< Earlier Kibitzing · PAGE 2 OF 2 ·
Later Kibitzing> 
Apr2208   ex0duz: I'm thinking 14.♕xf6 is the start of blacks flawed defensive strategy, since after 16.♘e3, white is subsequently threatening 17.♘g4, with a gain of time and a what seems to be a fairly strong attack/position. I guess thats part of the reason black further chose to weaken his king and pawn structure with 16.h5, and also possibly to stop any future h4h5 pawn storm ideas. Blacks position just detiorated rapidly after this move.. I also agree with most other posters in that he should have recaptured the f6 pawn with his knight, and vacate the d7 square for his other knight. That would also have saved him from having to play the just as equally questionable g7g6 defense. 

Apr2208   PinnedPiece: 21 Re7! What a way to defend the Knight on d5! Very insightful move. 

Nov1808   shintaro go: Another one of my favorite Wesley So games. 20 Nxd5! A second knight sacrifices leaves Black scrathing his head. Of course not 20..Qxd5? 21. Qxg6+ Kh8 22. Re7 is curtains. 

Feb2209   LaFreaK: in 12 moves,then how many variations in that 12 moves??
wew..
what an imagination! 

Apr0209   ChessEscudero: Nice knight sacrifice by the chess prodigy 

Oct1510   licuan: <PinnedPiece: 21 Re7! What a way to defend the Knight on d5! Very insightful move.> yup, a decoy move! 

Sep2313   Patanjali: <PinnedPiece: 21 Re7! What a way to defend the Knight on d5! Very insightful move.> 21. Re7 works, but 21. Ne7+! Kg7 22. f5!!.. works faster. Now, IF 22...gxf5, THEN, 23. Rxf5 Nf6 24. Rg5+ and the black queen falls after Kh8 Ng6+ Kg8 Ne5+. 

Feb1117
  iking: the attack came from nowhere, here and there .. 

Mar1818   Cheapo by the Dozen: I went with Nxd5. The flaw was that Black could play the knight fork one move earlier than I envisioned, forcing a queen exchange. 

Mar1818   Walter Glattke: I would prefer 19.Nxd5 exd5 (Qf5?? Ne7+!)
20.Rxe6 Qf5 21.Rxg6+ Kh8 22.Qxd4+ less "insane" 

Mar1818
  ChessHigherCat: This is what I came up with after pfutzing around on WinBoard for 15 minutes, which counts as a serious analysis for a patzer like me (I just noticed that "patzer" has now been removed from the acceptable words in the spellchecker, what kind of racist nonsense is that?): 19. Nxd5 exd5 20. Re6 Nc5 21. Rxf6 Nxd3 22. Rxg6+ Kh8 23. f5 Ne5 24. Rh6+ Kg7 25. Rxh5 Rh8 26. Ne6+ Kg8 27. Rg5+ Kf7 28. cxd4 Ng6 29. fxg6+ Kxe6 30. g7 Rg8 31. Rf8 Nd7 32. Rxa8 Rxa8 33. g8=Q+ Rxg8 34. Rxg8 

Mar1818
  ChessHigherCat: Wow, that was before So became so soso! 

Mar1818
  Richard Taylor: I didn't try to solve this. Most of these Sunday ones are too time consuming and difficult. But a great attack by Wesley So! 

Mar1818   Walter Glattke: 19.Nxd5 exd5 20.Re6 Nc5 21.Qxg6+! Qxg6
22.Rxg6+ Kh8 23.Re1 Nbd7 24.Re7 Nf6
25.cxd4 Ne4 26.Ne6 threatens 26.b5!?
27.Rh6+ Kg8 28.Rg7# 

Mar1818   RKnight: <Walter> I too looked at lines starting from 19 Nxd5. The problem with your analysis is that instead of 23...Nbd7 black can play 23...Ra7 or 23...Ne4 and it is not evident how to continue the attack. 

Mar1818
  malt: Started with
19.N:d5 ed5 20.Re6 Nc5 21.Q:g6 Q:g6 22.R:g6+ looks equal, then went for 19.cd4 (stopping the attack on the Knight and preventing ...Nc5 so as to play 20.N:d5) after 19...Nc6 20.N:d5 ed5 21.Re6 Q:d4+
Gone for 19.cd4 Nc6
20.N:e6 Q:e6 21.N:d5 Qf7 22.Qc4 Rc8 23.f5 b5 (23...gf5 24.R:f5 Q:f5 25.Ne7+ ) 24.Q:b5 far as I got, did not see 19.N:e6 

Mar1818   Walter Glattke: Yes, after those moves White can get only some pawns for a Knight, but no threatenings. 

Mar1818
  whiteshark: After <19.Nxd5?!> click for larger viewBlack to move
1) =0.00 (25 ply) 19...exd5 20.Re6 Nc5 21.Rxf6 Nxd3 22.Rxg6+ Kh8 23.Ne6 Rf7 24.Rg5 dxc3 25.Rxh5+ Rh7 26.Rxd5 Rd7 27.Rh5+ Rh7 28.Rd5 2) +14.28 (25 ply) 19...Nc6 20.Nxf6+ Rxf6 21.Rxe6 Nc5 22.Qc4 Nxe6 23.Qxc6 Nxg5 24.Qxa8+ Rf8 25.Qc6 Nh3+ 26.gxh3 Rf7 27.cxd4 a4 28.Qxb6 a3 29.bxa3 Kg7 30.Qb5 Rf5 31.Qe8 Rf7 32.Qe5+ Kh6 33.Qh8+ Rh7 34.Qe8 Rg7 35.Qb5 Kh7 36.Qg5 Rf7 3) +14.35 (24 ply) 19...Qg7 20.Rxe6 Rf7 21.Rxg6 Nf8 22.Rxg7+ Rxg7 23.Nxb6 Ra6 24.Qc4+ Kh8 25.Qc8 Kg8 26.Ne6 Rf7 27.Qxb8 dxc3 28.Nxf8 Rxf8 29.Qb7 Rxb6 30.Qxb6 cxb2 31.Qg6+ Kh8 32.Qxh5+ Kg8 33.Qxa5 Rf7 34.Qe5 Rf8 35.Qxb2 Kh7 36.Qc2+ Kh6 37.g4 Rf7 38.g5+ Kg7 39.Qc3+ Kg6 1.0 minute analysis by Stockfish 9 v010218 

Mar1818
  thegoodanarchist: I wondered what to do if Black continues on with 30...Rf7, Δ 31.Re1+ Kf8. But White doesn't have to play 32.Re1+... He can play 32.Qg8+ + 

Mar1818
  whiteshark: After <19.Nxd5?! exd5 20.Re6 Nc5> click for larger viewWhite to move
1) =0.00 (29 ply) 21.Rxf6 Nxd3 22.Rxg6+ Kh8 23.f5 dxc3 24.Rh6+ Kg7 25.Rg6+ Kh8 2) =0.00 (28 ply) 21.Qxg6+ Qxg6 22.Rxg6+ Kh8 23.cxd4 Ne4 24.Rc1 Nd7 25.Rh6+ Kg8 26.Rg6+ Kh8 3) 5.43 (28 ply) 21.Qb5 Nxe6 22.Qxd5 Qf5 23.Qxe6+ Qxe6 24.Nxe6 Rf6 25.Nxd4 Nc6 26.Nf3 Rxf4 27.Re1 Rf6 28.Ng5 Kg7 29.Nf3 Rd8 30.h3 b5 31.Kf2 b4 32.a3 Rd2+ 33.Re2 bxc3 34.bxc3 Rd1 35.Re1 Rxe1 36.Kxe1 Re6+ 37.Kd2 Kf6 1.0 minute analysis by Stockfish 9 v010218 

Mar1818
  agb2002: White is one pawn down.
Black threatens dxe3.
The first idea that comes to mind is 19.Nxd5 exd5 20.Re6: A) 20... Nc5 21.Qb1 (21.Qxg6+ Qxg6 22.Rxg6+ Kh8 23.cxd4 followed by Rh6+ and Rxh5, and three connected passed pawns look like a very reasonable compensation for the knight) 21... Qf5 22.Rxg6+ Kh8 23.Rh6+ A.1) 23... Kg8 24.Qd1 (threatens Qxh5 and Qxd4)
A.1.a) 24... dxc3 25.Qxh5 looks winning (25... Ra7 26.Rg6+ Rg7? 27.Qh7#). A.1.b) 24... Qg4 25.Qxd4 and the threat Qh8# wins. A.1.c) 24... Ra7 25.Qxh5 with the threats Rg6+ and Rh8+ looks winning for White. A.2) 23... Kg7 24.Rh7+
A.2.a) 24... Kg8 25.Qd1 as above (25... Qxh7 26.Nxh7 Kxh7 27.Qxh5+ followed by cxd4 and White will resume the attack against the black king). A.2.b) 24... Kg6 25.Qd1 as above (25... Qxg5 26.fxg5 Rxf1 27.Qxf1 Kxh7 28.Qf7+ Kh8 29.g6 wins). A.2.c) 24... Kf6 25.Qd1 again.
B) 20... Qxe6 21.Nxe6 Rf6 22.Nxd4 +  [Q+P vs r+n]. 

Mar1818   njchess: I thought this was pretty easy for a Sunday. Given Black's position, the goal for White is {Qxg6+. Therefore, 19. Nxe6 Qxe6 20. Nxd5! Qf7 (Qxd5? 21. Re7 Rf7 Qxg6+ +) 21. Re7 Qf5 22. Qxd4 + The rest is tactics. 

Mar1818
  cormier: click for larger viewAnalysis by Houdini 4: d 23 dpa done
1. + / = (0.52): 14...Nxf6 15.Nf2 Bxd3 16.Qxd3 Qd6 17.Rae1 h6 18.Re2 Nc6 19.Rfe1 Rfe8 20.Ne5 cxd4 21.Nxc6 Qxc6 22.Qxd4 Rac8 23.Nd3 Ne4 24.g3 Qb7 25.Ne5 Rf8 26.Kg2 Rf5 27.c4 g5 28.cxd5 exd5 29.Rd1 gxf4 30.Qxd5+ Qxd5 2. + / = (0.66): 14...Rxf6 15.Bxa6 Nxa6 16.Nf2 Rff8 17.Rae1 Nc7 18.Qe3 a4 19.a3 Rae8 20.Ng4 Nf6 21.Nge5 Ne4 22.Nc6 Qd6 23.Nfe5 Rc8 24.Qd3 Rf5 25.c4 Rcf8 26.Rxe4 dxe4 27.Qxe4 cxd4 28.Nxd4 

Mar1818   stst: Late, but give the obvious line, harass the Q:
19.Nxd5 exN (no better continuation)
20.Re6 Qg7 (else Qd8, 21.Qxg6+ losing quicker)
21.Qg7 Rxg6 pinning Q, K ===> Q gone soonsee if I miss anything... 

Mar1918   stst: <.. would prefer 19.Nxd5 exd5..> Agreed!!
engine analyze the two choices:
Nxd5 vs
Nxe6 


< Earlier Kibitzing · PAGE 2 OF 2 ·
Later Kibitzing> 