Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Very easy mate to solve.

Author: leonid

Date: 05:18:16 12/23/01

Go up one level in this thread


On December 22, 2001 at 21:45:09, Heiner Marxen wrote:

>On December 22, 2001 at 20:23:43, leonid wrote:
>
>>On December 22, 2001 at 17:21:52, Heiner Marxen wrote:
>>
>>>On December 22, 2001 at 09:21:12, leonid wrote:
>>>
>>>>Hi!
>>>>
>>>>This one is very accessible:
>>>>
>>>>[D]2qQQq2/r6P/b1PqqR1N/r6B/n1BRQQ1B/q6k/q1NKNP1q/bq4qn w - -
>>>
>>>According to Chest this is a mate in 9.  And yes, it is easy: just 101 seconds
>>>on my K7/600:
>>
>>Hi, Heiner!
>>
>>Your time look like very good. When I saw your time I wanted to go back and see
>>my for nine moves for brute force, but finally forgot to try this. Had very busy
>>day with my work that I brought home.
>>
>>At least, time for brute force up to 8 moves is:
>>
>>Moves           Time          Branching factor             NPS
>>
>>4               0.82 sec                                   91k
>>                              6.6
>>5               5.43 sec                                   71k
>>                              5.8
>>6               31.8 sec                                   59k
>>                              4.97
>>7               2 min 35 sec                               51k
>>                              4.13
>>8               10 min 41 sec                              51k
>>
>>Cheers,
>>Leonid.
>
>My timing:
>
>#  1      0.00s                 0kN           0.87          1-         0
>#  2      0.00s                 0kN           1.00          1-         0
>#  3      0.02s                 1kN [ 15.01]  0.96         70-         0
>#  4      0.14s [  7.00]       11kN [  8.31]  1.06        546-         0
>#  5      0.83s [  5.93]       55kN [  5.14]  1.28       3548-         0
>#  6      4.06s [  4.89]      253kN [  4.60]  1.61      19237-         0
>#  7     13.95s [  3.44]      825kN [  3.26]  2.91      67658-         0
>#  8     42.52s [  3.05]     2414kN [  2.93]  5.04     205998-         9
>#  9    101.03s [  2.38]     5642kN [  2.34]  7.81     487627-      4549
>
>Once more our EBF is quite comparable, and even the nodes/second.
>I'm a bit puzzled how your program does that without a hash table?
>
>Obviously your move ordering is a bit better, while my basic speed
>(as seen from the small depthes) is better.  The hash table does not make
>that big a difference.
>
>Sometimes I have seen (from experiments with Chest) that a bad move ordering
>is in part compensated by the hash table, and a good move ordering greatly
>decreases the additional effect of the hash table.
>This effect seems to be involved here.

Probably our chess solving part is very close in response, even if your initial
speed is much better. How we ever come to this closeness is a mystery for me.
Few times I tryed to find some visible explanation for this but each time I
could see later that my explanation was wrong. Fact is that our two solvers
stays completely aside from main part of mate solvers in "heavy positions". I
remember one the most striking coincidence in one positions that I will write
here.

[D]8/qkqqb2p/4n2P/1NRRqNQK/1B2b3/1QQQnQQP/4p2Q/qnqrr2B w - -

This position I wrote around two months ago and when I tried it I was almost
sure that I have some bug in my code. I went later to your program and could see
that our programs responded in almost identical way on it. Strange thing in it
is terrible jump in branching factor between 7 and 8 move. My time was:

Move         Time            Branching factor          NPS

4            0.32 sec                                  41k
                             6.1
5            2.03 sec                                  73k
                             7.1
6            13.9 sec                                  73k
                             16.5
7            3 min 49 sec                              215k
                             73.5
8            4 hours 41 min 55 sec      mate found.

Your program had identical gramatic jump in branching factor just on the same
place.

Cheers,
Leonid.





>Cheers,
>Heiner



This page took 0.01 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.