Computer Chess Club Archives




Subject: Re: Spike-Rybka

Author: G√ľnther Simon

Date: 03:24:37 12/29/05

Go up one level in this thread

On December 29, 2005 at 06:05:22, Vasik Rajlich wrote:

>On December 28, 2005 at 16:07:00, Eelco de Groot wrote:
>>Hello Vasik,
>>Being just a weak 1300 clubplayer, I think I nevertheless would agree with you
>>that White is better and I would not like to have to play with the black pieces
>>in this position. The main characteristic of the position seems to me that Black
>>has very few plausible moves (even looking superficially) and White has many? If
>>you had some sort of static evaluator that could count the number of
>>superficially plausible moves somehow, that would give White a headstart here.
>>No idea if that could be constructed though in a static eval but with a 1 ply
>>search plus qsearch it would be different already? For Black I see only
>>plausible moves Qg5, Qh5, Rh6 (backward Rookmoves should maybe get a penalty
>>unless they attack something or defend an attacked piece, in this case the Rh6
>>defends f6 but the pawn is already well protected. The black bishops can go to
>>the backrank but maybe that also should not count, unless there they attack
>>something or defend something that needs defending. pawn e6 can go forward to
>>e5, but that gives up two very important fields on d5 and f5 for the the White
>>Knight on e3. Practically every other Black move loses the piece or creates a
>>double pawn after Queen exchange etc.
>>I could do some sort of similar calculation for White but I think it is obvious
>>that White has many more plausible choices for every single piece. Even if you
>>discount moves backward for pieces other than Knights, every piece from White
>>can still move. Well, you asked for some "static" evaluation of the position, I
>>tried to give one. But I don't know if it is possible to implement such a
>>"plausible mobility" evaluator...
>> Best Regards,
>> Eelco
>I think it's a question of the potential of the position. Black is optically
>pretty good, but the question is what's next. It's not easy to figure this out,
>for humans or computers ..
>>[D]8/1b2b1k1/pp1ppp2/nPq5/P1P1P3/2QRNPP1/8/1N2KB1r b - - 0 1

Humans' sense for 'optical advantages' always had a lot of flaws ;)
I would say it relies too much on learned patterns that overwhelm
much deeper aspects of complicated positions, like this one.
I believe chess players neurons just get mad about the first row
with W: Nb1, Ke1, Bf1 and B: Rh1 and the following emptyness in
the second row. Probably this is already enough to cover that
White is a pawn up and Blacks bishops don't bite :)

Best regards,

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.