Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: An amazing checkmate. Can your program find it?

Author: Will Singleton

Date: 16:25:24 12/22/99

Go up one level in this thread


On December 22, 1999 at 18:37:23, Enrique Irazoqui wrote:

>On December 22, 1999 at 17:50:32, Djordje Vidanovic wrote:
>
>>On December 22, 1999 at 14:10:56, Dann Corbit wrote:
>>
>>>Have a look at this tremendous checkmate Chest found from the famous 1924 New
>>>York Chess Club championship:
>>>2r3k1/5ppp/7r/Q7/3P1p2/1N3Pnq/PP3K1P/R5R1 b - - acd 10; acn 535351375; acs
>>>15279; bm Ne4+; ce 32750; dm 9; id "C.A.P. 812153"; pv Ne4+ fxe4 Qe3+ Kf1 Rxh2
>>>Rxg7+ Kxg7 Qe5+ f6 Qe7+ Kg6 Qxf6+ Kxf6 e5+ Kg6 e6 Qf2#;
>>>
>>>I will be amazed if any general purpose program (or even most GM's) can find it.
>>
>>Another super fast mate in 10:  Chessmaster 7000 comes up with -9.90 (mate in
>>10) after only 15 seconds.  I wish I had the Tasc 2.1 CD handy (at a friend's
>>place), I am almost sure that its time would be unbeatable.  What's this about
>>mate in 9 then?  All the progs see the same first five or six plys as Chest.
>>The divergence is only later, with Nd2. I can see how Zarkov's line differs, but
>>then Zarkov did not see the mate in 9, it only found a mate in 13, or something
>>like that (99.70 :)):
>>
>>Zarkov 4.30
>>
>>PV: Ne4+ fxe4 Qe3+ Kf1 Qf3+
>
>Qe3+ is the shortest way to mate, I think.
>
>> Ke1 Rxh2 Rxg7+ Kxg7 Qe5+ f6 Qe7+ Kg6 Nd2 Rh1+ Nf1
>>Rxf1+ Kd2 Rxa1 Qxf6+
>>
>>Djordje
>
>It is a mate in 9 in all variations. The programs I tried find at best a mate in
>10. The fastest to pick Ne4+ so far has been Rebel Century: 4 seconds at ply 5,
>mate in 10 in 79'', ply 8. But no mate in 9 found. :(
>
>Enrique

Just guessing, but that's probably because it's not all that important to find
the shortest mate.  My evidence for such a wild hypothesis is that none actually
find the mate in 9.  Perhaps it would weaken overall play if time was spent on
tracking down the absolute shortest mate in every position.  Any mate, as long
as it is a valid mate, is sufficient.

One problem with my hypothesis is that, if you don't always (or most always)
find the shortest mate, then you may run the risk of missing a mate forever,
since there may exist always a longer mate than a mate in 1.  But in my
experience, that risk is just about non-existent.

Will




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.