Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: A test position from Fisher-Geller

Author: Uri Blass

Date: 12:21:26 09/16/01

Go up one level in this thread


On September 16, 2001 at 15:19:10, Uri Blass wrote:

>On September 16, 2001 at 14:29:45, Peter Berger wrote:
>
>>On September 16, 2001 at 09:02:20, Uri Blass wrote:
>>
>>>On September 16, 2001 at 07:52:45, Peter Berger wrote:
>>>
>>>>3.) I suspect many programs will see the possible draw after a long search -
>>>>1.Qf2 should be just as good as 1.Qf1.
>>>
>>>I am not sure about it.
>>>
>>>After Qf2 black can play Ng4 and not Nxe4.
>>>Fisher suggested also Ng4 in the book but the difference is that
>>>after Qf1 Ng4 the queen does not need to move.
>>>
>>>Deep Fritz also found Qf2 but later prefered Qf1 because
>>>it did not like Qf2 Ng4.
>>>
>>
>>Let's have a look:
>>
>>1.Qf2 Ng4
>>
>>a.) The defensive experiment:
>>
>>2.Qe1 Ne5 3.Rf4
>>
>>Very difficult to judge - probably better for black.
>>
>>b.) The endgame experiment:
>>
>>2.Qf3 Ne5 3.Bxc3 Qxe4 4.Bxe5 Qxe5
>>
>>Looks won for Black.
>>
>>As 1.Qf1 Nxe4 is won for White I agree that 1.Qf1 is better than 1.Qf2.
>>
>>But the major quality of Ng4 doesn't seem to be the tempo here anyway but the
>>much better square for the knight,so how about
>>
>>1.Qf1 Ng4
>>
>>I wasn't able to find anything leading to a major advantage for White - looks
>>pretty even - I can't look at "60 memorable games" - is this line covered?
>
>Yes
>19...Ng4 20.Bxc3 Qb7 21.Qf4 is the proposed line in the book.
>I did not analyze this line to check if 19...Ng4 is the best move.
>
>I know that Fisher had errors in his analysis later in the game(he was probably
>wrong to believe that 20.a3 is losing) so I do not assume that fisher is right
>without checking it for hours by a computer.
>>
>>I don't think it is a really good testposition - the programs get Qf1 sooner or
>>later but not for the right reason ( look at Yace's eval - just like Fritz).
>
>At least they can see positive evaluation of 0.03 so they believe that Qf1 is
>better for white.

Correction.
Yace found 0.04 pawns and not 0.03 pawns based on Dieter's post.

Uri
>
>Uri



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.