Sergash: I checked the game with Komodo 9.2 64 bits.
Apparently the first imprecision in the game was <7.Ne2?! Nxf4 8.Nxf4 c6 > and Calrsen already has the better position! Instead, White could have played 7.Bg5 h6 8.Bh4 = Ahmad Najjar (2245) vs Nabil Saleh (2180), Asian Championship 1996 in Dubaï (United Arab Emirates), 0-1 <12...e4??> A terrible mistake!  click for larger view12...f5! and now :
A) 13.Nxe5 Bxe5!
A1) if 14.Qh5? (threatening Qh5-h7#) Bxh2+! 15.Qxh2 (not 15.Kxh2?! Qc7+! 16.f4 Qg7! 17.Rf3 (or 17.Qg5 Rf6! 18.Bh5 Be6 ) Nd7 18.Rg3 Nf6 19.Bh7+! Kh8 20.Rxg7 Nxh5 21.Rc7 Ng7! 22.Bg6 Kg8!  A2) 14.dxe5 Kg7 15.Bh5 Qh4 16.Be2! Be6! 17.f4 Rh8! 18.h3 Na6!  B) 13.dxe5! dxc4 14.Qe2! Qe7 15.Rad1 =
<13.Nd2??>
Funny how a mistake is often answered by another mistake! 13.Nh4! (with the terrible threat of Qd1-h5) Re8 14.Qh5 Kf8 15.f3! winning. <13...f5??>
Wow! The only playable move was 13...Bh6 / with a big and possibly winning advantage! <14.cxd5??>
Only move: 14.Qh5 Rf6 15.f3! Qd6! 16.c5! Qe6 17.Bh7+ Kf8 18.fxe4! Qf7! 19.Qxf7+ Rxf7 20.Bg6! Bh6 21.Rae1 Re7! 22.e5  <14...cxd5??> Played the recapture automatically withought much thought? 14...Qg5! winning, as the white bishop is trapped. <15.Rc1?? Nc6 >
Only move was again 15.Qh5 (threat Qh7#) Rf6 16.f3 Qd6! 17.Bh7+! Kf8 18.fxe4! Rh6 19.Rxf5! Bxf5 20.Qxf5+ Bf6 (or 20...Ke8 21.Qc8+! Qd8 22.Qxb7 Nd7 23.Bf5 Rd6! ) 21.e5 Qd7 22.Rc1! Qxf5 23.Bxf5 Bg7! 24.Rc8+! Ke7! / = |