Author: Thomas Mayer
Date: 21:23:23 01/27/04
Go up one level in this thread
Hi Russel, maybe you should look at the code a bit differently. What is the main objection of IID ? Simple -> you want to have a move for your move-ordering if you did not get one out of your Hashtables. Now with most engines you do not have a move when score is below alpha. So he must research. E.g. in Quark I must not research because I have almost ALWAYS a move to store, so I will find a move in my hashtables after the research... Anyway I had a bug in my IID and thanks to that discussion I get back to my IID code and found that bug... :) When does IID helps most ? Well, when the hashtables are full and already searched positions get overwritten... (at least I think so that it helps in those positions most) Since some time I have seen that Quark has problems when the hash is full - huger problems then others... after some first tests with the new code it seems that these problems did get smaller now... Greets, Thomas
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.