Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Game from US correspondence championship finals

Author: Robin Smith

Date: 11:37:57 03/16/00

Go up one level in this thread


On March 16, 2000 at 13:36:18, Dann Corbit wrote:

>Here is the ending of the game, analyzed at long time control (at least 7 full
>moves are searched for each of the  positions).  The interesting thing to me is
>that *except* for your 17. ... Qd5! *all* the other moves were the ones chosen
>by the computer.

The finish of the game involved a very forcing sequence, so it is not surprising
that computers chose these moves.  It is intersting that in the starting
position the computers analysis starts with f5.  This move must be mostly for
the purpose of preventing Be4, but Qd5 does this as well, and with other added
benefits.  The computer does not see at this point that accepting the exchange
sac is unsound.

>
>[d]r1bq1rk1/pp3ppp/4p3/8/1bNP4/3B4/PP2KPPP/2RQ3R b - - acd 15; acn 1002531606;
>acs 36008; ce 62; pv f5 Qa4 a5 a3 Be7 Ne5 Qb6 Rc2 Rd8 f4 Qxd4 Qxd4 Rxd4 Rhc1 Rd8
>Rc7;
>
>It is interesting that even in the last position, the chess program does not see
>the problem.  However, I think it was about to become aware, because there was a
>fail low.  Perhaps a 24 hour analysis would diagnose the position accurately.

If guided through to the end of the VERY forced sequence of the "if moves" I
sent my opponent, the computer would likely fail low big time after 24 hours.

Robin



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.