Author: leonid
Date: 19:08:46 05/02/01
Go up one level in this thread
On May 02, 2001 at 21:41:00, Heiner Marxen wrote:
>On May 02, 2001 at 20:12:09, leonid wrote:
>
>>On May 02, 2001 at 16:24:15, Paul wrote:
>>
>>>On May 02, 2001 at 16:15:41, leonid wrote:
>>>
>>>>On May 02, 2001 at 12:54:56, Paul wrote:
>>>>
>>>>>On May 02, 2001 at 11:58:03, leonid wrote:
>>>>>
>>>>>>Hi!
>>>>>>
>>>>>>If you want to solve some mate that is completely moderate, then this is one.
>>>>>>Position is not very simple but have nothing of deep or complicated. It stay
>>>>>>just in middle.
>>>>>>
>>>>>>[D]qQR1qQrk/1r1qQqqn/3N1nbn/1Q2qQNR/3Q4/bq3q2/QBB2Q1Q/K7 w - -
>>>>>>
>>>>>>Please indicate your result.
>>>>>>
>>>>>>Thanks,
>>>>>>Leonid.
>>>>>
>>>>>Hi guys!
>>>>>
>>>>>Pretz just found a mate in 10 ...
>>>>>
>>>>>10:36 WM10 09 Ndxf7+ Bxf7 Nxf7+ Qbxf7 Qxg8+ Qfxg8 Qxg8+ Qgxg8 Qfxf6+ Qfxf6
>>>>>Qfxf6+ Qxf6 Qdxf6+ Nxf6 Bxf6+ Qg7 Rxe8+ Qxe8 Q5xe8#
>>>>>
>>>>>Could become shorter with longer thinking time though ...
>>>>
>>>>Hi!
>>>>
>>>>And by what way you found 10 moves deep? By brute force (perfect search that see
>>>>everything) or selective (quick)?
>>>>
>>>>Salut,
>>>>Leonid.
>>>
>>>I only have a selective search ... so that's what I used this time. :)
>>>
>>>Well ... of course I can switch off nullmove, all extensions ... etc, but that
>>>really won't cut it. You need some clever stuff to make a frute force search
>>>fast, and I haven't got that ... but that's what we've got Heiner for. ;)
>>
>>I asked you this because I expected that this time you went by brute force since
>>this position is 10 moves deep. So, your result is 100% right. You have good
>>program!
>>
>>This position happened to have very accessible branching factor for brute force.
>>To my surprise, it took only 38 min 25 sec to say that mate in 9 do not existe.
>>With hash at work it could be even shorter.
>>
>>I solved this position, just like you, by 10 moves selective search.
>
>Hi pals,
>
>Chest happens to agree completely. In 51.6 mins (K7/600, 350MB) it detects
>two solution moves for a mate in 10. So Pretz is not as bad as you
>expected, Paul ;-) It has quite a good hit rate! Here are 2 PVs:
Hi, Heiner! Good result even this time. Mine stopped just on the first mate, so
now I could indicate two. Simetime during verification all this data is very
handy.
Good night!
Leonid.
>Ndxf7+ Bxf7 Nxf7+ Qbxf7 Qxg8+ Kxg8 Rxe8+ Qxe8 Qxh7+ Nxh7 Q5xe8+ Nf8
>Q8xf8+ Qxf8 Rg5+ Kh8 Qxh6+ Qh7 Qhxh7#
>
>Ngxf7+ Bxf7 Nxf7+ Qbxf7 Qxg8+ Kxg8 Rxe8+ Qxe8 Qxh7+ Nxh7 Q5xe8+ Nf8
>Q8xf8+ Qxf8 Rg5+ Kh8 Qxh6+ Qh7 Qhxh7#
>
>Again, they are basically equal. Captures over captures.
>Timing info:
>
># 1 0.00 0.87 1- 0
># 2 0.00 1.00 1- 0
># 3 0.03 0.94 111- 0
># 4 0.16 1.06 782- 0
># 5 0.58 1.31 2513- 0
># 6 3.01 1.40 14280- 0
># 7 16.35 1.71 81503- 0
># 8 86.06 2.04 431966- 0
># 9 553.16 2.38 2589797- 316
># 10 3098.17 2.79 13605811- 4908464
>
>Again, lots of mate-in-one in the opposite direction: 12.5%.
>
>CU,
>Heiner
>
>
>>Paul, do you have, by accident, Web address of Genius 2 ("free Uri's version).
>>Something happened to mine when I wanted to see this position on Genius 2. I use
>>Genius 2 (Genius 4 I have) because I remember that it was without any bugs in
>>mate solving and because Rebel 10 have limit of 100 moves. Meantime, will try to
>>see if address is still here in Chess Club old pages.
>>
>>Salut,
>>Leonid.
>>
>>
>>>Groetjes,
>>>Paul
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.