Author: Bas Hamstra
Date: 11:00:28 07/10/01
Go up one level in this thread
On July 10, 2001 at 13:44:26, Rich Van Gaasbeck wrote: >When a search is extended when it hasn't reached "quiescence", what exactly >that mean? Does it mean that no captures are possible? Is so, it would seem >that seaches would be extended considerably. The idea of quiescence is that it is very difficult to evaluate a position accurately when you are in the middle of some wild combination. Therefore you search for a quiet position first. On top of the normal full width search you do a limited search of only captures and/or checks. If no fast material wins are possible the evaluation is more reliable. However, if you do this, about 90% of the nodes are searched in the qsearch. So it appeared efficient to restrict the qsearch by trying to exclude nonsense captures. >Alternatively, does it mean that the only captures possible are of pieces of >lower value than the capturing piece? Is this is so, then it seems that you >would always stop searching right before the beginning of a combination (which >to me seems like a very non-quiet position). No, what you mention is not commonly used. But there is a better way to skip dumb captures: before trying N x p you examine what pieces of both sides attack the pawn. If it's "more defended than attacked" you skip that pawncapture. Bas.
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.