Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: A test position for programs(best move -easy, target-to see mate)

Author: blass uri

Date: 04:23:41 03/08/99

Go up one level in this thread



On March 08, 1999 at 02:19:57, Dann Corbit wrote:

>On March 06, 1999 at 14:59:14, blass uri wrote:
>
>>3r1r1k/2n1Qp1p/1q2pNpP/p3P3/1ppp4/3P2P1/PP3Pb1/R3R1K1 w - - 0 1
>>
>>Ne8 is mate in 10.
>>
>>It is easy for programs to find Ne8 but it is hard to find the mate.
>>
>>CM6000 sees mate in 10 in 3 minutes on pentium200MMX.
>>My other programs are significantly slower and I do not know their time.
>Crafty failed after 15 minutes on a PII 300 (and I have most of the tablebase
>files).  CM6000 must be doing some very clever singular extensions, since 20 ply
>is quite a distance to cover in 3 minutes.  What is the correct mating sequence?

 1. Ne8 Kg8
 2. Nxc7 Rfe8
 3. Nxe8 Ra8
 4. Nf6+ Kh8
 5. Qxf7 Qb7
 6. Nd7 Qxd7

Is CM6000's line.
It does not show more moves
after it
 7.Qxd7 Rg8
 8.Qf7 is a simple mate in 3

CM6000 with ss=10 can see mate in 10 in 26 seconds

I asked some personalities of CM6000 and most of them could find mate in 10 in
time close to 3 minutes:

some times:

Anand 3:14
Botvinik 2:51
capablanca 3:00
Euwa 3:06
Fisher 3:07
karpov 1:23
Kasparov 1:24
polgar 2:56

only alekhine failed to see mate in 10 and found only mate in 11 after 8
minutes.

Uri




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.