Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: One mate to solve.

Author: leonid

Date: 18:20:09 02/28/02

Go up one level in this thread


On February 26, 2002 at 17:18:51, leonid wrote:

>On February 26, 2002 at 08:04:56, Heiner Marxen wrote:
>
>>On February 25, 2002 at 12:43:33, leonid wrote:
>>
>>>On February 25, 2002 at 10:47:42, Heiner Marxen wrote:
>>>
>>>>On February 23, 2002 at 19:29:04, leonid wrote:
>>>>
>>>>>[D]5B2/pRNRQN2/p2qqrkr/K2p3p/P2q3n/P2bbnnq/PPQPNN2/5B2 w - -
>>>>>
>>>>>Please indicate your result.
>>>>>
>>>>>Thanks,
>>>>>Leonid.
>>>>
>>>>Chest confirms that there is no mate-in-10.  Hence, the already found
>>>>solutions in 11 moves are shortest possible.  I'll followup, if depth==11
>>>>is completed...
>>>
>>>Thanks, Heiner!
>>>
>>>Your help is all the time valuable. This positions look like very easy and with
>>>more, or less normal branching factor up to depth 10. At 10 it jump
>>>dramatically. At least, it is true for my program. If mine looked 9 moves only
>>>in 23 min. and had previous branching 6.65, it not responded for 10 moves even
>>>after 8 hours. This signify that branching factor jumped more that two times.
>>>
>>>Cheers,
>>>Leonid.
>>
>>Hi Leonid,
>>
>>Chest finds a unique key move for depth 11 (Athlon 1500+, 2.7 hours, 128MB hash)
>>
>>PV: Nh8+ Kf5 Bxh3+ Qg4 Qxd3+ Ne4 Nxe6 Bxd2+ b4 Bxb4+ Rxb4 Rxe6 Qf7+ Ref6 Rxd6
>>Ng2 Rxd5+ Ne5 Rxe5+ Kxe5 Qfd5#
>>
>>Here is my timing info:
>>#  4      0.03s                 3kN [  5.34]  1.04        459-         0
>>#  5      0.17s [  5.67]       13kN [  4.27]  1.20       2522-         0
>>#  6      0.70s [  4.12]       53kN [  3.95]  1.64      11254-         0
>>#  7      2.63s [  3.76]      216kN [  4.10]  2.22      45584-         0
>>#  8     13.78s [  5.24]     1468kN [  6.79]  2.56     249668-         0
>>#  9    107.31s [  7.79]    15276kN [ 10.41]  2.59    1704722-      6749
>># 10   1318.46s [ 12.29]   187860kN [ 12.30]  2.64   20453322-  17254089
>># 11   9864.46s [  7.48]  1420383kN [  7.56]  2.70  151896018- 148696785
>>
>>So, it does jump with its EBF, but not as dramatically as for you.
>>In the rest of my statistics I also do not see anything unusual or dramatic.
>>Dunno, what kind of effect you are experiencing, sorry.
>
>Hi, Heiner,
>
>I don't know exactly how much my strange jumping in branching factor was really
>for this position, since I had no patience to see all 10 moves deep. The last
>time I put my computer to work at night and 8 hours later I disconnected it.
>Then branching factor should be already around 22. Big surprise after previous
>7.
>
>For depth that I look, it was:
>
>Moves        Time            Branching factor         NPS
>
>4            0.219 sec                                49k
>                             4.75
>5            1.043 sec                                43k
>                             4.89
>6            5.38 sec                                 43k
>                             4.4
>7            22.5 sec                                 50k
>                             9.01
>8            3 min 24 sec                             77k
>                             6.65
>9            22 min 44 sec                            111k
                              20.79
10            7 hours 52 min 43 sec                    128k

>
>I will try to put this position once again this Friday for 10 moves. I do go to
>work in the morning, so I expect that computer could work entire night and up to


Just today I put this position for 10 moves. I found that it took 7 hours and 53
min. Probably, the last time I was very close to end the search.

Cheers,
Leonid.


>midday. I hope that nobody will use it at my absence. I am curious to see how
>much time this position take. I do remember other very strange, probably much
>more impressive position where branching factor went even to 115.
>
>Conditions are usual. Celeron 600Mhz. No hash.
>
>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.