Author: Bruce Moreland
Date: 11:24:01 06/22/98
Go up one level in this thread
On June 22, 1998 at 13:56:57, Don Dailey wrote: >I have never used the counter idea. I thought about it and assumed >it would not be very effective. But I would like to give a try >to satisfy my curiousity. The reason I didn't try it is because >once a move got put in the "winner" slot, even if it was not >a very popular move, it might never get bumped since it is possible >for other moves to constantly "knock each other out." > >Move A count = 10 >MOVE B count = 1 > >Move B comes along, replace move C new count 1 >Move C comes along, replace move B new count 1 >Move B comes along, replace move C new count 1 >Move C comes along, replace move B new count 1 When you open a node, clear the killers at the next ply. So at the top of search, you zap these, and the first time you try a candidate move, the killers at that candidate's depth are fresh. This seems insane but maybe it isn't. I've been doing this for so long (I didn't invent it), that I haven't questioned, but now that I think about it I can think of some other experiments. bruce
This page took 0.01 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.