Author: Bruce Moreland
Date: 09:58:25 10/03/99
Go up one level in this thread
On October 03, 1999 at 11:28:19, Frank Phillips wrote: >I have been implementing SEE in the quiescent search, which slow the nodes per >second down but decreases the time taken to reach a fixed depth and so is a win >overall. > >The SEE discards from further consideration (in the quiescent search) those >captures which apparently loose material, by working through the sequence of >captures on the target square. All standard stuff – I think. > >One of the problems with this approach is that a capture sequence focussed on a >particular target square may appear to win material but, in fact, does not >because one of the capturing pieces is pinned against a piece of higher value >and so cannot move. > >Do other just use SEE to order the sequence of moves or throw them out as futile >– as Crafty does if I understand the code. Using SEE to only order moves >largely negates its advantage. It is OK to make some mistakes, if the result is better play in the vast majority of cases. Mine works like Crafty. bruce
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.