Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Deep Junior mated in 39 moves

Author: Jeroen van Dorp

Date: 03:39:38 05/11/01

Go up one level in this thread


On May 11, 2001 at 03:23:39, Dann Corbit wrote:

>Though the event tag is a hint, I think explaining the time control is very
>important.  Programs do funny things when they are in a horrible rush, just like
>we do.

Speaking about time management.

I recently 'arranged' a 40/40+40/40+40 game between Hiarcs 7.32 with black and
Shredder 5 under the ChessBase GUI. After playing some games I have the
uncomfortable feeling Shredder loses a lot of the strenght it has under its own
GUI, but that aside. I let Hiarcs use its own opening book, and Shredder the one
coming with Fritz, (general.ctg)

I haven't got the game at hand now, but the setup resembled something roughly
like this:
[D] 5r1k/4n1pp/2p1pq2/2Pp4/1P6/2QP4/3N1rPP/1R3RK1 w - -
(again, please don't use this for calculation purposes, it's far from exact, but
to give you a rough impression of the situation)

Around move 40, Hiarcs loses a pawn, and Shredder is capable of pushing a
resulting passed pawn on the b-file enough to bring Hiarcs in serious trouble.

I searched for a reason why this happened.

On move 31, Hiarcs decides to play 31…Qf6? (question mark by me- see rough
position above) thus giving Shredder the opportunity to swap queens 32. Qxf6.
Having a strong lineup of its two rooks and queen at the f-file, I think Hiarcs
shouldn't spill its chances. Especially not because in the game pawn g7 is a bit
of a weak spot, and the king cornered at h8.

 I let a few other engine crunch on the situation, and all came up with 31…h6 or
31…h5.
Hiarcs also decided for h6, but after more than four minutes of analysis it
switched during *ONLY 27 seconds* to Qf6 and then again to h6, not to return to
that Queen move again for the minutes after that.

It must have been *in that short time frame of less than 30 seconds of
evaluation* that Hiarcs looked at the time and decided to play this move, and
IMO lost the game. With the queen and a rook out of the way after the swap,
Shredder plays quickly Ra1 followed by Ra7. Supported by its knight, it seizes
the hanging black pawn and enters the black back ranks.

For a moment, the engine gave that move a better evaluation score, and looking
at its clock, decided it had to play now.
Well, no shocking discovery at all, but I thought it to be fun to observe the
importance of time management. I think, before swapping queens, I would have
taken more time to find out the soundness, and maybe catching up with that extra
5 minutes later in the game.
But automated chess players have their own rules…..

J.



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.