Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: HIARCS Palm to be released early June

Author: Steve Maughan

Date: 15:01:40 05/20/04

Go up one level in this thread


Evgeny,

>Hmm, I was simply saying - hiarcs is now chessbase product, so I would prefer
>to buy palm hiarcs from chessbase too

Hiarcs is not a ChessBase product.  They only distribute a version.  The code is
owned by Mark Uniacke, the author.  This is the same for Chess Tiger and many
other engines

>, and not from some _unknown_ site.

This is the authors website so I wouldn't call it unknown.  I do acknowledge
that it's not that clear who is running the website and they could make it
clearer.

>But _statistical significance_   :)
>ask Chesstiger author he will tell you about statistical significance ..

Actually I know quite a bit about statistical significance and most of these
results *are* statistically significant.  You can download a small program from
my website that gives the level of statistical significance of a results.
Basically since the victories are so convincing the number of games does not
need to be that high to give statistical relevance e.g. we can say with >99%
certainty that a 10-0 victory over Mephisto Risc 1 shows that Palm Hiarcs is
better than MR1.  We can also say that PH is better than the other two dedicated
machines with >90% certainty.

However I would also like to see Hiarcs v Tiger v Genius v PFritz 2.

>I didn't say chessbase publish statistics - but I trust SSDF , don't you ?

I do!

BTW I assume you wouldn't buy Chess Tiger for the Palm either since there is a
ChessBase version and the Palm verison is only sold on some strange website.

Regards,

Steve



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.