Author: Ralf Elvsén
Date: 10:16:52 11/21/01
Go up one level in this thread
On November 21, 2001 at 11:20:25, Robert Hyatt wrote: >On November 21, 2001 at 04:33:10, Ralf Elvsén wrote: > >>On November 20, 2001 at 21:09:37, Gian-Carlo Pascutto wrote: >> >>>On November 20, 2001 at 21:02:45, James Swafford wrote: >>> >>>>So you're arguing my original point. After some consideration >>>>I think either approach is fine. >>> >>>I think it's silly not to limit if the limit doesn't cost >>>anything and may save you ass someday or in some freak >>>positions. >> >>I think it's silly to limit if the absence of a limit doesn't cost >>anything and may save you ass someday or in some freak >>positions. >> >>If one has a SEE maybe one can use the value of this function at some >>stage of course to limit the qsearch. >> >>Ralf >> >> > >Dont' search captures that appear to lose material. That and the fact that >each capture removes one piece is limit enough to keep the q-search reasonably >small... Yes, I agree. I have never noticed the "overhead", so why not search to the bitter end? The SEE-thing was just an idle thought. Ralf > >>> >>>>Or are you proposing something else? Perhaps a different depth or >>>>something dynamic? >>> >>>I use max quiescent depth = nominal search depth * 2 >>> >>>-- >>>GCP
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.