Author: Albert Silver
Date: 06:16:13 05/21/03
Go up one level in this thread
On May 20, 2003 at 21:32:01, John Merlino wrote:
>On May 20, 2003 at 17:33:14, Albert Silver wrote:
>
>>>Johan does know the SKR settings, as he and I discussed them (and several other
>>>settings that had some success in user tournaments). I think it is most accurate
>>>to say that he was not CONVINCED that these settings were better than the
>>>default.
>>
>>Unless this is being extremely diplomatic, it says that he isn't convinced that
>>they aren't better either. I'm sure he has his own testing methods, so why not
>>subject this personality to them? Did he test it at all?
>
>I do not know for certain if he tested these personalities.
>
>>>Johan spends a great deal of time testing his engine, and his testing results in
>>>the settings that are used in the program. Johan then usually will test out his
>>>theories/suspicions in tournaments such as Leiden. Ultimately, it would be silly
>>>to say that somebody else knows The King better than (or even as well as) Johan.
>>
>>Maybe, maybe not. Ed Schroeder has often sent out contests for users to try to
>>improve on his settings, or find settings with an edge in specific areas, and he
>>certainly tested his program enough. This would seem to show that perhaps he
>>doesn't believe it's impossible for a dedicated user to find an improvement.
>>
>> Albert
>>
>>>jm
>
>Kurt's tests, while very thorough, were for a particular time control on
>specific hardware (although I honestly can't remember if more than one CPU type
>was used). Kurt's tests showed very well that, under these specific conditions,
>his settings were as good as or better than any other engine available.
>
>However, I would venture to say that Johan is probably concerned more about "the
>typical Chessmaster user", meaning that he would be very reluctant to grant
>improvement to one type of setting at the expense of another type of setting
>that someone else prefers to use.
>
>Personally, I think it's surprising that Johan's testing methodologies are
>coming into question immediately after he wins a big tournament, simply because
>he decided to use his own settings rather than others that have been shown to be
>successful under certain circumstances. Clearly, he seems to know what he is
>doing, no?
>
>jm
Of course Johan deserves hearty congratulations for his fine victory but truly a
single tournament is even less "proof" that the settings are the best. I don't
mean to imply they're bad by any means, just that the number of games is too
insignificant to support this. Much less than Kurt's.
I also didn't mean to imply Johan's methodologies were poor, but settings are a
very tricky matter and there are a LOT of possibilities. I shared the example of
Schroeder's contests to show that other highly respectable authors believe a
user may stumble on a better set of settings. But a simple look at the case of
CM8000, whose default settings were consistently shown to be less than ideal
does prove this point. You point out (in another thread) that Johan has taken
the changes in King Safety and Selectivity now present in CM9000 from user
personalities which also lends evidence that he too acknowledges users may find
something he initialy missed, no?
Albert
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.