Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Regarding the Kramnik match: is CCC so deeply asleep??? :) :)

Author: Christophe Theron

Date: 10:43:38 04/19/01

Go up one level in this thread


On April 19, 2001 at 13:15:42, Chris Carson wrote:

>Great Post!
>
>1.  No book given to Kramnik (not part of the engine).


Yes.



>2.  Separate positional and book learning.  Kramnik has only the
>    raw engine + updates.


Yes.



>3.  Prepare book and learning for program using PGN from Kramnik games.


Yes!



>4.  Have program save all games played by Kramnik and e-mail back to author
>    so that knowledge can be used to update program, captured in learning.


Yes.



>5.  Have personality settings for the program.  More attacking, less attacking,
>    anti-GM, anti-Kramnik, ...


Yes.



>6.  Create on the fly personality changes for the program.  Kind of a multi-
>    personality program.  This could be manual, semi-automated, or automated.
>    Personalities not given to Kramnik, just the raw engine.


YES!!!

This is my favorite (and the one I was thinking about initially).

I call this POLYMORPHISM.

It is nowadays possible to write a program with an always changing style, with
little variations in playing strength.

It is even possible to change the personality settings between each move
(without any action from the operator).

Kramnik can get this program months ago, he will never be able to repeat a won
game twice.

He might find successful opening lines, but how can the organizers prevent the
operator to tell the program to play different lines? They cannot dictate the
opening lines choice.




>7.  Do a lot of mis-leading minor changes to the program and send to Kramnik,
>    then do one big bang change at the end that is the final engine.


Why not. Sounds border line, but it is a legitimate defence against a stupid
requirement.



>8.  Add a timer to the engine that makes it 50% (or what ever) slower and then
>    remove the timer in the final release.


Why not.



>Will Kramnik have an 8-way system to practice on, if so, have the program
>utilize only one processor (slowly) and then have the final program utilize
>all 8.
>
>I am not sure these are ethical, but they may be legal under the rules.  :)



The rule itself is not ethical.

I would not refuse to play, but I would use the same level of ethics, exactly.
:)



>9.  Hire Christophe Theron to consult with the author of the program or
>    deliver Chess Tiger 14 as the final version.  What a surprise that
>    would be.  ;)


:)



    Christophe



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.