Author: Robert Hyatt
Date: 08:56:29 10/04/04
Go up one level in this thread
On October 04, 2004 at 10:37:50, Uri Blass wrote:
>On October 04, 2004 at 10:28:29, Robert Hyatt wrote:
>
>>On October 04, 2004 at 10:23:52, Tony Hedlund wrote:
>>
>>>Another black defeat in sicilian for Crafty! Are you sure it's 10% Bob? The
>>>logfiles show 39%.
>>
>>Actually it is set at 50%. Not sure when it was changed.
>
>What about learning?
>Is the probability constant and not changed based on results?
>
>Uri
That "play percentage" comes from the bookc.bin/books.bin and is not influenced
by learning at all at the present (probably should be but it isn't). This was
done at the request of a couple of IM and GM players that wanted to train
against specific openings and not have learning cause it to stop playing those
openings. As a result, if you use the ! or {play nn%} options in the book pgn
input, then those openings will be played at the specified frequency no matter
what the learning outcome...
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.