Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Nolot #6 Solved

Author: Joshua Lee

Date: 06:32:49 01/18/03

Go up one level in this thread


On January 17, 2003 at 15:46:38, Jeremiah Penery wrote:

>On January 17, 2003 at 14:54:30, Aaron Gordon wrote:
>
>>On January 17, 2003 at 12:45:45, Robert Hyatt wrote:
>>
>>>On January 17, 2003 at 10:42:33, Joshua Lee wrote:
>>>
>>>>bm axb5 - NOLOT #6
>>>>rnbqk2r/1p3ppp/p7/1NpPp3/QPP1P1n1/P4N2/4KbPP/R1B2B1R b kq - 0 1
>>>>
>>>>Analysis by Fritz 7:
>>>>
>>>>1...Nd7 2.Nd6+ Kf8 3.Ng5 Nb6 4.Ndxf7 Kg8
>>>>  =  (-0.16)   Depth: 6/35   00:00:00  366kN
>>>>1...Nd7--
>>>>  =  (0.12)   Depth: 7/28   00:00:00  391kN
>>>>1...Nd7 2.Nd6+ Kf8 3.Nxc8 b5 4.cxb5 Rxc8 5.Qxa6 cxb4
>>>>  ²  (0.69)   Depth: 7/28   00:00:00  427kN
>>>>1...axb5!
>>>>  ²  (0.66)   Depth: 7/37   00:00:02  989kN
>>>>1...axb5!
>>>>  ²  (0.50)   Depth: 7/37   00:00:02  992kN
>>>>1...axb5!
>>>>  =  (0.19)   Depth: 7/37   00:00:02  999kN
>>>>1...axb5 2.Qxa8 Na6 3.h3 Nc7 4.Qa5
>>>>  =  (0.09)   Depth: 7/37   00:00:02  1004kN
>>>>1...axb5--
>>>>  ²  (0.37)   Depth: 8/31   00:00:02  1124kN
>>>>1...axb5
>>>>  ²  (0.37)   Depth: 8/31   00:00:02  1133kN
>>>>1...axb5 2.Qxa8 Na6 3.cxb5 Nc7 4.Qb8 cxb4 5.axb4 f6 6.h3
>>>>  ²  (0.50)   Depth: 9/30   00:00:06  2430kN
>>>>1...axb5--
>>>>  ±  (0.78)   Depth: 10/35   00:00:12  5037kN
>>>>1...axb5
>>>>  ±  (0.78)   Depth: 10/35   00:00:12  5093kN
>>>>1...Nd7!
>>>>  ±  (0.75)   Depth: 10/35   00:00:13  5464kN
>>>>1...Nd7 2.Nd6+ Kf8 3.Nxc8 b5 4.cxb5 Rxc8 5.h3 axb5 6.Qd1
>>>>  ²  (0.69)   Depth: 10/35   00:00:13  5799kN
>>>>1...Nd7 2.Nd6+ Kf8 3.h3 Nb6 4.Qe8+ Qxe8 5.Nxe8 Kxe8 6.hxg4 Bd4 7.Nxd4 exd4
>>>>8.bxc5 Nxc4
>>>>  ²  (0.50)   Depth: 11/40   00:00:35  16021kN
>>>>1...Nd7 2.Nd6+ Kf8 3.Nxc8 Ndf6 4.Qa5 Qxc8 5.h3 Bd4 6.Nxd4 cxd4 7.Kf3 Ne3 8.Bxe3
>>>>dxe3
>>>>  ²  (0.66)   Depth: 12/39   00:01:01  28254kN
>>>>1...Nd7!
>>>>  ²  (0.37)   Depth: 13/42   00:02:50  80884kN
>>>>1...Nd7 2.Nd6+ Kf8 3.Nxc8 Bd4 4.Nxd4 exd4 5.Nd6 d3+ 6.Kxd3 Nf2+ 7.Kc2 Nxh1
>>>>8.Nxb7 Qe7
>>>>  ²  (0.31)   Depth: 13/42   00:03:13  92105kN
>>>>1...Nd7 2.Nd6+ Kf8 3.h3 Nb6 4.Nxb7 Bxb7 5.Qb3 Bd4 6.Nxd4 exd4 7.hxg4 Qf6 8.bxc5
>>>>  ²  (0.47)   Depth: 14/43   00:06:54  200200kN
>>>>1...Nd7 2.Nd6+ Kf8 3.h3 Nb6 4.Nxb7 Bxb7 5.Qb3 Bd4 6.Nxd4 exd4 7.hxg4 Qf6 8.bxc5
>>>>  ²  (0.47)   Depth: 15/47   00:19:29  569978kN
>>>>1...Nd7--
>>>>  ±  (0.75)   Depth: 16/48   00:38:08  1119117kN
>>>>1...Nd7 2.Nd6+ Kf8 3.h3 Nb6 4.Nxb7 Bxb7 5.Qb3 Nd7 6.hxg4 Bg3 7.Rh3 Bf4 8.Kf2
>>>>  ±  (0.81)   Depth: 16/51   01:09:25  2057428kN
>>>>1...axb5!
>>>>  ±  (0.78)   Depth: 16/51   01:14:45  2209222kN
>>>>1...axb5!
>>>>  ²  (0.62)   Depth: 16/51   01:18:18  2310468kN
>>>>1...axb5 2.Qxa8 Qc7 3.h3 Bd4 4.Ra2 Nf6 5.Nxd4 cxd4 6.Kf3 bxc4 7.Qa5 Qxa5 8.bxa5
>>>>  ²  (0.41)   Depth: 16/51   01:28:34  2601340kN
>>>>1...axb5 2.Qxa8 Qc7 3.h3 Bd4 4.Ra2 Nf6 5.Nxd4 cxd4 6.Kf3 bxc4 7.Qa5 Qxa5
>>>>  ²  (0.34)   Depth: 17/50   03:11:19  5594783kN
>>>>1...axb5 2.Qxa8 Qc7 3.Rb1 Bd4 4.bxc5 Qxc5 5.Qxb8 0-0 6.Rb4 Nf2 7.Nxd4 Nxh1
>>>>8.Qxe5
>>>>  ²  (0.37)   Depth: 18/55   07:10:51  12610578kN
>>>>
>>>>(Lee, Kingpawn 07.01.2003)
>>>
>>>
>>>That doesn't look "solved".  IE .37 is not winning anything.  The score seems to
>>>be
>>>dropping iteration by iteration in fact... which causes the best move to change,
>>>but maybe
>>>not for the right reason.
>>
>>How do we know that the person that played the move in the first place didn't
>>see something else and then see the winning line later a few moves after playing
>>axb5? The GM could have been playing through for the same reasons Fritz was and
>>only later found that it was infact winning.
>
>It was a big home preparation line.  It's not all _that_ hard for a computer to
>think that axb5 is the best move in the position, but finding the next move,
>Bd4, is probably too difficult for any computer today.  Certainly, they won't
>find it with any kind of decent score.  The line Fritz gives there is pretty
>pointless, and isn't going to win anything.


I was mpressed that it found the move and kept it for seven hours. It's not a -+
score but it did find the move and i hven't heard of anybody else saying that a
program found the move. In reality to see Bd4 or to see a plus score i think i
figured out that a program would have to search to a (fritz equivalent) depth
somewhere 30ply or more maybe 25.

Next week i'll let it try again on my new computer




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.