Author: David Blackman
Date: 23:46:04 03/01/00
Go up one level in this thread
On March 01, 2000 at 17:14:29, Dave Gomboc wrote: >On March 01, 2000 at 13:56:49, John Coffey wrote: > >>The transposition table will tell us if a move is more desired than the other >>moves. So will iterative deepening and iterative iterative deepending. In both >>cases it seems necessary to sort the moves from most prefered to least prefered. >> What is the method for doing this? Is it just a standard sort routine? >> >>When doing move generation on a position that we haven't seen before, could >>we do things like sort captures to the top of the list, especially when a more >>valuable piece is being captured. >> >>John > >Sometimes people generate only some moves, in the hopes that there will be an >early cutoff and they won't have to generate the rest of them. > >Sometimes people generate all the moves, but use selection sort for the first >few moves, because there's no point in sorting the whole list if there's an >early cutoff. > >When your program notices that a cutoff isn't likely to happen, you might want >to generate the rest of the moves, or switch to a sort with better complexity. Or not bother sorting the rest, just do them in whatever order. If your pretty sure you won't get a cuttoff, the order doesn't matter much, because you will probably search them all anyway.
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.