Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: One stupid(?) null-move question

Author: Christophe Theron

Date: 05:21:19 08/10/98

Go up one level in this thread


On August 10, 1998 at 04:59:04, Robert Hyatt wrote:

>Basically null-move fails whenever there is a zugzwang.  Which means that
>not moving is the best move.  There are few such positions when there are
>pieces on the board, but they do exist.  And in such positions, each time
>such a situation arises, the null-move fails high, not because the position
>is good for the side on move, but because the position is so bad that doing
>nothing is actually the best move...
>
>and no amount of searching will get around that, unless the null-move
>algorithm is modified in some way...  It is possible, however, to reduce
>the frequency of occurrences, but it means turning null-move off at selected
>nodes...


Here is a position where (I think) white is in zugzwang after Qd1!

Both sides have queen and bishop and many pawns, so it's quite unusual to be in
zugzwang here:

8/2b2pk1/6p1/p2p3p/P2P1P1P/2P3P1/2qB2QK/8 b - - ; Qd1!

The position is not quite clear, and it could be that Qxa4 is the best move
(Tiger finds it after 20 seconds, but I don't know if it is correct). Qd1 seems
to lead to a draw.

I would like to have the opinion of experts on Qd1. Is there really a zugzwang
here? Here there programs that fail to find Qd1 or Qxa4?

If you have a zugzwang position with a lot of material on the board, please post
them here. Thanks!


    Christophe



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.