Computer Chess Club Archives


Search

Terms

Messages

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

Author: Heiner Marxen

Date: 14:19:28 02/06/01

Go up one level in this thread


On February 06, 2001 at 16:55:44, Uri Blass wrote:

>On February 06, 2001 at 12:02:30, Heiner Marxen wrote:
>
>>On February 06, 2001 at 10:17:01, Paul wrote:
>>
>>>On February 06, 2001 at 08:59:10, Heiner Marxen wrote:
>>>
>>>>I have not yet tried with side to move switched ;-)
>>>
>>>I have :) ... I also get a mate in 7, is much more difficult tho :)
>>>
>>>Bxe4 B5xe4 Rb1+ Bxb1 Rd8 Bxc7+ Bxc7 Bxa7+ Kxa7 Ra5+ Ba6 Rgxa6+ Kb8 Bxc7#
>>
>>I like Rb6# more  :-)
>>
>>Well, didn't that mate the wrong side?
>>
>>With side to move switched to white Chest does not find a mate in 11,
>>in only 80 seconds (K6-3/400 30MB hash).  Therefore I suspect, this time
>>switching side to move is not much fun ;-)
>>(except in the way you handle it :-)
>
>chess playing programs may be faster than chest in this case.

Sure, that happens all the time.

>They can prove that there is no mate.
>
>Evaluation of mate against white at the end of the iteration is a proof that
>white has no mate.
>
>Proving that white has no mate in 11 is not a proof that white has no mate.

Sure!  I took it just as a hint, not as a proof.  Sorry for not being clear.

Chest has a bit support to find a mate against the side to move, by just
trying to solve a series of mates 1 ply deeper (option -x):
I just tried it with depth=6:

After Rh8 - h1  ...
        No solution in 6 moves.
After Rb3 - b1  ...
        No solution in 6 moves.
After Bc8 - b7  ...
        Solution (in 2 moves):
        Be5 : c7
After Bc8 - a6  ...
        Solution (in 3 moves):
        Be3 : a7
After Bc8 - d7  ...
        Solution (in 3 moves):
        Be3 : a7
After Bc8 - e6  ...
        Solution (in 6 moves):
        Bf5 : e6
        Be5 : h8
After Bc8 : f5  ...
        Solution (in 5 moves):
        Rg6 - g8
After Ba8 - b7  ...
        No solution in 6 moves.
After Ba8 - c6  ...
        No solution in 6 moves.
After Ba8 - d5  ...
        No solution in 6 moves.
After Ba8 : e4  ...
        No solution in 6 moves.
After Ba5 - b6  ...
        Solution (in 6 moves):
        Rg6 : b6
After Ba5 - b4  ...
        Solution (in 2 moves):
        Be5 : c7
After Ba5 - c3  ...
        Solution (in 2 moves):
        Be5 : c7
After Ba5 : d2  ...
        Solution (in 2 moves):
        Be5 : c7
After Rh8 - g8  ...
        Solution (in 4 moves):
        Rg6 : g8
After Rh8 - f8  ...
        Solution (in 6 moves):
        Bd2 : a5
After Rh8 - e8  ...
        No solution in 6 moves.
After Rh8 - d8  ...
        No solution in 6 moves.
After Rh8 - h7  ...
        Solution (in 5 moves):
        Rg6 - g8
After Rh8 - h6  ...
        Solution (in 5 moves):
        Rg6 - g8
After Rh8 - h5  ...
        Solution (in 5 moves):
        Rg6 - g8
After Rh8 - h4  ...
        Solution (in 5 moves):
        Rg6 - g8
After Rh8 - h3  ...
        Solution (in 5 moves):
        Rg6 - g8
After Rh8 - h2  ...
        Solution (in 5 moves):
        Rg6 - g8
After Rb3 - c3  ...
        No solution in 6 moves.
After Rb3 : d3  ...
        Solution (in 6 moves):
        Be5 : c7
After Rb3 - a3  ...
        Solution (in 4 moves):
        Be3 : a7
After Rb3 - b4  ...
        No solution in 6 moves.
After Rb3 - b5  ...
        No solution in 6 moves.
After Rb3 - b6  ...
        Solution (in 3 moves):
        Be5 : c7
After Rb3 - b7  ...
        No solution in 6 moves.
After Rb3 - b2  ...
        No solution in 6 moves.
After Ra7 - b7  ...
        No solution in 6 moves.
After Ra7 - a6  ...
        Solution (in 2 moves):
        Be5 : c7
Time (virt) = 236.180 sec (ca. 3.9 min)

Still no proof of anything, but you can try to find a proof with Chest.
With more depth and time it might find a mate after all legal moves.

Also, when processing an EPD batch, Chest first tries to find a direct mate,
and if not found, it tries to prove the side to move is lost, in the way
indicated above.

>Uri

Heiner



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.