Author: Don Dailey
Date: 09:38:55 12/18/97
Go up one level in this thread
On December 18, 1997 at 08:54:32, Bas Hamstra wrote: > >On December 18, 1997 at 08:24:12, Dan Homan wrote: > >>On December 18, 1997 at 07:36:12, Bas Hamstra wrote: >> >>>Dan Homan wrote: >>> >>>>On move ordering.... >>>> >>>> Bob's point about using killers after captures because they are faster >>>>than generating the rest of the moves (for use with the history >>>>heuristic) >>>>is an interesting one. That must be bit-board thing, because I don't >>>>see a good way of doing that in my current framework... I can see >>>>how it would be quite a bit faster if I didn't have to generate all the >>>>moves at each ply. >>> >>>I might be mistaken, but wasn't the point of searching captures first >>>that it >>>ont he whole causes more cutoffs? >>> >>Yes, if you cause a cutoff, you don't need to generate the rest of >>the moves. I generate all the moves at the start, which is slower. > >Yes, if you search captures first it gives greater probability on a >cutoff than searching killers first. > >Bas Hamstra. Bob's suggestion is an intelligent one. Even if the killers do not reduce the node counts significantly they certainly will speed up the program if he can avoid move generation. They still perform their function as killers without the calculation and move generation time of the history heuristic. It's ironic that it's greater value in this case is a free speedup instead of a node reduction, but that is no reason not to take advantage of it. - Don
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.