[Event "FIDE World Championship 2021"] [Site "Dubai"] [Date "2021.12.07"] [Round "9.1"] [White "Nepomniachtchi, Ian"] [Black "Carlsen, Magnus"] [Result "0-1"] [ECO "A13"] [WhiteElo "2782"] [BlackElo "2855"] [PlyCount "78"] [EventDate "2021.11.25"] [EventType "match"] [EventCountry "UAE"] [Source "ChessBase"] [SourceQuality "1"] 1. c4 e6 2. g3 d5 3. Bg2 d4 $5 (3... Nf6 4. Nf3 dxc4 5. Qa4+ Bd7 6. Qxc4 c5 7. Ne5 Qc8 {1-0 (49) Nepomniachtchi,I (2792)-Duda,J (2738) Zagreb (Blitz) 2021}) 4. Nf3 Nc6 5. O-O Bc5 6. d3 Nf6 7. Nbd2 a5 8. Nb3 Be7 9. e3 dxe3 (9... e5 10. exd4 exd4 11. Re1 O-O 12. Ne5 Nxe5 13. Rxe5 c6 {1-0 (45) Gabuzyan,H (2587) -Bellahcene,B (2497) Krasnaya Polyana RUS 2021}) 10. Bxe3 Ng4 11. Bc5 O-O 12. d4 a4 13. Bxe7 Qxe7 14. Nc5 a3 $5 15. bxa3 $5 (15. b4 Nxb4 16. Rb1 b6 17. Rxb4 bxc5 18. Rb5) 15... Rd8 (15... Rxa3 16. Nb3) 16. Nb3 Nf6 17. Re1 Qxa3 18. Qe2 h6 19. h4 Bd7 20. Ne5 Be8 21. Qe3 Qb4 22. Reb1 Nxe5 23. dxe5 Ng4 24. Qe1 Qxe1+ 25. Rxe1 h5 $5 26. Bxb7 Ra4 $5 27. c5 $4 c6 $1 28. f3 Nh6 29. Re4 Ra7 30. Rb4 Rb8 31. a4 Raxb7 32. Rb6 Rxb6 33. cxb6 Rxb6 34. Nc5 Nf5 35. a5 Rb8 36. a6 Nxg3 37. Na4 c5 38. a7 Rd8 39. Nxc5 Ra8 (39... Ra8 40. Nb7 Bc6 41. Nd6 Nf5 42. Nc4 Bxf3 43. Nb6 Rd8 $19) 0-1 [Event "FIDE World Chess Championship 2021"] [Site "?"] [Date "2021.12.07"] [Round "9"] [White "Nepomniachtchi, Ian"] [Black "Carlsen, Magnus"] [Result "0-1"] [ECO "A13"] [WhiteElo "2782"] [BlackElo "2856"] [Annotator "samsh"] [PlyCount "78"] [EventDate "2021.??.??"] [SourceVersionDate "2021.12.07"] {The ninth game saw Nepo avoid 1.e4 for the first time in the match. He actually got quite a nice position early on, but after White was somewhat outplayed to reach an equalish position, the game ended abruptly in one move.} 1. c4 e6 2. g3 d5 3. Bg2 d4 {I was a little surprised to see this move, which sort of invites a kind of Benoni with two tempi less: one because of colors reversed, and a second because e6-e5 will come in two moves rather than one. The move is surely fine objectively, but it felt like it would lead to more dynamic positions than I would expect someone to want with the black pieces and a two-game lead. [During the press conference, Carlsen shared that this was his preparation although he was having trouble recalling all the subsequent lines.]} (3... Nf6 {This would be more typical. I don't really think White has anything better than transposing into a Catalan. With Magnus taking the white side of the Catalan twice already, I assume he would have checked it pretty closely. So it was a bit surprising to see 3...d4 instead.}) 4. Nf3 Nc6 5. O-O Bc5 {This is a pretty unusual setup for Black, but it makes some sense. Black is never too worried about b2-b4 since he can always meet a2-a3 with a7-a5. And the bishop obviously does a good job of lessening the punch of e2-e3, which is the only real way for White to try to open the game.} 6. d3 Nf6 7. Nbd2 {Nepo was still playing pretty fast, if not immediately. I suspect he had checked this line a little but that he had not reviewed it too much and spent more of his efforts on other moves, but he still knew what to do.} a5 8. Nb3 Be7 9. e3 dxe3 10. Bxe3 {Up to here, Magnus was taking some time, but I think he was still in his preparation. He certainly knew where to put his pieces and that he should not be castling right away.} Ng4 $1 {Black harasses the e3-bishop. This is quite a good decision because White's best chance for an edge is to fix his one bad piece with Nf3-d4. This will be harder to pull off now that the bishop is destabilized.} (10... O-O {This looks like a more natural move, but White should be a bit better after} 11. Nbd4 $1 Bd7 12. Nxc6 Bxc6 13. d4 $14) 11. Bc5 O-O {Nepo took a long think here, aiming to find the best way to coordinate his pieces.} 12. d4 $1 {This is not an easy decision. First of all, White loses any hope of playing Nb3-d4 later, and also his knight will be forced to c5 and unable to retreat after. It takes a lot of confidence in your calculation to play this way.} a4 13. Bxe7 Qxe7 14. Nc5 {Now, Magnus needs to find a tough only-move in a very direct position to keep things under control. His ability to coordinate his pieces in a solid but uncomfortable situation is out of this world, and it really showed in just how easily he held these uncomfortable Spanish positions. Here, it is a different story. Black has to be very direct and calculate close lines in a shaky way.} a3 $6 {This is a very human move, but it's also a serious mistake. White now has to be very energetic, but if he finds the way, he is clearly better. One thing I have often said is that Magnus is certainly not perfect, but if you want to beat him, you do have to be accurate. He rarely gives chances, and when he does, you have to be accurate all the way through.} (14... e5 $1 { This was the only way for Black, and even then, he remains a little worse.} 15. Re1 $1 Rd8 $1 {Now, White has a few tempting options, but none of them work too well.} 16. b4 $1 (16. Nxa4 Qb4 $1 17. b3 exd4 $1 {After a precise sequence, Black gets his pawn back, and looks more or less okay to me.}) (16. h3 $2 Nxd4 $1 {Black emerges ahead, as the c5-knight is hanging.})) 15. bxa3 $2 {Already now, the machine still gives some 0.2 for White, but in practice, I actually prefer Black. His position feels very easy to play, and White is the one with long-term weaknesses. The a-file will cause endless annoyances.} (15. b4 $1 { This was the way, and White is clearly better after some complications.} Nxb4 { This has to be the most critical try for a human to worry about.} (15... Rd8 { This is the machine's choice, but after} 16. Rb1 {White is clearly better.}) 16. Rb1 $1 b6 (16... Nc6 17. Nxb7 Rb8 18. Na5 $1 $16 {This is another only-move that White needs to see before pushing b4. It is not too hard, but having to find a bunch of easy-ish only-moves all in advance to justify playing the right way can be tough.}) 17. Rxb4 bxc5 18. Rb5 $1 {An important only-move that White has to see before pushing b2-b4. Black cannot take on d4 and will lose his c5-pawn. The game then continues, but White is surely better due to his superior structure.}) 15... Rd8 16. Nb3 Nf6 17. Re1 Qxa3 18. Qe2 h6 {This is exactly the kind of position where the a-file is a huge problem for White. He needs to play Ne5, and to do this, he needs to secure the d4-pawn. As such, he would very much like to get his a-rook to d1, but this would drop the a2-pawn.} (18... Bd7 {I would have preferred simply tucking the bishop away on e8, as the inclusion of h4 and h6 could allow some vague chances on the kingside. What Magnus did is fine too, of course.}) 19. h4 Bd7 20. Ne5 Be8 {Now, I was somewhat surprised White did not take on c6.} 21. Qe3 {This is a solid and reasonable move, though I certainly would have tried something else.} (21. Bxc6 $5 Bxc6 22. g4 $5 {The machine is totally unimpressed with this idea, and claims Black is just better. But in human terms, this feels very scary. g4-g5 is not messing around and all of Black's pieces are on the queenside—I would be worried about getting mated. When I was following the game without the computer, I thought this could be very powerful. It would be interesting to know if Nepo saw this option and (correctly) chose to reject it—or if it did not occur to him.} Nd7 $1 {Not an easy decision as this seems to just be begging for Nxf7.} 23. Nxf7 $1 {Let the fun begin!} (23. Nxc6 bxc6 24. g5 hxg5 25. hxg5 Nf8 $1 {Black prevents g6, and once his knight sits on g6, any queen and rook lift to the h-file doesn't do much either.}) 23... Rf8 $1 {This is a very tough move.} (23... Kxf7 24. Qxe6+ Kf8 25. d5 Ba4 26. Re3 {Apparently, Black is fine here. I would be terrified.}) 24. Nxh6+ gxh6 25. Qxe6+ Kh7 { The machine claims Black is better here, but it looks very unclear to me, and it is hard to get here in the first place.}) 21... Qb4 $1 {A high-class move. As soon as White's queen no longer defends a2, this move is possible without allowing Rad1. Furthermore, it is prophylaxis against the plan of Qe3-f4 and g4-g5.} 22. Reb1 (22. Qf4 {This looks scary, but after} Rxa2 $1 {it's all over already.}) 22... Nxe5 $1 {I was surprised to see this move—it felt a little inhuman, but it is also best.} 23. dxe5 Ng4 $1 {It takes some guts to play this move since the knight is nearly trapped, but it is correct.} 24. Qe1 (24. Qf4 $2 h5 $1 {The knight can escape to h6. After} 25. Bf3 Qc3 $1 {Black's pieces are plenty active, and White can't win a pawn so easily.}) 24... Qxe1+ 25. Rxe1 h5 26. Bxb7 Ra4 27. c5 $4 {I have a bit of sympathy for this decision, but I still have to give it two question marks. I suspect what happened is that Nepo saw the idea of c6, realized it is not a big deal because the bishop is easily defended by Nc5, and then sort of forgot about it. Once c4-c5 comes though, the c5-square is no longer available, and the bishop is straight up lost. [During the press conference, Nepo admitted he missed the idea of ...c6 entirely. He mentioned that ...c6 is easy to miss because it is never possible while the white pawn is on c4 because White would have Nc5 to defend the bishop.]} (27. f3 {This should be preferred, but after} Nh6 28. Be4 Nf5 { White will not keep his extra pawn, and a draw looks super likely. Still, this is obviously better than what actually happened.}) 27... c6 $19 {I have to give Nepo a little credit; he forced Magnus to find some only-moves later on, but they were all routine enough.} 28. f3 Nh6 29. Re4 Ra7 30. Rb4 Rb8 31. a4 Raxb7 32. Rb6 $1 {a4-a5-a6 is coming.} Rxb6 {This is the only move, but hardly rocket science.} 33. cxb6 Rxb6 34. Nc5 {White gets the c5-square and the a-pawn is coming, but Magnus still has it under control.} Nf5 35. a5 Rb8 36. a6 Nxg3 $1 {White will not be able to force the pawn through without moving the knight off of c5. Once this happens, c6-c5 comes and the bishop gets to c6. White can win a piece back but will be down several pawns and still have no hope.} 37. Na4 c5 $1 {The bishop gets to c6 and that is the end of that.} 38. a7 Rd8 $1 {The last important move, and it is all over.} 39. Nxc5 (39. Rb1 { This would have been more challenging, inviting Black to play Bc6, but I'd be beyond shocked if Magnus did not play it right.} Ra8 $1 (39... Bc6 $2 {I suppose I could see a human messing this up.} 40. Rb8 Rf8 41. Nxc5 $1 {Black cannot prevent Nd7 and the tables have turned.}) 40. Rb8 Rxa7 41. Rxe8+ Kh7 42. Nxc5 Nf5 {White got his piece back, but is obviously done for.}) 39... Ra8 { Nothing more to say here. 0-} 0-1