Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Did someone tested the programs of today in deeper blue-kasparov gam

Author: Vasik Rajlich

Date: 04:55:17 04/30/05

Go up one level in this thread


On April 29, 2005 at 09:45:49, Peter Berger wrote:

>On April 25, 2005 at 22:56:24, Mike Byrne wrote:
>
>>On April 25, 2005 at 12:44:22, Uri Blass wrote:
>>
>>>[D]r1bqkb1r/pp1n1pp1/2p1Nn1p/8/3P4/3B1N2/PPP2PPP/R1BQK2R b KQkq - 0 8
>>>
>>>Some programs like White here(shredder or Junior)
>>>Some programs prefer black(for example Yace or Movei)
>>>
>>>It may be interesting to know if the programs that prefer white can win with
>>>white against the programs that prefer black at 120/40 time control.
>>>
>>>I have no time for testing it now but if other people are interested in doing it
>>>then I will be interested to see the games.
>>>
>>>Uri
>>
>>
>>The move 7...h6 is considered a huge mistake in top GM human play -- but it was
>>reported much later after the match that several grandmaster were able to win
>>this position with black against the top computer programs of the day.
>
>I disagree with this assessment. It is not 100% clear if the sac is even
>correct. It was still a bad decision to play this line against a computer of
>course.
>
>8. ...Qe7 is considered to be a clear mistake, although it's not _that_ clear
>either.
>
>An interesting starting point for further analysis should be 8. ...fxe6 9. Bg6+
>Ke7. One of the little published analysis I could find by Granda Zuniga even
>favours black.
>
>The reason this hasn't played much might be more of a psychological nature. This
>whole kind of position is not to the taste of the typical Caro Kann player, and
>why go for it if there is such a natural move as Bd6 availlable. And of course
>the line has been discarded due to the Deep Blue game too.
>
>In practice white seems to score extremely well, but it might be one of the
>lines that get revived once a top GM gets interested to have a closer look.

Apparently the PCs from 1997 liked to follow up the piece sac by playing Re1,
Qe2 and Qxe6, trading queens in order to decrease the material deficit. I guess
this discovery was something Kasparov was keeping in his pocket for the last
game with black.

This is the tricky thing with relying on an anti-computer approach. If you guess
right, the computer looks really bad. If you guess wrong, you look really bad.

Vas



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.