[Event "Wch Match Dubai 2021"] [White "Carlsen,Magnus"] [Black "Nepomniachtchi,Ian"] [Site ""] [Round ""] [Annotator "Anish Giri"] [Result "*"] [Date "2021.12.03"] [PlyCount "271"] 1. d4 Nf6 2. Nf3 d5 3. g3 {A sneaky move order, trying to go for a Catalan-like setup, but not exactly, as White delays the move c4.} e6 {Black had plenty of options, but not surprisingly Ian sticks to his repertoire and offers a transposition into the Catalan.} 4. Bg2 Be7 5. 0-0 0-0 6. b3 {The point of Carlsen's move order, trying to sidestep the main line with 6.c4, which is what happened (through the conventional Catalan move order) in game 2.} c5 7. dxc5 {Here too 7.c4 transposes into some kind of theory. With this move, Magnus is hoping to take Ian into the relatively unknown territory.} Bxc5 {Given the reply, 7...Nc6!? also deserved attention.} 8. c4 { More natural is 8.Bb2, but it takes a lot to surprise a well prepared top player who fights for the World Championship crown and so another unusual move followed.} dxc4 9. Qc2 $5 {Creative opening play by Magnus. The c4 pawn is not being recaptured yet.} Qe7 {Protects the c5 bishop and forces White to recapture the pawn, or so you would think.} 10. Nbd2 $5 {Very creative piece of preparation, sacrificing the pawn.} Nc6 $1 {Ian wants to have none of it and doesn't accept the sacrifice. From now on, for some reason, Magnus started taking quite a bit of time. Strange, given that so far, all of Ian's moves have been the top choice of the engines.} 11. Nxc4 b5 {Alternative was 11...e5 or the passive 11...Bd7, but this adventurous move got a lot of praise by the commentators.} 12. Nce5 Nb4 $1 {The point, now Black avoids losing material on the diagonal.} 13. Qb2 {There is no action around the h7 square (which could be the case if the bishop had been on b2 and the e5 knight not in the way, threatening Ng5 and Bxf6) and so the queen is better placed here than on b1.} Bb7 14. a3 {An exciting deviation would have been 14.Bg5!?.} (14. Bg5 h6 15. Bh4 {Here Black can go after the h4 bishop with immediate 15...g5!? or 15... Bxf3 16.Bxf3 g5!? and the eventual complications lead to some sort of equality as well, as always, but the lines are much more exciting than after 14.a3.} )Nc6 15. Nd3 {Keeping more pieces on the board.} (15. Bg5 {Is quite drawish, for example} Nxe5 16. Qxe5 Bxa3 17. Bxf6 gxf6 18. Qxb5 {And now Black can just give up one of the bishops to simplify.} Rab8 19. Qa4 Bxf3 20. Bxf3 Bc5 {A position like that is totally fine for Black, as with the full control over the dark squares, there is little danger to the somewhat open king.} )Bb6 16. Bg5 (16. a4 {Seemed like a natural inclusion at first, but likely Magnus disliked the simplifications after} Rfd8 $5 17. axb5 Nd4 { And it looks like mass exchanges are likely.} )Rfd8 {Useful move, there is no need to push White to take on f6.} 17. Bxf6 {White takes anyway, as Nd4 was coming.} gxf6 {An interesting choice, perhaps stemming from some ambition. Queen trade would have also been fine. Sometimes in those endgames White can claim a good d3 knight against a "bad" bishop, but with Black pieces so active and bishop being on b6, there is little talk of White fighting for anything.} 18. Rac1 Nd4 {Alternative 18...e5!? would be very double edged, weakening the f5 square, but it isn't easy to exploit that, as 19.Nh4 can be met with 19... Nd4!.} 19. Nxd4 Bxd4 20. Qa2 Bxg2 {There were many ways to play this, but Ian chooses to activate the queen and send it to e4.} 21. Kxg2 Qb7+ 22. Kg1 Qe4 23. Qc2 {Sort of grabbing the c-file. From now on Rac8 is always an option, though it's never neccessary.} a5 24. Rfd1 Kg7 {This allows 25.e3!?, instead Black could have gone for 24...f5 when after 25.e3 the bishop has some nice squares to retreat to- f6/g7.} 25. Rd2 {Commentators like 25.e3, using the fact that the bishop has no comfortable retreating square and 25...Be5 26.Qe2 Bd6 27. Nc5! gives White some risk free pressure. Magnus maybe disliking 25...Rc8 or for some other reason, chose to keep the status quo.} Rac8 {This is absolutely not neccessary, but it did lead to chaos, so the fans should be grateful to Ian.} 26. Qxc8 Rxc8 27. Rxc8 Qd5 28. b4 a4 29. e3 {First critical moment. 29. Bb2 forces the matters, but Ian chooses against it.} Be5 $6 {There was a reasonably easy draw in 29...Bb2, but maybe Ian didn't want to look for it from a position of weakness.} (29... Bb2 30. Rc5 Qd6 31. Rxb2 {Here this is a must.} Qxd3 32. Rbc2 {Another only move, otherwise White can't keep the queenside pawn from falling, as 32.Ra2?? is 32...Qb1+!.} Qxa3 33. Rxb5 { and now the natural 33...Qb3 is met by 34.Rc1! which is still within a draw, but the easiest is to setup a perpetual.} Qa1+ 34. Kg2 Qb1 {And the threat of a3-a2 combined with the Qe4-Qb1 perpetual ideas secure Black a draw.} )30. h4 $5 {Magnus realizing that Nepo has no follow up, makes a useful move and asks Ian if he wants to play Bb2 after all, but now in a worse version.} h5 $6 { Ian makes another waiting move. But so does Magnus.} (30... Bb2 {was still alright here.} 31. Rc5 Qd6 32. Rxb2 Qxd3 33. Rbc2 Qxa3 34. Rxb5 Qa1+ 35. Kg2 Qb1 {the h2 square is there, but the king can't escape.} 36. Rc3 Qe4+ 37. Kh2 Qb1 {intending Qf1.} )31. Kh2 $5 Bb2 $2 {Actually exactly here, this idea that was always a bail out for Black, loses.} 32. Rc5 Qd6 33. Rd1 $2 {The win was very hard to spot and calculate, but it was there. The sequence Magnus goes for, gives up a pawn and it is not entirely clear what he missed there (perhaps Qd7! idea).} (33. Rcc2 $1 {The move itself is not obvious, giving up both queenside pawns, but the attack is devastating.} Bxa3 34. Nf4 $1 Qxb4 { and now White goes all out for the king.} 35. Rd7 e5 36. Nxh5+ Kg6 {You have to see what your follow up is here, otherwise the whole sequence makes little sense.} 37. Rc6 $1 {The knight can go, as the Black king finds itself in the mating net and Black's pieces are unable to setup a defense.} Kxh5 {and now both Rxf6 as well as Rxf7 first just lead to checkmate eventually, as the king is basically inside a mating net.} 38. Rxf7 )(33. Rxb2 {was though a safe option, leading to a drawn endgame once again.} Qxd3 34. Rbc2 {here besides Qxa3 Black can also go 34...Qf1!?, when I don't see how White untangles, but Black also doesn't have any reasons to play this for a win.} )Bxa3 34. Rxb5 Qd7 $1 35. Rc5 e5 {Somehow the b4 pawn never got captured. Either Ian didn't see it hanging, or was afraid of some discovery on the queen after Bxb4 Rcc1.} (35... Bxb4 $1 36. Rcc1 Be7 {Here White is on the defensive.} 37. Ne5 Qb5 38. Rd7 Qxe5 39. Rxe7 Qb2 40. Rf1 a3 41. Rd7 a2 42. Rdd1 {and 2 rooks should hold here, but White is very passive and the a-pawn is alive.} )36. Rc2 $2 {Still giving up the b4 pawn. Instead, the rather loose looking 36.e4! is pointed out by the engine, intending 36....Bxb4 37.Rd5, while if 36...Qd4 then 37.Ra5!.} Qd5 $2 {Ian starts to drift. He could have collected the b4 pawn.} (36... Bxb4 $1 37. Rcc1 {Now White would be OK, but Black has a strong sequence, saving the extra pawn.} Ba3 $1 {Not giving White time for a break.} 38. Ra1 Qg4 $1 {The queen steps out of the x-ray.} 39. Rd2 Be7 {and Black could well be winning here, as he can combine the a-pawn with f5-f4 ideas. Very unpleasant for White, though very often you see that there are some fortress ideas eventually.} )37. Rdd2 Qb3 38. Ra2 e4 {This is panic. It loses, but also in general, it is incredibly risky to kill the mobility of the kingside pawns.} (38... Bxb4 {would probably hold, but not in a fun way.} 39. Rdb2 Qxd3 40. Rxb4 f5 41. Rbxa4 Qf1 {With pawn on e5 and not on e4, Black has f4 push when neccessary and White is probably not able to pose any real problems here, though he can move around forever, without any risk.} )(38... f5 {is the engine way to hold it, but it's a bit far fetched.} 39. Nxe5 Bc1 40. Re2 a3 41. Rac2 Bb2 42. Nc4 Bf6 {with zeros.} )39. Nc5 Qxb4 40. Nxe4 $2 {Magnus was probably happy to survive the time trouble and that too, with what looks like a promissing position. Instead, he had 40.Rdc2!!, winning the a4 pawn and reaching a won endgame.} (40. Rdc2 $1 f5 41. Nxa4 Qxa4 42. Rc3 Qd1 43. Rcxa3 { This endgame looks winning. The idea is that White can smoke the queen out of f1 by using zugzwang. A random example line:} f6 {White can anyway provoke this push later.} 44. Ra4 Kg6 45. Rd4 Qf1 46. Rd8 Kh6 47. Rd5 Kg6 48. Rad2 { This is the zugzwang position. Once the queen moves, White goes Kg2, the queen is out and then White will join the rooks to target something, for example the 7th rank, or an f-pawn. Black probably has to push f4 and the conversion will take some time, but it feels inevitable and the high engine evaluation suggests the same.} )Qb3 {Good move, Black stays in the game, as he keeps his a-pawn.} 41. Rac2 Bf8 42. Nc5 (42. Rc8 {is fun, but there is not more than a draw, after} a3 43. Rdd8 a2 44. Rxf8 a1=Q {Black king is almost in a mating net, but not quite. White doesn't have a mate here, only different kinds of perpetual checks.} )Qb5 43. Nd3 a3 {Black gets the passer quite far, protected by the f8 bishop. It is now hard for white to do much, though obviously it is White that can press without any risk here.} 44. Nf4 Qa5 45. Ra2 Bb4 46. Rd3 {White is not in time to pick up the h-pawn with Rd5 and then prevent Qa4-Qb3 counterplay.} Kh6 47. Rd1 Qa4 48. Rda1 Bd6 49. Kg1 Qb3 50. Ne2 Qd3 51. Nd4 Kh7 52. Kh2 {White can move around endlessly, but Nc2 is not a threat, due to Be5. It seemed as though Black has achieved a fortress, but suddenly he decided to give White a chance to change the nature of the position.} Qe4 $6 {Possibly a miscalculation. There was no need to trade the a-pawn for the h-pawn and the arising position is not too pleasant to defend, even if still holdable.} 53. Rxa3 $1 Qxh4+ 54. Kg1 Qe4 {moving to g4 seems more accurate, as it allows a pawn trade.} (54... Qg4 55. Ra4 Be5 56. R1a2 { alternatively 56.Kg2 is met with either 56...Qd7!? or 56...h6!? 57.Rh1 Bxg3! which holds after 58.fxg3 Qe4+! and some further complications.} (56. Kg2 h4 57. Rh1 Bxg3 58. fxg3 Qe4+ 59. Nf3 Qc2+ $1 60. Kf1 Qd3+ $1 (60... Qxa4 $2 61. Rxh4+ $1 )61. Kf2 Qc2+ $1 {an important line, that is very hard to spot.} )Qd1+ 57. Kg2 h4 $1 {Trading the weak h-pawn, definitely an achievement for Black.} 58. Nf5 hxg3 59. Nxg3 {Black is still suffering, but he is closer to a draw I would feel than before h4-hxg3.} )55. Ra4 $1 Be5 56. Ne2 Qc2 57. R1a2 Qb3 58. Kg2 {Also Kh2 deserved attention here.} (58. Kh2 Kg6 $1 {The move to hold on, but Black might not have such an easy task in the rook endgame.} 59. Nf4+ Bxf4 60. gxf4 {though most likely it holds with patient defense.} )Qd5+ 59. f3 {Intuitively one wouldn't be thrilled to push the pawn forward, creating the weaknesses, but Carlsen wanted to dislodge Ian's bishop and found a safe spot for his king on f2. Still, 59.Kh2 was definitely a worth alternative, trying to maneuvre around with ideas like Ng1!?.} Qd1 {Black could push f5 right away, but he chooses not to commit yet.} 60. f4 {In the following phase of the game a lot of random moves are happening. Black is trying to tie white's pieces to the weak e3 pawn, not to allow White to coordinated his forces and collect the pawn weaknesses, or even worse, deliver a checkmate somewhere.} Bc7 61. Kf2 Bb6 62. Ra1 Qb3 63. Re4 Kg7 64. Re8 f5 65. Raa8 Qb4 66. Rac8 Ba5 67. Rc1 Bb6 68. Re5 Qb3 69. Re8 Qd5 70. Rcc8 Qh1 71. Rc1 Qd5 72. Rb1 {A lot of shuffling back and forth, Black must stay alert.} Ba7 { Understandable to keep the bishop on the diagonal hitting the e3 pawn, but there are some concrete issues now. Alternative 72...Bd8! seems to have been better.} (72... Bd8 73. Re5 Qd3 74. Rbb5 h4 75. Rxf5 hxg3+ 76. Nxg3 Bh4 { Black is holding here, as the White king is too exposed.} )73. Re7 Bc5 74. Re5 {Suddenly Black has to make only moves.} Qd3 75. Rb7 {The queen far from h1, allowing White to create all kinds of threats with the rooks.} Qc2 76. Rb5 Ba7 {Black finds a way not to lose the f5 pawn for nothing, but the ensuing position is extremely unpleasant, even though objectively drawn.} 77. Ra5 Bb6 78. Rab5 Ba7 79. Rxf5 Qd3 {Black managed to create some play now, as Bxe3+ is there, but now we reach another kind of endgame.} 80. Rxf7+ $1 Kxf7 81. Rb7+ Kg6 82. Rxa7 {We get a terribly unpleasant endgame for Black, which, with perfect defense should be holdable. A lot of shuffling ensues.} Qd5 83. Ra6+ Kh7 84. Ra1 Kg6 85. Nd4 Qb7 86. Ra2 Qh1 87. Ra6+ Kf7 88. Nf3 Qb1 89. Rd6 Kg7 90. Rd5 Qa2+ 91. Rd2 Qb1 92. Re2 Qb6 93. Rc2 Qb1 94. Nd4 Qh1 95. Rc7+ Kf6 96. Rc6+ Kf7 97. Nf3 Qb1 98. Ng5+ Kg7 99. Ne6+ Kf7 100. Nd4 Qh1 101. Rc7+ Kf6 102. Nf3 Qb1 103. Rd7 Qb2+ 104. Rd2 Qb1 105. Ng1 {Magnus decides for the Ne2, Rd4 and e4 setup.} Qb4 106. Rd1 Qb3 107. Rd6+ Kg7 108. Rd4 Qb2+ 109. Ne2 Qb1 110. e4 Qh1 111. Rd7+ Kg8 112. Rd4 Qh2+ 113. Ke3 h4 {Opening more files should easen Black's defensive task.} 114. gxh4 Qh3+ 115. Kd2 Qxh4 116. Rd3 Kf8 117. Rf3 Qd8+ 118. Ke3 {Now as rightly pointed out by Magnus, getting the knight to g3 was of a great importance for him. Therefore Black had to give a check on b6, inviting the knight to a more central square, but giving White king less safe space on the board.} Qa5 $2 119. Kf2 $1 Qa7+ 120. Re3 {Now Magnus gets quite some hopes, as he has a plan of pushing the pawns forwards and with the knight on g3 and king on f3, his king is quite well positioned.} Qd7 121. Ng3 Qd2+ 122. Kf3 Qd1+ 123. Re2 Qb3+ 124. Kg2 Qb7 125. Rd2 Qb3 126. Rd5 Ke7 127. Re5+ Kf7 128. Rf5+ Ke8 129. e5 Qa2+ 130. Kh3 Qe6 {The losing move according to the tablebases. Keeping the queen behind, would still hold objectively, but by this point it was no easy task. Practically speaking, I feel on move 118 a major mistake happened.} 131. Kh4 {Next is Nh5 and the White king gets to hide in front of the pieces, leaking in via g5-h6.} Qh6+ 132. Nh5 Qh7 133. e6 $1 { The clincher, using the fact that the rook can't be taken because of the fork.} Qg6 134. Rf7 Kd8 135. f5 Qg1 136. Ng7 {A very elegant win. White runs the king up the boards to g8, where it hides from checks and then e7-e8 decides. What a game!} *