Author: Chris Carson
Date: 11:20:53 02/07/03
Go up one level in this thread
On February 07, 2003 at 13:31:54, Mogens Larsen wrote:
>On February 07, 2003 at 12:57:24, Chris Carson wrote:
>
>>I see support for the Super GM (2700+) case, I do not see any data/results for
>>the below 2700 case that you state. This is a broad range of players and a
>>large spread of ratings. Lots of games, some tournament and some matches.
>
>The important distinction is the match conditions, ie. similar to those given to
>Kasparov and Kramnik as initially mentioned in my reply to Chessfun. (The
>financial arrangement aside) Games from the past are virtually irrelevant in
>this respect.
>
>Whether "weak" GMs would have a chance push computer programs below 2700 without
>preparation is a different question. I don't think that is impossible either if
>the incentive is there. Older data may not be as suggestive as one might think,
>because the parameters are vastly different.
>
>Then there's still the scientific test, ie. the matchup without preparation,
>where the engine is stripped of artificial help like opening books and endgame
>tables. I'm sure that there are those that doesn't consider this to be the
>scientific test. That's their problem IMHO.
>
>There's plenty of life left in the human vs. chess program encounter if it isn't
>manipulated by commercial interests IMO. I hope that doesn't sound too much like
>Rolf ;-).
>
>Regards,
>Mogens
I think you have one different question here, so I will re-state it and then
talk about it.
Question: What is the Elo Rating of a Chess Engine (no book, tablebases, ...)?
or How many points do opening books, TB, ... add to a program?
This is a different question than the one I have been interested in. I am (and
most people, I think) interested in the strength of the entire programs/hw (with
books, ...). The Engine strength could be determined, but funding for this
would be very hard to find (perhaps impossible unless you use your own funds).
However, perhaps the "average" chess player might be interested to know how
"valid" is the engine analysis (ELO strength)? Good question, easy to answer,
hard to get funding/tournaments/matches. Perhaps test positions provide some of
this.
The spurious variables (motivation, conditions, ...) that you suggest may exist,
however, I have done some analysis on this and find no significant difference
(at a 95% confidence level) using ad-hoc analysis and SPSS (stats program) data
analysis. It could be there, but it is unlikely and funding for additional
tournaments, matches and analysis is unlikely. The events would have to be set
up so that a very large difference in motivation, conditions and the other
variables were present. Small and medium effects can be measured with existing
data and there was no significant difference.
I do this for free (although this is my profession, behavioral research), I am
not paid by anyone for chess research. If the matches could be set up to test
these "variables"/issues you raise, then I will be happy to provide input on how
to create the tests and do the analysis (no charge). The GM's, tournament
officials and programmers may or may not volunteer time.
Best Regards,
Chris Carson
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.