Baku Final G3: Svidler blunders, Karjakin wins

[Event "FIDE World Chess Cup 2015"] [Site "Baku"] [Date "2015.10.03"] [Round "7.3"] [White "Svidler, Peter"] [Black "Karjakin, Sergey"] [Result "0-1"] [ECO "B53"] [WhiteElo "2727"] [BlackElo "2762"] [Annotator "Ramirez Alvarez,Alejandro"] [PlyCount "60"] [EventDate "2015.??.??"] [EventCountry "AZE"] 1. e4 c5 {Certailny no Spanish will do when you must play for a win, but even in the Sicilians White can try to play a relatively dry variation.} 2. Nf3 d6 3. d4 cxd4 4. Qxd4 $5 {Of course not the main line, but completely possible. Honestly I expected some king of early c3 Sicilian, but Qxd4 with the idea of going into a Maroczy structure also seems like a good idea.} a6 5. c4 Nc6 6. Qe3 Nf6 7. h3 g6 8. Nc3 Bg7 9. Be2 Nd7 (9... O-O {is certainly the main line. There have been many games from this position, though most end in White wins.}) 10. Rb1 Nde5 11. O-O O-O {The Maroczy structures all share similarities. White is trying to prevent Black from creating any counterplay, which Karjakin desperately needs. For this reason Black chooses to put his pieces in such a way to allow an early f5, even if it weakens his position.} 12. Rd1 Nxf3+ 13. Bxf3 f5 14. exf5 Bxf5 15. Be4 {The more pieces are traded off, the happier Svidler is. If he is able to develop his bishop on c1 it is difficult to imagine that Black can create something concrete.} Qd7 16. Nd5 Qe6 17. Bxf5 Qxf5 18. Bd2 Rae8 19. Bc3 {Still trying to trade pieces, though this one is a bit too straightforward. White had other moves that didn't sideline his knight. } (19. Qe1 e6 20. Ne3 Qf7 21. Ng4 $5 {Trying to trade pieces on h6.}) 19... e6 20. Nb6 d5 21. Bxg7 Kxg7 22. Qc5 {This looks natural as Black doesn't want to advance the d-pawn and trade queens.} Rf6 23. b4 Ne5 24. cxd5 Nd3 {Trying to go for some counterplay, but it isn't much yet.} 25. Qe3 Nxf2 $6 {Objectively a bad move, but Karjakin can't afford to trade queens.} 26. Rf1 Qe4 27. Rbe1 ( 27. Rfe1 $1 $18 {Is unnatural but much stronger, with the threat of Qxf2 (with the other rook this is not a threat)} exd5 28. Qxf2 Rxf2 29. Rxe4 {in the case of the b-rook moving to e1, Black here would have Rxf1+, turning the tables.}) (27. Qxe4 $6 Nxe4 28. Rxf6 Kxf6 29. dxe6 Kxe6 {gives Black some slight edge in an endgame.}) (27. Qc3 {Also looks strong, preserving all the pins, threatening Nd7 for example.}) 27... exd5 $6 28. Rxf2 $2 (28. Qc3 $1 {and the game is basically over. Black simply loses material, the variations are not even difficult to calculate.}) 28... Qh4 29. Qd2 $4 (29. Qxe8 Qxf2+ 30. Kh2 Qxb6 31. Re7+ Kh6 32. Rd7 {will probably end in a draw. Black's king is too weak to win.}) 29... Rxf2 {To me it is unclear what Svidler missed. Of course White is just lost in this position as the R on e1 is hanging in the case of Qxf2.} 30. Qc3+ d4 {White is simply down massive amounts of material and is getting mated.} 0-1