Commentary: 2017 U.S. Championship, Round 10 (Zherebukh - Nakamura)

Here we continue with the theme of Caro-Kann Classical wins by Black at the top level, with this game featuring dynamic opposite-side attacking play by GM Hikaru Nakamura from the 2017 U.S. Championship. Unlike the 2020 commentary game in which he lost as White to Awonder Liang, here Nakamura plays actively, even aggressively, but in a way that is in tune with the position's characteristics rather than attempting to force an attack. His opponent, GM Yaroslav Zherebukh, had been having a great tournament up to that point, but apparently was caught out by Nakamura's opening choice and subsequent play.

White's problems stemmed from his choice on move 14 to prematurely exchange minor pieces. This resulted in the h-file being opened for Black's rook, and he was also able to start pushing around White in the center, seizing the initiative and conducting a punishing attack. Nakamura's execution of it is particularly instructive. The variations show the interplay of various tactical ideas, revolving around fixing White's vulnerabilities around the king position - particularly the h1 and f2 squares - and taking advantage of loose/awkward piece placement. A devastating win in only 21 moves, showing that the Caro-Kann Classical can be a fighting choice as well as a solid one for Black.

[Event "U.S. Championships Men 2017"] [Site "Saint Louis"] [Date "2017.04.08"] [Round "10"] [White "Zherebukh, Yaroslav"] [Black "Nakamura, Hikaru"] [Result "0-1"] [ECO "B18"] [WhiteElo "2605"] [BlackElo "2793"] [Annotator "ChessAdmin/Komodo Dragon"] [PlyCount "42"] [EventDate "2017.??.??"] 1. e4 c6 2. d4 d5 3. Nc3 dxe4 4. Nxe4 Bf5 5. Ng3 Bg6 6. Nh3 {this has the same idea as the "book" N1e2, namely deploying the knight to f4. It does not seem to have independent significance and is not mentioned in theory books, although it is interesting to see that it is roughly equivalent in the database - actually with a few more games - to N1e2.} Nf6 (6... e6 {is the other main response.}) (6... h6 {is the third most popular, but scores very well for White. There is no need to prevent the knight from going to g5 and preparing to retreat the Bg6 loses time for Black.}) 7. Nf4 Nbd7 (7... e5 { has been the traditional antidote to this variation since the Tal-Botvinnik matches in the 1960s, but the line can be drawish, so likely was avoided by Nakamura for this reason. Here's a high level example:} 8. Nxg6 hxg6 9. dxe5 Qa5+ 10. Bd2 Qxe5+ 11. Qe2 Qxe2+ 12. Bxe2 Nbd7 13. O-O O-O-O 14. Rad1 Nb6 15. Bc1 Bd6 16. Rfe1 Kc7 17. h3 Rhe8 18. Bd3 {1/2-1/2 (18) Radjabov,T (2735)-Anand, V (2799) Morelia/Linares 2008}) 8. Bc4 Qc7 9. O-O e6 10. c3 {while structurally sound, this is a rather passive move.} (10. Re1 {is a more typical White idea in this position, pressuring the e-file and threatening some tactical ideas if Black gets careless with development.} Bd6 {and now White can sac on e6, but it only maintains equality if Black defends correctly. However, it could certainly be a good practical try, Ivanchuk lost a blitz game against Duda in this line.} 11. Rxe6+ fxe6 12. Nxe6 Qb6 13. Qe2 Ke7 { and now the defensive idea is to play ...Rae8, perhaps after a4 and ...a5 are inserted. That way the king can flee via d8 and not lock the rook out of the game.}) 10... Bd6 11. Qf3 {Defending the Nf4. Also, by placing his queen on the h1-a8 diagonal, he influences e4 and d5 and eyes b7 in the event Black goes for the typical ...c5 break, which Nakamura does play immediately.} c5 { after an exchange on c5, recapturing will help activate Black's minor pieces. White could also push the d-pawn in response, but Black does not have any problems there.} 12. dxc5 (12. d5 {there are not in fact any database games with this line. White players likely do not want to give up the center so easily.} e5 $5 {may be the easiest way to neutralize White's chances.}) 12... Bxc5 (12... Nxc5 {no one plays this way, since the knight will just have to go back.} 13. Bb5+ Ncd7) 13. Bb3 {White takes care of his hanging bishop, otherwise ...Bxf2+ is threatened.} O-O-O {by castling opposite sides, Black signals further that he is looking for winning chances. It is also the engines' top choice. While Black's king position looks more open, he has effective local superiority on the queenside, as White is much better positioned for action on the kingside. The half-open c-file and queenside pawn majority are also not immediate problems, as it will be tough for White to mobilize anything there. Finally, White's dark-square bishop is currently out of the game, something which is corrected shortly, but still leads to problems. } 14. Nxg6 $2 {positionally, this allows for the bishop to come to a more active square, but ignores the even more significant extra activity that Black gains as a result, most notably through the opening of the h-file. Tactically, White needed to prepare the piece exchange more, as now he gets pushed around and gives Black the initiative.} (14. Be3) (14. Qe2) 14... hxg6 $17 {this is exactly what Black wanted with queenside castling, now the Rh8 is ideally placed to exert pressure on White's king position.} 15. Bf4 e5 {now Black is firmly in control.} 16. Be3 e4 {I think it likely that White underestimated or missed this move, attacking the queen with tempo. The Ng3 is pinned against the mate threat on h2, is the problem.} 17. Qe2 {abandoning the kingside to its fate.} (17. Qf4 $5 {would put up more resistance on the kingside.} Bd6 18. Qg5 Ne5 $19) 17... Bxe3 (17... Rxh2 {is also possible immediately, but trading bishops simplifies the calculations.}) 18. Qxe3 (18. fxe3 {would avoid the threat of a K+Q fork on h2 and e3, but has its own problems, namely that the Ng3 will no longer be protected after the sacrifice on h2.} Rxh2 {and the rook still cannot be taken, due to} 19. Kxh2 Rh8+ 20. Kg1 Qxg3 {with a winning attack, thanks to the Nf6. Eliminating it is not enough, either.} 21. Rxf6 Nxf6 22. Bxf7 Kb8 {and Black wins material after ...Qh2+, thanks to the hanging Ra1. Trying to defend with Qf2 loses to the deflection tactic ...Rh1+}) 18... Rxh2 { this thematic sacrificial breakthrough idea should always be looked at when the other rook is available to immediately come to the h-file.} 19. Rfd1 { nothing better than to give White's king a path to run.} (19. Kxh2 $2 Ng4+) 19... Rdh8 {Black is now simply up material with a strong attack and dominating position. He does not have an immediate combinational win, but White is going to have to drop material or worsen his king position considerably.} 20. Qxa7 {re-establishing material equality at least, in the hopes that Black cannot carry through with his attack.} (20. Rxd7 Nxd7 21. Rd1 Kb8 22. Qxe4 {and now Black can win easily by entering a material-up endgame with} (22. Bxf7 {and White looks like he might have some compensation, but a nice sacrificial winning line for Black is} Ne5 23. Be6 Nf3+ $1 24. gxf3 Qxg3+ 25. fxg3 Rh1+ 26. Kf2 R8h2#) 22... f5 23. Qe7 Nc5 24. Qxc7+ Kxc7 $19) (20. Kf1 Kb8 {and now White has no good defense, as Black's knights can mobilize further to c5, e5 and g4 and White's king is too open. For example} 21. Qd4 Ne5 22. Bc2 e3 {similar to the game} 23. Qxe3 Nfg4 $19) 20... e3 $1 {now the threat is ...Qxg3, and the queen cannot be taken because of the mate on h1; the e3 pawn covers the f2 flight square.} 21. Qxe3 (21. Qa8+ Nb8 22. Rd4 R2h4 23. Rad1 (23. Rxh4 exf2+ 24. Kxf2 Rxh4 $19) 23... e2 $19) (21. Rd4 Qxg3 22. Qa8+ Qb8 23. Rc4+ Kd8 24. Qxb8+ Nxb8 25. Rd1+ Nbd7 $19) 21... Ng4 {because the knight replaces the black pawn in covering the f2 flight square, the ... Qxg3 threat is renewed, along with mate threats on h1, and they can no longer be met without White losing his queen.} 0-1