Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: perpetual check recognition

Author: Ulrich Tuerke

Date: 13:59:43 01/30/04

Go up one level in this thread


On January 30, 2004 at 13:22:45, Andreas Herrmann wrote:

>On January 30, 2004 at 12:55:45, Ulrich Tuerke wrote:
>
>>Andreas, are you also checking for repetitions when you are in quiescence search
>>?
>>
>>Uli
>
>Hi Uli,
>
>no i don't, because i search only capture and promotion moves inside the qsearch
>and there can't be a repetition. Perhaps i will try to generate also check moves
>in quiescence search in the future. Have never tryd it sofar.

I think that you are right: in case you have captures and promotions only, no
repetition can happen. Comet also generates some checks in quies search.

In fact, I try a little trick: in case the same piece gives 3 checks in series
without a capture in this line, I cut the line after the 3rd check and return a
zero score, provided this series is not too close to the root - and also
depending on the iteration number. The required distance to the root grows with
the iteration number. This helps in some rare cases to detect a perpetual one
ply earlier. But one has to be very careful with this because it implies a
certain risk.

Uli
>
>Andreas



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.