Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Mchess Pro 7.1

Author: Thorsten Czub

Date: 13:20:46 12/13/97

Go up one level in this thread


>You are in no trouble!   I will not get angry if someone disagree's
>with me.  Maybe I will even change my mind if I like your arguments.

Thanks. I will hopefully do the same with your arguments.


>I still say you  cannot do this with any kind of reasonable accuracy.

Than "he" asks himself what "he" has done over the whole years at home.
Self betrayal ?

>It depends on what you are trying to evaluate about the student.

The score of the game is the amount of writing errors the pupil has
done.
The WAY he has written the text is the QUALITY of the game. I get most
of my "feelings" about a data-source from THE WAY.

>Woops, I'm showing my age!

:-) Those days were undoubtably GOOD days. And I ate almost any data in
those years too.
And much better: people like Richard, Marty, Don=you were my holy famous
persons BEHIND the product i loved. I found out later that these
mysterious persons behind the product are very sympathical too. That has
even increased my old feelings that the respect with that I had anaylsed
the products in those old times was right.


>Yes, this is a mistake in my opinion.  Sometimes the "lower rated"
>player
>can win more than his fair share of games against a higher rated player
>because his style is "right" for that one opponent (or class of
>opponents.)


I agree here. I have to. Thats my point of view. The style is a part of
the strength. The score has an influence on the strength. But this
influence is very limited. Because the score IS somehow the strength.
The style has also an influence on the strength. But a much more
important.

>But I do not believe this affect is very strong.  I doubt it could ever
>stretch someones rating more than 50 points or so (WARNING: I'm taking a
>wild
>subjective guess here)  against some opponent.

This is indeed the master-question. How big can the style influence the
strength.


>I think a lot of psychology is happening here.

But the style has also much influence on 2 different computer-programs.
The mechanisms fit or don't fit to a certain degree and I can feel it. I
can feel that program A will always have this problem with opponent B
and can estimate an overall score from watching a few samples.

>The whole ELO rating system is based on the idea that strength really is
>absolute.


Tarrasch thought there is always at least ONE best move.
Your above statement concerning ELO and Tarraschs idea about chess and
the moves in each position is - from my point of view - as wrong as the
idea that there is ONE smallest unit in an atom. This deterministic
concept is wrong. There is NO smallest particle. And no best-move. This
is - you guessed it - my idea about chess and compurerchess and
chess-strength and ELO.


>I think the intransitivity is very strong here.  Girl B's performance
>is probably strongly related to the person she is kissing!
>
:-) Touchee ! I cannot top this !
You are right and I only wanted to explain that there IS no absolute
best kissing girl, therefore there is no absolute best-chess move in a
position.
These concepts were state-of-the-art in a mechanistical world, a few
hundred years ago. They are old-fashioned and only socialogically
explainable.


>Human judgement is notoriously imprecise.   I think you have very
>little chance of really saying with any degree of confidence that
>player A is better than B unless the difference is large.

Guess an ESP-guy could find out about how the dices will fall.
WHat to the scientists now try: they put him into a room, with nothing
else into it and force him to do it over and over. NOBODY could
reproduce good results. If you are an ESP'er or a lover of your wife: if
somebody puts you into a room and commands: Make it to her, and make it
once mal, and again and again. Unless we have enough data to say: this
has statistical relevance !
I am sure we would all have the same problems in showing how potent we
are (under these negative circumstancers) like an ESP guy would
have in a faradayic room, with no atmoshere and 3 scientists and any
kinds of machines arround him to find out: it will not work...
How would you stay your man under these circumstances ?

THIS is the reason we normally say: human judgement is not very precise.
Because we all know how subjective it is.
But - and this is the magic thing with it, although it is subjective and
difficult to do, we can do it. And we do it all the day ! I drive since
I am 18. I drive normally much. And I have never made an accident nor
bumped into any car. Not with a single scratch. How would a computer
have had to LEARN driving a car ? How would a computer have had to
handle driving different kinds of cars, different kinds of areas,
different kinds of weather (snow, rain, dark, in PARIS !!, and and and).
And although you say human my actions are notoriously imprecise, I have
never had a car accident before. How does this fit together. Again LUCK
??
Over 12 years ??
Please again: I believe ANYBODY knows this and IS able to do the same
thing. Without beeing a genius. It is the normal unbelievable stuff our
body/brain does although it is fuzzy ! We can all drive cars although
sometimes the circumstances are very very difficult. And this is not
only LUCK or God...


You will have problems to explain all this luck with a Tarrasch point of
view of the world. Or another determinstic point of view.


>You can definitely say you like this programs style much better, or
>that it tends to be more sacraficial or more positional etc.   But
>you have to decide in advance what you are trying to evaluate.
>
>-- Don

I am the same opinion Robert M. Pirsig is in Zen and the Art of
motorcycle maintenance: you can measure quality !




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.