Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Any alternative ways of evaluating positions ?

Author: Werner Kraft

Date: 10:24:35 05/09/05

Go up one level in this thread


I have not tried out the surreal numbers yet - it sounds VERY surreal to me...
topology - I have browsed through the book " You are a mathematician " -
Topology is mainly about rubber - sheet surfaces - warping sth , changing angles
etc , but not changing the quality / feel of an object.
A google search brought the topic of a " Martian Chess Game ".

The book -  " You are a mathematician " is an excellent introduction .

I am an almost an  idiote savante , in regards to mathematics, and very autistic
about things like " Group Theory " . But my few remaining working brain cells
guided me today to the British Library - and I did a sear on Group Theory .

Example :   " Computational experimental Group Theory :  AMS ( that should be
American Mathematical Society )  - ASL  joint special session , interaction
between logic, group theory and computer science. January 15 - 16th , 2003
Baltimore, Maryland. "

Or, even better :  " Complexity theory for Lie-group solvers , 1999 "

And , if that does not beat Deep Blue , let's try :
"Modelling and signal processing for fault diagnosis . Colloquium on modelling
and signal processing for fault diagnosis.  September 1996 , IEE ."

And, before I forget it :  " Advances in |HOPF  algabra " !!

Now , what can I make out of it - can we take take chess positions, where
player feels intuitively , that there is something wrong ,  let the computer
work out, what it could be ...
Then we could compare the results of many such analysis and ....
we find patterns for fault analysis !?

So - the human says - in this positions seems to be something wrong - just
cannot figuere it out what it is .
The computers says - you are absolutely right - bit just in 64.3 % of the cases
... now let's have a look at YOUR patterns -
always one you thing , playing Nf4 is wrong - in reality that is the correct
move!
I think that leads anywhere... :-(





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.