Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: New SSDF-list

Author: Don Dailey

Date: 20:50:22 02/24/98

Go up one level in this thread


Hi Thorsten,

Do you ever read your own posts?   I just read all of your stuff since
the rating announcement and am embarassed for you.   I think your posts
are the most emotional and least informative of all the ones I read.

It's very clear you love computer chess but you do not seem to look at
ANYTHING with any kind of objectivity.  I just recently realized that
I always look forward to reading your stuff but I am becoming
conditioned to ignore any conclusions you arrive at.  In fact if I
discovered that I agreed with you on something I would probably
seriously
rethink my point of view!   OK,  I'm kidding about this but you
get the point!

It's pretty obvious you really hate the Fritz people and I don't really
understand (or care about) the politics involved.  I do know that Fritz
is a strong program and would like to remind you that it is the current
world champion.   There is certainly enough evidence to at least believe
it belongs near (if not AT) the top.  Why don't you reserve judgment for
a while and see what happens?

But keep the posts comming, I always look forward to reading them!

Your friend,

Don



On February 24, 1998 at 16:52:00, Thorsten Czub wrote:

>>Time will tell.  I disagree.  You can only do so much computing and
>>still
>>hit an NPS value of X.  Hiarcs is clearly slow.  So I assume it is doing
>>a
>>lot at each node.  Ditto for MchessPro.  But if you want to call Rebel
>>"knowledged" then we have to call Fritz "knowledged" as well since it is
>>obviously whacking rebel and everyone else fairly soundly.
>
>
>I cannot follow this logic.
>Fritz outsearches Rebel BECAUSE rebel is slower than Fritz5.
>If you watch how rebel plays and how fritz5 plays I wonder that you call
>FRITZ knowledged and not rebel ?
>I can only take the main-lines and the games as my evidence. You can
>call the RESULTS as your evidence.
>But the results don't tell us much about HOW the games have been played.
>The main-lines and evaluations DO IN FACT tell me much about HOW rebel
>works and how fritz5 works.
>Fritz5 is stupid, rebel9 is knowledged. Not the way Hiarcs or Mchess is
>knowledged (they do it more static) but relatively to fritz.
>
>This is only my little opinion. If you believe more in Fritz, stay free
>to do it. I will not use Fritz anyway for my personal analysis or
>email-chess, it would be a major mistake to use Fritz5 because it is a
>dump program.
>
>
>
>> And
>>"knowledge"
>>is a term I would use *very* loosely when talking about Fritz.  It seems
>>to
>>have "good knowledge" but I suspect it is really just "quite finely
>>tuned
>>general/simple knowledge" based on the speed it searches.
>
>?!! Now you sound like myself. Right. Fritz has simple knowledge.
>
>
>
>  Of course,
>>there's
>>nothing to say that finely tuned simple knowledge can't produce a true
>>GM
>>one day either...
>
>Shit x 1000 is still shit. Although it could kill a GM, e.g. when you
>throw 1000 tons of shit over Karpov, he will maybe not only die but also
>lose the game therefore. ONE method to win a game of chess ! But not the
>method I like.
>Sorry.
>
>Sorry. Prejudices. You have yours, I have mine.



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.