Author: Mridul Muralidharan
Date: 23:47:21 01/13/05
Go up one level in this thread
On January 13, 2005 at 18:16:49, Dieter Buerssner wrote: >On January 13, 2005 at 16:36:47, Dann Corbit wrote: > >>>You can easily perform your own experiments with patzer and yace. You will see >>>benefit at every time control. >> >>I should mention that my bitbase experiments also had tablebase files activated. >> I do not know if this is important or not, and probably experiments should be >>also run with bitbase files only. > >While Yace bitbase code is designed with the idea to work (in most cases) >without TBs, much of this is not implemented (and perhaps never will be). For >example, say KRKN. A forced line to won KRKN position is found by search (with >the help of bitbases). Now the pos is on the board. If it is difficult to win >(which really might be the case in KRKN), Yace won't win in without TBs. Probing >TBs only at the root will be enough, however (it is possible to set this in Yace >- most conveniently when run as UCI engine). In typical computer match setups, >such a problem could easily be hidden. <snip> Assume the opponent is an engine with >TBs. It will probably resign ... > >Regards, >Dieter This is where yace will have tough time with my progs - they will never resign in this situation , since a) eval sucks - so wont see the mate coming (no mate patterns in eval) , b) I will not probe TBs when it is lost endgame :) - what is the point ! lose faster ?! ;) (currently - no TBs period - so even better against yuo :D ) Just kidding. Hope you got to your data recovered ! WIll try to catch you in icc soon Mridul
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.