Author: Robert Hyatt
Date: 20:09:34 12/31/00
Go up one level in this thread
On December 31, 2000 at 14:46:06, Georg v. Zimmermann wrote: >>But note that I do not count quiescence searching at all since move ordering >>is a lot simpler there. And note that I only count this on nodes that fail high >>on some move. If it fails high anywhere, in increment a counter. If it fails >>high on the _first_ move searched, I increment another counter (I don't count >>the null-move as a move here, for obvious reasons). > >Thanks for the clarification, I was wondering about how to meassure this only a >few days ago. > >Do you remember a position with very low firstmove_cutoff rates ? What was the >lowest percentage you observed in recent crafty versions ? Did you ever test the >distribution of fh on all possible moves , not only the first ? > >Best regards, > >Georg v. Zimmermann I did test the distribution, which is how I came up with the decision on where to give up on history moves and just take the rest in the order generated. And no, I don't have a good example of a bad fail high percentage, but I just did a quick check over a dozen games and the lowest number I saw was 88%...
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.