Author: Robert Hyatt
Date: 20:58:58 11/01/99
Go up one level in this thread
On November 01, 1999 at 23:39:01, William Bryant wrote: >As my search gets more complex, I begin to understand more of these >elements. I could use some help on the implementation of IID > >As I understand it, >While searching the pv root node (while no score has yet to be backed up and >alpha and beta have not yet been changed from there root values) if there is >no best move, need to search at a reduced depth and an find one. > >if the value from this reduced depth is <= alpha, research >at a reduced window, then using the move that gets backed >up at pv[ply][ply]; > >if it is >alpha and <beta -- then the move to use should be backed up on the >pv array and found at pv[ply][ply]. > >if the value is >= beta, what ever move generated the cutoff should be used >as the best move. > >Is this close. Any (preferably detailed) explinations of IID would be >appreciated. > >William >wbryant@ix.netcom.com sounds perfect to me... note that it is _possible_that you won't get a move, so be prepared. Because you could be mated, stalemated, one move beyond a three-fold repetition, etc...
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.