Author: Christophe Theron
Date: 12:22:33 11/01/03
Go up one level in this thread
On October 31, 2003 at 19:56:15, Uri Blass wrote:
>On October 31, 2003 at 17:50:06, Gian-Carlo Pascutto wrote:
>
>>On October 31, 2003 at 17:15:44, Uri Blass wrote:
>>
>>>Unfair advantage?
>>>If you force kasparov to play lines that anand played some years ago then I
>>>think that you give advantage to the opponent.
>>
>>The comparison makes totally no sense whatsoever - Kasparov has his own
>>good book.
>>
>>>The king is using an old book.
>>>I do not think that it is unfair advantage but disadvantage relative to the >case of using a good book.
>>>
>>>Even using a good new book is a disadvantage if the book does not fit the style
>>>of the program.
>>>
>>>If an engine can top the ssdf with a book of another engine that it says
>>>something very good about the engine.
>>>
>>>It says that the engine is better than the second engine because it can get
>>>better rating inspite of using a book that may not fit its style.
>>
>>Utter nonsense. The list tests combinations of books and engines, and the
>>books have large intrisic values that are not related to their style. See
>>the discussions of Jeroen Noomen and Alex Kure in the past.
>
>It is better instead of discussion if the same engine will be tested with
>different books so we can see the advantage of general.ctg relative to another
>book.
>
>
>>
>>I'll turn your statement around: if an engine gets a middle ranking with
>>general.ctg it's actually a very weak engine because the book represents
>>a signficant advantege (for example over amateur programs).
>
>You need to define middle ranking and significant advantage before I can reply
>but the only way to know is to test.
>
>If general.ctg give significant advantage relative to amatuer book then the only
>way to convince me about it is by testing.
>
>If you add another entry of Yace or ruffian with General.ctg to the ssdf and it
>gets clearly better rating then it is going to prove your point.
>
>Uri
And until this point has been proved either way, it is not wise to allow CM9000
to use the Fritz book. The influence of this book is in my opinion an unfair
advantage. In your opinion it is the opposite. In any case, it invalidates the
test in the sense that the test will not give any useful information about the
real strength the the CM9000 PRODUCT (The King engine running inside the CM9000
GUI with its own book).
Christophe
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.