NetNews Usenet Archive 1992 #31/NN_1992_31.iso/spool/rec/games/chess/11896

Internet Message Format | 1992-12-22 | 3.4 KB

Path: sparky!uunet!pipex!bnr.co.uk!uknet!strath-cs!sdonalds
From: [e-mail] (Scott Donaldson)
Newsgroups: rec.games.chess
Subject: Re: TO THE GM's/IM's FROM DON
Summary: A way to measure "How good is Fischer" ?
Keywords: Fischer, Computer
Message-ID: <[e-mail]>
Date: 22 Dec 92 15:35:57 GMT
References: <[e-mail]> <[e-mail]> <[e-mail]>
Sender: [e-mail]
Organization: University of Strathclyde
Lines: 60
Nntp-Posting-Host: fleming-06.cs.strath.ac.uk

In article <[e-mail]> [e-mail] writes:
>In article <[e-mail]>, [e-mail] (Mark Ginsburg) writes:
>> In article <[e-mail]> [e-mail] writes:
>>
>> When chess games are analyzed, a very strong computer can point out better
>> moves.  We don't need a [potentially biased] human to do so.  These
>> corrections do not take away from the match participants, yet they do
>> provide instructional value.  It's funny to see the analysis reduced to
>> what is perceived as personal attacks.  Chess moves are just that, chess
>> moves.  They are not an attempt to get "inside of Fischer's mind" or know
>> "what his strategy was."
>
>Yes, perhaps a computer can sit back and do calculations and determine
>strengths and weaknesses of positions. And, if Bobby Fischer had been playing a
>computer, one could argue that he should have made move X. But he was playing a
>human who already had great respect for his skill at doing the impossible with
>the pieces. Maybe Fischer made less than excellent moves at times because he
>wanted to make his opponent spend time examining some trial structure of the
>pieces.
>
>Reducing chess to math is to leave out 99% of the game.
>


Deep Thought at the moment is gradually crawling towards the 2600 mark. It
would give a solution to the continuing argument on 'chess perfection' if
DT continues to climb the rating list, approaching world champion standard.
Then debates of the type 'was Fischer at his best better than Alekhine ?'
could be resolved by feeding DT the games of the players in question to
analyse.

By analysing each move a "DT correlation" could be returned. That is, for
example, -- "In this game Fischer played the DT best moves 87% of the time.
Or "Alekhine played this endgame with a 97% DT rating". Until this time
(alas - or fortunately, depending on your viewpoint) we can only
speculate on the skills of the top players. For, by definition of being the
best players, at certain points only they can find the "best" moves.

Even when (if) a computer reaches this point of near perfection, it is still
arguable whether some of the spectacular combinations from previous years
(the immortal and evergreen games, eg.) could be found by a computer, without
leaving an imbalance in another part of its game. But there must be a distinct
possibility of this being possible in particular phases of the game
(following on Browne's (?) lead with BELLE analysing endgames)

Maybe the techies will comment ?


Scott
  ~


**********************************************************

[e-mail]             "Free the mind
------------------------              and you can achieve
Computer Science                      dream control"
University of Strathclyde

**********************************************************