Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: evaluating KNNN-K

Author: Rafael Andrist

Date: 06:43:11 11/18/01

Go up one level in this thread


On November 18, 2001 at 06:03:57, Uri Blass wrote:

>On November 18, 2001 at 05:51:03, Rafael Andrist wrote:
>
>>According to the knnnk.tbs file, ~ 85,3% of the possible KNNN-K positions are
>>won. Is this enough to consider KNNN-K as a general win? Or can at least be said
>>that there's a high chance for a win?
>>
>>Rafael B. Andrist
>
>It is enough to consider KNNN vs K as a general win.
>
>I doubt if there is a KNNN vs K drawn position when the weaker side cannnot
>force capturing the knight or getting a stalemate in a few moves.
>
>Uri

Thanks. I the meantime I checked some positions in the TB's and saw that even
the ones where Black has a positional advantage are won for White if there isn't
a forced capture or a stalemate.

e.g.
[D]N6N/8/8/4k3/8/8/8/K6N w - - dm 20;
(Black to move draws because of capturing a knight)

or
[D]8/8/5k2/8/8/8/NK6/NN6 w - - dm 17;

Because the win is not easy to find by search but not impossible, I think I will
use an value between +2 and +4 [pawns] for this kind of positions.

Rafael B. Andrist



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.