Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Problem Solved!!!

Author: Vincent Diepeveen

Date: 12:00:05 10/10/00

Go up one level in this thread


On October 10, 2000 at 10:14:47, Steve Maughan wrote:

>Vincent,
>
>>i have 85 to 86% there, so not much better.
>
>Hmm - interesting - I had thought that 90% - 92% was the 'standard'.  What move
>order do you do?  I do
>
>1) Hash
>2) Good Captures
>3) Killer x 2
>4) History
>5) The Rest
>
>>what is your fliprate: chance that a move which in hashtable was stored as
>><= alfa node (or pv node) that it gets >= beta now?
>
>What is the significance of this?  Is it just the proportion of Hash moves that
>produced a cutoff?

it's more significant as your cutrate if you talk about tree size at
bigger depths (though must take nullmoe and hashtable transpositions
and qsearch also into account) as you don't need to research the same
trees then. where with cutmove you search usually a relatively
small tree.

>I am using quite a simple hash replacement algorithm - maybe it could be
>improved.  I have two tables one is replaced if there is a deeper search the
>other is replaced in all other sitiuations.  Do you think multiple hit attempts
>adds much - I cann't see it myself.
>
>Steve



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.