Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Two mate positions to solve.

Author: Vincent Diepeveen

Date: 08:46:23 09/08/00

Go up one level in this thread


On September 08, 2000 at 11:14:45, leonid wrote:

>Hi!
>
>Please solve two mate containing positions.
>
>First is almost from the book. Just changed it a little bit to make it search
>for mate three moves deeper.
>
> rnb3kr/pp3pp1/7p/3q3N/3pN3/Q7/PPP2KPP/R1B1R3 w

00:00 4 (0) 1 0.03 Nh5xg7 Kg8xg7
00:00 9 (0) 1 3.09 Nh5-f4
++ h5-f6
00:00 215 (0) 1 13.01 Nh5-f6 g7xf6 Ne4xf6 Kg8-g7 Nf6xd5
++ e4-f6
00:00 2576 (0) 1 MATE07 Ne4-f6 g7xf6 Re1-e8 Kg8-h7 Re8xh8 Kh7-g6 Rh8xh6 Kg6-f5 R
h6xf6 Kf5-e4 Qa3-f3 Ke4-e5 Bc1-f4
00:00 2965 (0) 2 MATE07 Ne4-f6 g7xf6 Re1-e8 Kg8-h7 Re8xh8 Kh7-g6 Rh8xh6 Kg6-f5 R
h6xf6 Kf5-e4 Qa3-f3 Ke4-e5 Bc1-f4
00:00 8144 (0) 3 MATE07 Ne4-f6 g7xf6 Re1-e8 Kg8-h7 Re8xh8 Kh7-g6 Rh8xh6 Kg6-f5 R
h6xf6 Kf5-e4 Qa3-f3 Ke4-e5 Bc1-f4
00:00 20791 (121) 4 MATE07 Ne4-f6 g7xf6 Re1-e8 Kg8-h7 Re8xh8 Kh7-g6 Rh8xh6 Kg6-f
5 Rh6xf6 Kf5-e4 Qa3-f3 Ke4-e5 Bc1-f4
00:01 47605 (121) 5 MATE07 Ne4-f6 g7xf6 Re1-e8 Kg8-h7 Re8xh8 Kh7-g6 Rh8xh6 Kg6-f
5 Rh6xf6 Kf5-e4 Qa3-f3 Ke4-e5 Bc1-f4
00:01 107618 (121) 6 MATE07 Ne4-f6 g7xf6 Re1-e8 Kg8-h7 Re8xh8 Kh7-g6 Rh8xh6 Kg6-
f5 Rh6xf6 Kf5-e4 Qa3-f3 Ke4-e5 Bc1-f4
00:02 243251 (175) 7 MATE07 Ne4-f6 g7xf6 Re1-e8 Kg8-h7 Re8xh8 Kh7-g6 Rh8xh6 Kg6-
f5 Rh6xf6 Kf5-e4 Qa3-f3 Ke4-e5 Bc1-f4
00:06 801049 (254) 8 MATE07 Ne4-f6 g7xf6 Re1-e8 Kg8-h7 Re8xh8 Kh7-g6 Rh8xh6 Kg6-
f5 Rh6xf6 Kf5-e4 Qa3-f3 Ke4-e5 Bc1-f4


>I don't know how much this position could take if solved by brute force. When I
>found that 6 moves took already on my program 2 min 15 sec., looked for mate in
>7 only by selective search.
>
> 2rkr3/n1ppp1n1/2q1q3/NN1N1NN1/2NRN3/8/2R1R3/3K4 w

00:00 7695 (0) 1 19.43 Na5xc6 Na7xc6 Ng5xe6 Ng7xe6
00:00 20794 (0) 2 19.43 Na5xc6 Na7xc6 Ng5xe6 Ng7xe6
00:03 88618 (0) 3 20.08 Na5xc6 Na7xc6 Ng5xe6 Ng7xe6 Rd4-d2
++ b5-a7
00:07 224840 (0) 3 36.50 Nb5xa7 Re8-h8 Na7xc6 Qe6xc6 Na5xc6 Kd8-e8 Nf5xg7 Ke8-f8
 Nd5xe7 Kf8xg7 Ne7xc8 d7xc6
00:19 1008731 (153390) 4 33.19 Nb5xa7 Re8-g8 Na5xc6 Qe6xc6 Na7xc6 Kd8-e8 Nf5xe7
00:44 2934632 (182173) 5 35.31 Nb5xa7 Re8-g8 Na5xc6 Kd8-e8 Nf5xg7 Rg8xg7 Ng5xe6
Rg7-g1 Re2-e1 Rg1xe1 Kd1xe1 d7xc6 Na7xc8 c6xd5 Ne6xc7 Ke8-d7 Nc7xd5 Kd7xc8
01:47 8292162 (998663) 6 46.44 Nb5xa7 Re8-h8 Na5xc6 Kd8-e8 Nf5xg7 Ke8-f8 Ng7xe6
d7xe6 Ng5xe6 Kf8-e8 Na7xc8
04:00 17961437 (3290552) 7 MATE08 Nb5xa7 Re8-h8 Na5xc6 Kd8-e8 Nf5xg7 Ke8-f8 Ng7x
e6 d7xe6 Re2-f2 Kf8-g7 Rf2-f7 Kg7-h6 Rc2-h2 Kh6-g6 Nd5xe7
05:19 28956711 (5705477) 8 MATE07 Nb5xa7 Re8-h8 Nf5xg7 Rh8-h1 Kd1-d2 Rh1-d1 Kd2x
d1 Qc6-a8 Ng5xe6 d7xe6 Nd5-f6 Qa8-d5 Na7-c6
10:46 74456309 (9178260) 9 MATE07 Nb5xa7 Re8-h8 Nf5xg7 Rh8-h1 Kd1-d2 Rh1-d1 Kd2x
d1 Qc6-a8 Ng5xe6 d7xe6 Nd5-f6 Qa8-d5 Na7-c6


>This fantasy position I tried to do as accessible for every program as possible.
>Looked for number of nodes in the ply to be never more that 100.
>
>In this position (also created today) I don't know exact number of minimum moves
>that could lead to the mate. I know only that it is between 7 and 8. Was too
>long for me to see this by brute force. 6 moves revision took already 48 min
>(AMD 400Mhz). Mate is solved only by selective search 8 moves deep in 0.1 sec.

>Please, indicate as much about all your main parameters as you can. Try not

150mb hash. i can't search without it. Parallellism works bad without
hashtable. This is the version as it searches at the icc/fics/gics
from tomorrow and further.

>forget your program name, hash size, selective or brute force search, speed of
>your computer and so like.
>
>Will be especially interested to know what impact on your search make your hash
>table. I don't have hash in my program to try. When I solved one position
>recently on Rebel 10, difference was spectacular. For 6 moves search, almost 6
>fold. Rebel's hash was 28 M. With other comercial programs I can't see this that
>clearly. They almost never permit to put hash size to zero.
>
>Thanks,
>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.