Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Great Kasparov Interview (part 1)

Author: Steven Edwards

Date: 12:32:10 04/06/05

Go up one level in this thread


On April 06, 2005 at 11:50:03, Robert Hyatt wrote:
>On April 05, 2005 at 19:36:11, Ted Summers wrote:

>>http://www.chessbase.com/newsdetail.asp?newsid=2309

>Hardly worth the read.  Especially the part about "show me the printouts".  I'd
>be happy to show them to him, since I have them all here on my laptop.  Of
>course he could have them as well if he'd just take the time to download them.

I am in full agreement with Bob on this point.  It is unbecoming to a world
champion for him to baselessly and persistently make claims of fraud without an
honest examination of alternative explanations.

>But that would eliminate his only "complaint".

Kasparov also complains that the results in question are irreproducible due to
the dismantling of DB.  Perhaps so, but this is a concern which could of and
should of have been made known prior to said disassembly.

>I wonder if he has any idea how completely foolish he looks when he makes that
>statement over and over, that they would not show him the output???

Oh, I think he knows.  But he also knows that only a very small minority of his
audience could fully understand the significance of a traditional iterative A/B
searcher log output.  Even if he reviewed the log, he could claim fakery.
Ridiculous, of course, but how many could effectively rebut him in front of the
vast majority of chessplayers?

On the other hand, it could conceivably be possible that a GM chessplayer
working in secret with a very highly skilled chess programmer could produce a
fake A/B log for DB that might fool a lot of people.  But not everyone.  And any
fake log would have to be consistent with real log files.  Would any sane person
at IBM take the risk of detection of even a very clever fraud?  That is simply
unbelievable.

Perhaps there is a lesson here to be learned for chess program authors.  Would
it not be worthwhile to optionally produce a real time search narrative, even
for a traditional iterative A/B searcher, that would eliminate all doubt?



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.