Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: I think Chris Carson has been vindicated!!!!!!! Tiger Achieves grandmast

Author: Uri Blass

Date: 03:18:04 07/08/01

Go up one level in this thread


On July 08, 2001 at 04:50:27, David Dory wrote:

>On July 06, 2001 at 08:28:55, Otello Gnaramori wrote:
>
>>On July 06, 2001 at 06:07:13, David Dory wrote:
>>
>>
>>>When the GM's have the incentive to wallop the micro's, and yes I mean $$$$$,
>>>(see Hyatt's post below this one regarding "Anti-computer playing..."). THEN
>>>we'll see the REAL ability of the programs versus the GM's.
>>>
>>>Ask yourself just how hard are you going to train yourself to beat a micro when
>>>the micro can't win the prize$$$$, anyway?
>>>
>>
>>So you are saying that actually the computer are stronger than GM's...and you
>>are stating that it's due to the lack of preparation in "anticomp" techniques ?
>
>What I'm saying is that the real test (Is the program == GM) has not yet been
>made clear. There are a number of reasons, such as:
>
>  1. It's a brand new program - GM has no/few games of the programs to study
>     over for preparation. Deep Blue vs. Kasparov is a classic example.

We are not talking about Deep blue but about commercial programs when there are
a lot of ssdf games to study.

>
>  2. GM's are not competing against the program, because the programs can't win
>     any prizes/money anyway.

GM's can get better ranking relative to other humans by beating the machine and
better ranking means more money so they still have motivation to win against the
machine.


 If GM's are only competing for the championship
>     against other GM's, guess who they will prepare for, and fight hardest
>     against??

GM's have a lot of humans to play against them.
They cannnot prepare against everyone of them for a long time so usually they
prefer to play the lines that they know to play without special preperations for
different opponents.

>
>Naturally, I'd like to see the programs have equal footing with the humans in
>the majority of chess titles and championships. Only when these conditions are
>met can we really determine the relative strengths and weaknesses of both the
>programs and the humans.

programs played in the israeli league when the team had the right to choose the
player to play against the machine and inspite of this bad conditions programs
could do performance that was close to 2500(Rebel's performance was 2541).

I believe that if the average GM had to play under the same conditions when the
opponents can choose the player to play against him(her) then (s)he would do
worse performance than the programs in the Israeli league.

I guess that having the right to choose the opponent is worth not less than 100
elo also against humans.

If you remember also that today's programs are better than they were 1.5 years
ago
then it seems clear to me that programs are GM strength.

Uri



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.