Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: new thoughts on verified null move

Author: Miguel A. Ballicora

Date: 17:20:17 11/23/02

Go up one level in this thread


On November 23, 2002 at 18:43:46, scott farrell wrote:

>On November 23, 2002 at 11:19:39, Miguel A. Ballicora wrote:
>
>>On November 23, 2002 at 08:11:37, scott farrell wrote:
>>
>>[snip]
>>
>>>I know that measuring nodes for a fixed search depth sounds like good science,
>>^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>>
>>IMHO, it is a big mistake when using tactical positions.
>>Nodes to solution should be a better parameter.
>
>Which is largely what I was getting too, but also making a note that a solution
>can be found WITHOUT having to complete the ply in question.

Yes, I have used this procedure a few times. It takes a little more time
sometimes but I think that it has less "noise". It is very important to select
good positions. Those when it is clear that the program understand what's going
on when it "sees" the "best move" without any doubt. The problem is how to
handle the data later. Doing a simple average is very wrong IMHO. If anybody is
interested, I can post what I think it is best, I am in a rush now and I need to
explain it in detail to be understood.

Later,
Miguel


Miguel



>
>>
>>Miguel



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.