Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: One mate to solve.

Author: leonid

Date: 16:20:49 03/02/02

Go up one level in this thread


On March 02, 2002 at 18:31:18, Heiner Marxen wrote:

>On March 02, 2002 at 14:33:53, John Merlino wrote:
>
>>On March 02, 2002 at 08:21:10, leonid wrote:
>>
>>>[D]3Q3R/Q2b2Q1/1Q1q1Q2/1qQqQq1B/Nr1q1rNB/2qkq3/1qnbnq2/Q2K2QR w - -
>>>
>>>Please indicate your result.
>>>
>>>Thanks,
>>>Leonid.
>>
>>Chessmaster 8000 says Mate in 10 in 2 seconds:
>>
>>Time	Depth	Score	Positions	Moves
>>0:02	1/3	Mate10	164719		1.Nxf2+ Rxf2 2.Nxb2+ Qxb2 3.Bxe2+
>>					Qxe2+ 4.Qxe2+ Rxe2 5.Qfxd4+ Nxd4
>>					6.Q7xd4+ Rxd4 7.Qcxd4+ Qdxd4 8.Qbxd4+
>>					Qdxd4 9.Qaxd4+ Qxd4 10.Qgxd4#
>>0:06	1/4	Mate10	317014		1.Nxf2+ Rxf2 2.Nxb2+ Qxb2 3.Bxe2+
>>					Qxe2+ 4.Qxe2+ Rxe2 5.Qfxd4+ Nxd4
>>					6.Q7xd4+ Rxd4 7.Qcxd4+ Qdxd4 8.Qbxd4+
>>					Qdxd4 9.Qaxd4+ Qxd4 10.Qgxd4#
>>0:27	2/5	Mate10	1265606		1.Nxf2+ Rxf2 2.Nxb2+ Qxb2 3.Bxe2+
>>					Qxe2+ 4.Qxe2+ Rxe2 5.Qfxd4+ Nxd4
>>					6.Q7xd4+ Rxd4 7.Qcxd4+ Qdxd4 8.Qbxd4+
>>					Qdxd4 9.Qaxd4+ Qxd4 10.Qgxd4#
>>2:05	2/6	Mate10	5480662		1.Nxf2+ Rxf2 2.Nxb2+ Qxb2 3.Bxe2+
>>					Qxe2+ 4.Qxe2+ Rxe2 5.Qfxd4+ Nxd4
>>					6.Q7xd4+ Rxd4 7.Qcxd4+ Qdxd4 8.Qbxd4+
>>					Qdxd4 9.Qgxd4+ Qxd4 10.Q7xd4#
>>
>>jm
>
>Chest confirms this to be a shortest possible mate (congrats to CM!), and also
>the symmetric solution found by Sjeng (no surprise from the initial nearly
>symetric position) in 2.3 hours on K7/600 with 350 MB hash:
>
>PV: Nxf2+ Rxf2 Nxb2+ Qxb2 Bxe2+ Qxe2+ Qxe2+ Rxe2 Qgxd4+ Qdxd4 Qfxd4+ Qbxd4
>Qgxd4+ Qxd4 Qcxd4+ Rxd4 Qbxd4+ Nxd4 Q7xd4#
>PV: Nxb2+ Qxb2 Nxf2+ Rxf2 Bxe2+ Qxe2+ Qxe2+ Rxe2 Qgxd4+ Qdxd4 Qfxd4+ Qbxd4
>Qgxd4+ Qxd4 Qcxd4+ Rxd4 Qbxd4+ Nxd4 Q7xd4#
>
> Nxb2+ Qxb2 Nxf2+ Rxf2  Bxe2+ Qxe2+ Qxe2+  ...+6
>                              Rxe2  Qexd4+ ...+6
>                  Qxf2  Bxe2+ Qxe2+ Qxe2#
>       Rxb2 Nxf2+ Rxf2  Bxe2+ Qxe2+ Qxe2+  ...+5
>                              Rxe2  Qexd4+ ...+5
>                  Qxf2  Bxe2+ Qxe2+ Qxe2#
> Nxf2+ Rxf2 Nxb2+ Qxb2  Bxe2+ Qxe2+ Qxe2+  ...+6
>                              Rxe2  Qexd4+ ...+6
>                  Rxb2  Bxe2+ Qxe2+ Qxe2+  ...+5
>                              Rxe2  Qexd4+ ...+5
>       Qxf2 Bxe2+ Qxe2+ Qxe2#
>
>Chest's timing info shows an EBF around 6:
>#  3      0.03s [  3.00]        1kN [ 12.62]  0.95         93-         0
>#  4      0.15s [  5.00]       10kN [  8.13]  1.12        308-         0
>#  5      1.20s [  8.00]       82kN [  7.82]  1.29       2813-         0
>#  6      7.29s [  6.08]      483kN [  5.91]  1.79      15452-         0
>#  7     44.32s [  6.08]     3024kN [  6.26]  2.40      99438-         0
>#  8    248.53s [  5.61]    17283kN [  5.72]  3.13     597059-         0
>#  9   1155.17s [  4.65]    83584kN [  4.84]  4.42    2797265-       535
># 10   8042.53s [  6.96]   526938kN [  6.30]  6.01   25077142-  16329266

Thanks, Heiner!

Now I see that this is mate in 10. Probably, will put later mine to search mate
in 10 in the night. It could take hours, after my time for 9 moves. I see that
Chess Master found better by searching shorter mate in 10. Mine found by
selective only mate in 11 in 2 seconds. Will look if few additional versions for
selective will take this position in 10. This position was created last Summer
and then I had few selective versions for search.

My brute force, for my usual conditions:

Moves        Time        Branching factor         NPS.

4            0.93 sec                             112k
                         7.35
5            6.86 sec                             110k
                         8.23
6            56.53 sec                            98k
                         7.2
7            6 min 47 sec                         106k
                         6.62
8            44 min 57                            109k
                         6.03
9            4 hours 31 min 27 sec

Cheers,
Leonid.


>Cheers,
>Heiner



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.