Tactical Blind Spots? (Learning from a Bxh7+ sac that should have gone wrong.)

May 13, 2011, 6:25 PM |

I recently accepted a random challenge here on Chess.com from someone 400 points lower rated.  Analysis shows that I should have lost this game against good defense, but instead I mated my valiant opponent in 15 moves.  The point of showing this game is not to show off but to consider some interesting questions about practical chess decisions.  Here's the game:

Now I initiate the Bxh7+ move.
After Ng5+, ...Kg8  ...Kh8  and  ...Bxg5 all lose.  I looked at ...Kg6 because I remembered reading in Vukovic's "The Art of Attack in Chess" that it sometimes works, and I decided it was dangerous enough for Black that I could live with it if he played it.  But what about ...Kh6(!!)  This move wins.  It looks totally insane but there is no way I can find for White to capitalize on the Black king's weird position.  See next diagram:
In the ...Kh6 line, Black appears to be winning, and my computer agrees, but it might take a master player to prove it.  Black's king is mighty exposed out there on the h-file, and there might be chances for White if Black makes even a slight inaccuracy.  Here's what this game left me wondering:
  • How could I have looked at every legal reply to Ng5+, and totally missed ...Kh6?  I had 2 days per move, for Pete's sake.
  • If you had seen ...Kh6 before Bxh7+, would you still play it?
  • If you had seen ...Kh6, and correctly evaluated that White might be busted, would you still play it hoping to come out on top in a complicated position where Black's king is stranded on a forward square?
  • Most importantly:  Is the prospect of winning a nice attack worth losing due to an unsound sacrifice?
Let me know your thoughts.  I learned a lot from this little game.  Thanks for reading.