defense: 28… d5! Now 29 exd5 Rxe3 30 fxe3 Rd2 is a little annoying in the other direction, so Firouzja chose 29 e5 Qe6 30 h3 Rc4 31 Rd3 Re4 and the contestants proceeded to the armageddon game.
Round 10: Praggnanandhaa-Nakamura
Another dud of a round, with three uneventful draws. Just as in the Candidates, Nakamura needed to win, but he once again failed to get anything from the opening. Naka’s choice against the Ruy Lopez, 5… Bc5, is not common, and the position above is new to me, but it wasn’t to Pragg, who has defended the black side of this position twice in the past few years. Playing someone’s opening against them is an interesting psychological trick for certain tournament situations, but perhaps better when you need a draw rather than a win. In any event, the position was so clearly even that Carlsen felt comfortable drawing early against Caruana and coasting to victory.
Thanks for reading; there are some other fun events going on with some interesting games that I’ll try to tackle next week!
Rapid games are fun to watch live, but I find them a little hard to take seriously when doing analysis, even the kind of lazy computer-aided analysis in this post. All ten positions here are from the classical games.
This was particularly egregious in Ju Wenjun’s final round game, as the commentary team was raving about her 100% accuracy score around move 25, a measure that says more about how complex the game was than anything else. (And might get lead to some subtle, not-subtle attacks from Kramnik on social media.)
Well-known to me at least. NM Roger Poehlmann, longtime Berkeley Chess School coach, loved to play this position with either color against the kids at the BCS Summer Camp, drawing each time with black and winning each time with white. They had a hell of a time figuring it out. That said, a friend rated about 2100 took a draw on the stronger side of this position in a tournament, so not even experienced players necessarily know it.
A feature which I assume exists to give them enough time to play the rapid game if necessary.