Author: Robert Hyatt
Date: 07:24:11 06/14/98
Go up one level in this thread
On June 14, 1998 at 01:27:48, Stuart Cracraft wrote: > >On June 13, 1998 at 13:49:42, Peter Kasinski wrote: > >>The advantages are obvious. >>What are the possible negative consequences? >>What is the right way to implement it? >> >>Thank you kindly, as always. >> >>PK. > >Perhaps we could expand on your post by talking generally >about "razoring" and what it's all about. Wasn't this coined >by Beal and is really more an English aphorism? What's >the rest of the world call it? > >--Stuart I have a vague memory it was coined by Kent/Birmingham years ago. It is a mechanism for reducing the depth along branches that are not worth searching. When you get to a node within one ply of the quiescience search (at ply 7 in an 8-ply search, for example), and the current score is bad enough (below alpha), for each move you can ask "is there any hope that this move can recover the lost material?" If the move is a non-capture, non-check, non-tactical-threat, then just drop right into quiesce rather than another full-width ply before going to the capture search. The only trick is to be *sure* that the move you are reducing the depth on cannot recover the missing material. Works well, *if* you control what you razor...
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.