Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: KR against KQ is draw without tablebases

Author: Robert Hyatt

Date: 18:01:51 01/27/03

Go up one level in this thread


On January 27, 2003 at 20:06:37, Ricardo Gibert wrote:

>On January 27, 2003 at 16:38:54, José de Jesús García Ruvalcaba wrote:
>
>>On January 27, 2003 at 16:15:32, Daniel Clausen wrote:
>>
>>>On January 27, 2003 at 16:00:40, José de Jesús García Ruvalcaba wrote:
>>>
>>>>On January 27, 2003 at 12:42:09, Daniel Clausen wrote:
>>>>
>>>>>On January 27, 2003 at 12:35:50, Georg Langrath wrote:
>>>>>
>>>>>>I tried to let Chessmaster 9000 play against itself in endgame KRKQ. The side
>>>>>>KR with tablebases and the side  KQ without. KQ-side wasn?t nearer checkmate in
>>>>>>move 50 than in move 1, even with 3 minutes per move.  It is rather amusing to
>>>>>>look at. I think that it is impossible for a computer to solve without special
>>>>>>knowledge or tablebases as in the case KBNK.
>>>>>>Georg
>>>>>
>>>>>Now do the experiment again but let a strong human play for the KQ-side and see
>>>>>whether the result is any different. ;)
>>>>>
>>>>>Sargon
>>>>
>>>>A strong player will win this ending after the second or third try if he has not
>>>>practised before. It is not really difficult, even a weak player like I can win
>>>>against a computer with tablebases after some practise.
>>>>José.
>>>
>>>Wasn't there this experiment with some GMs who failed to win when they played
>>>against the TBs?
>>
>>Yes, it was this. GM Walter Browne could not win against Belle. But he tried
>>again after a few days and won quite easily. I think he could have won the same
>>day (after a few tries) if he were not so shocked.
>>José.
>
>
>Since you seemed to be inventing an explanation I had never heard before, I got
>curious and decided to do a little internet searching on the matter. I found the
>following link which thoroughly contradicts your characterization that WB "won
>quite easily" in the rematch:
>
>http://gopher.quux.org:70/Archives/usenet-a-news/NET.chess/82.01.07_sri-unix.458_net.chess.txt
>
>The ending is tough, even for a GM against good defense. To think otherwise is
>to underestimate it.
>

It is tough for a human, I agree.  I was at the "rematch" and if I recall,
GM Browne mated on move 50, not anywhere near the optimal number and nearly
turning a win into a draw...

And it might be that he won the rook on move 50, rather than mating...  That was
a _long_ time ago.

But for computers, it really should be trivial...

>
>>
>>>Or am I mixing this with another endgame?
>>>
>>>Sargon



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.