Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Nolot test revisted

Author: Jeremiah Penery

Date: 22:49:30 01/12/00

Go up one level in this thread


On January 12, 2000 at 18:45:55, Vincent Diepeveen wrote:

>On January 12, 2000 at 18:27:23, Albert Silver wrote:
>
>>On January 12, 2000 at 16:11:31, Bruce Moreland wrote:
>>
>>>   DT       Ferret   Chop     Dark Thought
>>>            4x450    500      500mhz 21264
>>>   -------- -------- -------- ------------
>>>1   6:00:00  1:56:54           0:12:25
>>>2   0:02:00  0:00:33
>>>3
>>>4   2:30:00  0:25:46
>>>5   2:00:00
>>>6
>>>7   6:00:00  1:27:00
>>>8            1:30:46
>>>9
>>>10  0:02:00  0:00:29  0:50:00
>>>11  0:05:00  0:00:13  0:12:00  0:04:32
>>>
>>>I looked at Shep's site, which includes results mainly on slow hardware (P6/200
>>>or 233).
>>>
>>>http://sccs.8m.com/nolot.html
>>>
>>>Many programs find #1 in under 20 minutes.
>>>
>>>CM5555 finds #2 in something under six hours.
>>>
>>>Rebel 10 found #10 in a couple minutes.
>>>
>>>Mine finds #2 quickly due to extension successes.  I'm surprised it found #8 at
>>>all, the score though was only +0.75.
>>>
>>>I will include more information as it becomes available.  I'll run #5 for a day
>>>or two.
>>
>>Just a question: is it sure they are all 100% correct? I never checked this
>>myself, but I recall someone (in fact I think it was you) mentioning in RGCC
>>some time back that Nolot wasn't absolutely sure about all of them.
>
>The axb5 sacrafice is more a longterm sacrafice, which i don't see,
>but the idea for black is to blow up the center which is possible if the
>queen of white is at the other side of the board (a8).
>
>the Nxg5 instead of bxg5 move is very difficult but i don't doubt it
>wins in the end, i lack just several tenths of positional eval when searching
>quite deep for DIEP to play it. All other moves i could verify to be correct,
>though from Ng5 it's a big mystery to me what line(s) DIEP needs to find in
>order to see it.
>
>The second problem (Rxc5) is really mating extensions dependant. I've had
>versions of DIEP which extended quite a bit threats which failed high
>within seconds at 7 ply. Later it needed 8 ply to find it, and that
>continuesly was a ply more till it needed 10 ply. Then i turned
>off the threat extension, plan to turn it on real soon and test
>again at NOLOT.

What do you think of #9, Ng5?  Here is the original analysis:

1.Ng5!! hxg5 2.hxg5! Rac8 3.Nf6!! Nb8
  (3...gxf6 4.gxf6 Rfe8 5.Qh5 Kg8 6.Rxc5! Bg6! 7.Qh4 Bxc5 8.Be4 Ne7 9.Kg2 Qd5
   10.Bxd5+-)
4.Qh5 Bxf6 5.gxf6 gxf6 6.Rxc5 Rxc5 7.Be4 f5 8.Kg2 Rg8 9.Rh1 Rg7 10.Bh6 Nd7
11.Bxg7+ Kxg7 12.Qxh7+ 1-0

Is it certain that all these moves are optimal?  I don't really see any computer
being able to find this one...



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.