Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Nullmove-Traps

Author: Stan Arts

Date: 09:44:00 04/17/04

Go up one level in this thread


On April 17, 2004 at 12:18:40, Eric Oldre wrote:

>I experienced a similar situation in a test game last week. My engine had 2
>queens a rook and a few pawns to the opponent's king and bishop, but it refused
>to mate.
>
>I had logic to not use null move in the endgame but due to the amount of
>material one side had it did not register that is was the end game.
>
>My solution was to change the function which detected when it was the end game.
>Instead of using all the material on the board, it would base the decision on
>the amount of material the side with less material had. so if either side is
>down to it's last few pieces, it would stop doing null move.
>
>It seemed to do the trick for the particular board position I had the problem
>with, but i'm not sure what if any additional problems it may cause...
>
>Have other people used this approach with any luck? (i'm sure someone has at
>least tried it)
>
>Eric Oldre
>

Hi Eric

When you are looking at material to deside if it's safe to do a nullmove, you
only need to look at the material of the side that's going to do the pass/
nullmove, the amount of material of the other side isn't all that important,
at that point. For instance : If one side has a king, queen, a bishop and a
bunch of pawns, and the other side a lonely king, then it's sort of safe to do
nullmoves for the side with all the material, but not for the lonely king.
Where the king might be forced into a near mating position by having to move,
but would be safe one or several moves if not. That would be the danger of a
nullmove then for instance. However the side with all the material will
probably be worse when it doesn't move, and then nullmove is beneficial, and
not dangerous.

My program Neurosis P3-550 48 MB hash , on the position Joachim gave.
It uses a different form of nullmove, and is not affected by the zugzwangs in
this position.

Legal moves: 22   Static score: -1092
Ply Score Time     Nodes      Best move and expected line
------------------------------------------------------------
2   -1092 6        6281       Qe8+ Kh4 Qg6
3   -1092 6        7672       Qe8+ Kh4 Qg6
4   -1092 44       40423      Qe8+ Kh4 Qg6 Kh3 Qf5+ Kg3 Qg6+ Kh2 Qd3 Qg6 Qd3 Qg6
5   -1092 127      116639     Qe8+ Kh4 Qg6 Kh3 Qf5+ Kg3 Qg6+ Kh2 Qd3 Qg6 Qd3 Qg6
6   -1092 517      481303     Qe8+ Kh6 Qf7 Kg5 Qg8+ Kh5 Qe8+ Kh6 Qf7 Kg5 Qg8+
Kh5
6   9993  605      559337     Qg7 < + > Kh4 Qg6 Kh3 Qg5 Kh2
6   9993  610      566481     Qg7 <M4> Kh4 Qg6 Kh3 Qg5 Kh2 Qh4+
7   9993  852      784274     Qg7 <M4> Kh4 Qg6 Kh3 Qg5 Kh2 Qh4+
Total nodes:1416603 n/sec:90100  (Q-nodes:10% max depth:12)

Goodluck and greetings
Stan







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.