Moves preceding a blunder may make a lot of sense.
I know from experience.
Yehuda Gruenfeld White
Andrew Soltis Black
Lone Pine 1979
Modern Defense B06
1 e4 g6 2 d4 ♗g7 3 ♘c3 d6 4 ♘f3 a6 5 ♗e2 b5 6 0-0 ♗b7
I wanted to win the e-pawn, after 7...b4 8 ♘d5 e6.
7 ♗g5 ♘d7
White was poised for 7...♘f6 8 e5! dxe5 9 ♘xe5, with advantage (9...0-0 10 ♗f3).
8 ♖e1 h6 9 ♗h4 ♘b6
I rejected the consistent 9...g5 10 ♗g3 e6 because 11 e5 dxe5 12 ♘xe5 is a bit worse for me.
10 a3 c5
But 10...g5 11 ♗g3 ♘f6 was reasonable (12 ♕d3 ♘h5 and ...♘xg3).
11 e5 cxd4
Yes, it’s risky to open the position when uncastled. But his center is collapsing.
12 ♕xd4 dxe5 13 ♕e3 ♕c7 14 ♗g3! f6?? 15 ♘h4! ♔f7
Now I expected 16 ♗d3 f5, when 17 ♗xe5 ♗xe5 18 ♕xe5 ♕xe5 19 ♖xe5 ♘d7 looked defensible.
But instead there came:
16 ♕d3! f5
Or 16…e4 17 ♘xe4 ♕d7 18 ♘d6+! and 19 ♕xg6+.
17 ♘xg6!
Murderous (17...♔xg6 18 ♗h5+! mates).
17...♕d7 18 ♗h5 ♕xd3 19 ♘f4+ resigns.
It’s 19...♔f8 20 ♘e6 mate or 19...♔f6 20 ♗h4 mate.
Eduardas Rozentalis White
Alexander Wojtkiewicz Black
Geneva 1990
1 e4 g6 2 d4 ♗g7 3 ♘f3 d6 4 ♘c3 a6 5 ♗e2 b5 6 0-0 ♗b7 7 ♗g5 ♘d7 8 ♖e1 h6 9 ♗h4 c5 10 a4 cxd4 11 ♘xd4 bxa4 12 ♖xa4 ♘c5 13 ♖b4 ♘f6 14 ♘d5 e6 15 ♖xb7! (15...♘xb7 16 ♘c6) exd5 16 ♘c6 ♘xb7 17 ♘xd8 and wins.
Question 99: Why not 14…♘xd5 15 exd5 ♗xd5 ?