What a strange and exciting round at the Norway Chess tournament! After four rounds Sergey Karjakin was very close to being the runaway winner, and by round 6 it seemed that the contest was between him and Magnus Carlsen, with the other eight players relegated to a secondary tournament of their own. Not any more! Coming into round 8, the penultimate round, Karjakin had 5.5 points, Carlsen 5 and no one else had more than 4. At that point it seemed almost impossible that someone else could win the tournament, but now it seems well within the realm of possibility.
The first game to finish was the round's only draw. Topalov-Aronian was a Karpov Variation Nimzo-Indian that saw Topalov come up with a good new plan with 21.Be3 (rather than 21.Ba3, which had been played before) and then 23.a4 and 24.a5. This put a little pressure on Black, but Aronian's decision to sac a pawn for play enabled him to hold without undue difficulty.
The shortest decisive game in terms of moves was Nakamura-Radjabov, which was won by the American; the logical result, given the trend through the middlegame up to around move 33, but then strange things happened. The game grew increasingly wild, with Nakamura trying to give mate while Radjabov sought to crash through the center. Whether due to the complicated nature of the position or from an attempt to play on Radjabov's severe time pressure (or some mixture of the two) Nakamura started to err (34.Nf5! would have kept a winning, or very nearly winning advantage), and had Radjabov played 36...Bxc4 he would have been on top. When he played 38...Bxc4 two moves later, however, it was the wrong time, and Nakamura was winning again - this time for good.
Taking a few more moves but (I think) finishing slightly sooner than Nakamura-Radjabov was Anand-Hammer. The game was very messy early on, with even the world champion admitting that he was both at sea and missing various tactical possibilities. Nevertheless, he kept his head together, and while Hammer may have missed some small chances, Anand took advantage when a big one came his way. 20...Rd8 was a serious mistake, and 21.Nxf7! was a crusher. With the win, Anand got to 5 points, tying him with Carlsen's score before the round.
As it turned out, that was also Carlsen's score after the round. He lost to Wang Hao in the same way that he usually beats people: he keeps on playing, and then an equal position gradually turns into a slightly better position, which turns into a pawn up, which turns into a win. Carlsen flirted with an edge with the white pieces, but after 23 moves the position was simply equal. Here Carlsen played the double-edged 24.Nd6. It's a good move, and an ambitious one too, but the danger is that the knight is too committed, and can't get out. That's what ended up happening. Carlsen played 29.c5 to cement it, and after an inevitable ...Bxd6 cxd6 the pawn would likely drop, as it finally did on move 34. Even after losing the pawn, Carlsen probably should have drawn the ensuing rook ending. Wang Hao suggested that 52.f4 would have drawn, and the engine "claims" that 56.h5 would draw and that 64.Kf2 was White's last chance to defend. After 64.Kg2? d3, it was definitely over, thanks to the nice tactical trick that finished the game. Carlsen had missed and Wang Hao had foreseen the cute 79...g3+!, which wins the queen: 80.Kxg3 Qg1+ followed by 81...Qh1+ and 82...Qxh8.
This meant that Karjakin could have won the tournament with a win against Peter Svidler, or at least guaranteed himself a tie for first overall (with the guarantee of nothing worse than a blitz playoff in case someone caught him) with a draw. Svidler had prepared the line he chose with White for some time, but only spotted the idea with 9...d5 that morning. He was unhappy, as he felt that it killed the line, but as it hadn't been played he consoled himself that Karjakin wouldn't know it. Sure enough, Karjakin had found it too and played it, and had he followed up with 11...Nd6, Svidler felt he would have nothing, that Karjakin's approach would have killed the line for White.
11...Nxd2 was no disaster though, but it allowed Svidler to sharpen the position and soon obtain a serious advantage. Both sides made errors (Svidler's 20.Qh5? instead of 20.Qg4; Karjakin's 30...Qb6? rather than 30...dxc4 [Svidler's explanation is that Karjakin intended 30...Qb6 31.cxd5 Bd2, only to realize a move too late that 32.Re7 (or even 32.Bxh7+ Nxh7 33.Re7) wins on the spot.]), but the general flow was in Svidler's favor. When the time control was made Svidler only had two bishops for a rook and two pawns, but what bishops! Practically speaking, Karjakin's situation was extremely difficult, and the bishops finally swallowed him alive. Objectively, he could have held with 47...Ra8 or the bolder 47...Rc2, and a move later he still might have been able to save the game with 48...d3. (48...d3 49.Qg6 a3 50.Bxh6 Qe7 51.Bd2 Qxe6! 52.Bxc3 d2 53.Bxd2 a2 54.Bc3 a1Q 55.Bxa1 [what a rapacious bishop!] 55...Qe1+ 56.Kg2 Qxa1, with a likely draw.) It's one thing to work things out moving pieces or (especially) with an engine, but at the board Black's plight is nearly hopeless, and the decisive error was 48...a3. Svidler finished in style, the key move being 53.Kh3! (Without that, it may still be drawn.)
There's one round left, and for those of you want to see it live, be forewarned that it starts three hours earlier than usual. Here are the last round pairings (scores are in parentheses):
- Aronian (4.5) - Carlsen (5)
- Wang Hao (3.5) - Anand (5)
- Hammer (1.5) - Nakamura (4.5)
- Radjabov (2.5) - Svidler (4.5)
- Karjakin (5.5) - Topalov (3.5)
Just think: if Aronian draws or wins, Anand draws, Nakamura and Svidler win and Karjakin loses we can have a five-man blitz playoff! Half the field is still in the running for first place, with three players having an especially good shot at it. Still, Karjakin has the best chances, both because he leads and because he has White. Will he do it? We'll see starting in five hours.