Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: The importance of opening books -- a simple experiment

Author: Uri Blass

Date: 07:56:29 02/18/05

Go up one level in this thread


On February 18, 2005 at 10:34:55, Arturo Ochoa wrote:

>On February 18, 2005 at 02:26:43, Uri Blass wrote:
>
>>On February 18, 2005 at 02:14:15, Uri Blass wrote:
>><snipped>
>>>the games in CCT are not more important than other games as long as movei is an
>>>extremely weak engine that is probably even weaker than glaurung because it is
>>>near the level of gothmog and gothmog is weaker than glaurung based on Tord).
>>
>>Note also that if you want to see some evidence that movei is near the level of
>>gothmog then you can look at
>>http://www.husvankempen.de/nunn/aegt1002.txt
>>
>>Gothmog 2549
>>Movei   2533
>>
>>Difference are inside the statistical error.
>
>First, the rating is inflated.

Not important.
The point is not to compare with humans.

> Second: are the games run in blitz time controls?

No blitz time control

40/40 ponder off on 2ghz cpu
see http://freehosting.hostrave.com/p/iggor/aegt/AEGT_draft.txt.


>Third, can it be consider valid for the topic to use a common tyny opening book
>as the nuun one?

part of the games of movei were from nunn position when part of the games were
with original book.

There was no big difference

Movei scored in the middle of the table both in nunn positions in queen class
with its own book and with nunn positions.

see results in
http://www.husvankempen.de/nunn/queen3.htm

Movei's result with its own small book and learning:

10/22+11.5/22+12/22+12/22+9.5/22=55/110

Movei overall score is 107.5/220( means 52.5/110 in the nunn positions)
It does not seem that movei small book is inferior relative to books of other
programs at similiar level.

Uri



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.