Author: Sune Fischer
Date: 05:39:23 06/14/04
Go up one level in this thread
On June 14, 2004 at 01:58:29, Kurt Utzinger wrote: >On June 14, 2004 at 01:06:04, Derek Paquette wrote: > >>On June 14, 2004 at 00:42:00, Uri Blass wrote: >> >>>On June 14, 2004 at 00:18:08, Derek Paquette wrote: >>> >>>>Hi there, i read your post about Petrovich vs S8, and you wondered if 'prefer >>>>open positions' was selected that it would do better, >>>> >>>>well i did that, vs a guy named FAther on playchess server, he's well known for >>>>making a stone wall defence, HOWEVER in addition to that, i also made another >>>>tweek to the personality, changing these things >>>> >>>>King Safety 94 >>>>pawn structure 96 >>>>pawn structure (endgame) 96 >>>> >>>>this combined with the open setting allowed Shredder 8 to....draw, >>>>actually a lose, because it was a 5min game and he would have won on time, >>> >>>I do not see how can white win on time against a chess program unless you play >>>manually but chess programs were not made to play manually. >>> >>>I think that black is winning and shredder's evaluation is correct. >>> >>>Black has a rook for a knight so I guess that the 50 move rule can convince >>>black to push the pawns forward and even sacrifice a pawn if it is needed and >>>win the game. >>> >>>Uri >> >>Father routeenly wins against dual xeon and opetron systems, >> >>he wins on time, he just moves his king back and forth, and the program runs out >>of time, and loses >> >>he's famous for that on playchess >>makes everyone get really pissed, but it works > > Blitz without increment is not the right way > to measure something. > Kurt Everybody has their own idea about some specific settings that must be used or the test is invalid. I don't buy those excuses, I think blitz with no increment is perfectly fine for testing how it handles in blitz with no increment. And blitz with increment is better for testing how it handles in blitz with increment. Etc.. -S.
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.