Author: Mike S.
Date: 01:17:43 06/07/03
Go up one level in this thread
On June 07, 2003 at 02:32:24, Jim Bodkins wrote: >(...) or (and I have seen this) it blunders badly late in the >middle game and is immediately forced into a disadvantaged endgame where it >loses. This happens more often than I would like actually. But this seems to be >one of the more common means of losing a game. Sounds plausible, because in the endgame itself (an equal position given), King is usually considered to be one of the strongest engines. (The other involves a serious >weaknes in finding solutions deep in the endgame even with a corrected KNNK >EGDB. It's embarrassing when theking playing a K and P vs a K allows the other K >inside it's king and loses it's pawn. How does that happen? Even I could play >that sequence correctly :) but theking is as likely to blow that as to get it >right). This sounds as if either your EGTBs are broke (did you generate them with the FEG 3.23 again after the second bugfix?), or these were positions which were draws anyway, IOW where it didn't matter anymore if King keeps the pawn or not. You could cross check that with the Nalimovs to see if that really were blunders missing an win (I wouldn't expect it). >It is fun experimenting though. (I have sworn off CB which means I may have >played my last game vs CB engines. Unless they are (shredder maybe) played at >ICS). My impression was, that especially Fritz 8 may be much more vulnerable to King 3.23 than Shredder 7. I've won a few games against Fritz 8 users which had much faster hardware, on Playchess.com. But for these games, King could use Fritz 7's big opening book... although, for most games I have even limited the book depth with the new option, to give King the chance to create one of his surprising gambits or the like (but that option didn't work always - I don't know why - so the complete F7 book variants were used sometimes, too). I used my CM9K "modified" Personality, this is Sel. 12, King Safety 150/150, and 32 MB Hash (as a kind of "easy compromise" of many successful settings), rest default. Regards, M.Scheidl
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.