Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: draw detection anomalies

Author: Will Singleton

Date: 09:07:39 11/23/05

Go up one level in this thread


On November 23, 2005 at 09:37:30, Charles Roberson wrote:

>
>   I've experimented with both methods. My results for method 2 were identical
> to yours and were caused by an horizon effect. Simply put: the draw wasn't seen
>because the second repeat was delayed beyond the horizon by a series of moves.
>This is the same problem as not seeing a mate due to a long series of delaying
>moves/sacrifices. Of course, various types of extensions reduce the
>probabilities of mate delays being an issue but think of it without the
>extensions.
>
>     However, there is a third option which is used by some of the others. The
>idea is to have a no progress penalty. So use method two, but when you see the
>first rep assign a no progress penalty. This may help. I had a conversation with
>Tony Werten (Xinix) about this, I believe he uses it.
>
>   Charles

I don't know if assigning a penalty for a first rep will work, it might have
unintended effects.  I do have a no-progress penalty, but it's just for the case
when the king is getting pushed all over the board without repetitions.

Will



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.