Author: Tony Werten
Date: 23:56:17 01/20/02
Go up one level in this thread
On January 20, 2002 at 23:32:48, Robert Hyatt wrote: >On January 20, 2002 at 20:03:45, Bas Hamstra wrote: > >>Using PVS, I don't use an aspiration window for quite a while now. As far as I >>can see it hardly matters. How is your experience? Data? >> >>Best regards, >>Bas. > > >It makes a significant difference. Set up a wild position where there are a >zillion mates but none are forced. The aspiration window will let them get >pruned, while the full-window (on the first move) will make you track down >every last mate score. In fact, a tier of fail low/high windows at the root >work even better although I have not gotten around to putting this back into >my aspiration search... This works even better for me since a lot of my extensions are a-b bound. A infinity window would fire them all up at once ( well, half of them). Using a series of windows (.2 pawns,1 pawn,knight,queen,checkmate) takes care of this. It takes a bit longer when something big is happening, but at the moment my bestmove fails with more than a knigh, it's time to spend extra time anyway. BTW most of fail high/low are within 1 pawn. Tony
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.