[Event "World-ch Nepomniachtchi-Ding"] [White "Ding,Liren"] [Black "Nepomniachtchi,Ian"] [Site "Astana"] [Round "12"] [Annotator "Shankland,Samuel L"] [Result "1-0"] [Date "2023.04.26"] [WhiteElo "2788"] [BlackElo "2795"] [PlyCount "75"] {[%evp 0,75,61,61,0,5,44,27,80,39,38,22,60,45,44,31,31,34,84,88,97,12,28,42,49,33,9,-20,-25,-21,1,0,41,41,43,84,42,26,40,-41,13,-19,1,4,8,-7,39,2,107,-48,15,0,0,-30,86,32,38,23,105,105,123,144,205,163,181,127,182,150,176,180,319,393,467,467,467,480,709,709]} 1. d4 Nf6 2. Nf3 d5 3. e3 {Ding's opening choice in game 12 would have been quite surprising if we had not know about Richard Rapport serving as his second. This move definitely is the kind of thing that Richard likes to try. It's surely harmless objectively, but it can be a bit annoying for Black to face if he does not know what he is doing.} c5 (3... e6 { I never saw any issue with continuing copycat. It would be interesting to know what Ding had in mind here.} )4. Nbd2 cxd4 5. exd4 Qc7 {This is an odd looking move, but it is the computer's top choice and Nepo played it immediately, so I'm quite sure it was still preparation.} 6. c3 Bd7 $2 {But this was a big surprise to me, and Nepo played it so fast! It's a bad move and White has an easy edge with Nf3-e5 followed by Nd2-f3. I assume Ian must have confused his preparation somehow.} 7. Bd3 $6 {Obviously this is not a terrible move, but it misses out on a big chance.} (7. Ne5 {This looks incredibly natural.} Nc6 8. Ndf3 $14 {White has a fantastic version of the Exchange Caro-Kann and looks pleasantly better to me.} )Nc6 8. O-O Bg4 {Black has survived the worst of it, but this move clearly indicates that Bc8-d7 was wasted time.} 9. Re1 { This is a bit slow.} (9. Qa4 $1 {I like the machine's recommendation. Ne5 comes next.} )e6 10. Nf1 Bd6 {Black should be fine now.} 11. Bg5 $6 { I really don't like the way Ding handled the early phase of this game. He certainly made up for it later on, but I think trying to double Black's pawns and play for mate is a bit too optimistic here. Black's king should be safe enough on g8.} (11. Ng3 {I'd think aiming to take the bishop pair with h2-h3 next would give White a little pull.} )O-O 12. Bxf6 gxf6 13. Ng3 f5 $1 { This was a good decision from Nepo. Shutting down the d3-bishop makes it unlikely that he will be checkmated.} 14. h3 Bxf3 15. Qxf3 {Now, the only issue Black is that his king's cover is somewhat compromised. He has no other weaknesses. He certainly will not be checkmated if he can maintain his f5-pawn, and as such, White only has two serious ideas at his disposal. He can either aim to sacrifice a piece on f5, or he can hope to push g2-g4 to tear open the diagonal.} Ne7 $1 {This is definitely the safest move, and the best one from a human perspective.} (15... Kh8 {My computer gives a slight not to putting the king in the corner and preparing for Rg8. I won't beat the machine, but it seems insane to me to allow White to take on f5.} 16. Nxf5 $1 exf5 17. Qxf5 f6 18. Re6 $1 {A computer might survive here, but it looks terrifying to a human.} )16. Nh5 Kh8 $1 {Black anticipates g2-g4 and gets ready to bring his rook to g8.} 17. g4 Rg8 18. Kh1 Ng6 $1 {An excellent decision from Nepo, albeit a not too difficult one. Black points out that gxf5 is not happening, and as such, he will be able to secure his kingside.} 19. Bc2 $6 {I'm not really sure what this move was about.} (19. gxf5 $2 Nh4 $1 {White will find himself mated by a rook on g2 sooner rather than later.} )(19. Rg1 {This looks much more natural to me than the game continuation.} )Nh4 $1 20. Qe3 Rg6 $1 {The tables have turned completely. Black is ready for f5-f4 and to bring his rook to h6, when his king will be totally safe and only White will have trouble on the kingside. The machine claims Black is nearly winning, and around here, it really felt like Nepo was going to become the next World Champion.} 21. Rg1 f4 $1 22. Qd3 Qe7 23. Rae1 Qg5 {White should be pretty helpless to stop Black's pieces from coming to the kingside and giving mate. Next up will be Rag8, Rh6, and some day, f5 will come. Ding lashed out here, aiming to complicate the game. This was not optional, it clearly had to be done, but ultimately, it also should not have worked.} 24. c4 dxc4 $1 25. Qc3 (25. Qxc4 Nf3 {White loses material.} )b5 26. a4 {After his weird opening hiccup, Nepo had played a spectacular game up to this point, and he should be ready to bring the point home and all but close out the match. While his next move is not a mistake per se, and in fact the computer gives it as best, it was the first sign of the tide turning.} b4 $2 {I don't care that my computer claims this is best, it is absolutely the wrong move, and I could just have easily given it a second question mark. White grabs a pawn and all of a sudden has room for his pieces to work with on the queenside. The whole match was a back and forth affair, with missed chances and mistakes from both sides. The players were extremely well matched and it definitely could have gone either way. Among all of the ample interesting moments, my suspicion is that some years down the road, this move is the one Nepo will regret the most.} (26... a6 $1 $19 { This is the move anyone would play in a bullet game, and it is absolutely winning. White does not have a shred of counterplay, his pieces are all kept out, Black is a clean pawn up, and the kingside play will end the game in due time.} )27. Qxc4 Rag8 (27... Nf3 {The machine claims, this is absolutely over. It's not so easy in human practice!} 28. Qc6 (28. Bxg6 Qh4 $1 {The point. White has three hanging pieces, and Black's only winning move is to not take any of them.} )Nxe1 $1 29. Qxa8+ Rg8 30. Qe4 Nxc2 31. Qxc2 {The computer says, Black wins here. To an human eye, this is not obvious at all.} Qh4 $1 32. Qd3 f5 $1 {It takes a lot of super precise moves to bring the point home. I can't fault Nepo for not finding Nf3, but I do think b5-b4 was a horrible decision.} )28. Qc6 {Now, Black is still winning, but his task has gotten so much harder.} Bb8 $2 {This not only misses the win, but also allows White a winning chance of his own! The position is incredibly complicated and rich, and its unsurprising to me that even 2800 level players made a lot of mistakes. } (28... Nf5 $3 {This was the only winning move. Black clears the h4-square for his queen.} 29. Rd1 {Still, even this does not look so clean.} (29. gxf5 $2 Qxh5 $19 )Qh4 $1 30. Qf3 Ng3+ $1 31. Rxg3 fxg3 32. Bxg6 {Now, it's hard to get this far in your calculations, and even if you do, it's easy to get frustrated when you can't find mate here. Indeed, there isn't one. Black has to win an endgame.} Qxh3+ $1 33. Kg1 gxf2+ $1 34. Kxf2 Qxf3+ $1 35. Kxf3 fxg6 $1 {This is a very obvious move once the position is on the board, but a hard one to see from afar. Normally when calculating a line like this, you'd think of the more positionally sound recapture with the h-pawn. That would leave White with very good drawing chances.} (35... hxg6 $2 36. Nf6 $1 $15 {This knight is very annoying and Black needs to be careful of mating threats on the h-file. White should not lose with good defense.} )36. Nf6 Rf8 $1 37. g5 Be7 $1 {It took a lot of tough moves, but eventually Black wins. I think an on-form Nepo early in the tournament would have found this sequence, but he was a bit off this game, and has always struggled at the end of long events.} )29. Qb7 (29. Bxg6 $1 {This was a winner. The point is that after} hxg6 {White's knight actually is not so easy to capture, as the opening of the g-file will be dangerous for Black. He can ignore everything Black is up to and proceed along with his own play.} 30. d5 $3 {White blasts through on the e-file. The e-pawn cannot move without allowing a check on f6 to transition into a winning endgame, and the knight cannot be taken.} gxh5 {What else?} 31. dxe6 $1 { White's rooks enter the game, and Black is done for. This is obviously a computer line and it proved too counterintuitive for either player to notice, but it's a nice one all the same.} )Rh6 $2 {On the last 3 ply, both sides missed wins. It is clear at this point that it's anyone's game, and Nepo ended up making the last mistake.} (29... Nf5 $1 {Again, this move was strong.} )30. Be4 Rf8 31. Qxb4 $1 Qd8 32. Qc3 Ng6 {This is a very bad sign for Black. He is only slightly worse still, but the trend of the game for the last several moves has been horrible. He is a pawn down, his knight is retreating and he won't give mate, and the queenside has been opened.} 33. Bg2 Qh4 34. Re2 f5 $4 {I don't know what this move was about. The pressure of a high stakes game can get to anyone, and fatigue after 12 rounds is very real. Still, this move is surprising and hard to explain.} (34... Bd6 {Some move like this one would be normal enough. White is a bit better but the game goes on, and Black is not without counterplay, or saving chances thanks to the opposite colored bishops.} )35. Rxe6 $18 {Just like that, the game is over. Losing a pawn is the least of Black's concerns. The real problem is he just lost the only hope he had of keeping the a1-h8 diagonal closed. next up is d4-d5+, and Black is absolutely checkmated.} Rxh5 36. gxh5 Qxh5 37. d5+ Kg8 38. d6 {Nepo resigned. This must have been the most critical game of the match, as it easily could have all but ended Ding's hopes of winning the match if it had gone differently at one or two moments. As is, it blew the match wide open, evening the score with just two games to go.} 1-0