Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Position analysis - Fritz 8 took 8 minutes to find move!

Author: Uri Blass

Date: 11:08:21 09/03/03

Go up one level in this thread


On September 03, 2003 at 13:56:30, Slater Wold wrote:

>On September 03, 2003 at 09:51:56, Uri Blass wrote:
>
>>>>>>[D]5krr/1b2qp2/pp1ppp2/2n2Pbp/P1RNP2R/2N2BP1/2P1Q2P/7K w - - 0 26
>>>>>
>>>>>q3 didn't get this in ~15 minutes.
>>>>>
>>>>>00:10:08.9	-0.40	15	510937279	fxe6 fxe6
>>>>
>>>>fxe6?
>>>>
>>>>I did not think that it can be good because of Bxh4
>>>>I thought that the only alternative that programs may consider is Rxh5
>>>>
>>>>The main line suggest that q3 even does not expect Bxh4
>>>>
>>>>I analyzed the position after fxe6 Bxh4 for some minutes with Fritz8 and saw
>>>>nothing for white(Nd5 -1.41/13,-1.56/14,-1.56/15)
>>>
>>>In 2 hours, it goes back to Rxh5.
>>
>>I do not know what you use but it seems crazy.
>>In the Bh6 problem it does not expect black to capture the bishop and evaluates
>>it as advantage for white.
>>
>>In the second problem it likes to sacrifice material for king attack that does
>>not exist.
>>
>>I suspect that you use something to solve test position and not to play games
>>and I am not surprised if it fails in my tests.
>>
>>My test positions were about tactics and I guess that overevaluating king safety
>>is not going to help.
>>
>>Uri
>
>I am using the q3 personality of Rebel 12.0.
>
>It is notorious for being *extremely* good for test positions, and *extremely*
>bad at actually playing chess.

I think that the main reason is that test position are not about  finding good
moves but often about good moves that get positional compensation for
material.

The first good moves was really a sacrifice but that Rebel personality did not
evaluate it correctly

The second good move was a case when the relatively good move simply keep the
position close and you need deep search to understand that the move that was
played in the game is bad and the move that was played (Rxh5) wins some pawns
for the piece in the right line and it is not easy to see that it is losing.

I remember that the idea of that Rebel personality was to reduce the relative
value of pieces so it is easy for it to find sacrifice of a piece for pawns but
in this case the move of the opponent win piece for pawns and it does not help
that Rebel personality to find the move.

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.