Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Test Your Positional Play

Author: Vincent Diepeveen

Date: 16:59:39 09/03/01

Go up one level in this thread


On September 03, 2001 at 19:42:39, Jeroen van Dorp wrote:

>Uri, I think you miss the point every time they ask you this question.

>Basically the question is "do you think it's fun to generate moves with a
>computer instead of yourself?" and you answer "Computer moves are better than
>most human moves. You should use a computer to find the best moves."
>
>But it's no answer to the question.

Suppose chess is 10^40 positions which a computer would need to search
to reveal some truth.

Obviously you can let a program analyze forever, but YOU the human
must take a decision which line to play or not to play, because
many lines the difference is like 0.01 for the computer but for
the game it's a crucial position. Also when is a move refuted and
when not.

When it gives -1.0?  Well it can be a local minimum, perhaps
the score goes up after that line because it's based upon a wrong
evaluation from the computer, who knows? Especially in anti positional
games the local maxima and local minima are very common things!

So you go try out moves decide which move is good and which move is
interesting to investigate.

For guys like Uri the thing is not so simple, in fact, most strong
chessplayers are very bad in this: building up a tree of interesting
moves and analyzing those in a very sound way.

That's not so easy as it looks like. In fact i'm pretty sure that
nearly no one can do it.

For sure is that if you want to make moves yourself, you should
join 40 in 2 tournaments yourself. Plenty of time to decide to make
a move yourself of course then.

>Obviously you don't think it's fun to play chess with just your own brains, so
>you enter a competition in which computers - sometime aided by a human - play
>chess with each other.
>
>You should say "I think it's fun to play chess in this way, and as it's allowed,
>I enter the competition in this way."
>It's a perfectly reasonable answer.
>
>And other people -like me- could answer "Well, I think it's boring to be like an
>automaton. I'd rather find a weaker move myself than the strongest move by my PC
>program." Neither is right or wrong. We just have a different approach to chess.
>
>I think there's a point in catering both "sides". Maybe you do like an
>accidental human vs. human game. But not in correspondence chess. Good. If it's
>allowed (and it *is* in your competition) there's nothing wrong with that.
>
>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.