Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Example for the potential of the new Botvinik-Makaroff-Threat-Extensions

Author: Robin Smith

Date: 08:42:07 10/02/03

Go up one level in this thread


On October 02, 2003 at 09:43:19, Tim Foden wrote:

>On October 02, 2003 at 08:49:10, Joachim Rang wrote:
>
>>Recently Pierre posted a very beautiful position, in which white can win in an
>>obviously lost position.
>>
>>http://talkchess.com/forums/1/message.html?318847
>>
>>All engines need either very much time to solve this or can't solve it at all.
>>There are a few exceptions of programs, which do not consider underpromotions,
>>since black can prolong being mated by underpromotions. So Tao, Amateur and
>>Aristarch see a mate where there actually is no mate (yet).
>>
>>But SmarThink 0.17a can solve this in a few seconds on depth 8 even with the
>>right evaluation! Older versions of SmarThink can't solve this, so I think it's
>>due to the new "same thread extensions" which described Sergei here:
>>
>>http://talkchess.com/forums/1/message.html?318833
>>
>>By the way, the posted position is a mirror position to the so-called "Georgian
>>peaseant", which was also posted here long ago. Consequently SmarThink solves
>>this too on the same depth (8)!
>>
>>Here are the positions and the Analysis:
>>
>>Chessbase Puzzle:
>>
>>[d]8/k3P3/p3K2n/1n3p2/1N4b1/1P1p1p2/8/4B3 w - - 0 1
>>
>>SmarThink 017  WB1 111 MB:
>>D5  00:00  2,05   b4c6 a7b7 c6a5 b7b6 a5c4 b6c5 e1f2 b5d4 f2d4 c5d4 e7e8 f5f4
>>D5  00:00  0,00   b4c6 a7b7 c6a5 b7b6 a5c4 b6c5 e1f2 c5b4 f2e1 b4c5 e1f2 c5b4
>>f2e1 b4c5 e1f2 c5b4 f2e1 b4c5 e1f2 c5b4
>>D6  00:00  -0,01   b4c6 a7b7 c6a5 b7b6 a5c4 b6c5 e1f2 c5b4 f2e1 b4b3 c4d2 b3c2
>>e7e8 b5c7
>>D6  00:01  -4,29   b4c6 a7b7 c6a5 b7b6 a5c4 b6c5 e1f2 c5b4 f2e1 b4b3 c4d2 b3c2
>>e7e8 b5c7 e6d7 c7e8 d7e8
>>D6  00:01  -4,28   b4d3 b5c7 e6d7 f3f2 d3f2 a7b6
>>D6  00:01  -2,83   b4d3 f5f4 e6e5 g4h5 e5f4
>>D7  00:01  -2,84   b4d3 f5f4 e6e5 g4h5 e5f4 h6g8 e1f2 a7b8 f4g5 h5f7
>>D7  00:01  -2,97   b4d3 f5f4 e6f6 b5d6 e1f2 a7a8 e7e8 d6e8 f6g5 e8d6 g5h6 d6e4
>>D7  00:01  -2,96   e7e8 b5c7 e6d7 c7e8 d7e8 g4h5 e8d7 h5f7 e1f2 a7b7 b4d3 f7b3
>>d3c5
>>D7  00:02  -2,96   e7e8 b5c7 e6d7 c7e8 d7e8 g4h5 e8d7 h5f7 e1f2 a7a8 b4d3 f7b3
>>D7  00:02  -2,95   e1f2 a7b7 b4d3 b5c7 e6f6 c7e8 f6g6 h6g8 d3c5 b7b6 g6f7 g8f6
>>c5e4
>>D7  00:02  -2,81   e1f2 a7b7 b4d3 b5c7 e6f6 c7e8 f6g6 h6g8 d3c5 b7c6 g6f7 g8f6
>>c5a6 f5f4 a6b4 c6d6 f7f8
>>D8  00:03  -2,82   e1f2 a7b7 b4d3 b5c7 e6f6 c7e8 f6g6 h6g8 d3c5 b7c6 g6f7 g8f6
>>c5a6 f5f4 a6b4 c6d6 f7f8 g4f5
>>D8  00:03  -2,96   e1f2 a7b7 b4d3 b5c7 e6f6 c7e8 f6g6 h6g8 d3c5 b7c6 g6f7 g8f6
>>c5a6 f5f4 a6b4 c6d6 f7f8 g4f5
>>D8  00:04  -2,95   b4d3 f5f4 e6f6 b5d6 f6g5 g4f5 d3f2 h6f7 g5f4
>>D8  00:04  -2,65   b4d3 f5f4 e6f6 g4d7 d3e5 d7e8 e1f2 a7b7 f6g7 h6f5 g7f8 b5d6
>>e5f3
>>D8  00:08  -2,64   b4c6 a7b7 c6a5 b7b6 e1f2 b6a5 e7e8 b5c7 e6d6 c7e8 d6c5 d3d2
>>f2d4
>>D8  00:10  0,36   b4c6 a7b7 c6a5 b7b6 e1f2 b6a5 e7e8 b5c7 e6d6 c7e8 d6c5 d3d2
>>f2d4
>>D9  00:12  0,35   b4c6 a7b7 c6a5 b7b6 e1f2 b6a5 e7e8 b5c7 e6d6 c7e8 d6c5 d3d2
>>f2d4
>>D9  00:19  0,08   b4c6 a7b7 c6a5 b7b6 e1f2 b6a5 e7e8 b5c7 e6d6 c7e8 d6c5 d3d2
>>f2d4
>>D10  00:39  0,09   b4c6 a7b7 c6a5 b7b6 e1f2 b6a5 e7e8 b5c7 e6d6 c7e8 d6c5 d3d2
>>f2d4
>>D10  01:08  0,23   b4c6 a7b7 c6a5 b7b6 e1f2 b6a5 e7e8 b5c7 e6d6 c7e8 d6c5 d3d2
>>f2d4
>>D11  01:57  0,24   b4c6 a7b7 c6a5 b7b6 e1f2 b6a5 e7e8 b5c7 e6d6 c7e8 d6c5 d3d2
>>f2d4
>>D11  03:24  0,65   b4c6 a7b7 c6a5 b7b6 e1f2 b6a5 e7e8 b5c7 e6d6 c7e8 d6c5 d3d2
>>f2d4
>
>Maybe I'm missing something here, but the final position from this PV is...
>
>[d]4n3/8/p6n/k2K1p2/3B2b1/1P3p2/3p4/8 b - -
>
>Which GLC reckons is a mate in 6 for Black...
>
> 10   8.642 +Mate06  19660k  Kb5 8. b4 d1=Q 9. Ke5 Qd2 10. Bc5 Kc6 11. Ke6 Qe2+
>                               12. Be3 Qxe3#
>
>So I'm confused... are you saying that Nc6 is the solution, or the wrong move?
>
>Cheers, Tim.  :)

You have the wrong position. In your diagram, the White king should be on c5,
not d5.

Robin



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.