Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Qv3PN

Author: martin fierz

Date: 03:06:06 08/08/03

Go up one level in this thread


On August 08, 2003 at 05:54:56, martin fierz wrote:

>On August 08, 2003 at 05:14:27, Uri Blass wrote:
>
>>On August 08, 2003 at 04:16:34, martin fierz wrote:
>><snipped>
>>>uri's argument of 3+3*1 < 9 is far too simplistic, in endgames against a lone
>>>queen the question is whether you can construct a fortress like
>>
>>fortress are the exception and in most practical cases the queen is stronger.
>
>of course
>
>>If you want to add knowledge about exceptions you can do it but without special
>>knowledge it is better to tell a program that the queen is stronger because in
>>the opposite case it is going to blunder more often.
>>
>>Uri
>
>of course again. i just added these fortress positions as some other examples
>where 9-6 is not a win. the main point (which i probably didn't make clear
>enough...) was about the Q-N+3P position:
>your comment on that position was:
>
>"A queen equal more than knight and 3 pawns and bonus for passed pawns is not
>enough compensation"
>
>which i find a very simplistic evaluation of that position. that's why i
>included the other position which is practically unchanged with the pawn on f4
>instead of f6, and already black has enough compensation. it's just not at all
>obvious that either of the two positions is drawn or lost, certainly it's not a
>question of 6<9 - it's a question of how much compensation exactly these passed
>pawns give black, which is VERY hard to see!
>
>cheers
>  martin

PS: if you move the entire original position one square toward the white side,
it's a draw too - and white has to give a perpetual in order not to lose. i
really don't think it's clear at all on general principles that white is winning
in the original position. interesting question though - how far advanced do the
passers have to be so that there is enough compensation with connected passers +
piece vs queen? is there some kind of general rule for this?

cheers
  martin



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.