Author: Slater Wold
Date: 21:16:36 06/16/02
Go up one level in this thread
On June 16, 2002 at 23:38:38, Uri Blass wrote: >On June 16, 2002 at 14:47:06, Telmo Escobar wrote: > >> I examined this game with several CB engines and all of them, in due time, >>evaluate 12.Nxg5! as winning for White. Below are the relevant lines. >> Main one, in my view, is 12.Nxg5 hxg5 13.Bxg5 Nbd7 14.Qf3 a6 15.Qg3 Kf8 -as >>played in the game- and now 16.f4! > >Some analysis of Movei(p200 128 Mbytes after 16.f4) > >10 -11 171868 45737870 c7b8 e4e5 d6e5 a1e1 d8b6 g5h6 f8e7 f4e5 e7d8 d5d6 e8g8 >11 -54 435241 110025087 c7b8 e4e5 d6e5 a1e1 d8b6 f4e5 f6h5 d5d6 h5g3 f1f7 f8g8 >f7d7 e8e6 c4e6 g8h8 h2g3 c8d7 e6d7 b6b2 >11 -54 566140 145189288 c7b8 e4e5 d6e5 a1e1 d8b6 f4e5 f6h5 d5d6 h5g3 f1f7 f8g8 >f7d7 e8e6 c4e6 g8h8 h2g3 c8d7 e6d7 b6b2 >12 -153 805022 211276281 c7b8 e4e5 d6e5 a1e1 b7b5 g5h6 f8e7 d5d6 e7d6 f4e5 e8e5 >f1f6 d7f6 e1e5 d6c6 g3f3 c6b6 c4f7 b8e5 f3a8 >12 -152 1065835 283549797 a8a7 >12 -150 1263166 335461482 a8a7 e4e5 d6e5 a1e1 b7b5 g5h6 f8e7 f4e5 d7e5 f1f6 e7d7 >f6f7 e5f7 g3g4 d7d6 g4g6 c8e6 e1e6 >12 -150 1469994 393603511 a8a7 e4e5 d6e5 a1e1 b7b5 g5h6 f8e7 f4e5 d7e5 f1f6 e7d7 >f6f7 e5f7 g3g4 d7d6 g4g6 c8e6 e1e6 > > >Movei does not like b5 and prefers Ra7 but clearly does not like black's >position. > >Movei agreed with the previous moves and gave a small advantage for black so it >is clear that it saw the tactics. > >Uri I believe that this position exploits the fact that *most* programmers give a bonus for captures. My program (which I just recently started working on again) has NO positional knowledge, and is strictly based on piece value. It will not look at the capture. If I give a normal, small bonus for a capture, it chooses the capture/sac. The difference between the two is about .15 pawns. Which, if I give a .5 pawn bonus for a capture, is enough to cover the spread. In all honestly, I don't know for sure if it's a *good* move (the capture/sac). But IMO, it is not the best move.
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.