[Event "2018 World Chess Championship"] [Site "London"] [Date "2018.11.09"] [EventDate "2018.11.09"] [Round "1"] [White "Fabiano Caruana"] [Black "Magnus Carlsen"] [Result "1/2-1/2"] [ECO "B31"] [Annotator "chess.com"] [PlyCount "229"] [CurrentPosition "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1"] {Greetings to all! This is Sam Shankland, and I'll be annotating most of the games of the World Championship match here for Chess.com. I was on Magnus Carlsen's team for his previous matches, but this one I have no role in whatsoever, so I can speak much more freely as a full spectator without any inside information. I was pleased to see that the first game featured some real fighting chess, and hope this trend continues in the upcoming games.} 1. e4 c5 {The first surprise. Caruana is mostly a 1.e4 player, but the Sicilian is a rare guest in Carlsen's repertoire.} 2. Nf3 Nc6 3. Bb5 {White avoids the main lines of the Open Sicilian that might arise after 3.d4. Nowadays at the top level, the Rossolimo is as good a chance to fight for an edge as any of White's other tries.} 3... g6 4. Bxc6 dxc6 5. d3 Bg7 6. h3 Nf6 7. Nc3 Nd7 {Presumably an improvement over their previous encounter from some years ago.} (7... b6 {Magnus chose this move in a game against Caruana back in 2015. He won very nicely, but the opening could have gone a lot worse.} 8. Be3 e5 {And now instead of castling, Caruana could have more less won the game with} 9. Nxe5! Nxe4 {When the silent killer} 10. Qf3! {Is a serious problem for Black to deal with}) 8. Be3 e5 9. O-O (9. Qd2 {White can also consider remaining flexible with the king. Sometimes it might want to go to the queenside.}) 9... b6! {I like this idea. Black prepares for Nf8-e6 before castling.} 10. Nh2 Nf8 {White more or less has to play f4 now- otherwise he will never manage once Black has played Ne6.} 11. f4 exf4 12. Rxf4 Be6!? {This was the first seriously notable moment of the game. I was very surprised Black would not want to put the knight on e6, but Carlsen had other ideas. His plan will be to play h6, g5, and Ng6. Then, depending on how White plays, he will castle on which side he feels is best. Around here I joined the live broadcast and said I already would prefer the black side of the position, though White is objectively not worse yet.} 13. Rf2 h6! 14. Qd2 {White develops another piece and pressures the h6 pawn, discouraging Black from castling kingside.} (14. Nf3 {The machine's proposal doesn't make any sense to me. What does the knight do here? Caruana's choice is much more human and was the move I expected during the game.}) 14... g5! {Another strong move. Black simply prepares to improve his pieces further, gaining space on the kingside and preparing for Ng6.} 15. Raf1?! {This was the first moment I really did not like Caruana's choice. While he is certainly not worse yet, it feels wrong to me to keep this square from the knight. It must be said that my computer disagrees with me.} (15. Nf1 {My idea during the broadcast was to try to bring the knight to g3 and f5. This is not a bad idea, but the wrong execution.}) (15. Ne2!? {Probably this enterprising pawn sacrifice is worth attention. White is ready for c3, Ng3, and Nf5.}) 15... Qd6 16. Ng4 {The computer hates Caruana's move here, but it's very consistent with his previous play. White eyes the squares on the weakened f-file. [Editor's note: This is the move Caruana criticized the most of his.]} 16... O-O-O 17. Nf6 {The knight is well placed, but soon removed.} 17... Nd7! 18. Nh5 Be5! {Around here I really thought Magnus would win the first game. White is horribly passive and will face serious issues on the kingside, which should soon open. The position became objectively lost at some point, but Caruana still must be praised for defending this position.} 19. g4! {The first good decision. White stops g5-g4 from Black, and it will not be easy for him to get h6-h5 through.} 19... f6 20. b3 Bf7 21. Nd1 Nf8? {Magnus got a little too creative for his own good. Even without the f6 pawn, his position remains strong, but there was no reason to give White anything for his suffering.} (21... Rhg8 {Looks simplest to me. Black prepares for Bxh5 followed by g5-g4. White faces a long and difficult defense.}) 22. Nxf6 {Now White at least has a pawn for his suffering, though his position soon deteriorated further.} 22... Ne6 23. Nh5 Bxh5 24. gxh5 Nf4 25. Bxf4 gxf4 26. Rg2? {Caruana had made some inaccuracies before to get into a mess, but this was the only decision I really did not understand. Even very strong players make mistakes and this is totally normal, but this particular one is hard to understand. White needed to give the exchange, at which point I might even prefer his position.} (26. Rxf4! Bxf4 27. Qxf4 $44 {White has exchanged off the annoying dark squared bishop, has 2 very solid pawns for the exchange, and most importantly, his miserable knight that ewas dominated by the f4 pawn can come to e3 and then f5. The computer screams equal but I think any human would prefer the white side.}) 26... Rhg8 27. Qe2 Rxg2+ 28. Qxg2 Qe6! {A high-class move. Black will soon seize the g-file with Rg8.} 29. Nf2 Rg8 30. Ng4 Qe8! {Now White is in really bad shape. If he could save the h5-pawn he would probably be better, but once it is captured, h6-h5 will boot yhe g4-knight off its post and the g-file will open. In addition to the position to being objectively bad, it is much harder for White to play than for Black.} 31. Qf3 Qxh5 32. Kf2 Bc7 33. Ke2? {It's hard to criticize Caruana since his position was just difficult, but this was not best.} (33. e5 Kb7 34. Nf6 {Offered more saving chances, but it is very hard to play like a machine all the time.} 34... Qh4+ 35. Ke2 Rg3 36. Qxf4 Qxf4 37. Rxf4 Bxe5 38. Rf5 Bxf6 39. Rxf6 Rxh3 40. Rf7+ {White should probably hold.}) 33... Qg5! {Simple and strong. h5 is coming, the g-file will open.} 34. Nh2 h5 35. Rf2 {Up to this point Magnus had played quite a good game, though I imagine he will be mad at himself for Nf8. But around here, he started to drift. The move he played is good enough to win, but he missed the most incisive continuation.} 35... Qg1? {The first mistake is the one that makes the win more difficult, not the one that makes it no longer winning.} (35... Qg7! {Black was best advised to invade with his queen on the dark squares. White's position looks beyond saving to me.}) 36. Nf1 {The next few moves were marred by time pressure, but I think Magnus will not see this is a valid excuse for letting White off the hook.} 36... h4? (36... Qg7! {Was again strong}) 37. Kd2 (37. e5! {Would have opened White's queen. It's very hard to do this since it allows Black to make good use of the open e-file, but it was indeed best.} 37... Rg5 38. Nd2! Rxe5+ 39. Ne4! {Blocking the rook again, like it did from g4, but this time there is no pawn thrust to boot the knight away.} 39... Qc1 40. Qxf4 Qxc2+ 41. Qd2 {White is suffering, but has very good defensive chances.}) 37... Kb7 38. c3 Be5 39. Kc2 Qg7? {Paradoxically, the move Magnus needed to play previously now lets White off the hook. The position has changed and now it will not be the queen, but the bishop, that invades the queenside- a much less scary prospect for the first player.} (39... a5 {The best move was b5, but this looked the most human to me. Just make a move, keep the king safe, and wait for time control to find the winning way.} 40. a4 Rg6 {Something like this would allow Black more time to find a breakthrough, and White has no counterplay to worry about.}) (39... b5 {According to the machine, this move wins convicningly. And I don't really understand why. It's a hard move to grasp and almost impossible to find in time pressure.}) 40. Nh2! {Caruana has survived the worst of it and now has more time. The position remains unpleasant, but very defensible.} 40... Bxc3 (40... Qg3 {This might have been a better try, but it's a tough move to play. After} 41. Ng4 {Black must give an exchange with} (41. Qe2! Qxh3 42. Nf3! {Should be preferred. White is worse but hanging in there.}) 41... Rxg4! 42. hxg4 h3 43. Qe2 {And apparently this position is good for him, but this is hard for a human to do.}) 41. Qxf4 Bd4 42. Qf7+! Ka6 43. Qxg7 Rxg7 44. Re2! {The e-pawn secures enough counterplay to make a draw.} 44... Rg3 45. Ng4! (45. e5? Bxe5) 45... Rxh3 46. e5 Rf3 47. e6 Rf8 48. e7 Re8 49. Nh6 h3 50. Nf5 Bf6 {The last several moves have been absolutely forced. Now the best Black will ever get is a pawn up rook ending with h3 and e7 trading, which definitely should be a draw.} 51. a3 b5 52. b4 cxb4 53. axb4 Bxe7 54. Nxe7 h2 55. Rxh2 Rxe7 {I think one of Magnus' biggest mistakes of the game was not offering a draw on the spot here. He definitely is not going to win the game, he will be furious with himself for letting the full point slip, and giving him more time to emotionally recover before playing tomorrow probably would have been a better choice than playing on for 50 more moves, hoping for the 0.1% chance of winning to come into fruition.} 56. Rh6 Kb6 57. Kc3 Rd7 58. Rg6 Kc7 59. Rh6 Rd6 60. Rh8 Rg6 61. Ra8 Kb7 62. Rh8 Rg5 63. Rh7+ Kb6 64. Rh6 Rg1 65. Kc2 Rf1 66. Rg6 Rh1 67. Rf6 Rh8 68. Kc3 Ra8 69. d4 Rd8 70. Rh6 Rd7 71. Rg6 Kc7 72. Rg5 Rd6 73. Rg8 Rh6 74. Ra8 Rh3+ 75. Kc2 Ra3 76. Kb2 Ra4 77. Kc3 a6 78. Rh8 Ra3+ 79. Kb2 Rg3 80. Kc2 Rg5 81. Rh6 Rd5 82. Kc3 Rd6 83. Rh8 Rg6 84. Kc2 Kb7 85. Kc3 Rg3+ 86. Kc2 Rg1 87. Rh5 Rg2+ 88. Kc3 Rg3+ 89. Kc2 Rg4 90. Kc3 Kb6 91. Rh6 Rg5 92. Rf6 Rh5 93. Rg6 Rh3+ 94. Kc2 Rh5 95. Kc3 Rd5 96. Rh6 Kc7 97. Rh7+ Rd7 98. Rh5 Rd6 99. Rh8 Rg6 100. Rf8 Rg3+ 101. Kc2 Ra3 102. Rf7+ Kd6 103. Ra7 Kd5 104. Kb2 Rd3 105. Rxa6 Rxd4 106. Kb3 Re4 107. Kc3 Rc4+ 108. Kb3 Kd4 109. Rb6 Kd3 110. Ra6 Rc2 111. Rb6 Rc3+ 112. Kb2 Rc4 113. Kb3 Kd4 114. Ra6 Kd5 115. Ra8 {Both sides are probably unhappy with this game. Caruana for getting pushed around and barely surviving with White in the first game, and Carlsen for letting a full point slip. In general Magnus is the one I might worry about tilting a little more, but I think at least for now, both players will come tomorrow ready to play and having put this first game behind them. I look forward to seeing what may come!} 1/2-1/2 [Event "2018 World Chess Championship"] [Site "London"] [Date "2018.11.10"] [EventDate "2018.11.09"] [Round "2"] [White "Magnus Carlsen"] [Black "Fabiano Caruana"] [Result "1/2-1/2"] [ECO "D37"] [Annotator "chess.com"] [PlyCount "97"] [CurrentPosition "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1"] {The second game of the match was much more in line with the kind of thing I expect in early rounds than the first one was. Unsurprisingly Black equalized without much trouble- this tends to be pretty easy to do with months of preparation time and your opponent not knowing what was coming. But unlike the first game, it never really strayed far from equality. Rounds 3 and 4 when each player plays their second game with White, knowing what may be coming, tend to be the time people seriously start fighting for advantages early on.} 1. d4 Nf6 2. Nf3 d5 3. c4 e6 (3... dxc4 {Caruana does not go into the QGA, which he has relied on heavily for the last couple years.}) 4. Nc3 Be7 {The classical QGD is not Caruana's main choice as of late, but he certainly has played it off and on in the past, so it can't be considered a huge surprise.} 5. Bf4 O-O 6. e3 c5!? {The first interesting moment. Caruana chooses a less common move than the typical Nbd7, which he has played a lot in the past with reasonable success.} (6... Nbd7 {Is much more common.}) 7. dxc5 Bxc5 8. Qc2 Nc6 9. a3 Qa5 10. Rd1 {Thus far, there have been some minor surprises, but this position is still well known and Black is supposed to be doing all right. Nakamura and Anand have played the Black side a fair amount, and with perfectly reasonable results. But nearly every recent game has continued with Re8, and Caruana chose something else.} 10... Rd8!? {Not a novelty, but definitely not a common move, and I can't find any games played by anyone over 2600.} (10... Re8 {Is a much more normal move. The main line continues with} 11. Nd2 e5! 12. Bg5 Nd4! 13. Qb1 (13. exd4? exd4+ 14. Ne2 Ng4 {And Black wins}) (13. Qc1 Bf5 14. Bxf6 Nc2+ 15. Ke2 Nd4+ {Is a well known draw, which was featured in several games, including a Carlsen-Nakamura encounter from 2017.}) 13... Bf5 14. Bd3 Bxd3 15. Qxd3 Ne4! {And White has not been able to show an advantage as of yet.}) 11. Be2 {Undoubtedly caught by surprise, Carlsen chooses a safe option. We will see in future games if this line is repeated and e goes for the more principled move.} (11. Nd2 {Is of course the critical try. This is the main move against Re8, and makes just as much sense now because the e6-e5 thrust loses power. White threatens Nb3 and can quickly get a good position if Black is not careful. It would be interesting to know what Caruana had in mind here, and we may well find out in subsequent games. My best guess would be he thought simplifying with} 11... dxc4 12. Nxc4 Rxd1+ 13. Qxd1 Qd8 14. Qxd8+ Nxd8 {Leads to a drawn ending. The machines give white a slight plus but the symmetry and Black's lack of weaknesses likely mean he should hold. Only future games will be able to tell if this is truly the case.}) 11... Ne4! {Black takes his chance to compromise White's queenside structure.} 12. O-O Nxc3 13. bxc3 h6 {Caruana was still playing instantly here, suggesting he was still in preparation. Black's position is pretty safe.} 14. a4 Ne7! {Another good move, and undoubtedly still prep.} 15. Ne5?! {Magnus is asking for trouble with this move, but if he plays accurately he is not going to be in any real danger.} (15. Rb1 {The computer initially claimed White is better here, but it's just wrong. Following} 15... Bd7 16. Ne5 Qxa4 17. Qxa4 Bxa4 18. Rxb7 dxc4 {The position is obviously drawn.}) 15... Bd6 {The most principled move, and a very forcing one. White has to be accurate not to be worse, but his moves are not too hard to find and I doubt Magnus was particularly worried.} (15... f6 16. Ng6 e5 {Was also possible, trying to maroon the bishop on g3. This probably would have been my choice, when White's moves seem harder to find than they were in the game. Still, after} 17. Bg3 Be6 18. cxd5! (18. Nxe7+? Bxe7 19. cxd5 Bxd5! {And Black is pleasantly better, since Rxd5 no longer takes control of the a2-g8 diagonal}) 18... Nxd5 (18... Bxd5? 19. Rxd5! Nxd5 20. Bc4! {Black somehow just loses as he cannot escape the pin. Kh7 would walk into several deadly discoveries.}) 19. Bc4! {Simple and strong, but also not too difficult. White should have enough counterplay to maintain the balance, though I still would take Black in a practical game, and I think this may have been a more promising option than Caruana's choice.} 19... Qb6 {Qc6 will follow and Black can play for an advantage, hoping to make good use of his better structure on the long term.}) 16. cxd5 Nxd5 17. Bf3 (17. Nxf7 {White could have played this kooky move. Flashy as it may be, it certainly would not have changed the result. Still, it would have looked nicer for Magnus than what he got in the game.} 17... Kxf7 18. Bxd6 Rxd6 19. Bh5+ Kg8 20. e4 Nf4 21. Rxd6 Qxh5 22. Rd8+ Kf7 23. Rfd1 b6 {The position is dynamically balanced and neither side can do much.}) 17... Nxf4 18. exf4 Bxe5 {White has to be a little precise here, but should hold easily.} 19. Rxd8+! (19. fxe5? {Is less accurate and Black can claim an edge with} 19... Bd7! 20. Bxb7 Rab8!) 19... Qxd8 20. fxe5 {Black is not well developed enough to make good use of White's weakened queenside.} 20... Qc7 (20... Bd7 21. Rb1 {Should also be a draw promptly}) 21. Rb1 Rb8 22. Qd3 Bd7 23. a5 {The machines hate this move, but they are wrong. It just forces a draw.} 23... Bc6 24. Qd6 Qxd6 25. exd6 Bxf3 26. gxf3 Kf8 27. c4 Ke8 {The ending is a little dangerous for White if Black can get the king to c6, though even then it's very likely still a draw. Magnus forced matters right away.} 28. a6! b6 29. c5! Kd7 (29... b5? {Not to be recommended! If Black could get the king to c6 he would just win with the b-pawn shoving. But he is a tempo too slow and loses after the simple} 30. c6) 30. cxb6 axb6 31. a7 Ra8 32. Rxb6 Rxa7 {Black will win the d6-pawn, but the position is a dead draw. In fact, White's doubled f-pawns make the defense even easier than it would be with a "healthier" structure of f2-g3-h4 because the king has a permanent happy home on g2, and Black can never get his pawns going past the fourth rank without massive trades.} 33. Kg2 e5 34. Rb4 f5 35. Rb6 Ke6 36. d7+ Kxd7 37. Rb5 Ke6 38. Rb6+ Kf7 39. Rb5 Kf6 40. Rb6+ Kg5 41. Rb5 Kf4 42. Rb4+ e4 43. fxe4 fxe4 44. h3 Ra5 45. Rb7 Rg5+ 46. Kf1 Rg6 47. Rb4 Rg5 48. Rb7 Rg6 49. Rb4 {Caruana understandably did not feel like trying further- he would never win. The match is of course wide open after the first two games. Still, I'd say the chessplaying trend is favoring Magnus, since he is the only one who has had any chances to strike yet, but the psychological advantage has to be on Caruana's side after today as he easily equalized with Black just like Magnus did, but unlike Magnus, he did not blow half a point. It will be interesting to see what happens on Monday, and if the Sicilian will be repeated. It certainly could, but I would not be surprised at all to see either side deviate.} 1/2-1/2 [Event "2018 World Chess Championship"] [Site "London"] [Date "2018.11.12"] [EventDate "2018.11.09"] [Round "3"] [White "Fabiano Caruana"] [Black "Magnus Carlsen"] [Result "1/2-1/2"] [ECO "B31"] [Annotator "chess.com"] [PlyCount "97"] [CurrentPosition "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1"] {Game 3 was a pretty quiet affair. White was probably a touch more comfortable coming into the early middlegame, and by the end, Black had some vague chances. But there was never a moment that I really believed either side could generate serious chances.} 1. e4 c5 2. Nf3 Nc6 3. Bb5 g6 4. Bxc6 dxc6 5. d3 Bg7 {Unsurprisingly, Caruana was the first to deviate from game 1, where everything went wrong for him.} 6. O-O (6. h3 {In the first game, White chose h3, serving as prophylaxis against Bg4 and remaining flexible with the rest of his pieces. This time, he declared his intentions earlier and castled right away.}) 6... Qc7!? {Magnus chose a very rare move. I am not at all convinced that it is superior to Nf6, which is far and away the most common choice, but it keeps in line with his willingness to mix things up a bit early on to try to get Caruana out of his book and comfort zone.} (6... Nf6 {Has been the main move for a long time. Black plans Nd7 and e5, much like he played in game 1, and challenges White to find a reason it doesn't work as well without h3 included.}) 7. Re1 {I never really understood such positions, but Caruana's plan makes a lot of sense. He wants to provoke Black into advancing e7-e5 so that he can then hope to play on the queenside with a2-a3 followed by b2-b4 without any problems on the long diagonal.} (7. h3 {The machine prefers this move, transposing to a similar kind of position to game one. Perhaps this is a better version since Black has prematurely played Qc7 (In that game the queen turned out to be better placed on d6), but I still can't imagine Black having any major problems.}) 7... e5 8. a3 {White's play is not particularly subtle, but very sensible. b4 is coming, and if Black has to take on b4, he is usually left with an unpleasant position. The rook on a1 will open up, and a lot of Black's cerntal control will disappear. Magnus spent some time here and may have been out of his concrete preparation, but his subsequent play suggested that he was very comfortable with the main ideas of the position.} 8... Nf6 9. b4 O-O {Not fearing bxc5.} (9... cxb4? 10. axb4 {This would give White exactly what he wants, and he looks pleasantly better.}) 10. Nbd2 (10. bxc5 {Grabbing the pawn was obviously a critical move, but Black is not without counterplay.} 10... Nd7! {The only way for White to hang onto the pawn now is fo go for} 11. Be3 {But after} 11... f5! {Black is getting King's Indian style counterplay. For example, the computer claims White is better after} 12. Ng5 Nf6 13. Bd2 {But I absolutely do not buy it. Following} 13... f4! {The position is closed, the extra c5-pawn hardly matters, and h6+g5 is on the way. Magnus won a game in similar style against Caruana in Tata Steel 2015.}) 10... Bg4!? {This is a very direct move. It's probably not best in a higher chess sense, but it simplifies the position a lot, to one that Magnus was probably quite sure he would not lose.} (10... cxb4 {According to the machine, this was the way to go. But I really do not mind Magnus' choice. Humans are predisposed not to take on b4, as every bit of chess wisdom known about the Rossolimo suggest that Black should leave his pawn on c5 if at all possible.}) 11. h3 (11. bxc5 {Was again possible, but now after} 11... Nd7 {Black wins the pawn back on the spot.}) 11... Bxf3 12. Nxf3 cxb4! {Normally this move is wrong, but the position is different because Black can immediately get the a7-a5 advance in. This trades off some pawns and neutralizes the potentially problematic open c-file.} 13. axb4 a5! 14. bxa5 Rxa5 {The computer claims White is slightly better here, but I really don't buy it. Black is very solid, has no weaknesses, and a clear plan for all of his pieces.} 15. Bd2 (15. Rxa5 {This may have been a better try. Still, after} 15... Qxa5 16. Bd2 Qc7 17. Qa1 Nd7 18. Qa7 Rb8! {Black overprotects b7, and he is ready to play Bf8-d6 (or c5 if allowed). Once the bishop is on d6, b5 will come, expelling the queen. Of course White is on the more pleasant side of equality but I really struggle to imagine Magnus having much trouble hanging on.}) 15... Raa8 16. Qb1 Nd7 17. Qb4 Rfe8! {Black stops any Qe7 ideas and clears the f8 square for his bishop} 18. Bc3 b5 {Around here Caruana took a lot of time before swapping all the rooks. I don't love his decision but as I said in the beginning, I don't think anyone ever had a chance.} 19. Rxa8 (19. Qb2 {According to the machine, this was a superior choice, with White holding a massive 0.09 advantage. It would almost certainly not have changed the final result.}) 19... Rxa8 20. Ra1 Rxa1+ 21. Bxa1 Qa7 22. Bc3 Qa2 23. Qb2 Qxb2 24. Bxb2 {This endgame is a little more pleasant for Black due to his extra space and potential to possibly make an outside passed pawn. Still, I'd say the odds of beating a 2800 are close to zero.} 24... f6 25. Kf1 Kf7 26. Ke2 Nc5 27. Bc3 Ne6 28. g3 Bf8 29. Nd2 Ng5! {I like this move. If for no other reason, it gives me a chance to shamelessly plug my book! In "Small Steps to Giant Improvement" (which you can buy a signed copy of from my website...), I spent a lot of time talking about how pawns do not move backward, and that there can be very good reason to provoke an opponent's pawn forward. Here, Magnus has played Ne6-g5, forcing the pawn to h4, where it will be worse placed.} 30. h4 Ne6 {Black's 2 "lost" tempi do not matter at all, but now the pawn on h4 will be fixed on a dark square. In the event of some pawn trades, this pawn could become a weakness.} 31. Nb3 h5 32. Bd2 Bd6 33. c3 c5 {Magnus came up with the plan of transferring the king to the queenside. It's not a bad idea, but I had another one.} (33... g5!? {With this move, Black poses the question to white of what he wants to do about the pawn tension. The pawn on h4 is indeed misplaced on the dark square, and he has to either allow gxh4, which will leave him with a long-term targetable weakness (Nf8-g6, Ke6, f6-f5, Be7 is a good way for Black to attack it), or he has to take on g5 to allow Black an outside passed pawn. Neither option is particularly appetizing, but the position is so simplified that I think in either case, Black will not have nearly enough to win.} 34. Be3 {Probably allowing gxh4 is the lesser evil.} (34. hxg5 fxg5 35. Kf1 c5 {White should probably hold just by sitting, but Black has a real plan to play both h4 and b4, making passed pawns on opposite sides of the board.}) 34... gxh4 35. gxh4 Nf8 36. Na5 {White has enough counterplay to hang on}) 34. Be3 Ke7 35. Kd1 Kd7 36. Kc2 f5 37. Kd1 fxe4?! {After this it's absolutely drawn.} (37... f4 {Was a better try, though again I really doubt it would have changed the result.} 38. Bd2 (38. gxf4? exf4 39. Bc1 Be7 {Black wins, and we see a point of Magnus' previous play to provoke the pawn to h4.}) 38... g5! {White now cannot allow hxg4 followed by Be7.} 39. hxg5 Nxg5 {The pawn tension between g3 and f4 is in Black's favor since White never wants to take f4- it would leave Black with a deadly passed h-pawn. In addition, c5-c4 is a threat, undermining the e4 pawn, which cannot be defended by f2-f3, thanks to the previously mentioned pawn tension. Still, if White hangs tight, I don't see a great plan for Black, and the defense is pretty straightforward.} 40. Ke2 c4 41. Na1 Kc6 42. Nc2 {It's hard to imagine White losing.}) 38. dxe4 c4 39. Nd2 Nc5 40. Bxc5 Bxc5 41. Ke2! {The last important move, but also a very easy one. I could imagine Caruana blundering with f3 if he had almost no time on his clock (though I still think it would be extremely unlikely), but on move 41, this would never happen.} (41. f3? {Would weaken White's position too much, and indeed he loses after} 41... Bb6! {Since he is in zugwang.} 42. Ke2 (42. Kc2 Bf2! 43. Nf1 Be1! {And white is permanently paralyzed as both his knight and king are stuck defending pawns. Black will break through with King to c5, and then b4.}) (42. Nf1 Kc6 43. Ke2 b4! {Setting the pawns in motion.} 44. cxb4 Kb5) 42... Ba5 43. Nb1 Kc6 44. Kd2 Kc5 45. Kc2 b4 46. cxb4+ Bxb4 47. Nd2 Kd4 {Black wins.}) 41... Kc6 42. Nf1 b4 43. cxb4 Bxb4 44. Ne3 Kc5 45. f4 exf4 46. gxf4 Ba5 47. f5 gxf5 48. Nxc4 Kxc4 49. exf5 {I think Magnus played quite a bit better than Caruana today, but the position was so dry (and initially a little unpleasant for Black, albeit very defensible) that he never had any serious chances. Still, since he did not fully equalize, I am curious to see if he will repeat the Sicilian again in game 5.} 1/2-1/2 [Event "2018 World Chess Championship"] [Site "London"] [Date "2018.11.13"] [EventDate "2018.11.09"] [Round "4"] [White "Magnus Carlsen"] [Black "Fabiano Caruana"] [Result "1/2-1/2"] [ECO "A29"] [Annotator "chess.com"] [PlyCount "67"] [CurrentPosition "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1"] 1. c4 {Magnus decided not to repeat d4 from the first game. This might imply that his team was unable to find anything satisfactory against Caruana's Rd8 in the QGD. I think more likely is that Magnus wants to play different first moves with White to get a rough sense of what Caruana has prepared, and once he knows what he will face if he opens with d4, c4, e4, etc, he will know which area he feels most confident targeting.} 1... e5 2. Nc3 Nf6 3. Nf3 Nc6 4. g3 {The 4 knights English has been very topical as of late.} 4... d5 (4... Bb4 {Has been a hot trend, with Caruana taking both sides of this position at the recently concluded Olympiad. Undoubtedly Magnus was ready for it.}) 5. cxd5 Nxd5 6. Bg2 Bc5!? {This funky move definitely is not the mainline, but it's reasonably common and has been played by some strong players in the past couple years.} (6... Nb6 7. O-O Be7 8. a3 {Has been considered the mainline forever. The position is strategically very rich and there is a lot to discover still. Both sides have fair chances to outplay their opponent.}) 7. O-O O-O 8. d3 Re8 {Again, Caruana chooses a less common move.} (8... Bb6 {Is the mainline, though it has to be said there are not too many games played. The position after} 9. Nxd5 Qxd5 10. b4 e4 {Is complex, though Black seems to be holding his own.}) 9. Bd2 Nxc3 10. Bxc3 Nd4 11. b4 Bd6 12. Rb1! {I like this move- it seems very human. Caruana started thinking here, though he may have still been in preparation and just trying to remember. It's hard to imagine Black not realizing Rb1 is a serious try.} (12. Nxd4 {Obviously this was an option, but Magnus eschewed it, probably thinking it would fizzle out to a draw right away. Caruana was still playing instantly so he had to know that it would be prepared.} 12... exd4 13. Bxd4 Bxg3 14. hxg3 Qxd4 15. Qd2 {Optically the position looks nice for White as he can play e3 and d4, putting his pawns on dark squares to complement the light squares bishop, but the lightly weakened kingside does matter. After something like} 15... c6 16. e3 Qf6 17. d4 h5! {Black will get counterplay against White's king. It's very understandable that Magnus was not keen to enter such a position against an obviously prepared opponent.}) 12... Nxf3+ 13. Bxf3 a6 14. a4 c6 15. Re1 {I have very little to say about the game after this move. Magnus' only chance to imbalance the position was to go for b5.} (15. b5 cxb5 {I'm sure this is the move that Magnus was concerned about.} (15... axb5 {It may have been Caruana's plan to play a position like this one. He already showed in game 2 that he was happy to play a slightly uncomfortable endgame straight out of the opening if he is confident that he can hold it and knows how to put his pieces and how to neutralize the edge.} 16. axb5 Bd7 17. bxc6 Bxc6 18. Bxc6 bxc6 19. Qc2 Bf8 {Undeniably, White is slightly better due to the weak pawn on c6. But how realistic are his winning chances? Black will plant his queen on d5, where she will be the boss of the board, and the e2-e4 advance will make d3 a long term target. The open files on the queenside means Black's rooks will become active if White is unwilling to make some trades, and White has no obvious way to make a second weakness in Black's position. Of course White has the moral advantage, but I really think this should be a draw.}) 16. axb5 a5 {The passed pawn is a long term trump, but I think White is still slightly more comfortable. Following} 17. Qa4! {It's very hard to permanently secure the a5 pawn, and if the b7-b6 advance doesn't drop a rook at some moment, it will always weaken the c6 square. White has some pleasant pressure. Still, after something like} 17... Qe7 18. Rfc1 Bf5! {With e4 on the way, the position will rapidly simplify and I think a draw is still by far the most likely result.}) 15... Bd7 {Now the b4-b5 plan is prevented, and it is very hard to come up with a credible idea for either side to play for. Soon enough, the pieces were swapped.} 16. e3 Qf6 17. Be4 (17. Bg2 {Magnus could have kept more tension by not allowing Bf5, but he definitely is not better and still doesn't have a great plan.}) 17... Bf5 18. Qf3 Bxe4 19. Qxf6 gxf6 20. dxe4 b5! {The last important move of the game. Black does not allow White to play b4-b5, and he fixes the b4-pawn on a dark square. His fractured kingside hardly matters without queens, and once the Black king advances to g6 to prevent any Kf5 plans, nobody has anything left to play for.} 21. Red1 Bf8 22. axb5 axb5 23. Kg2 Red8 24. Rdc1 Kg7 25. Be1 Rdc8 26. Rc2 Ra4 27. Kf3 h5 28. Ke2 Kg6 29. h3 f5 30. exf5+ Kxf5 31. f3 Be7 32. e4+ Ke6 33. Bd2 Bd6 34. Rbc1 {It was hard to expect anything else after 20...b5. Caruana clearly solved his opening problems with Black today, but maybe Magnus could have put more pressure by playing b5. I'll be interested to see the next couple games after tomorrow's rest day. I have a funny feeling they will be more tense than the last few have been.} 1/2-1/2 [Event "2018 World Chess Championship"] [Site "London"] [Date "2018.11.15"] [EventDate "2018.11.09"] [Round "5"] [White "Fabiano Caruana"] [Black "Magnus Carlsen"] [Result "1/2-1/2"] [ECO "B31"] [Annotator "chess.com"] [PlyCount "66"] [CurrentPosition "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1"] {The fifth game produced the fifth draw, but unlike the last few encounters, this one was much more notable and interesting. It was the first time Magnus at least seemed to get caught in the opening to some extent, and the first time I wondered if he might actually get himself into trouble. Still, when pressured to find some difficult moves, he managed to keep his head above water without much trouble.} 1. e4 c5 2. Nf3 Nc6 {Magnus clearly is sticking with the Sicilian, but each game is getting tougher for him. I will be interested to see if he pivots in game 8 or if he sticks with it all the way through, because the trend is not going his direction.} 3. Bb5 g6 4. O-O {A new move for the match.} (4. Bxc6 {Carana chose to take on c6 in the previous two encounters.}) 4... Bg7 5. Re1 e5 (5... Nf6 {I would be interested to know what Caruana had in mind here. I would not be surprised if he was hoping to transpose back to the Bxc6 lines now that Black has committed to Nf6, and can't play the Qc7 stuff he tried previously.} 6. Bxc6 $5 dxc6 7. d3) 6. b4 $5 {An enterprising, aggressive, and very principled move. Black does not need to do much to equalize, and the c5 pawn promises him strong central control. If he could play Nge7 and get castled, he might even have the more comfortable game. But for the moment, he is behind on development, and White tries to break down the center and open the position before that happens.} (6. c3 $2 { Continuing quietly with a normal developing move will not do.} Nge7 7. d3 O-O { I already prefer Black. He has an easy plan to play d6, h6, and eventually f5, launching a kingside attack, and I can't think of much for White to do.}) 6... Nxb4 $1 (6... cxb4 {Is also possible, but after} 7. a3 $1 {White has excellent compensation in a weird Sicilian-Benko hybrid. If Black is too insistent about keeping his extra pawn, he can get into massive trouble:} bxa3 $2 (7... b3 $1 { Is to be preferred, leading to rough symmetry and good chances to equalize, although it still looks unpleasant to me.}) 8. Bxa3 Nge7 9. Bd6 $1 {Black is already more or less lost. His pieces have no good route to get into the game and Nc3-d5 is coming.}) 7. Bb2 {White is down a pawn but his lead in development is very dangerous, as the game shows.} a6 (7... Qb6 {This was also possible, but after} 8. Na3 $1 a6 9. Bf1 Nc6 10. Nc4 Qc7 11. c3 $1 {Black cannot prevent d2-d4, when the position opens up faster than he will enjoy. The machine thinks he is okay, but it looks terrifying for a human.}) (7... Qc7 {Is also too materialistic. After} 8. c3 $1 Nc6 9. d4 $1 {White's lead in development should soon become a big problem for Black when the center opens.}) 8. a3 $1 axb5 9. axb4 Rxa1 10. Bxa1 {Caruana was still blitzing his moves and clearly in preparation. This was the first moment in the match where I thought he might have chances to score a full point. Black's center is coming under serious fire and he is not too close to castling yet.} d6 (10... Ne7 11. bxc5 O-O {Getting castled as soon as possible comes at a heavy price. White wins a central pawn, and the game.} (11... d6 $1 {Transposes to the game}) 12. Bxe5) 11. bxc5 Ne7 $1 (11... dxc5 $2 {Of course it is senseless to lose the e5 pawn and fall further behind on development.} 12. Bxe5 {Black is in big trouble}) 12. Qe2 $5 {I'm not sure what this move was about.} (12. cxd6 {The computer is pretty insistent that this is the best try. after} Qxd6 13. d4 $1 exd4 14. Bxd4 O-O 15. Bxg7 Qxd1 16. Rxd1 Kxg7 {White has a slightly better endgame due to his better structure, but I think it should be holdable. Perhaps Caruana reached the same conclusion and chose to play the "objectively inferior" Qe2 (according to the machine) hoping to make use of the practical chance that Black was in a position where a mistake would be more costly.}) 12... b4 $1 { Material is equal, and Black cannot afford to lose a pawn. This must have been an unsettling moment for Magnus, but he found the best move.} 13. Qc4 {Caruana was still blitzing, but I think he must have confused his preparation or forgotten something. He started thinking quite a bit after Black's next move, which is the first choice of the machine! I can gaurantee that while preparing this line, if one reaches this position, they would certainly check Qa5 as it's the computer's recommendation and the most obvious move. So it is very surprising to me that Caruana would blitz up to here, and only then start thinking. I have a feeling he got confused or misremembered something, but one can never truly know what went on.} (13. Qb5+ Nc6 {Leads nowhere for White}) 13... Qa5 $1 {Simple and strong, but has to be well calculated. Black hits the a1 bishop, defends the b4 pawn to prepare for Bc8-e6, and pressures the c5-pawn. Play now takes on a very forced nature.} 14. cxd6 {Entirely forced.} ( 14. Bb2 $2 Be6 {White loses the c-pawn, and probably the game.} 15. Qe2 Qxc5) ( {Ed note: At press conference, Caruana mentioned that} 14. d4 {"Is ridiculous and probably losing for White" -- Caruana.} Qxa1 15. cxd6 Nc6 16. d5 {and it doesn't look like it should work, perhaps after} Na7 {but the computer actually likes White after} 17. Qxb4) 14... Be6 $1 {An important resource.} 15. Qc7 Qxc7 16. dxc7 Nc6 $1 {Black has a very healthy position in the longrun. He can play Kd7, take the c-pawn, bring a rook to the a-file, and enjoy the bishop pair. White's only chance is to try to make something of his lead in development before Black consolidates- the typical story of any Sicilian.} 17. c3 {This doesn't do much and Magnus is absolutely fine now.} (17. d4 $5 { Blasting open the center at all costs was worth a shot, but Black should equalize with precise play. The moves are very forcing and not that difficult.} exd4 18. Rd1 d3 $1 {Now things get messy, albeit briefly.} (18... Kd7 { Probably holds, but won't be fun.} 19. Nxd4 Nxd4 20. Bxd4 Bxd4 21. Rxd4+ Kxc7 22. Rxb4 Ra8 {With no weaknesses, active pices, and chances to play for a 4 vs 3 rook ending in the worse case scenario, I suspect Black would eventually hold this. But he will suffer a lot first.}) 19. Bxg7 dxc2 20. Rc1 Rg8 $5 {The machine gives a slightly psychotic line, which is fun, but also unnecessary.} ( 20... cxb1=Q {The human move equalizes easily.} 21. Rxb1 Rg8 $11) 21. Bd4 Nxd4 22. Nxd4 cxb1=Q 23. c8=Q+ $1 {This does not change the result, but a double queening is amusing.} (23. Rxb1 Kd7 {Is also a draw}) 23... Bxc8 24. Rxb1 Kd7 25. Rxb4 $11) (17. Rd1 $1 {This was probably the best chance to make trouble. Still, I think Black should hold.} Kd7 18. d4 Bg4 $6 (18... exd4 19. Nxd4 { Black is forced into the 18...Kd7 line of the 17. d4 variation, with an unpleasant but probably holdable position.}) 19. d5 (19. dxe5+ Kxc7 {Is better for Black, as he will regain the e-pawn with a lot of activity}) 19... Nd4 20. Nbd2 $1 Rc8 {The machine calls this equal but it looks a little unpleasant for Black to me. I'd worry about the protected passer on d5.} 21. Nc4 $1) 17... Kd7 18. cxb4 Ra8 19. Bc3 Kxc7 {White won the b4-pawn, but it will be a hard pawn to maintain. He certainly is not better as the short and long term weakness of the b4-pawn and the horrible knight on a1 provide Black with excellent compensation.} 20. d3 Kb6 (20... b5 $5 {Magnus could have tried fixing the pawn on b4 as a long term weakness, and eventually try to pick it off with some combination of Ra4 and/or f6 followed by Bf8. This would have given him some vague chances to play for an edge, but I don't think it would be enough to win.}) 21. Bd2 {From here on out, the game follows a very forced line that chops all the pieces off.} Rd8 22. Be3+ Kb5 23. Nc3+ Kxb4 24. Nd5+ Bxd5 25. exd5 Rxd5 26. Rb1+ Kc3 (26... Ka4 27. Rxb7 Rb5 {Is also a draw}) 27. Rxb7 Nd8 $1 28. Rc7+ Kxd3 {Black has won a pawn, but his minor pieces are stuck and passive, plus his king is a little vulnerable, so he has no serious winning chances.} 29. Kf1 h5 (29... e4 $2 {Setting the pawn majority in motion is not to be recommended!} 30. Ne1#) 30. h3 Ke4 {The simplest route to a draw. White should take the f7 pawn before Black gets the king far enough back to consolidate and hoping to push with 4 vs 3.} 31. Ng5+ Kf5 32. Nxf7 Nxf7 33. Rxf7+ Bf6 {An interesting game, and probably the first seriously unsettling one for Carlsen. I'll be interested to see if he repeats the Sicilian again in game 8, or if he pivots to something else, as he has thus far done in his two white games.} 1/2-1/2 [Event "2018 World Chess Championship"] [Site "London"] [Date "2018.11.16"] [EventDate "2018.11.09"] [Round "6"] [White "Magnus Carlsen"] [Black "Fabiano Caruana"] [Result "1/2-1/2"] [ECO "C42"] [Annotator "chess.com"] [PlyCount "160"] [BlackElo ""] { Yesterday was the first game where I thought Carlsen was under any real pressure, and today was the first day where he was in serious danger of losing. I don't know who will be less happy with today's turn of events. Much like game one, one player should not be encouraged to be lucky to survive with white, and the other player can't be happy about not taking one of what will probably be very few chances to score a full point. Beating 2800+ guys is tough! } 1.e4 { Magnus opts for 1.e4 for the first time. I wondered if Caruana would want to play the Petroff, sticking with his main and best opening but eschewing any surprise value. I mostly expected something else since he has at least so far ditched the QGA, but it looks like at least for now, he is happy to stick with his guns. } 1...e5 2.Nf3 Nf6 3.Nxe5 d6 4.Nd3 { This is a trendy new move as of 2018, but I can't imagine its the most testing try. } ( 4.Nf3 { Is the main move. There follows } 4...Nxe4 { And now both d4 and Nc3 are serious tries to fight for an advantage. Caruana has faced both these moves many times in the past couple years. } ) 4...Nxe4 5.Qe2 Qe7 6.Nf4 Nc6!? { Initiative a funny forced sequence, but both players were undoubtedly ready for it. } ( 6...Nf6 { Also looks fine. I have a hard time imagining White getting anywhere in these symmetrical endings. Still, boring as they appear, they can liven up. Caruana's game with Kramnik in the Candidates saw a quick queen trade as well, and a full point was passed back and forth more than once as the position got complicated and the players got low on time. } ) 7.Nd5 Nd4! 8.Nxe7 Nxe2 { The position is really weird, but funnily enough, it seems balanced. White cannot make use of the extra tempo with both knight ransacking the other side's back rank. } 9.Nd5 ( 9.Bxe2 Bxe7 { Is obviously equal } ) ( 9.Nxc8 Nxc1 { Leads nowhere for White either } ) 9...Nd4 10.Na3 ( 10.Nxc7+?! { Grabbing the rook leads to trouble } 10...Kd8 11.Nxa8? ( 11.Bd3! Nxf2 12.Kxf2 Kxc7 { Should be equal } ) 11...Nxc2+ 12.Kd1 Nxa1 { The loose f2 pawn will be White's undoing. } ) 10...Ne6 11.f3 N4c5 12.d4 Nd7 13.c3 { Both players were still playing fast. Black's knights have been booted all the way backwards from their prevously impressive posts on e2 and e4, but he retains a very solid position and doesn't have any weaknesses. } 13...c6 14.Nf4 Nb6 15.Bd3 d5 16.Nc2 { White's moral advantage of slightly better coordination is more or less meaningless. When I woke up and this position was on the board, I expected another fast draw, not to have to wait 5 hours until after lunch to write about the game! } 16...Bd6 17.Nxe6 Bxe6 18.Kf2 h5 19.h4 Nc8 20.Ne3 Ne7 21.g3 c5 22.Bc2?! { This move was the start of Carlsen's troubles. Bringing the bishop to b3 to pressure the d5 pawn might seem natural, but it was a decision he later likely regretted as the bishop became a target and didn't accomplish very much. } ( 22.Ng2 { Magnus foung this maneuvre later, and he should have done it here as well. White still retains a moral advantage that is very, very unlikely to yield a full point. } ) 22...O-O 23.Rd1 Rfd8 24.Ng2 { White makes the positional threat of dxc5 followed by Be3 to make an isolated d5-pawn. I was not at all surprised to see Caruana play cxd4 in response, but I found it a little strange he did not make the move immediately. } 24...cxd4! 25.cxd4 Rac8 26.Bb3 Nc6! { The knight is coming to a5, when is becomes clear the bishop on b3 is doing more harm than good for White's position. } 27.Bf4 Na5 { Now White has some problems to solve. } 28.Rdc1 { Very necessary. } ( 28.Bxd6? Nxb3! 29.axb3 Rc2+! { An important in-between move. Black takes the second rank and White is in trouble. } ) 28...Bb4 29.Bd1 Nc4 30.b3 Na3! { This was Black's only serious try. The knight will be good on b5, and he is hoping to use the a2-pawn as a long term weakness. } ( 30...Nd6 { I guess this could transpose, but it does allow Bxd6. } 31.Rxc8 ( 31.Bxd6!? { I am not convinced that the bishop pair matters here since White has no bad pieces. If his bishop was on e3 he would be much worse, but as is, it is on the right color complex and will not magically move to a dark square. With moves like Bf3 and Ne3 coming I can't imagine White having any serious issues. } ) 31...Rxc8 32.Rc1 ) 31.Rxc8 Rxc8 32.Rc1 Nb5 33.Rxc8+ Bxc8 34.Ne3 Nc3 35.Bc2 ( 35.a4 { In light of the problems White faced in the game, it may have been worth considering to play a4 to avoid the weakness on a2. He voluntarily gave up the light squared bishop later anyway. } 35...Nxd1+ 36.Nxd1 f6 { Black is slightly better and will press, but I have a hard time seeing him win. Ne3, Bf4, Kd3 looks very hard to break down. } ) 35...Ba3! { Preventing a2-a4. It will be very hard to hold onto this pawn. } 36.Bb8 a6 37.f4 Bd7 38.f5 Bc6 { Now Nxa2 is an honest to goodness threat. } 39.Bd1 { A sad necessity, also highlighting why I think a4 earlier on may have been worth considering. } 39...Bb2 ( 39...Nxd1+!? 40.Nxd1 { Leads to a similar ending to the one after 35.a4, but here the pawn is fixed on a2, and f5 is vulnerable too. I still think it should be a draw but White's defense will be tougher. } ) 40.Bxh5 Ne4+ ( 40...Nxa2!? { I would have certainly gone for this one, hoping to use the potential for an outside passed pawn. Caruana's choice is not bad of course. } ) 41.Kg2 Bxd4 42.Bf4 Bc5 { White's position should be defensible. What Magnus chose to do next made his life tougher, which is the first step towards losing the game. } 43.Bf3 Nd2 44.Bxd5!? ( 44.Nd1 { In hindsight, this may have been a better choice. I do think the game move should hold, but White's margin for error was very slim. } 44...Nxf3 45.Kxf3 d4+ 46.Ke2 Be4 47.g4 { White is worse but hanging in there. Nf2-d3 is a real plan. } 47...Bb1 48.Kd2! { The pawn is taboo. } 48...Bb4+ ( 48...Bxa2? 49.Kc2 ) 49.Kc1 Bd3 50.Nf2 Be2 51.g5 Be1 52.Ne4 Bxh4 53.Bg3 { This computer line shows White holding, but I can understand why Magnus would not be keen to enter it. If he was confident the piece down position was a draw, one cannot berate him for making that choice. } ) 44...Bxe3 45.Bxc6 Bxf4 46.Bxb7 Bd6 47.Bxa6 Ne4 48.g4 Ba3! { An important move, without which White would hold easily by playing a2-a4 and keeping the bishop defending b3. Now White must lose one of his queenside pawns. } 49.Bc4 Kf8 50.g5 Nc3 51.b4! { Magnus correctly chooses to leave himself with the pawn the furthest away from the kingside. } 51...Bxb4 52.Kf3 { This is a very difficult endgame to play, judge and assess, but I'll share my general impressions. The bishop on c4 lasering down on the f7-pawn means that sooner or later, Black will be forced to play f7-f6. Then, White probably will not want to take the pawn unless forced to- he still has a plan to bring the king to g6 to attack the base of the pawn chain, since he can play gxf6 anytime. If White takes on f6 and Black can recapture with the pawn, he will never be able to take Black's last one. } 52...Na4 53.Bb5 Nc5 54.a4 f6 { As previously mentioned, Black was forced to play f6. Now, I think his best winning plan is to try to force the g5/f6 pawn tension to resolve. If white ever plays g6, Black can blockade the a-pawn with his knight, bring the bishop to the c1-h6 diagonal to let the king join the fight, and he will probably win. White should never let this happen and if forced to resolve the tension he must take on f6, though this is also not ideal. } 55.Kg4 ( 55.g6? { A move like this illustrates my point nicely. } 55...Bd2 56.h5 Ke7 57.a5 Kd6 { Black will win the a5-pawn and the game. White needs to leave the kingside pawns as they were to keep some annoying counterplay with a potential passer that Black will need to pay attention to. } ) 55...Ne4! { Black ups the pressure on the g5-pawn. He really wants it to move. } 56.Kh5 { Now Kg6 is on the agenda. Black's king is clearly needed to defend the kingside, and if Kg6 comes next, it will be an easy draw. } 56...Be1! { Black freezes White's king by attacking the h4-pawn, preventing it from going to its desired destination. } ( 56...Bd2? { It would be foolish to only think about Black's plan. White gets the king to g6 and draws easily since g5 cannot be taken. } 57.Kg6! fxg5 58.hxg5 { And now Bxg5 would allow Bd3, while Nxg5 would allow f6. The game is drawn. } ) 57.Bd3 Nd6 58.a5! { I like Magnus' decision. The a-pawn was never a threat to queen and could only serve as a distraction for Black's pieces, so he takes his one and only chance to force Kg6 through, even at the cost of a pawn. } ( 58.Kg4 { The machine prefers to wait. I don't know if Black can win, but he has serious attempts. For instance, } 58...Nf7 { Is a good start. Black wants Bd2 and to force a resolution of the pawn tension, giving him one fewer weakness to worry about, the security that the f6-pawn will never be captured, and the g7-square for his king. } 59.Bc2 ( 59.Bc4? Ne5+ ) 59...Bd2! 60.gxf6 gxf6 61.Kf3 { This position might be a draw, but I am not as confident in this as I am about the one with the king on g6. } ) 58...Bxa5 59.gxf6! ( 59.Kg6? Bd8! { And White will not be left with a passer. After } 60.gxf6 Bxf6! 61.h5 Nf7! { Black prevents further trades, and with no passed pawn, there is no counterplay. Black will eventually win. } ) 59...gxf6 60.Kg6 Bd8 { At this point, I think Black will win if he can get his king to h8. Then Ne5+ would be a valid plan to kick out the White king. } 61.Kh7! ( 61.Bc2? Kg8 62.Bb1 { The knight will take its time to get to e5, but it does get there. } 62...Nf7? { Too hasty. } ( 62...Nb5 63.Bd3 Nd4 64.Bc4+ Kh8 65.Bd5 Ne2 66.Bc4 Nf4+ { f4 might be a fine square too. As long as Black can bring his king up to stop and ultimately capture the h-pawn, he should win. In order to stop this, White absolutely must keep the black king off of the g-file. } 67.Kf7 Kh7 ) 63.Ba2 ) 61...Nf7 62.Bc4 Ne5 63.Bd5 Ba5 64.h5 { White is not happy about his pawn being forced onto a light square, but there was nothing he could do about it anyway since Black could always play Be1 if he so desired. Still, I think this position is just a draw. White leaves his king on h7 until the end of time. If the Black king ever leaves, Kg7 followed by h5-h6 becomes a serious idea. And the only way to dislodge the king is to try to bring the knight to g5. As long as White's bishop is on the a2-g8 diagonal at that time to avoid the coming Kg8, I cannot see a plan for Black. } 64...Bd2 65.Ba2 Nf3 66.Bd5 Nd4 ( 66...Ng5+ 67.Kg6 Bc3 { As I mentioned, this gets Black nowhere since the king cannot come to g8. h6 draws on the spot, but even if White sits with Ba2, there isn't much for Black to do. As soon as the knight moves, Kh7 comes again. } ) 67.Kg6? { Technically this move does lose in a higher sense, and the monster machines point out mate in 64 or something like this. It is well outside of human capability for Black to prove it, even for a 2800+ player. Still, I think it was the wrong idea to leave h7 on simple grounds. IT was the best square for White's king and should not leave unless forced to. } 67...Bg5 68.Bc4 Nf3 { I would not feel right awarding this move a question mark since the win was some computer nonsense that was clearly impossible for a being with flesh and bones. } ( 68...Bh4! { According to the monster, this was the only winning move. The point is that White is in a deadly zugzwang. If he ever advances h6, the pawn eventually gets captured. If the bishop leaves the a2-a8 diagonal, Kg8 comes. And if the bishop leaves the a6-f1 diagonal, Ne2-f4 is winning. } 69.Bd5! { The most resilient try. } ( 69.h6 Nf3! { White loses after Ne5+ } 70.Ba2 ( 70.h7 Ne5+ 71.Kh5 Kg7 72.Ba2 Bg5 ) 70...Ne5+ 71.Kh5 Bg5 72.Bb3 Nd3 { Nf4 comes next. The pawn will be lost. } ) 69...Ne2 70.Bf3 ( 70.Kh7 Bg5! { Somehow White cannot hold the pawns. } ) 70...Ng1! { Another impossible only move. If Caruana could make it here by some miracle I'd give him good chances to find the rest of his way, but Bh4 and Ne2-g1 both feel completely outside of human capability. } 71.Bd5 ( 71.Bg4 Kg8! { Black got his king to g8, and will be able to put White in zugzwang to get the knight out. } 72.Kh6 Bf2 73.Kg6 Bd4 74.Kh6 Be3+ 75.Kg6 Bg5! { White must either play h6 or let the knight escape. } 76.h6 Kh8 77.h7 Bh4 78.Kh5 Be1 79.Kg6 Bc3 80.Kh6 Bd2+! 81.Kg6 Bg5! { The same Bg5 makes a second zugzwang. The knight gets out. } ) 71...Bg5! { And White is un zugzwang again. h6 will always lose, but he also cannot allow the bishop to leave d5. } 72.Kh7 ( 72.h6 Nh3 ) ( 72.Ba2 Nh3 { Nf4 will win the pawn } ) 72...Ne2! { And White's bishop runs out of squares. } 73.Ba2 ( 73.Bb3 Nd4! { Wins the f5-pawn } ) ( 73.Be6 Nf4 ) ( 73.Bc4 Nf4! ) 73...Nf4! { Forces h5-h6. } ) 69.Kh7 { It's a technical draw again. } 69...Ne5 70.Bb3 Ng4 71.Bc4 Ne3 72.Bd3 Ng4 73.Bc4 Nh6 74.Kg6 Ke7 { Black can bring the king, but f5 is impossible to capture without letting the h-pawn run. } 75.Bb3 Kd6 76.Bc2 Ke5 77.Bd3 Kf4 78.Bc2 Ng4 79.Bb3 Ne3 80.h6 Bxh6 { The ending was fascinating to study, though Magnus's only serious danger came from an impossible computer line. Still, he did come much closer to losing today than he had in any other game yet. Now he has tried 3 different first moves in his first 3 white games. He will play white again in game 7, and my guess is we will see him repeat a previous choice for the first time, hoping to come with better analysis now that he has more of an idea what to prepare for. The second half of the match should produce some more interesting games as the pressure increases and there is more to play for in every game. } 1/2-1/2 [Event "2018 World Chess Championship"] [Site "London"] [Date "2018.11.18"] [EventDate "2018.11.09"] [Round "7"] [White "Magnus Carlsen"] [Black "Fabiano Caruana"] [Result "1/2-1/2"] [ECO "D37"] [Annotator "chess.com"] [PlyCount "79"] [CurrentPosition "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1"] 1. d4 {As I predicted, Carlsen goes back into a line that he had seen previously, and came with a hopeful improvement in mind. Eventually one runs out of good first moves to try!} 1... Nf6 2. Nf3 d5 3. c4 e6 4. Nc3 Be7 5. Bf4 O-O 6. e3 c5 7. dxc5 Bxc5 8. Qc2 Nc6 9. a3 Qa5 10. Nd2 (10. Rd1 {Carlsen chose this in game 2. Caruana countered with the very unusual} 10... Rd8 {Which Magnus presumably did not find anything good against in the days between games 2 and 7. Otherwise he likely would have tried this again.}) 10... Qd8!? {Another extremely rare move. Caruana definitely was well prepared with this c5 line.} (10... Be7 {and}) (10... Bb4 {Are the mainlines. I had marked the latter as the best move in my own analysis, though it was old work and could definitely be wrong.}) 11. Nb3 Bb6 12. Be2 (12. Rd1 {I thought White might want to make a useful developing move and hope to recapture on c4 in one move- this is hardly an uncommon occurrence in Queen's Gambit positions. But Black does seem fine after the accurate} 12... Qe7! {Pointing out that White cannot comfortably take on d5.} 13. cxd5? (13. Be2 dxc4 14. Bxc4 Rd8 {Would be prudent, but I think Black is easily equalizing.}) 13... Nxd5! 14. Nxd5 exd5 {The pawn is taboo as White is too far behind on development (in fact he is already worse.)} 15. Rxd5? Be6 16. Rd3 Rac8 {The goons are coming and White will be crushed long before he can get castled.}) 12... Qe7 13. Bg5! {The most challenging move. Black now has to be pretty accurate to maintain a tenable position, but it's not such a difficult task and Caruana had no problem managing.} (13. cxd5 Nxd5 14. Nxd5 exd5 15. O-O d4 {Would promptly lead to a draw. Magnus' choice is more ambitious.}) 13... dxc4! {This will allow Black's kingside pawns to be compromised, but it turns out not to be an issue as he will get a lot of counterplay.} (13... h6? 14. cxd5! {Favors White.}) 14. Nd2 (14. Bxc4 Bd7 {Is nothing for White. His pieces don't coordinate very well.}) 14... Ne5! {Basically the last important move of the game. Black protects the c4 pawn so that White will have to make some concessions to get it back if he wishes to compromise the black kingside.} (14... Bd7? {An innocent move like this one could lead to some trouble after} 15. Nxc4 Bc7 16. Ne4 {When Black cannot prevent his kingside pawn cover from being disturbed.}) 15. O-O (15. Nce4 {A move like this would definitely mess up the Black kingside, but it comes at a heavy price as c4 will be hard to recapture. Following} 15... Bd7 16. Nxc4 {White has taken the c4 pawn back and if he could consolidate and chop on f6, he'd have some chances. But after} (16. Nxf6+? gxf6 17. Bh6 Rfc8 {Favors Black, as White will not be able to take back on c4 and I don't think his attacking chances are so great.}) 16... Nxc4 17. Bxc4 Rfc8! {Black has enough activity to hold the balance.} 18. O-O Bc6! 19. Nxf6+ gxf6 20. Bh4 {If White had one move to consolidate he would be somewhat more comfortable, but after} 20... Bd5! 21. b3 Bxc4 22. bxc4 {White is left with pawn weaknesses of his own, and the position is balanced.}) 15... Bd7 16. Bf4 Ng6 17. Bg3 Bc6 18. Nxc4 Bc7 {Someone more imaginative than me might find something interesting to say about the rest of this game, but I certainly can't. It's a tribute to the 30-move rule that the players did not shake hands right here.} 19. Rfd1 Rfd8 20. Rxd8+ Rxd8 21. Rd1 Rxd1+ 22. Qxd1 Nd5 23. Qd4 Nxc3 24. Qxc3 Bxg3 25. hxg3 Qd7 26. Bd3 b6 27. f3 Bb7 28. Bxg6 hxg6 29. e4 Qc7 30. e5 Qc5+ 31. Kh2 Ba6 32. Nd6 Qxc3 33. bxc3 f6 34. f4 Kf8 35. Kg1 Ke7 36. Kf2 Kd7 37. Ke3 Bf1 38. Kf2 Ba6 39. Ke3 Bf1 40. Kf2 {I'll be interested to see what happens tomorrow, and if we get a fourth Rossolimo or not.} 1/2-1/2 [Event "2018 World Chess Championship"] [Site "London"] [Date "2018.11.19"] [EventDate "2018.11.09"] [Round "8"] [White "Fabiano Caruana"] [Black "Magnus Carlsen"] [Result "1/2-1/2"] [ECO "B33"] [Annotator "chess.com"] [PlyCount "75"] [CurrentPosition "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1"] 1. e4 c5 2. Nf3 Nc6 {Magnus once again sticks with his Sveshnikov/Rossolimo repertoire, though he really got away with one today. It will be interesting to see if he decides to keep going with it even though his positions out of the opening are getting more and more dangerous.} 3. d4 {Caruana enters the mainline. With the Sveshnikov being implied very heavily in game 1, his team has had a lot of time to come up with ideas.} (3. Bb5 {Was Caruana's choice in his first 3 white games. I think he took the right choice to go into the riskier open Sicilian mainlines, as his best chance to win the match is in regulation. I also think the positions suit his style better than Magnus's style.}) 3... cxd4 4. Nxd4 Nf6 5. Nc3 e5 6. Ndb5 d6 7. Nd5!? {This is a very forcing move that leads to a defined and unbalanced pawn structure right away.} (7. Bg5 {Would lead to the Sveshnikov mainlines.}) 7... Nxd5 8. exd5 Nb8 9. a4! {I always felt running the a-pawn the most critical try in pawn structures like this one. More often than not it comes from a Najdorf with Be3 and an early Nd5, but White will definitely appreciate leaving the c-pawn at home in certain positions.} 9... Be7 10. Be2 O-O 11. O-O Nd7 {This position is not nearly as common as the Sveshnikov mainlines, but it's still reasonably well known and there have been plenty of games played. Caruana's next move was the one that got Magnus thinking.} 12. Bd2!? {A rare choice, but a very logical move. White prioritizes developing his pieces- what could be more natural. In a lot of cases, White will play f4, and the bishop will stand well on c3 to pressure Black's center. This certainly turned out to be the case during the game.} (12. Kh1 {For better or for worse, this has been the most common move in the past. It doesn't make much sense to me. I can't imagine avoiding Qb6 coming with check is worth a tempo. Caruana's move looks much stronger.}) 12... f5 (12... a6 {In light of how the game proceeded, Magnus could have considered stopping White from playing a5. First he can push the knight back, and then after} 13. Na3 a5!? {The position doesn't seem so bad to me. The machines like White but I don't see a great plan and Black's pieces have good squares to go to. It has to be said that this is one of the positions where it is very clear that remaining flexible with the c-pawn was in White's favor, as if the pawn were on c4, he would have no good way to open the queenside or deal with a knight sitting on c5 for the rest of the game. A sample line could go} 14. f4 f5 (14... exf4!? 15. Bxf4 Bf6 16. c3 Be5 {Also looks fine for Black to my unrefined and unintelligent human eye, but the machine disagrees.}) 15. Nc4 (15. c3 Bf6) 15... Bf6 {Black has decent counterplay and I do not mind his position at all. It would be interesting to know what Caruana had in mind had Magnus tried a6-a5.}) 13. a5 a6 {Not a bad move I guess, but it looks wrong to me. The knight is not particularly good on b5 and likely would want to reroute to a3 and c4 anyway, and Magnus now forces this to happen with a loss of tempo and weakens the b6-square at the same time. Magnus's idea was to play e4 without allowing Nd4, but still, I dislike the move.} (13... f4!? {It makes a lot of sense to prevent White from playing f2-f4 to contest the center and the kingside. The position is locked and Black looks like he should have decent attacking chances, for instance after} 14. Bb4 Nf6 15. c4 a6 16. Nc3 Bf5 {The machines like White but g5-g4 is a real plan. Black is not without counterplay}) 14. Na3 e4 15. Nc4 Ne5 {Magnus was clearly aiming for this position when he played a6. But it seems unpleasant for Black.} 16. Nb6! {Caruana was still in preparation and playing very quickly.} (16. Nxe5? dxe5 {Would give Black a lot of counterplay. I actually prefer his position.}) 16... Rb8 17. f4! {Black must not be allowed to play f4 himself} 17... exf3 18. Bxf3 g5? {This was asking too much. White is breaking through on the queenside too fast to expect the kingside attack to work.} (18... Bf6 {A solid move like this one looks fine to me.} 19. c3 Nd7! 20. Be3 Be5! {Black's pieces have found good squares. The machine slightly prefers White but I don't think Black has much to worry about.}) 19. c4! {Bc3 followed by c5 is a very easy plan to execute. The loose position of the e5 knight means White will not even need to defend the c5 pawn to get the advance through, and if Nxf3 ever comes, Black may well regret advancing g7-g5 with the open long diagonal. Black is in big trouble.} 19... f4 20. Bc3! {Another strong move. Apparently Sesse was claiming that White is absolutely winning here- this is far too optimistic an assessment in human play, there is a lot of potential counterplay to contain, much like one sees in a mainline King's Indian. But his position is definitely very promising.} 20... Bf5! {Magnus chooses the combattive move. He understood his position was "dangerous" though I would go a step further and just call it "bad". But he handled it absolutely the way he needed to- in a combattive manner looking for counterplay, as is the general plan in the Sveshnikov.} (20... Qc7? {Trying to prevent c5 is only a temporary fix. Black has a lost position on the queenside as c5 will eventually win the game. He needs to look for kingside counterplay.} 21. b4 Bf6 22. Rc1 {White should be winning. Even a simple plan like Bxe5 followed by Nxc8 leaves Black unable to get g5-g4 through or make any kind of counterplay while c5-c6 comes very fast.}) 21. c5 {Caruana found the best move. His position is close to winning technically, it's more his kind of game, and he had a huge time advantage. This was his first real chance to score a full point after only missing a completely impossible line in game 6, and he did not manage to come through. If he loses the match, he will likely look at this game as the turning point.} 21... Nxf3+ 22. Qxf3 dxc5 23. Rad1 Bd6 24. h3?? {I wavered between offering this move one question mark and two, eventually settling on the harsher evaluation, even though I normally only offer it for a move that squanders a full point rather than a half point. Time is absolutely of the essence. White is a pawn down and facing the bishop pair with an imbalanced structure. His position is strong, but he absolutely needed to look for the most direct moves. Magnus is now instantly fine after one slow move that was not at all in the spirit of the position, and he even said so himself after the game.} (24. Nc4! {Of course this move needed to be played. The idea is totally obvious and White did play it on the next move, only he needed to do so faster. Black cannot allow d5-d6 to come through so eventually his dark squared bishop will be exchanged, and then the attack is decisive.} 24... g4 {Undoubtedly Caruana was worried about this move, otherwise h3 would really be inexplicable. But White is faster.} (24... Qd7 {Black cannot simply move the queen to bring the sleeping b8-rook into the game on e8 the way he did after White played h3, because the queen lacks access to the g6 square- the only way to hold both h5 and d6. After} 25. Qh5 {Black can resign.}) 25. Qf2 f3 (25... g3 26. Qd2! gxh2+ 27. Kh1 {Is hopeless for Black. Rxf4 is a big threat, and} 27... Bg6 28. Rde1! {Offers no salvation as Re6 will come next.}) 26. Rfe1! {Black is hard pressed to deal with Nxd6 followed by Be5. He should lose.}) ({Also possible was} 24. Qh5 Bg6 25. Qh6 Rf7 26. Nc4 Bf8 27. Qh3 {as was given by Carlsen at the press conference. "It was very hard for me to evaluate but it felt like White is having all the fun… This felt like a dangerous moment to me."}) 24... Qe8! {The move came instantly. The queen will be perfect on g6, preventing Qh5, defending the d6 bishop, and preparing for Rbe8 to context the open e-file. Black is not absolutely fine and can even take over the game if White is not careful- he is a pawn up with the bishop pair after all.} 25. Nc4 Qg6 26. Nxd6 {Caruana messed up in a big way to let Black hold, but he was also professional and objective enough to make a draw when it was time.} (26. Rfe1 {Trying for a win at all costs was possible, but after} 26... Rbe8 (26... g4!? {Looks decent too. h3 did not stop anything and now even serves as a hook}) 27. Be5 Bxe5 28. Nxe5 Qg7 29. Qc3 g4! {Lots of counterplay}) 26... Qxd6 27. h4 gxh4! {The simplest. White must trade queens and then Black doesn't have a care in the world.} 28. Qxf4 Qxf4 29. Rxf4 h5 30. Re1 Bg4 31. Rf6 Rxf6 32. Bxf6 Kf7 33. Bxh4 Re8 34. Rf1+ Kg8 35. Rf6 Re2 36. Rg6+ Kf8 37. d6 Rd2 38. Rg5 {Caruana has had some good moments and been able to apply pressure here and there, and had that one impossible engine win in the piece up ending. But today was the only game thus far where he absolutely had it within his power to come home with a full point. Magnus does not give many people chances like this, and he might not give another one. Still, while Caruana can't be feeling too happy (or confident) that he missed his opportunity today, he is clearly outpreparing Magnus every game with both colors for several games in a row. If Magnus loses the match, I suspect it will be due to another opening mishap. I'm sure he notices this trend himself and I expect him to switch things up a bit in the final four games.} 1/2-1/2 [Event "2018 World Chess Championship"] [Site "London"] [Date "2018.11.21"] [EventDate "2018.11.09"] [Round "9"] [White "Magnus Carlsen"] [Black "Fabiano Caruana"] [Result "1/2-1/2"] [ECO "A29"] [Annotator "chess.com"] [PlyCount "112"] [CurrentPosition "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1"] {Game nine was likely a very frustrating one for Magnus. He came up completely dry out of the opening in all four of his first four white games, and then when he finally got a really nice position from what looked like a throwaway practical try, one careless moment immediately squandered all of his winning chances.} 1. c4 e5 2. Nc3 Nf6 3. Nf3 Nc6 4. g3 d5 5. cxd5 Nxd5 6. Bg2 Bc5 7. O-O O-O 8. d3 Re8 {Thus far, the game has followed the path of game 4.} 9. Bg5!? {A new idea. As far as I can see in my database, it has only been essayed once before, and by two low-rated players.} (9. Bd2 {Was Magnus' choice in the aformentioned game. He might have had a little chance to get b5 in, but in the game, Caruana held very easily. Clearly he was dissatisfied with his choice then and wanted something fresher.}) 9... Nxc3 10. bxc3 f6 {Caruana was still playing quickly here, implying he was in preparation. But soon enough, Magnus' idea was revealed.} 11. Bc1! {Absolutely the most challenging move. White has a very simple plan of playing Bb2 followed by d4 to take a large share of the center. Black is definitely fine but will have to make hard strategic decisions over the board -- exactly the kind of thing Magnus must have been hoping for. Soon enough, Caruana faltered.} 11... Be6 12. Bb2 Bb6 {I feel wrong criticizing this move as it is approved by the machine, but on a human level, I dislike it. d4 is clearly coming and cannot be prevented, and the bishop will be passive on b6.} (12... Bf8 {My instinct would be to tuck the bishop back on f8, a typical maneuver once 0-0 and Re8 has been played. I certainly don't think Black is worse.}) 13. d4 Bd5 (13... Na5 {Might be worth considering. Any Grunfeld player knows to abuse the c4 square in these structures. Still, it's of course easy to sacrifice other people's pawns and one cannot really fault Caruana for not wanting to allow dxe5 against a clearly prepared opponent.}) 14. Qc2 {White is not so subtly advertising that he wants to take the whole center with e2-e4. Black is definitely in some strategic danger as if White can achieve this goal, he should have a huge position.} 14... exd4!? {True to his style, Caruana goes with the most forcing continuation. But Carlsen was ready.} (14... e4 {Was also possible. The position is very unclear after} 15. Nh4 Na5 16. Nf5) 15. cxd4 Be4?! {Again, I feel a little odd criticizing a move the computer takes no issue with. But practically, it does put Black in a tough spot in short order.} (15... Qd6 {A quiet move like this one would not be too great. After} 16. e3! {It's hard to imagine how Black will get his b6-bishop back into the game. White can plan Rfc1, Ne1-d3, and he looks pleasantly better to me.}) (15... Nb4! {I think Black's best option was to try to force through c7-c5 at all costs. I really dislike passive pieces and buried bishops.} 16. Qa4 c5! 17. dxc5 Bxc5 18. a3 (18. Rad1 Bc6! $11 {Should equalize easily}) 18... Bc6! {An important move.} 19. Qb3+ Qd5! 20. Qxd5+ Nxd5 21. e3 {White has the somewhat more pleasant position, and this variation would be hard for a human to find. But Black has no bad pieces, which I think he should have placed a higher priority on.}) 16. Qb3+! {Before coming back to d1, Magnus provokes the bishop to d5 to deprive the queen of this square.} (16. Qd1 Qd5! {Gives Black a lot of counterplay with good pressure on the light squares}) 16... Bd5 (16... Qd5 {Trading queens would also fail to reach the desired goal} 17. Qxd5+ Bxd5 18. e3 {Black has no good way to solve the problem of the b6 bishop. He looks quite a bit worse to me. Rfc1 and Ne1 or Nd2 should leave White with a great position.}) 17. Qd1! {No early repetitions today. Black's position is far from critical, but one can see why Caruana felt the need to lash out immediately. If White is given the time to play e3 and bolster the d4 pawn, his play flows very naturally and Black doesn't have much counterplay.} 17... Bxf3!? {The machines hate this move, but from a human point of view, I think I might actually be a decent decision. The alternatives look quite unpleasant and this leads to position I do believe to be defensible.} (17... Qd6 {Trusty Stockfish suggests this is a decent try, among other do-nothing moves like Qe7 or Qd7 [Caruana mentioned 17...Qe7 and also the more active 17...Na5 as alternatives]. But simple development may not solve all of Black's problems. For instance, after} 18. e3 Rad8 19. Bc3 {Black has a very difficult position. His rooks on the open central files look nice, but lack targets. White can easily develop with Qc2, Rfc1, Nd2, etc and he can play against the b6 bishop and the potentially weak queenside pawns. The machine says that White's edge is much smaller than the opposite colored bishops position Caruana chose instead, but I am not at all sure that I agree.}) 18. Qb3+ Kh8 19. Bxf3 Nxd4 (19... Bxd4? 20. Rad1 {White wins immediately}) 20. Bxd4 Qxd4 (20... Bxd4 21. Rad1 {Again, the pin is decisive- Black cannot stop e2-e3.}) 21. e3 Qe5 22. Bxb7 Rad8 {White has a very pleasant position. His four-versus-three pawn majority on the kingside means his king will be much safer. This is very important in opposite colored bishop positions, and Magnus correctly assessed that his plan should be to open the black king. Conversely, Black's extra c-pawn on the queenside is of no consequence, also due to the opposite colored bishops. Coupled with the b6-bishop biting on granite and White's bishop having excellent light squares to work with, it seems like White would have good chances to play a long, grinding game and try to apply pressure for long enough for Black to crack. This is exactly the kind of chess Magnus excels at, and Caruana had a lot less time. I'm sure Magnus will not be pleased with himself for letting an opportunity like this slip, though it has to be said that the position is not technically won or anything like that.} 23. Rad1 Qe7 24. h4 {Of course. White plans to set his kingside majority in motion and try to open up the black monarch, relying on opposite colored bishops to increase his attacking chances. [Note that Carlsen criticized this move, instead preferring Bf3 and maybe Kg2.]} 24... g6 25. h5? {A bad move, probably based on a miscalculation. [Carlsen said if he didn't play h5 himself, he thought that Black's ...h5 and ...f5 and then "I don't really see how to make progress. It's just too solid.") White surely wanted to get h5 in, but he could not rush as now when Black takes, he is ready for h5-h4 very fast, opening up White's king as well. I think Magnus probably played h5 right away without bothering to prepare it because he was worried that Black would beat him to the punch and play h7-h5 himself, but in this case, the cure was worse than the disease.} (25. Kg2 {A slow buildup move like this one would ensure Black would be suffering for a long time. My guess is Magnus disliked that after} 25... h5! {He no longer has an easy plan to open the kingside, and Black can play f6-f5, then sit with Qf6 and Kg7. The position looks remarkably solid.} 26. Bc6 Rf8 27. Qc2 f5 28. Bf3 Qf6 29. a4 Kg7 {I think something like this was the best either side could hope for. Black's position is surely unpleasant and it's right up Magnus's alley to grind out edges like this, but at the same time he is extremely solid and it's hard to see a great plan for White. Note that the presence of the pawns on h4 and h5 (as opposed to if they were absent via an early h4-h5-hxg6, as Magnus cleary desired) means that White does not have the g3-g4 break at his disposal to open the up the black king. Still, I could see White eventually preparing f2-f4 and then e3-e4. It surely would take a long time to pull off, but Black is just sitting and waiting. Maybe it could be done? We will never know.}) 25... gxh5! {A very important resource. Caruana would have really struggled without this move.[Caruana called it a easy choice: "It wasn't such a difficult decision because…White's king becomes vulnerable as well."] Black butchers his own structure and will likely lose if the h5 pawn were to fall. But concrete calculation shows that h5-h4 will happen, helped in large part by the f6-f5 advance,} (25... Kg7 {If Black were to sit and hang tight, he would face a lot of issues. Following} 26. Bf3! {Black has lost his chance to play gxh5, and he can't even get f5 in due to the presence of h5-h6 at some inopportune moment. In addition, even if he eventually gets f5 in, White can play g3-g4 at the right time to open the king further. Black faces a long and very possibly unsuccessful defense, and is really regretting not being able to play h7-h5. A line like this illustrates why Magnus played h4-h5 himself so quickly so as to avoid h7-h5 by Black, unfortunately, he was one tempo too slow to make it work.} 26... f5) 26. Qc4 {Carlsen played this quickly, suggesting it was his plan in case Black were to take on h5. White prevents h5-h4 and is ready to scoop the pawn up next, but one more accurate move leaves Black in fine shape.} (26. Kg2? {Is a tempo too slow. If Black were to waste a move and allow Rf1-h1, he would be in awful shape. As is, he is immediately fine after} 26... h4! (26... Kg7? 27. Rh1 {Is an illustration of my point. White should win. But a tempo can mean a lot...})) 26... f5! {This looks scary at first glance, but the black king is still totally safe. The queen can easily come to g7 or f6 as needed and there will be no way to harass him on h8, and in the meantime, h5-h4 now cannot be stopped. As soon as the g3 pawn comes to h4, White will no longer have any winning chances as his own king will be open, and Black will have the plan of ...f5-f4 to blast open the b6 bishop, who has been very passive throughout the entire game.} 27. Bf3 h4! {A strong and important move, but also a very easy one. Black is fine.} (27... Kg7? {Again, if Black were to fall asleep at the wheel for one move, he would be easily losing when White grabs the h5-pawn.} 28. Bxh5) 28. Rxd8 Rxd8 29. gxh4 Rg8+ 30. Kh1 Qf6 31. Qf4 Bc5! {The dumb bishop will find a nicer square on d6.} 32. Rg1 Rxg1+ 33. Kxg1 Bd6 34. Qa4 f4 {Everything liquidates in short order.} 35. Qxa7 fxe3 36. Qxe3 Qxh4 37. a4 Qf6 38. Bd1 Qe5 39. Qxe5+ Bxe5 {White can claim the moral advantage due to his a-pawn, but with opposite bishops and even material, the position is still a very easy draw for Black.} 40. a5 Kg7 41. a6 Bd4 42. Kg2 Kf6 43. f4 Bb6 44. Kf3 h6 45. Ke4 Ba7 46. Bg4 Bg1 47. Kd5 Bb6 48. Kc6 {Black has to be somewhat careful here, but it's not particularly difficult. He just cannot let the c7 pawn fall, or for White to get Kb7 and a7 in with tempo lost. [Carlsen said he had been hoping to induce ...h5 to at least have a target.]} 48... Be3 {Simple enough. c7 cannot be taken.} (48... Bf2? {For instance, this would lose. After} 49. Kxc7 {White will go Kb7 next, play a7, win the bishop, and bring the king back. Of course, Caruana would fall for something this basic sometime after hell freezes over.}) 49. Kb7 (49. Kxc7 Bxf4+! {Check! No time for a7.}) 49... Bb6 50. Bh3 (50. a7 Bxa7 51. Kxa7 {Black only draws by one tempo here, but it is a very comfortable draw. White's problem is his bishop needs to perform three tasks: stop the h-pawn, stop the c-pawn, and control f5. It can pull off only two of these three, but not all. 3.} 51... c5! (51... Kg6? {If Black were to waste a tempo he would lose, as now White's king can stop the c-pawn and the bishop takes care of the f and h-pawns.} 52. Ka6 c5 53. Kb5) 52. Kb6 c4 53. Kc5 c3 {Sooner or later the bishop had to leave.} 54. Bd1 Kf5) 50... Be3 51. Kc6 Bb6 52. Kd5 Ba7 53. Ke4 Bb6 54. Bf1 Ke6 55. Bc4+ Kf6 56. Bd3 Ke6 1/2-1/2 [Event "2018 World Chess Championship"] [Site "London"] [Date "2018.11.22"] [EventDate "2018.11.09"] [Round "10"] [White "Fabiano Caruana"] [Black "Magnus Carlsen"] [Result "1/2-1/2"] [ECO "B33"] [WhiteElo "2832"] [BlackElo "2835"] [Annotator "chess.com"] [PlyCount "107"] [CurrentPosition "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1"] 1. e4 c5 2. Nf3 Nc6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 e5 6. Ndb5 d6 7. Nd5 {I didn't really expect a repeat of this line, even though Fabiano had achieved a great deal with it, simply because I thought it was mostly chosen because of the surprise value.} 7... Nxd5 8. exd5 Nb8 9. a4 Be7 10. Be2 O-O 11. O-O Nd7 12. b4 {We have an early deviation from Fabiano. It is very direct but at the same time it seems a bit strange to me to allow so happily the direct counter with a5.} (12. Bd2 {was seen in game 8. But no doubt Magnus was this time not only prepared for that move but also remembered the nuances!}) 12... a6 (12... a5 {is also fairly natural because the Nb5 could end up being misplaced in the long run but I'm guessing Magnus didn't like too much the possible pressure on the d6 pawn after say} 13. bxa5 Rxa5 14. Ba3) 13. Na3 a5 14. bxa5 ({I had wrongly assumed that} 14. c3 {was White's idea but it does look a bit slow. Black can just answer with} 14... f5 {as} 15. Nc4 b6 {and Black's queenside is very safe.}) 14... Rxa5 15. Nc4 Ra8 16. Be3 f5 17. a5 {Fabiano goes for the b6 square, and that idea was undoubtedly part of his preparation.} 17... f4 18. Bb6 Qe8 {Magnus played the most natural moves and now surprisingly Fabiano thought for a while, although probably not because he was out of preparation but more likely to try and see the position with a fresh eye with so many options available to him. Although from my point of view, it seems like White can run into so much trouble with his king that entering this line without having a very clear idea of what happens next seems very dangerous.} 19. Ra3 {This relocating of the rook via the 3rd rank is extremely natural but at the same time I realized I didn't see the upside of going into this sort position as it seemed to be neither safe nor objectively better or easier to play.} (19. Bc7 {would have been the direct approach. But Black can answer with} 19... Rf6 20. Re1 Rh6) (19. Re1 {was possibly a more subtle waiting move. If Black goes like in the game} 19... Qg6 (19... Rf6 {would still be my preferred choice}) 20. Bc7 {he doesn't have the same resource as in the game! Still his position is reasonable after a move like} 20... Nc5 21. Bh5 Qf6 22. Nb6 Ra6 23. Nxc8 Rxc8 24. Bb6 {but there at least White got rid of the light-squared bishop. I would even consider something like} 24... Bd8 25. Be2 Bxb6!? 26. Bxa6 Bxa5 27. Rxa5 bxa6 {with Black and hope to have enough compensations for the exchange thanks to the Nc5.}) 19... Qg6 20. Bc7 {White has to play this one day or another as he needs to restrain the Nd7. Now Nc5 would be answered by Nb6.} (20. Re1 Nf6 21. Bc7 e4! 22. Nb6 Ra7 {looks extremely dangerous for White.}) 20... e4 (20... b5! {was possible immediately. I didn't initially see it while I was casually watching the game but once you realize the option exists there's basically no other move that you'll consider!}) 21. Kh1 b5! {And Magnus goes for it. Not that there was anything wrong with any other attacking move on the kingside but this move seems the most challenging. Carlsen: "I thought for so long and I wasn't sure about it but I thought I just go for it and up the stakes even more. Either you win the game, or you get mated."} ({Caruana did consider this, but was expecting} 21... Kh8 {saying "I thought it actually asked White what the next move is. Many of my useful moves have problems."} 22. f3 (22. Re1 Nf6 {and then f2 hangs in some lines. (Caruana)}) 22... e3 23. g3) 22. Nb6 (22. axb6 {doesn't seem to work.} 22... Rxa3 23. Nxa3 f3! (23... Nc5 {was my first thought when looking at the position but playing slowly might be too much. White is definitely not resourceless after a move like.} 24. Nb5!) 24. gxf3 Ne5! {and the computer seems really happy about this position, which might not come as a surprise. White's best defence is} 25. Rg1 Qh6! 26. Nc4 (26. Qc1 Qh4) 26... exf3 27. Nxe5 (27. Bf1 Nxc4 28. Bxc4 Rf4 29. Qc1? (29. b7! {only move but after} 29... Bxb7 30. Qc1 Qf6 {Black is obviously much better.}) 29... Bg5! {threatening both the trivial Rh4 and Qxh2+ after} 30. b7 Qxh2+!! 31. Kxh2 Rh4+ 32. Kg3 Rh3#) 27... fxe2 28. Qxe2 dxe5 29. Bxe5 g6! {And while White retains practical chances the pair of bishops manages to contain the passed pawns on the queenside.}) 22... Nxb6 23. Bxb6 Qg5!? {A very interesting practical decision by Magnus. Basically every sensible move was possible in this position so it doesn't make sense to try and analyse them in depth but this approach of going straight at White's king is not only the most direct but also the most frightening in practice.} 24. g3 {I have to admit this move didn't cross my mind at all. My guess on why Fabiano went for this is that the advantage of this approach is that now Black cannot really choose how to proceed on the kingside and has only one setup available. So if he decided he could hold everything together he decided to go for it.} (24. Bxb5 {Of course the computer would play that. I can only assume Magnus intended} 24... Rf6 25. Re1 ({At the press conference Carlsen suggested} 25. f3 e3 26. g3 Rh6 27. g4 Bd8) 25... Rg6 26. Bf1 Bf5 {And while objectively this should be better for White it is also quite scary.}) (24. f3 {with an idea similar to the game, could have led to a very cute end} 24... e3 25. g3! Rf6 (25... fxg3 26. Rg1 {and there's no real attack anymore.}) 26. gxf4 Qh5! 27. Bxe3 Qxh2+!! 28. Kxh2 Rh6+ 29. Kg1 Rg6+ {and White is better advised to agree to the perpetual as} 30. Kf2?? Bh4# {is an immediate mate.}) 24... b4 25. Rb3 (25. gxf4 Qxf4 26. Rg3 {was also an option but Black is certainly pressing after} 26... Bh4) 25... Bh3 (25... f3 26. Bb5 Bh3? (26... Rf6 27. Re1 Qg4 28. Bf1 {and among others White has the idea of Rxe4 followed by Re3 (which could have been available on this move), and also in some moment e3 is available for his bishop to defend the h file.} (28. Rxe4 {might be even more practical although it's hard to say what this is after} 28... Qxe4 29. Re3 Qg4 30. Rxe7 Bf5)) 27. Re1 {is the reason why Magnus started with Bh3.}) 26. Rg1 (26. Re1 Bf6! (26... e3 {looks scary enough and might be the main reason Fabiano rejected this move, but White has the cold-blooded} 27. Rg1! {available.} 27... f3 28. Bxf3 exf2 29. Bxf2 Rxa5 30. Be3 {would just be equal now.}) 27. Rxb4 Bc3 28. Rxe4 fxg3! 29. hxg3 Bxa5! 30. Bxa5 Rxf2 {and here White barely saves a raw after} 31. Bg4 Bxg4 32. Rxg4 Qh5+ 33. Rh4 Qxd1 34. Rxd1 Rxa5) 26... f3 27. Bf1 Bxf1 28. Qxf1 {Any direct approach with the rook would fail as White's bishop can get to e3 and cover the crucial h6 square.} 28... Qxd5 (28... Rf6 29. Rxb4 Qh5 30. Be3! {and Black's attempt at an attack is easily repelled.} (30. Rxe4?? Rh6 31. h4 Bxh4 32. Rxh4 Qxh4+ 33. gxh4 Rxh4+ {would mate})) 29. Rxb4 {The dust settled and while White's king is clearly no longer in danger - there's an argument that he never was really in danger other than psychological, but even then it felt serious - Black has a beautiful pawn center and could be aiming at pushing his central pawns at the appropriate moment. Also even if the position is less complicated to navigate now both players were getting short on time around here.} 29... Qe6 30. Rb5! {An excellent move, even if rather obvious.} 30... Bd8 (30... d5 {would be strongly met by} 31. Qd1 Rf5 32. g4! Rg5 33. Be3 {and White is the side pressing.}) 31. Qe1 Bxb6 32. axb6 Rab8 {The position is around equal but there is still room for a big fight. White's obvious trump is his pawn on b6, while Black can count on his center and the weakened White king, which while not in immediate danger, can never easily join the center in case of an endgame.} 33. Qe3 Qc4 34. Rb2 Rb7 35. Rd1 {A bit of an inaccuracy due to Magnus' strong answer. But not even sure we can call it that as Fabiano kept things well under his control afterwards.} 35... Qe2! 36. Re1! (36. Qxe2 fxe2 37. Re1 Rxf2 38. Kg1 Rbf7! {while still being within drawing margin, would be ostensibly unpleasant to fac as White has no easy way to regain the pawn.} 39. Rbb1 (39. b7 Rf1+ 40. Kg2 Rxb7 41. Rxe2 Rxb2 42. Kxf1 Kf7) 39... d5! 40. c3 h6 {and b7 is always met by Rf1+}) (36. Qb3+?? Kh8 37. c4 {would trap the queen if not for} 37... Rxb6! 38. Rxe2 fxe2 {and Black wins}) (36. Qd4?? {would also be a serious attempt if not for the beautiful and winning} 36... e3! 37. c4 exf2! 38. Rxe2 fxe2 {and as Magnus pointed it out afterwards, the pawn on c4 prevents White from capturing the pawn with the fork Qc4+}) 36... Qxe3 37. Rxe3 d5 38. h4 {This is the most practical approach to get White's king more active. He aims to reach g3 via g4.} 38... Rc8 (38... Rf6 39. c4! dxc4 40. Rxe4 Rbxb6 41. Rxb6 Rxb6 42. Rxc4 {would liquidate into a draw.}) 39. Ra3! ({The attempt to liquidate everything fails:} 39. Rb5? Rxc2 40. Rxd5 Rxb6 41. Rxe4 Rxf2 {would lose on the spot}) 39... Kf7 (39... Rc6?? {would be a huge blunder because of} 40. Ra8+ Kf7 41. Ra7 Rxa7 42. bxa7 Ra6 43. Rb7+ {and now the king has no good square} 43... Kf6 (43... Kf8 44. Rb8+) 44. Rb6+ {wins}) 40. Kh2 (40. Ra7 Rb8 41. c4! dxc4 42. Rc2 {would have drawn immediately.}) 40... Ke6 41. g4 Rc6 42. Ra6! {A very important decision and the correct one. Fabiano correctly judged that with both rooks on the b file he wouldn't be able to obtain the necessary counterplay with his rooks.} (42. Rab3 {would still objectively hold but White is reduced to passivity after say} 42... h6 43. Kg3 (43. h5 Ke5 44. Rb5 g5 45. hxg6 Rxg6 46. c4! Rxg4 47. Rxd5+ Kf4 48. Rh5 {also looks scary but is holding.}) 43... g5 44. hxg5 hxg5 45. Rb1 Kd6 46. Rb5 Rc5 47. R5b2 Kc6 {and even if the computer claims} 48. c3! Rxc3 49. Rh1 {is a draw, you definitely don't want to get to that point if you can avoid it.}) 42... Ke5 43. Kg3 h6 44. h5 Kd4 {Maybe underestimating Fabiano's next move, Magnus gets into a bit of a scare.} ({Here I expected} 44... g5 {but both} 45. hxg6 ({or the immediate} 45. Rb5 Rxc2 46. Raa5 Rd2 47. Rb1 {and Black makes no progress.}) 45... Rxg6 46. Rb5 h5 47. Raa5 Rxg4+ 48. Kh3 {are easily holding as the rooks create more than enough counterplay.}) ({Carlsen mentioned} 44... Rb8 {when} 45. Rb5 {is interesting (Caurana).}) 45. Rb5! {White is more than in time to create counterplay and the king is now misplaced. Hard to figure what Magnus missed on that one, but he took his time and found the arguably best practical decision.} 45... Rd6! (45... Kc4 {was the move recommended as best by the now famous Sesse although} 46. Raa5 d4 (46... Rbxb6 47. Rxd5 Rb1 48. Re5 Rg1+ 49. Kf4 {even if this is a draw it looks downright scary.} 49... Rf6+ 50. Rf5 Rxf5+ 51. Rxf5 Rg2 52. Kxe4 Rxg4+ 53. Kxf3 Rg1 {and Black holds as the king will eventually manage to come back to the kingside.}) 47. Re5 {definitely looks scary as the king is so far away from the kingside. If Black loses his central pawns he might be lost in any rook endgame with his king cut. I also can't see an immediate way to draw even if I'm confident in the fact that it still exists.} 47... Rcxb6 48. Rxe4 Rb2 49. Ra4+ R7b4 {is possibly still holding but looks far too scary.}) 46. Ra4+ Ke5 47. Rab4 Ke6! {Here Fabiano saw no point to torture Magnus even though with colors reversed the game might still be going on at the time I'm typing those words.} 48. c4 {There is no real point in delaying that move as Black is aiming at Rc6 anyway.} 48... dxc4 49. Rxc4 Rdxb6 50. Rxe4+ Kf7 51. Rf5+ (51. Rf4+ Kg8 52. Rc5 {was arguably a try at keeping both pairs of rooks on the board. But in general Black always manages to eventually exchange them either by harassing the rooks (with two rooks on the same file like here, just moving endlessly the rooks) or arguably by staying passive on the last two ranks as White will have a hard time making progress anyway. Let's note though that} 52... Rf6?! 53. Rxf6 gxf6 54. Kxf3 {would be a downright draw with the pawn on f7, but with the pawn on f6 the additional weakness on the seventh rank makes it a tad more difficult to defend in practice.}) 51... Rf6 52. Rxf6+ Kxf6 53. Kxf3 Kf7 54. Kg3 {Fabiano doesn't see the point in keeping on playing, and I can't argue with that decision as White can't make any progress without exchanging pawns on the kingside, but then there's really not enough pawns remaining as it is well known for a few centuries! In any case we got a very interesting game today in spite of a peaceful result once again, and I felt like both players took some risky or principled decisions on purpose, which means they're not waiting for what looks like a probable tiebreak. So I wouldn't dismiss the possibility of a decisive result kicking in in the last two games. Also, the quality of the games seems very high to me in general especially considering they've been at it for almost three weeks now.} 1/2-1/2 [Event "2018 World Chess Championship"] [Site "London"] [Date "2018.11.24"] [EventDate "2018.11.09"] [Round "11"] [White "Magnus Carlsen"] [Black "Fabiano Caruana"] [Result "1/2-1/2"] [ECO "B31"] [Annotator "chess.com"] [PlyCount "110"] [CurrentPosition "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1"] 1. e4 {I was little surprised to see 1.e4 today, but then again, I have failed to predict basically any opening choice Magnus has made. Maybe that's why he is so hard to prepare for.} 1... e5 2. Nf3 Nf6 3. Nxe5 d6 4. Nf3 Nxe4 5. Nc3 Nxc3 6. dxc3 Be7 7. Be3 O-O (7... Nc6 {Was Caruana's choice in the infamous "shush" game in St. Louis in August, 2018.}) 8. Qd2 Nd7 9. O-O-O Nf6 {A somewhat unusual move, but not a bad one, and mentioned on one of Caruana's Chessbase files in the leaked video!} (9... c6 {Is definitely the main try. Caruana has played this move multiple times. I'm sure Magnus had some idea in mind.}) 10. Bd3 {Up until here Magnus was blitzing, but he started to think more after Black's next move. This I do not really understand...} 10... c5 {This is Black's most common move by far, and the top choice of the machine. I don't see how Magnus could have been prepared up to Bd3 and surprised when Black plays the best move according to both human and silicon practice.} 11. Rhe1 Be6 12. Kb1 (12. Bg5 {Is a more critical try. The machines offer it as a clear improvement and in practice it has fared well, including a fine win by Karjakin over Harikrishna. It must be the most challenging move, though obviously one has to believe Caruana would have been prepared for it.}) 12... Qa5! {This forces the queens off. Black has to be a little careful I guess.} 13. c4 Qxd2 14. Bxd2 h6 {This was the point where I joined the Chess.com live broadcast, and basically called the game perfectly up to the point where it was no longer interesting.} ({Caruana deviated from} 14... Rfe8 15. Ng5 Bd7 16. f3 Bc6 17. a3 h6 18. Ne4 Nxe4 19. Bxe4 Bxe4 20. Rxe4 Bf8 {as played very recently in Sasikiran,K (2671)-Miroshnichenko,E (2606) Saint Louis USA 2018}) 15. Nh4?! {This is very forcing, extremely direct, and basically just a draw. Black is in horrible shape if he doesn't play the most forcing moves at his disposal. Given this is Caruana's normal M.O. anyway, I can't imagine Magnus expected to get much.} (15. h3 {A patient move like this one was certainly a better try. White might go for Nh4 in the future now than Ng4 becomes less enticing, and g2-g4 is a plan as well. The computer even offers White a slight edge, but it's hard to imagine a 2800 losing with Black.}) 15... Rfe8 (15... Ng4 16. Nf5 (16. Ng6 fxg6 17. Rxe6 {Carlsen noted that he thought Black would be a bit worse here. We agreed during the live analysis. But it's hard to imagine Caruana subjecting himself to this when Rfe8 equalizes easily.} 17... Nxf2 18. Rde1 Nxd3 19. cxd3 Bf6 20. Rxd6 Bd4 21. Rde6 {And White will play Kc2 and one of the rooks to e2 if needed [Carlsen].}) 16... Rfe8 {Would transpose to the game} (16... Bxf5? 17. Bxf5 Nxf2 18. Rc1! {Nice and simple. I found this line with Robert and Danny during the live broadcast, and the knucklehead cliub figured out that White is pleasantly better. The knight is in serious danger of getting trapped.})) 16. Ng6 Ng4! {Strong, but also very straightforward.} 17. Nxe7+ Rxe7 18. Re2 {If White had a move to consolidate, he would be much better. But he doesn't.} 18... Ne5! {The bishop pair is halved, and everything liquidates in short order.} 19. Bf4 (19. Bc3 f6! (19... Nxd3?! 20. Rxd3 Rd7 21. Rg3! {Would be a little uncomfortable for Black, though I do think he should hold without a ton of trouble.}) 20. Bxe5 (20. b3 Nxd3 21. Rxd3 Rd8 {White cannot stop Kf7 followed by d5}) 20... dxe5 {I was not sure about this position at first, worrying about White being able to dominate the d-file. But Black is definitely fine, and if White plays Red2, f6-f5 followed by Kf7-f6 becomes an idea, when Black sets his healthy majority in motion.}) 19... Nxd3 20. Rxd3 Rd7! {The last accurate move ends the game. The d6 pawn could have been a target, and now it will be exchanged.} 21. Rxd6 (21. b3 Rad8 {With d6-d5 to follow, and absolute equality.}) 21... Rxd6 22. Bxd6 Rd8 23. Rd2 Bxc4 24. Kc1 b6 25. Bf4 Rxd2 {This was unnecessary as Black will likely lose a pawn, but with opposite colored bishops, it is still a very straightforward draw.} (25... Re8 {Was a simpler route to full equality. The text is hardly a mistake though.}) 26. Kxd2 a6 27. a3 Kf8 {The simplest. Black does not try to save his pawns, will lose one button, and have an easy fortress.} 28. Bc7 b5 29. Bd6+ Ke8 30. Bxc5 h5 {Once g6 comes, the kingside is permanently secured. The black king can come to c6 and White's extra pawn is of no value.} 31. Ke3 Kd7 32. Kd4 g6 33. g3 Be2 34. Bf8 Kc6 35. b3 Bd1 36. Kd3 Bg4 37. c4 Be6 38. Kd4 bxc4 39. bxc4 Bg4 {The only plan other than the one Carlsen tried is to play h4 and then hope for f2-f3, g3-g4, and then Kg5 to create a passed h-pawn. Black would have to be extremely cooperative for this to work.} 40. c5 Be6 41. Bh6 (41. h4 Bf5 42. Ke3 Be6 43. Kf4 Bf5 44. f3 Be6 45. g4 {Is an illustration of my plan in action. Black probably did not have to allow even this. But still, it's just a draw if he stops the king from coming to g5.} 45... Bd5 {With dead equality.} (45... Bc4? 46. Kg5 {And White wins with the h-pawn.})) 41... Bd5 42. Be3 Be6 43. Ke5 Bd5 44. Kf4 Be6 45. Kg5 Bd5 46. g4 hxg4 47. Kxg4 Ba2 48. Kg5 Bb3 49. Kf6 Ba2 50. h4 Bb3 51. f4 {Now f5 is coming. If Black takes it, the h-pawn goes. But it's not a problem.} 51... Ba2 (51... Bc2? {Too soon. Black needs to wait for f5 first.} 52. Kxf7) 52. Ke7 Bb3 53. Kf6 Ba2 54. f5 {If Black were to fall asleep at the wheel, fxg6 would win. But after} 54... Bb1! {Now Kxf7 is no longer an issue since it can be met with Bxf5.} 55. Bf2 {The last trick. White hopes Black will take on f5. But it's almost like hoping Santa Claus will bring you Christmas presents. Once you reach a certain level, you know it will never happen.} (55. fxg6 fxg6 {Is also a draw. Obviously Black should not take with the bishop to allow a second passed pawn.}) 55... Bc2 {Caruana finished the game with more time than he started with, and Carlsen with only three minutes less. There wasn't much to say, and if I thought Chess. com would pay me my standard ! million per game fee for leaving my annotations blank, maybe I would have done it! Game 12 should be interesting. The high pressure might induce more mistakes than normal and we could see more of a fight. One wonders if Carlsen will stick with the Sveshnikov for a sixth game in a row. Caruana is very likely to keep playing 1. e4.} (55... Bxf5? {Lets the h-pawn through.} 56. h5) 1/2-1/2 [Event "2018 World Chess Championship"] [Site "London"] [Date "2018.11.26"] [EventDate "2018.11.09"] [Round "12"] [White "Fabiano Caruana"] [Black "Magnus Carlsen"] [Result "1/2-1/2"] [ECO "B33"] [Annotator "chess.com"] [PlyCount "62"] 1. e4 c5 2. Nf3 Nc6 {Magnus sticks to his same Sicilian repertoire in the all decisive final game. One can only admire his willingness to enter sharp positions with it all on the line.} 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 e5 6. Ndb5 d6 7. Nd5 Nxd5 8. exd5 Ne7!? {The first surprise. Games 8 and 10 saw 8...Nb8} (8... Nb8 {And here Caruana was playing} 9. a4 {a very different plan from what we saw in the present game.}) 9. c4 Ng6 10. Qa4 Bd7 11. Qb4 Bf5 {Up to this point both sides were blitzing. Caruana now thought for a bit before ultimately going for the most testing move.} 12. h4!? {White aims to play against the g6 knight. The threat of h4-h5 is not particularly subtle, but a bit annoying for Black to meet since playing h7-h5 himself will leave him with a weak h5-pawn and potentially a problem on the g5 square.} 12... h5 {There had only been one human game up to this point, and h5 is formally a novelty. But, it was contested between top engines in the TCEC competition, so Magnus surely was familiar with the ideas. He was still blitzing out his prep at this point.} (12... a6? {would be well met by} 13. h5! {when White has an excellent position.}) ({[That was in fact played in the recent game Kramnik-Roganovic, Batumi Olympiad 2018 which continued} 12... a6 13. h5 Nf4 14. Nc3 Be7 15. Be3 Nd3+ 16. Bxd3 Bxd3 17. h6 {and White won in 42 moves. - ed.]}) 13. Qa4 Bd7 14. Qb4 Bf5 15. Be3 {No early draws this time!} (15. Bg5!? {may have been a more critical move. Still, I am sure Magnus would have been ready for it.} 15... Qb8 16. Be2 a6 17. Nc3 Nf4 18. Bf1 Be7 19. Bxe7 Kxe7 20. g3 Ng6 {and a draw after 59 moves in Houdini 6.03-Stockfish earlier this year.}) 15... a6 {The first move Magnus thought about.} (15... Be7!? {The machine thinks giving the a7-pawn is interesting. It might be right but it looks risky to me.} 16. Bxa7 O-O 17. g3 Be4 {Otherwise White just goes Bg2 with an extra pawn} 18. Rg1 Bf3 {Black has obvious compensation, but I'm not fully convinced it is enough.}) 16. Nc3 Qc7 (16... Be7 {Again, Black could have pitched the pawn to speed up his development. But I don't really mind Magnus' choice. He clearly understood the resulting middlegame better than Caruana.}) 17. g3 Be7 18. f3?! {The first dubious move of the game. Perhaps White was worried about e5-e4 then Ng6-e5 can come to solve the problem of the poorly placed g6-knight, but this strikes me as too defensive and too slow.} (18. Be2 {The computer flirts with weird checks on a4. This move looks by far the most natural to me.} 18... e4 19. Bb6 Qd7 20. Bd4 {White looks better to me, but of course it is still complicated.}) 18... Nf8! {One bad piece makes a bad position, and White's play has been slow enough that Black has time for Nf8-d7 before castling.} 19. Ne4 Nd7 20. Bd3 O-O 21. Rh2?! {White's second dubious decision. It's not such a silly way to develop the rook as it clearly belongs on the c-file. but it appears to articifial to me.} (21. O-O {Sometimes boring is prudent. I don't think White is better, but he is definitely not worse and has a fair plan to expand with c4-c5 eventually.}) 21... Rac8 (21... Bg6 {Preparing f7-f5 immediately was fine as well. I suspect Carlsen did not want to allow White to respond with} 22. g4 {Though here Black is also better after the accurate response} 22... Nf6!) 22. O-O-O Bg6! {f7-f5 is coming. Black is setting his central pawn majority in motion while White's queenside is badly stalled- c4-c5 looks impossible to ever pull off. I think Black is already seriously better. Note that Carlsen chose this moment to go for Bf5-g6 since White cannot comfortably play g3-g4 in response due to the hanging c-pawn.} 23. Rc2 (23. g4? hxg4! 24. fxg4 Bxe4! {The point. White loses the c4-pawn, and the game.}) 23... f5 24. Nf2 (24. Ng5 {The computers prefer this move by a wide margin, but I still would not want to be White after something like} 24... Bxg5 25. hxg5 e4! 26. Bf1 (26. fxe4 fxe4 27. Be2 Ne5 {Also looks awful}) 26... f4! {When the bishop on g6 becomes a monster.} 27. Bxf4 Nc5 {Black's attack is extremely dangerous.}) 24... Nc5 {Around here, I thought Magnus would win the game, and in fine style to close out the match after eleven draws. White's position looks miserable.} 25. f4 a5!? {This move is not best. But it doesn't really bother me- I think Black should still win.} (25... exf4 {Sesse claims opening the position was the best way. White will either lose the h4-pawn after} 26. gxf4 (26. Bxf4 {Or this can be met with} 26... b5! {When the queenside is blasted open. Technically this may be best, but White suddenly has a solid looking position, a good bishop on f4, and the g6-bishop is not inspiring. Magnus's choice is no worse in my opinion.}) 26... Bxh4 {and White has no meaningful compensation}) (25... b5 {This also looks very strong, and may be the most human way to bring in the full point. Black forces open the queenside. The machine's recommendation of} 26. Bxc5 dxc5 27. Qd2 Bd6 {Looks rather hopeless for White to me. Eventually he will lose on the dark squares.}) 26. Qd2 e4 {White is completely stuck, has no counterplay, no plan. But the position is locked. All Black has to do is get the b7-b5 advance through. If he manages, he should win. And he should manage.} 27. Be2 Be8 28. Kb1 Bf6 29. Re1 a4?? {Much like I did to Caruana for playing h2-h3 in game 8, I am giving an exceptionally harsh evaluation to Magnus for this move, which is totally not in the spirit of the position. His only plan to win the game is b7-b5. It should be enough if he manages under decent circumstances. But now, he has cleared the b4-square for White's queen! This makes his task significantly harder, if not impossible. In addition, the pawn on a4 deprives Black's knight and bishop access to this square, which both pieces would want to use. Magnus said this was prophylaxis against Bd4 due to Nb3, but it still doesn't come close to justifying allowing the queen to b4.} (29... Ba4! {This was the most convincing route to victory according to the machine, and Caruana mentioned it was what he was most concerned about. Black wins in all lines, for instance} 30. b3 (30. Bxh5 {This looks like a little bit of counterplay but Black wins the race after} 30... Bxc2+ 31. Qxc2 b5! 32. Bg6 bxc4 33. Qe2 Rb8 {Black mates first.}) (30. Rcc1 b5! 31. cxb5 Qb6 {The b-file opens, and Black should win.}) 30... Bxb3! 31. axb3 Nxb3 32. Qd1 a4! {And Black can just ignore White's extra piece, which is meaningless. b7-b5 comes next. Still, this is hard to do in an OTB game. I don't really mind Magnus not playing Ba4. Playing a4 was what I took issue with.}) (29... Rb8!? {A move like this one is what I would expect a human to do. b5 is coming and White can't do anything about it.} 30. Bd4 {A bad move, but what else? In addition to Nd3 winning material, this position is also just wretched after the simple} 30... Bxd4 31. Qxd4 b5 {Black should win, easily.}) (29... g6 {Even a patient waiting move like this might be enough.}) 30. Qb4! {Now how will b5 come? Black has no other way to open the position.} 30... g6 31. Rd1 Ra8 {Magnus has gotten a lot of grief for offering a draw here, from both club level online kibitzers to past World Champions. I'll spare him my full thoughts on the matter, but I will say I don't really understand his decision. He faces absolutely no risk ever and still has some plans at his disposal to try to get b5 through, with as Bd7, Rrb8, etc. Maybe White is holding, maybe he isn't. I'd need to look a lot longer to draw any conclusions. But I really think Black should have played, at least until time control. All that said, I still think he is a favorite in the playoff. His rapid and blitz skills have long been atop the world, and I do think he played better overall chess in the classical portion (if this can ever be said for a 6-6 tie with no decisive games). Still, while he is a favorite in the playoff, his victory is by no means gauranteed. Upsets can happen, and Magnus would do well to try not to be tilted by both starting and ending the match by drawing a winning position with Black.} 1/2-1/2 [Event "2018 World Chess Championship TieBreak"] [Site "London"] [Date "2018.11.28"] [EventDate "2018.11.09"] [Round "13"] [White "Magnus Carlsen"] [Black "Fabiano Caruana"] [Result "1-0"] [ECO "A22"] [Annotator "chess.com"] [PlyCount "55"] 1.c4 e5 2.Nc3 Nf6 3.g3 { Magnus tries a different move order, which makes quite a lot of sense when considering how Caruana was playing against the English previously. } 3...Bb4 ( 3...d5 { In the classical games, Caruana went for the reverse sicilian. But here, due to White avoiding Ng1-f3, after } 4.cxd5 Nxd5 5.Bg2 { The knight on d5 is hanging- by contrast, the position with Nf3 and Nc6 included, the knight is safe. As a result, Black cannot play Bc5, as he did in the 2 classical games. He has other options to fight for equality and probably succeed if well prepared, but it clearly avoids Caruana's prepared line. } 5...Bc5? ( 5...Nb6 { Would be normal } ) 6.Nxd5 ) 4.e4!? { Magnus plays an interesting move, aiming just to get his typical Rossolimo positions a tempo up. Considering how he outplayed Caruana in those games to some extent, it makes a lot of sense to aim for such a position where the extra tempo means White can never be worse. Objectively Black must be fine, but practically, this was excellent preparation. } 4...O-O 5.Nge2 c6 ( 5...d6 { I might prefer something like this, aiming to play on the dark squares. Still, after something like } 6.Bg2 Nc6 7.O-O { The machine calls it equal, but White has the very straightforward plan of sending the King's Indian death star with h3, d3, f4-f4, etc. Black is under some pressure and will need to find the right plan. I think he should be okay after Bc5 followed by Nd4. } 7...a5 8.d3 Bc5 9.h3 Nd4!? 10.Kh2 c6 11.f4 exf4 { A position like this one is probably okay. I played a very similar plan in a game with Lu Shanglei several years back. } ) 6.Bg2 a6?! { Too slow. } 7.O-O b5 8.d4! { Energetic and strong. I don't know if this was preparation or not, but it certainly looks good. Magnus is playing with a lot of force. } 8...d6 ( 8...exd4 9.Nxd4 bxc4 { Is a positional disaster after } 10.e5! { When Black looks absolutely lost despite the pawn up } ) 9.a3! { Again, not fearing the loss of a pawn. } 9...Bxc3 ( 9...Ba5 { In hindsight, this was likely a better try. } 10.b4 Bb6 11.dxe5 dxe5 12.Qc2 { I would take White, but the game goes on. } ) 10.Nxc3 bxc4 11.dxe5 dxe5 { Black's position is more or less lost if he is unable to save the c4-pawn, so he has to cling to it for dear life. } 12.Na4 ( 12.Be3 { Looks more natural to me. The knight was quite fine on c3. Of course it's easier to criticize with an engine running. } ) 12...Be6 13.Qxd8 Rxd8 14.Be3 Nbd7 15.f3 { Magnus' play has not been the most incisive in the past few moves, and Caruana now almost gets back into the game. } 15...Rab8! 16.Rac1 Rb3! { The c-pawn will eventually fall, but Black can compromise White's structure before it is captured. } 17.Rfe1 { Bf1 is coming, and it leaves Black in miserable shape. The move Caruana played is not terrible but I much prefer another option. } 17...Ne8 ( 17...Rdb8! 18.Bf1 c3! { Black pitches the c-pawn, but White will have to compromise his structure. Anything else would allow Bxc4 and likely an easy win for White. } 19.Rxc3 Rxc3 20.bxc3 { Black is a bit worse and will suffer, but I think it is probably holdable. } ) 18.Bf1 Nd6 19.Rcd1! { Now Black is in all sorts of tactical problems. } 19...Nb5? { Activity is always appealing and often correct, but now the position falls apart. } ( 19...Nb7 { Sad as it might be, a move like this may have been prudent. Black is under a lot of pressure but not falling apart just yet. } ) 20.Nc5 Rxb2 21.Nxe6 fxe6 22.Bxc4 { Black will not hold onto his extra pawn, and his position remains a trainwreck. } 22...Nd4 ( 22...Kf7 { Trying to defend e6 ignores the bigger and more subtle threat. } 23.Bxb5! axb5 24.Bg5! { White wins a piece, and the game. } ) 23.Bxd4 exd4 24.Bxe6+? { This is imprecise. } ( 24.Rxd4! { The pin is lethal. White has to spot a very difficult move, but then he wins. } 24...Kf7 { Black is hoping for Nd7-d5, when the rook on d8 cannot be comfortably taken due to Nxf3. But White wins with the astounding } 25.Kh1!! { Getting out of Ne5 just in time. } ( 25.Kf1 { Will not do because Black gets too much counterplay after } 25...Ke7 26.Red1 { Otherwise the knight escapes } 26...Rxh2! ) 25...Ke7 ( 25...Ne5 26.Rxd8 Nxf3 27.Re2 ) 26.Red1 { The immortal pin on the knight should win the game in short order. Trying to get it out with } 26...Rbb8 { Fails to } 27.Bxa6! { when the passed a-pawn will decide the game when all the rooks trade. } 27...Nc5 28.Rxd8 Rxd8 29.Rxd8 Kxd8 30.Bc4 { White should win. } ) 24...Kf8 25.Rxd4 Ke7 26.Rxd7+ Rxd7 27.Bxd7 Kxd7 { White is a pawn up, and if he could just include his king, he would win easily. But the king is cut off on the first rank, which complicates his task considerably. I'm not sure if the position is winning or not. } 28.Rd1+ Ke6 29.f4 c5 { The c-pawn is coming, and hard to deal with. } 30.Rd5 Rc2 31.h4 c4 32.f5+ Kf6 33.Rc5 h5 { Magnus let the win slip away. He cannot include his king in the action and hie pawns are firmly blockaded. Still, he clearly was in control the entire game, and is looking like the much better rapid player so far. } 34.Kf1 Rc3? { I've been told I have been too harsh with giving each player double question marks at some point in the match, so I will spare Caruana here. But this move is so completely wrong. Black absolutely cannot let the White king enter the game. } ( 34...c3 35.Ke1 Ra2 { Something like this draws easily. The king is simply too passive. } 36.Rxc3 Ke5 ) 35.Kg2! { Freedom! } 35...Rxa3 36.Rxc4 Ke5 37.Rc7! Kxe4? { The last mistake. Black still could have held with Ra2+, but this was a very hard move to find with no time on the clock and its unsurprising even a very strong player messed up in such a tough position. While this is technically the move that lost Black the game, I would be much more critical of Rc3, allowing the White king to escape. Often the bigger mistake is the one that makes your task difficult, not the one that ultimately tips the scale. } ( 37...Ra2+! { Black once again needed to fight to keep White's king passive. } 38.Kh3 Kxe4 39.Rxg7 Ra1! { Black is ready for Kf3 next to make counterplay. } 40.Rc7 Rf1! { Another only move, and Black holds by the skin of his teeth. This seems nearly impossible for a human to find when so low on time. } ) 38.Re7+! { Sealing the deal. White does not allow Black's king into f3. } 38...Kxf5 39.Rxg7 { Black cannot prevent the loss of the h5-pawn, and hie a-pawn offers no real hope of salvation. A back and forth game (it is rapid after all) but Magnus really seemed to be in the drivers seat the whole way. One game is obviously the smallest possible sample size, but if this is the kind of strength difference we can expect among them in rapid chess, his draw offer in round 12 becomes much more understandable. } 39...Kf6 40.Rg5 a5 41.Rxh5 a4 42.Ra5 Ra1 43.Kf3 a3 44.Ra6+ Kg7 45.Kg2 Ra2+ 46.Kh3 Ra1 47.h5 Kh7 48.g4 Kg7? { The position was beyond saying but this gives up the last practical chance. } ( 48...a2 { This is a decent try, but it does lose. } 49.Kh2! { It is important to keep the g2-square clear. } ( 49.Kg2 Kg7 50.g5?? ( 50.Kh2 { Still wins } ) ) 49...Kg7 50.g5 Kg8 ( 50...Rb1 51.Ra7+ Kg8 52.Rxa2 Rb5 53.Rg2! { White needs this move to save his pawns. } ) 51.Ra7 Kh8 52.h6! { again, keeping the g-pawn as the safe one. } ( 52.g6? { An innocent move like this one is not to be recommended... } 52...Rb1! 53.Rxa2 Rb5 { And it's a technical draw. white needed the Rg2 resource to save his less advanced pawn. By playing g6 instead, he allows a draw. } ) 52...Rb1 53.Rxa2 Rb5 54.Rg2! $18 { I really doubt Magnus was unaware of this idea and probably knew the Ippolito-Ganguly game, but it was the only thing worth trying for Black. } ) 49.Kh4 a2 50.Kg5 Kf7 51.h6 Rb1 52.Ra7+ Kg8 53.Rxa2 Rb5+ 54.Kg6 Rb6+ 55.Kh5 1-0 [Event "2018 World Chess Championship TieBreak"] [Site "London"] [Date "2018.11.28"] [EventDate "2018.11.09"] [Round "14"] [Black "Magnus Carlsen"] [White "Fabiano Caruana"] [Result "0-1"] [ECO "A22"] [Annotator "chess.com"] [PlyCount "28"] 1.e4 c5 2.Nf3 Nc6 { Another Sveshnikov. Magnus is clearly showing he is willing to fight with Black, even in a game where a draw suits him very well. } 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 e5 6.Ndb5 d6 7.Nd5 Nxd5 8.exd5 Ne7 9.c4 Ng6 10.Qa4 Bd7 11.Qb4 Qb8 ( 11...Bf5 { Was Magnus' choice in game 12 of the classical phase. He outplayed Caruana very convincingly but also might have been worse in the early middlegame. It's not surprising he was the first one to pivot. Still, Caruana was clearly ready. } ) 12.h4 h5 ( 12...a6 13.h5! { Again, this seems strong. } ) 13.Be3 a6 14.Nc3 { Now White would like to get Nc3-a4 in to abuse the weakened b6 square. This plan is much more effective now that Black's queen is on b8. } 14...a5 { Black is very insistent that White not get Nc3-a4 in. But this loses some time and sends the queen back home, where she can harass the kingside. Magnus was still blitzing so was likely still in preparation, but I really don't like his position. } 15.Qb3 a4 16.Qd1 Be7? { This move looks so innocent, but in fact its a serious error. The g6 knight now is really bad. } ( 16...Ne7! { It was best to try to reroute the knight to f5 right away. I am not sure Black is equalizing, but he is not in such terrible shape after something like } 17.Bd3 Nf5 18.Bg5 Be7 ) 17.g3 { Black looks much worse to me. } 17...Qc8! { I like Magnus' move a lot. He really has to be credited for his resilience. } ( 17...Nf8 { My instinct would be to play in the same manner Magnus did in game 12, rerouting the g6-knight to f6. But it looks really bad. For instance after } 18.Bd3 g6 19.O-O Nh7 20.b3! axb3 21.Qxb3 { White is breaking through on the queenside. Still, Black has some counterplay after } 21...Nf6 22.Rab1 O-O 23.Qxb7 Qd8 ) 18.Rc1 ( 18.Be2 { The machine prefers this direct approach. } 18...Bg4! { Black provokes the f2-pawn forward, where it will loosen the white kingside and restrict the e2-bishop. Pawns don't move backwards after all. I think some moron wrote a book about that recently... A book well worth buying! } 19.f3 ( 19.Rc1 { Would transpose to the game } ) 19...Bd7 { White is clearly better, but 0-0 followed by f5-f5 is clear counterplay. Especially given the quick time control, I prefer Carauana's choice for its simplicity. } ) 18...Bg4 19.Be2 Bxe2 20.Qxe2 Qf5! { Another good move } ( 20...f5 { I would be tempted to set the kingside mass in motion right away, but it's hard to find a great plan when White's pawn is still on f2 and Black just looks worse after } 21.Nb5 Qd7 22.Bg5 ) 21.c5? { This is far too rushed, and Magnus is fine again. } ( 21.O-O O-O 22.Nb5 { Still looks very pleasant for White. The g6 knight is permanently bad, e4 never seems to work, and White can slowly expand on the queenside while keeping counterplay under control. It's possible Caruana was concerned about } 22...e4 { Aiming to play Ne5 with a lot of counterplay, but White wins after } 23.Bd4! Ne5 { Otherwise Rfe1 wins the pawn on the spot. } 24.Bxe5! Qxe5 25.Rfe1! { When Black will lose the e4-pawn, since f7-f5 allows the devastating Nc7-e6. } ) 21...O-O 22.c6 { White needs a move like this one to justify the c4-c5 advance. If he can secure the pawn on c6 he should win easily, but he is a bit too slow. } ( 22.cxd6 { The machine prefers a much less principled move like this one. After } 22...Bxd6 23.O-O { White is likely fine, but certainly is not better, and has no credible plan or short or long term assets to try to win with. Given the match situation, it's very understandable that Caruana went for the more ambitious and double edged option. } ) 22...bxc6! 23.dxc6 Rfc8! { It's not easy to defend the pawn. The natural } ( 23...e4!? { Black gets a lot of counterplay with this move, but the text is much more convincing. } 24.O-O Ne5 { White's king will face some problems, and Nf3 followed by d5-d4 looks good as well. Even c6 is hanging! } ) 24.Qc4 ( 24.Nd5 { Is one tempo too slow after } 24...Qe4! { Had White already been castled here, he would basically win after Rfd1. Instead, he loses material on the spot. c4-c5 was just not quite fast enough to work, and White should have been more patient. } ) 24...Bd8! { Magnus is absurdly strong at rapid chess, he just finds good moves so easily. The c6-pawn is under control and Ne7 is coming. } ( 24...Qf3 25.Qd5! { Is a nice resource Magnus surely noticed, and correctly avoided. Still, after } 25...e4 26.Qxe4 Qxe4 27.Nxe4 Ne5 { He will win the c6-pawn and only be marginally worse. Black should not lose. } ) 25.Nd5 ( 25.O-O Ne7! { White loses the pawn, and probably the game } ) 25...e4! { Massive counterplay in the center is on the way, truly in the spirit of the Sveshnikov. White has a winning position in the longrun on the queenside, but he faces problems in the center along the way. The previously awful g6 knight is now about to enter the position with a bang. } 26.c7? ( 26.O-O Ne5 { Is an illustration of the problems White will shortly face if he is not very careful. } 27.Qb5 Qh3 { checkmate incoming. } ) ( 26.Bd4 { Was absolutely forced. Black has several appealing options. The simplest is } 26...Ne5 ( 26...Ra5!? { A maximalist would notice a move like this one. } 27.Ne3 Qf3! 28.O-O Bxh4! { I find it very unlikely White will survive in a rapid game, although the machine does think his position is defensible. } ) 27.Bxe5 Qxe5 { When Ra5-c5 should be enough to contain the c6-pawn. } ) 26...Bxc7 27.Nxc7 Ne5 { Despite an opening gone wrong, Magnus completely dominated the second game, much like he did the first. I thought he played somewhat better chess in the classical phase, but if 12 games still ended in 12 draws, you can't really claim he was the superior player. But the rapid has thus far been a totally different story, and the players are not on the same level. } 28.Nd5 ( 28.Qe2 Nd3+ 29.Kd2 Qa5+ { Is a rather simple finisher } ) ( 28.Qd5 { This was the best try, though Black still wins after the accurate } 28...Rab8! { When Black's threats are easily decisive. The simplest one is Rxc7 followed by a check (this was not previously possible because of the hanging rook on a8). } ) 28...Kh7 { White is done for. Magnus wins the second game in absolutely crushing fashion, and looks nearly certain to keep his title. } 0-1 [Event "2018 World Chess Championship TieBreak"] [Site "London"] [Date "2018.11.28"] [EventDate "2018.11.09"] [Round "15"] [White "Magnus Carlsen"] [Black "Fabiano Caruana"] [Result "1-0"] [ECO "B44"] [Annotator "chess.com"] [PlyCount "46"] 1. e4! {White can choose more or less any sensible first move (including 1. e4) in a normal game, or even a must-win game. e4, d4, Nf3, c4- all strike me as having similar chances of fighting for an edge. But if one specifically needs a draw with White, 1.e4 is absolutely the way to do it. Play the most forcing moves!} 1... c5 2. Nf3 e6 3. c4 {A bit of a weird move order. c4 only really works against the Kan, but not against the Taimanov} (3. d4 cxd4 4. Nxd4 a6 {Now c4 would make a lot more sense. Carlsen played this against Anand in their second match.}) 3... Nc6 4. d4 {Transposing back to the Taimanov} (4. Nc3 {This would have avoided the option Black chose in the game.}) 4... cxd4 5. Nxd4 Bc5! {Definitely the best move under the circumstances} (5... Nf6 6. Nc3 {Has long been known to end in sterile equality very quickly. This is not normally a problem from Black's point of view, but this game is a bit different...} 6... Bb4 7. Nxc6 bxc6 8. Bd3 {The position is very dry and very safe for White. Neither side has any good plans at their disposal to imbalance the game.}) 6. Nc2 (6. Nb3 {Is a better chance to fight for an edge, and White may well be better. But at least Black is keeping some tension in the game.}) 6... Nf6 7. Nc3 O-O (7... Qb6 {Black absolutely needed to disrupt White's coordination. after something like} 8. Qd2 Ng4 9. Nd1 O-O 10. h3 Nf6 11. Nc3 {Black is not better, and White can complete his development, but at least he doesn't have perfect coordination and a nice space advantage.}) 8. Be3 {Now White develops easily and even maintains a nice edge due to his extra space in the style of the Maroczy bind. My annotations won't be published until the game ends, but at this point, I find it very unlikely we will see a fourth game.} 8... b6 9. Be2 Bb7 10. O-O Qe7 11. Qd2 Rfd8 12. Rfd1 {White is incredibly solid. In such structures the best plan for Black is generall to try to force d7-d5 through, but all that does is liquidate the position.} 12... Ne5 13. Bxc5 bxc5 14. f4! {White takes even more space.} 14... Ng6 15. Qe3 d6 (15... d5 {"Objectively" this has to be the best move, and Black is almost equalizing. But it chops off all the pieces, and this is obviously unacceptable for Caruana.} 16. exd5 exd5 17. Qxe7 Nxe7 18. cxd5 Nfxd5 19. Nxd5 Nxd5 {I don't think Magnus will lose a position like this one very often...}) 16. Rd2 a6 17. Rad1 Qc7 18. b3 {White's play is simple and easy. Just keep everything protected and maintain the space advantage. Winning the game will not be easy, but his strong play in the first 2 rapid games earned him the right to sit tight here.} 18... h6 19. g3 Rd7 20. Bf3 Re8 21. Qf2 Ne7 22. h3 Red8 {Not much is happening. Black is struggling to find any kind of active plan.} 23. Bg2 Nc6 24. g4! {This is undoubtedly the best move, and Magnus must be given credit for being willing to somewhat open his king in a game where a draw clinches the World Championship. White expands his space advantage further.} 24... Qa5 25. Na4 Qc7 26. e5 (26. g5!? {White could have gone directly for the king, but one can hardly blame Magnus for looking for simplification and instand gratification. Had this been game 1, I doubt he would hesitate to play g5. And he'd likely have won.}) 26... dxe5 27. Nxc5 Rxd2 28. Rxd2 Rxd2 29. Qxd2 Ba8! (29... exf4 30. Nxb7 Qxb7 31. Qxf4 {Looks very safe for White.}) 30. fxe5 Qxe5 31. Nd7 {This looks like the easiest way to force a draw. All the pieces come off.} 31... Qb2 (31... Nxd7 32. Qxd7 Qg3 {Almost keeps some fight, but the bishop on a8 is lost after} 33. Qc8+) 32. Qd6 Nxd7 33. Qxd7 Qxc2 {Caruana makes the right choice to keep different pieces on the board. White's c-pawn is pretty dangerous and the bishop is better then the knight in it's abolity to send the pawn through, but White's king is open and the knight and queen can be a dangerous attacking pair. White is certainly not worse but Caruana has to take any chance he can get to imbalance the position at all.} 34. Qe8+ Kh7 35. Qxa8 Qd1+ 36. Kh2 Qd6+ 37. Kh1 Nd4 38. Qe4+ f5 39. gxf5 exf5 40. Qe3 {The g2 bishop is too strong. Its really hard to imagine how White can get mated.} 40... Ne6 41. b4! {Passed pawns must be pushed!} 41... Ng5 {Black clearly wants f4-f3- it's his only winning attempt. The move is objectively bad but in an absolute must-win scenario, you gotta do what you gotta do.} 42. c5 Qf6 43. c6 {White's c-pawn will decide the game long before any attack has a dream of coming through.} 43... Ne6 44. a4 Nc7 45. Qf4 Ne6 46. Qd6! {going for the 3-0 sweep. Magnus showed why he is one of the best players ever, making one of the best players in the world (who even fought him to a 6-6 tie in the classical portion) look completely uncompetitive in the rapid playoff. I was happy to see him look for complicated positions with the black pieces, and any future challengers would be best advised to try to beat him in regulation- his rapid play is absurdly powerful. I'd like to thank chess.com for inviting me to annotate the games, I did my best to bring the best and most objective analysis I could, and I hope I was able to shed some light on the games. Thanks to all for reading!} (46. Qe3 {One wonders if Caruana would have repeated moves here. We'll never know because Magnus wanted the whole cake.}) 1-0