Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Normal distribution no way for machines of diff. generations QED

Author: Rolf Tueschen

Date: 13:58:20 06/06/02

Go up one level in this thread


On June 06, 2002 at 16:43:21, Dann Corbit wrote:

>On June 06, 2002 at 15:31:08, Rolf Tueschen wrote:
>
>>On June 06, 2002 at 14:25:43, Dann Corbit wrote:
>>
>>>On June 06, 2002 at 09:10:16, Rolf Tueschen wrote:
>>>
>>>>On June 05, 2002 at 00:05:34, Dann Corbit wrote:
>>>>
>>>>>Since they are different hardware setups or different program versions,
>>>>>they are treated as different organisms.
>>>>
>>>>Now we can make a few conclusions. Here is one of the most important. We have a
>>>>principal difference between human chessplayers and machines. Next. We have a
>>>>principal difference between the generations of chess machines.
>>>>
>>>>I could already stop here, because from the above it is crystal clear that Dann
>>>>Corbit's explanations are a vain attempt. Because normal distribution is for
>>>>different individuals of the same organisms or "race". But - the different
>>>>generations of chess machines are different organisms. Completely new "limbs" or
>>>>"heads" are existing in newer generations. Hence you can't put them into the
>>>>same population for a normal distribution. Chess strength in human chessplayers
>>>>however is differentiated by degrees of strength between the weakest players to
>>>>the best. But there is no principal difference as far as the organism is
>>>>concerned. Period. Thank you.
>>>
>>>When someone takes measurements of bunnies in a field or crocodiles in a river
>>>or bears in the woods, they are talking about different organisms.
>>>
>>>I am afraid that mathematics and statistics are not your strong suit.
>>>
>>>You don't understand the math, the background, the methodology.  Quite frankly,
>>>you have no idea what you are talking about.
>>
>>Well, I won't debate about such ad hominems. If you think that you could vary
>>the different machines on different hardware and with completely different
>>"parts" like "learning" yes or no, and still get a reasonable normal
>>distribution for strength resp. performance, then fine, do what you must do, you
>>have the right to talk about my knowledge in whatever style you prefer, and that
>>is what makes the debate with you so interesting and 'telling' BTW.
>
>I have said nothing about you.  I have described your lack of understanding
>about statistics.  There is nothing wrong with being ignorant about something.
>Will Rogers said it best:
>"Everyone is ignorant.  Only in different areas."
>
>An Ad-hominem attack is an attack against the person and (more specifically)
>against their character.  I have made no such attack.

Of course you did. Simply because I made clear resp. I wrote in the earlier
parts of the debate that I know exactly the difficulties and theories of
statistics. That is why I asked you, out of astonishment, if you had experience
with statistics, then I explained that in stats it's very important to clarify
the parameters in _advance_. So, if you say that I don't know what I am talking
about, then this is a forbidden insult in the presence of the rules of this
forum. You can write pages about the errors in my postings, but you have no
right to state that I have no idea about what I'm talking about. This is
insultive. Like the people of SSDF you take ad hominems as replacement for lack
of arguments.

I still think that you have honest motivations, but it seems to be a question of
bafflement. Simply do it my way. Criticize me but then tell me what exactly was
weong and why.

Rolf Tueschen



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.