Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Junior 9 and Hydra vs Adams

Author: Robert Hyatt

Date: 09:34:52 06/24/05

Go up one level in this thread


On June 24, 2005 at 02:09:17, Dann Corbit wrote:

>On June 24, 2005 at 00:57:35, Christophe Drieu wrote:
>
>>Hello, i asked Junior 9 to analysed the third Hydra Adams
>>final position. Junior seems to have problem after depth 18. Is it a known
>>problem ?
>>
>>Regards
>>
>>[D]2r2rk1/1q2bpp1/3R1n1B/3ppN2/2p1P3/2P4P/1P3PP1/3QR1K1 b - - 0 1
>>
>>Analysis by Junior 9:
>>
>>1...g7xh6 2.Cf5xh6+ Rg8-h7 3.Td6xd5 Rh7xh6
>>  µ  (-1.40)   Profondeur: 3   00:00:00
>>1...g7xh6 2.Dd1-f3 Cf6xe4 3.Df3-g4+ Ce4-g5 4.Td6xh6 e5-e4
>>  =  (-0.11)   Profondeur: 6   00:00:00  23kN
>>1...g7xh6 2.Dd1-f3 Cf6xe4 3.Df3-g4+ Fe7-g5 4.Cf5xh6+ Rg8-h7 5.Dg4-f5+ Rh7-h8
>>6.Te1xe4 d5xe4 7.Df5xg5 Db7-b5
>>  ±  (1.16)   Profondeur: 9   00:00:00  615kN
>>1...Fe7xd6 2.Fh6xg7 Fd6-e7 3.Fg7xf8 Fe7xf8 4.e4xd5 Tc8-d8 5.d5-d6
>>  =  (0.22)   Profondeur: 9   00:00:01  960kN
>>1...Fe7xd6 2.Fh6xg7 Fd6-e7 3.Fg7xf6 Fe7xf6 4.Dd1-g4+ Rg8-h7 5.Dg4-h5+ Rh7-g8
>>6.Dh5-h6 Ff6-h8 7.Dh6-g5+ Rg8-h7 8.Dg5-h4+ Rh7-g8 9.Cf5-e7+ Rg8-g7 10.Dh4-g5+
>>Rg7-h7
>>  ²  (0.67)   Profondeur: 12   00:00:08  7567kN
>>1...Fe7xd6 2.Fh6xg7 Db7-b3 3.Dd1-d2 Cf6-h7 4.Fg7-f6 Tf8-e8 5.Cf5xd6 d5xe4
>>6.Cd6xc8 Te8xc8 7.Dd2-d7 Ch7xf6 8.Dd7xc8+ Rg8-g7 9.h3-h4 Db3xb2 10.Dc8xc4
>>  +-  (3.59)   Profondeur: 15   00:03:47  202881kN
>>1...g7-g6 2.Fh6xf8 Tc8xf8 3.Cf5xe7+ Db7xe7 4.Td6-a6 d5xe4 5.Dd1-a4 Tf8-d8
>>6.Da4xc4 Rg8-g7 7.b2-b4 De7-d7
>>  +-  (3.01)   Profondeur: 15   00:06:27  346121kN
>>1...g7-g6 2.Fh6xf8 Tc8xf8 3.Cf5xe7+ Db7xe7 4.Td6-a6 d5xe4 5.Dd1-a4 Tf8-c8
>>6.Te1-d1 Cf6-h5 7.Ta6-a8 De7-f8 8.Ta8xc8 Df8xc8 9.Da4-a7
>>  +-  (3.24)   Profondeur: 16   00:08:47  463306kN
>>1...g7-g6 2.Fh6xf8 Tc8xf8 3.Cf5xe7+ Db7xe7 4.Td6-a6 d5xe4 5.Dd1-a4 e4-e3
>>6.Te1xe3 Cf6-d5 7.Te3-g3 Cd5-f4 8.Da4xc4 De7-d7 9.Tg3-g5 Dd7-d1+ 10.Rg1-h2
>>Dd1-e1
>>  +-  (3.22)   Profondeur: 17   00:11:38  595169kN
>>1...g7-g6 2.Fh6xf8 Tc8xf8 3.Cf5xe7+ Db7xe7 4.Td6-a6 d5xe4 5.Dd1-a4 Rg8-g7
>>6.Da4xc4 Tf8-b8 7.b2-b4 Tb8-d8 8.Dc4-c6 Td8-d2
>>  +-  (3.27)   Profondeur: 18   00:18:11  900957kN
>>1...g7-g6 2.Fh6xf8 Tc8xf8 3.Cf5xe7+ Db7xe7 4.Td6-a6 d5xe4 5.Dd1-a4 Rg8-g7
>>6.Da4xc4 Tf8-d8 7.Dc4-c6 Td8-d2
>>  +-  (3.32)   Profondeur: 19   00:36:45  1782751kN
>>1...g7-g6 2.Fh6-g5
>>  +-  (3.62)   Profondeur: 20   01:44:25  777269kN
>>1...g7-g6 2.Fh6-g5
>>  +-  (3.92)   Profondeur: 21   04:09:25  3631450kN
>>1...g7-g6 2.Fh6-g5
>>  +-  (4.22)   Profondeur: 22   09:23:07  1898555kN, tb=36
>
>That's not a problem, it's just a hash table hit.


It looks like multiple fail-lows to me, followed by finally dropping the lower
bound to -MATE which makes the resulting tree impossibly large to search.
Apparently it is happy that g6 is the best move without actually knowing the
true score...



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.