Author: Peter Fendrich
Date: 10:51:31 02/01/99
Go up one level in this thread
On February 01, 1999 at 12:52:13, Bruce Moreland wrote: - snip - > >This is all a matter of definitions, and none of this changes program strength, >but if people are going to compare programs, they should compare them fairly. > >bruce That's true but really impossible... I don't know how node counting could be really 'fair' with different design of tree searching functions. I never count a QS node which doesn't generate any moves because that is nothing but a pure evaluation node. My programs history made me think like this. In an earlier version I could detect a situation like that even before calling the 'empty' QS node and did never call it but instead called the eval. The nodecount changes a lot with different approaches here... Once I had the idea to include the Qs search in the evaluation. That would give some options to narrow the alpha/beta window before calling QNode or entirely skip it. I'm sure that there are lots of other ideas around where the defintion of a node isn't clear, compared to other programs. The only case when I care about the node counts is when comparing different versions of my own program. //Peter
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.