Commentary: Tata Steel Masters 2018, Round 14 tiebreak (Carlsen - Giri)

In keeping with a thematic approach to commentary games, this next one features an English Opening with a b3/Bb2 development. It is a different structure than Tarjan - Kosteniuk, however, as Black (GM Anish Giri) here adopts a Queen's Gambit Declined (QGD) approach, while Carlsen uses a more central strategy with e3 instead of a double fianchetto for his light-square bishop.

This was actually a tiebreak blitz game, which however doesn't make it any less instructive for how Carlsen chose to strategize his play and the numerous positional decisions made along the way. The overall strategy for White was to get a comfortable game with no weaknesses and then keep pressuring the obvious Black targets. Giri as a result was always struggling for equality with less harmonious piece placement, not a position you want to be in regardless of the time control.

[Event "Tata Steel Masters TB"] [Site "Wijk aan Zee"] [Date "2018.01.28"] [Round "14"] [White "Carlsen, Magnus"] [Black "Giri, Anish"] [Result "1-0"] [ECO "A13"] [WhiteElo "2834"] [BlackElo "2752"] [Annotator "ChessAdmin/Komodo Dragon"] [PlyCount "109"] [EventDate "2018.??.??"] 1. Nf3 Nf6 2. c4 {now we are in an English Opening, unless White plays an early d4.} e6 3. b3 d5 4. Bb2 Be7 {QGD setup} 5. e3 {with this move, White chooses to exert more direct control over the center, particularly the d4 square, and develop his bishop accordingly.} O-O 6. Nc3 {this directly pressures d5 and is in keeping with the opening's focus on the center.} c5 { now that Black has increased his influence on the d4 square, White exchanges in the center.} 7. cxd5 {exchanging the c- for the d-pawn is a standard idea. It will give White a numerical advantage in center pawns, increase the scope of the light-square bishop, and open up the c-file for White's rook.} Nxd5 ( 7... exd5 {is just as frequent a choice here. After} 8. d4 {play will be similar, but with the extra pair of minor pieces.}) 8. Nxd5 exd5 (8... Qxd5 { has been played more often, but White has a much higher score in the database, at 68 percent. After} 9. Bc4 {and the queen retreats, White has a pleasant game with a small lead in development.}) 9. d4 {now White challenges with the pawn, to force an isolated queen's pawn (IQP) structure.} Qa5+ {a novelty. Carlsen however is fine with the queens coming off the board.} (9... Nc6 { is the more familiar way to play, with some more options for White.} 10. dxc5 ( 10. Be2) 10... Bxc5 11. a3 (11. Bd3)) 10. Qd2 {forced} Qxd2+ 11. Kxd2 Nc6 ( 11... b6 {is the engines' preference here. White would not want to go for a hanging pawns structure after a pawn exchange on c5, since he does not have enough firepower to sufficiently pressure the c5/d5 pawns.}) 12. dxc5 {we now have the IQP structure on the board.} Bxc5 13. Bb5 {this bishop development gives the option of exchanging on c6, to inflict a backwards c-pawn on Black, while allowing the king to go to e2 and not block the bishop after the upcoming check. Bd3 was also a good possibility.} Bb4+ 14. Ke2 Be6 { reinforcing d5, although this makes the bishop a "big pawn" in effect.} 15. Rac1 (15. Bxc6 {is the engines' preference. After} bxc6 16. Ne5 Rfc8 17. Rhc1 c5 18. Nd3 {White has a more concrete slight positional plus. In the game, Carlsen avoids committing himself, however.}) 15... Rac8 {"It's always the wrong rook" is a common refrain. Here, Black's Rf8 is less active and the engines prefer him committing it to the queenside. Perhaps he had thoughts of . ..Re8 and ...d4 at some point, however.} 16. Rhd1 {getting the other rook into the game and pressuring the d-pawn.} (16. Bxc6 $5) 16... Be7 {Black decides the bishop is not doing anything useful on b4 and retreats it. This also covers the g5 square, preventing White's knight from landing there.} 17. h3 { preventing ...Bg4} a6 {putting the question to the bishop. Now the exchange on c6 seems less favorable and White retreats it.} 18. Bd3 (18. Bxc6 Rxc6 19. Nd4 Rxc1 20. Rxc1 Rc8 $11) 18... Nb4 19. Bb1 {these types of retreats are common in master play. Here of course it protects the a2 pawn, but the larger positional point is that the scope of the bishop on the b1-h7 diagonal is not diminished, so White loses nothing by having the piece on the back rank.} Rxc1 20. Rxc1 Rc8 21. Rd1 {Carlsen again avoids committing himself to the major piece exchange. Black's rook cannot penetrate on the c-file and White's rook is doing good work pressuring the d-pawn again.} Nc6 {Black's strategic problem is that he has nothing very useful to do. He would like to liquidate the d-pawn, but White has an ultra-firm grip on the d4 square, blockading the pawn's advance.} (21... Nc2 {doesn't get Black anything.}) (21... g6 {might be a somewhat useful waiting move.}) 22. g4 {one different between masters and amateurs is that masters have a much better sense of when to advance pawns, particularly kingside ones. Here the pawn advance does not impact White's king safety and restricts Black by controlling the f5 square; note the role played by the Bb1. This seizure of territory will help a future f-pawn advance as well.} h6 {getting "luft" for the king and also controlling g5.} 23. Nd4 { physically blockading d4 and clearing the way for the f-pawn.} Nxd4+ 24. Bxd4 ( 24. exd4 $6 {would negate the whole point of White's strategic play against the IQP.}) 24... Ba3 {controlling the c1 square, which could theoretically be used by either Black's rook or bishop. This is still a case of Black not having much useful to do, however.} 25. f4 {in contrast, White is now seizing space.} f6 {this gives the bishop the f7 square to retreat to if necessary, in order to maintain its guard over d5. It does inflict long-term weaknesses on the kingside pawn structure, however, which become important later.} 26. Bg6 { immediately taking advantage of the hole left behind by the pawn advance. Now White could exchange off bishops after ...Bf7 and e8 is also controlled.} Kf8 { moving to centralize the king, now that we are essentially in the endgame.} 27. Kf3 {likewise advancing his king to a more influential square.} (27. Rg1 { is liked by the engines, with the plan of further advancing the kingside pawns with the rook pressuring the g-file.}) 27... Ke7 28. h4 {at this point White still has the easier game, but with the material left on the board, it looks pretty even. This can still be dangerous for the side without a real plan, however.} Bb4 29. Bd3 Bd7 {the bishop can now go to c6 if it needs to.} 30. e4 {Carlsen decides to simplify the situation with this pawn break. Time control may have been a factor here.} (30. Rg1 {would keep the IQP tension.}) 30... Bc3 {Black would be happy to trade off pieces on c3, giving him a nice place for the rook and getting rid of his relatively worse bishop.} (30... dxe4+ 31. Bxe4 b5 $5 {looks fine for Black, for example} 32. Bb7 Rc2 $11) 31. Bf2 {safely avoiding the trade.} Bc6 $6 {Komodo Dragon identifies this as the move which gives White an advantage. Let's see how.} (31... d4 $5 {looks like it would pose White more problems, as he cannot win the d-pawn immediately.} 32. Be2 ( 32. f5 $14) 32... b5 33. Bxd4 Bxd4 34. Rxd4 Rc2 {and now Black can recover the pawn, for example after} 35. a4 Rc3+ 36. Rd3 Rxd3+ 37. Bxd3 bxa4 38. bxa4 Bxa4 $11) 32. exd5 Bxd5+ 33. Be4 Bxe4+ 34. Kxe4 {White's positional advantages here are twofold: a better king position and Black's weaker kingside pawn structure, which White's king is threatening to penetrate and White's bishop has the potential to attack. Tactically, Black has to watch out for his Bc3 getting pinned against the Rc8.} Ke6 {this looks reasonable, but White's next move is obvious as well.} (34... Bb4 {is preferred by the engines, but White still has an edge after} 35. Rd4 {followed by Rc4, thanks to his more advanced and centralized king.}) 35. f5+ $16 Ke7 36. Rc1 {now the only way to un-pin the bishop is to protect the rook. There are multiple ways to do this.} Rc6 $2 { the losing move, as identified by the engines. Let's see why.} (36... Kd7 37. Kd3 Be5 38. Rxc8 Kxc8 39. Bc5 {is favorable for White, who has pawns he can target with his bishop, but would it be enough to win?}) 37. Kd3 Bb4 38. Rxc6 bxc6 39. Kc4 $18 Bd6 40. Bc5 {By now we can see Black's problems more clearly. The isolated pawns on the queenside must be defended from White's king, while the kingside pawns need to be defended against an incursion from White's bishop. White will benefit from the creation of a zugzwang situation as well. Exchanging pieces would simply give White a won K+P ending.} Kd7 (40... Bxc5 41. Kxc5 {and Black's a-pawn is doomed.}) 41. h5 $1 {Making the zugzwang even more apparent. White can therefore focus on increasing the pressure.} (41. Bxd6 $2 {this exchange does not work, because White's king is in a worse position and Black has the added resource of ...h5 to undermine White's pawn structure. For example} Kxd6 42. b4 h5 43. gxh5 Ke5 $11) 41... Bf4 (41... Kc7 $2 {now the bishop exchange works.} 42. Bxd6+ Kxd6 43. Kb4 $18 {it would take too long for Black to demolish White's kingside, as White gobbles up the queenside.}) 42. Bf8 Ke8 {hoping to trap the bishop if it captures on g7.} 43. Bc5 Kd7 44. Kb4 { heading to penetrate on the queenside.} Bd2+ 45. Ka4 Kc7 46. b4 {now the White king's way up the a-file is clear again.} Bf4 47. Bf8 {Black can no longer hold both sides of the board.} Kb6 (47... Kd7 48. Ka5 $18) 48. Bxg7 Bg5 49. Bf8 Bf4 50. Be7 Bg5 {Black attempts to hold out with a fortress. The problem is that if the bishop is exchanged, White gets a passed pawn and it's game over.} 51. Kb3 Kc7 52. Kc4 Kd7 53. Bc5 Kc7 54. Kd3 Kd7 55. Be3 {Black loses another pawn or sees White create a passed pawn now, which is losing either way.} 1-0