Author: Uri Blass
Date: 04:52:08 11/26/02
Go up one level in this thread
On November 26, 2002 at 07:44:06, David Rasmussen wrote: >On November 26, 2002 at 05:11:18, Uri Blass wrote: > >> >>Movei find nothing strange in that position and inspite of doing checks in the >>qsearch. >> >>The number of qsearch nodes is less than 50% of the number of nodes. >> >>From the first 38159802 nodes it has 17497978 qnodes. >> > >Mmm. Weird. I must have a bug or something. On the other hand, this bug can't be >too grave, since Chezzz have an OK record against other amateur engines running >on the same hardware. Chezzz's standard rating on ICC is about 2400 and it's >blitz rating is about 2700. > >I'm confused... > >/David I do not know if you have a bug. My qsearch is different than your qsearch and my evaluation is different so we search different trees. I also limit my qsearch to 7 plies but I do not think that this is the reason for the different results. the limit was done not to prevent cases when 98% of the time is in qsearch but to prevent caes when it needs a long time to finish the first ply. Maybe I may get a small gain by avoiding that limit but I do not plan to do it. Uri
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.