Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: If you like to solve real mate...

Author: Tony Werten

Date: 11:07:07 01/27/01

Go up one level in this thread


On January 27, 2001 at 11:09:10, leonid wrote:

>On January 27, 2001 at 09:10:05, Tony Werten wrote:
>
>>On January 27, 2001 at 08:53:07, leonid wrote:
>>
>>>Hi!
>>>
>>>If you like to solve forced mate position from real game, it is the one.
>>>
>>>[D]7r/p4R2/1p4nk/3qP3/P1pr4/6PB/2Q4K/5R2 w - -
>>>
>>>It is from game played in Hollannd in 1975 between Ernsteen and Berchem.
>>>
>>>Leonid.
>>
>>XiniX solves this one in 1 sec  1. Qxg6+ Kxg6 2. Rf6+  , Kg5 3. Rf5+  , Kg6 4.
>>Rf6+  , Kg7 5. Rg5+  , Kh7 6. Bf5+
>
>If you can mention what is this program and on what computer you solved, it will
>be appreciate.

XiniX is a chessprogram I wrote in Delphi. It finished 10th ( out of 14 ) at the
last dutch championship. After the first weekend I was on shared 3rd place, the
second weekend disaster stroke.

I had a little correction-for-depth-by-checkmates mistake. In the game against
Ant I had 3 queens against a pawn but my engine refused to mate, giving me a
loss on time ( I forgot to claim a draw ).

In another game I oversearched. ( new vocabalary ) I outsearched my opponent by
3 ply, went for what should be a pawn win, but when the game arrived there I had
an undeffendable pinned knight. I should have searched to 2 ply more or 3 ply
less and there wouldn't have been a problem.

Machine is an AMD K6-2 333 Mhz

Tony

>
>Time is good.
>
>
>>This way it is easy.
>>
>>The position in BT2630 is a couple of moves earlier. ( I never managed to solve
>>this one )
>>
>>[D] 2b1q3/p7/1p1p2kb/nPpN3p/P1P1P2P/6P1/5R1K/5Q2 w - -
>>
>>Tony
>
>And in how many moves this position expected to be solved?
>
>I tried it by selective search 13 moves deep but, after waiting around 6
>minutes, disconnected. In 10 moves selective could find nothing. By brute force
>even 5 moves lead me already to 5 minutes search. No mate existe in five.
>Thought that maybe it is some 20 moves position that is, anyway, beyond my
>actual program.
>
>If somebody would like to solve average positions (not too easy but not too
>complicated), below is one.
>
>[D]qn1kbrqn/n1nqn1bp/bN1N1N1N/B1N1Nn1B/N1N1N1N1/8/1R1Q1R2/3K4 w - -
>
>Mate existe for sure in 11 moves. Response was found by selective search and
>minimal number of moves is unknown.
>
>Leonid.



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.