Computer Chess Club Archives




Subject: Re: Does your program understand castling/en passant rights on 3x repetition

Author: blass uri

Date: 08:50:43 02/27/00

Go up one level in this thread

On February 27, 2000 at 09:46:22, Mike S. wrote:

>On February 27, 2000 at 03:04:05, Tom Kerrigan wrote:
>>I don't think exact adherence to these rules is necessary.
>No? You may want to try this position with your computer:
>Zepler # in 4 - [+1430.04e5e8]
>Die Schwalbe, 1929
>[D]r3k3/2Qp3R/1p6/1b2K3/4p3/8/5p2/8 w q - 0 1
>1.Kd4 (threats 2.Qe5+ and 3.#) [1.Qd6? 0-0-0] 1...Ra4+ 2.Ke5 Ra8 3.Qd6 1-0
>I was busy with this issue when a wrote a short article for CSS 4/1995. All the
>programs I tested then, didn't apply the correct rule (Fritz 3, Rebel 6, W Chess
>1, Genius 3 and Hiarcs 3). It was remarkable that the mistake happened in
>problem mode also.

This is really a problem but most programs are not for solving problems but for

> CB's Mate 2.0-engines handles it correctly.
>Btw., I'm sure programmers know the rule very well of course, but when it's
>disregarded it is been done to gain more search speed I think. I have never
>heard of any important, or not even unimportant, game where this would have
>But if this would happen (incorrect claim of a draw, program refuses to
>continue), I think there would be no special tournament rules for this and the
>unavoidable anger, dispute and scandal would follow.

There are clear rules that if the program does not do a move and claims a draw
incorrectly when there is no draw then it loses on time.

There will be no scandals  but only a loss.

I guess the practical chance that it happens in a game is less than 0.00001
and it is more productive to fix other problems.

 Such things usually arise
>in the most important game of a championship. I will be amused...

There is no reason that they arise in the most important game.
The chances are the same for every game.


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.