Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Genius' asymmetric search in example: TRY out !

Author: Thorsten Czub

Date: 11:32:26 12/31/97

Go up one level in this thread


>>If your program wins against genius, it cannot be weak !!

>Yes, this is your joy. When a test program beats Genius or Fritz. Then
>you are happy. Not of course anything to do with feelings about Genius
>or Fritz :)

Look, if you have a strong enemy or target (climbing the Mount Everest)
and now you try it, and you win it, than you are happy. I don't think
this has anything to do with feelings AGAINST the mount everest. Without
a strong enemy/target, life would be boring.


>Now, Thorsten, your position over genius was this:
>
>1. genius is a boring program because it is so accurate and not
>speculative.

The accuracy and the non-speculative behaviour are connected with the
asymmetric search.
This results in: do nothing but do it well.

>
>now position has changed to:
>
>2. genius is a boring program because it culls its own possibilties at
>deeper odd plies, and therefore plays less than best moves.

Culls ? Whats this ?
Deeper plies ??

No - it has a mechanism that follows more moves in the plies 2,4,6,8
than in the plies 1,3,5,7,9.
Maybe you misunderstood me.
Why DEEPER plies ???

>
>I think your first position was correct, and the secoond is not. Boring
>is probably too strong a word, and subjective anyway, but Genius is
>accurate, seems to contain very little in the way of speculative stuff;
>and this is what tends to lead to its playing generally quieter
>positions. If you want wild play, you have to do it with steering
>towards these positions by use of evaluation.

But genius CAN find the good moves if you shift the search 1 ply !
So the evaluation is maybe different in plies 2,4,6,8 than in plies
1,3,5,7,9. Or maybe the search is different.
What is right ?
The discussion shall find it out.


>Your second position makes no sense to me. If genius culls his own
>'quiet' moves at deeper odd plies and promotes / extends his active
>moves / lines; then you'ld expect relatively active lines to be found,
>and relatively active lines to be played, no ?

You statement does not remind me on my own statements. I have not said
this. Or ?


>I mean, if it did the opposite and culled the active moves, and extended
>the dull ones, then we really could argue for overall dullness, no ?

As I said, this IS NOT my point . I have never said this.

>
>Play style (Tal or Tarrasch) has to be evaluation based. Search is only
>about what can be found, at what depth, and in what time; not about
>style of play.

PLaying style has maybe not only to do with evaluation.
Always overseeing some best moves while defending against all threads of
the opponent has NOTHING to do with evaluation ! It is a selectivity
problem of the search and could maybe result in:
boring = non active playing style.



This page took 0.01 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.