[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: [Bug-gnubg] Gnubg's ratings and checker versus cube errors.
|
From: |
Misja Alma |
|
Subject: |
RE: [Bug-gnubg] Gnubg's ratings and checker versus cube errors. |
|
Date: |
Wed, 03 Dec 2003 17:55:21 +0100 |
I
don't understand why there should be different ratings for cube and checker play
errors? Is it perhaps not a good idea to base the rating on lost game/ match
winning chances regardless if it was a cube error or a checker
play error?
You
mentioned your money play series between GNU and SW. I'm curious how you
calculated the errors that both programs made; I mean, which program has
done the rollouts?
Misja
There's
always been a lot of discussion about how to assign ratings based
on
checker and cube errors and particularly their ratio.
Some have
commented that Gnubg is too harsh with it's ratings for cube
errors.
Based on my GNU v SW money play series both Gnubg and S4 have a
ratio in EMG where cube errors are ~35% the size of checker play errors.
(I
took the liberty of removing a S4 beaver blunder in a highly unusual position
to get this number).
Is there a way to use this as some kind of
benchmark and incoprporate this into Gnubg's rating
assignment?
Obviously there's Matchplay to consider as well.
I
just wanted to start off a discussion.
Michael