[ View Thread ] [ Post Response ] [ Return to Index ] [ Read Prev Msg ] [ Read Next Msg ]

BGonline.org Forums

Depreli comparison study

Posted By: neilkaz
Date: Tuesday, 9 November 2010, at 4:35 p.m.

In Response To: Guy returns to backgammon after not playing for four years; guy asks questions about eXtreme Gammon and XG Roller double/beaver position (Flaccus)

http://www.bgonline.org/forums/webbbs_config.pl?read=64993

What more can I say except to point out that on equivilent plies, XG evaluations are shown in this study to be more accurate than the other bots.

Anyhow note that as one would expect, XGR+ is significantly stronger than XGR.

We shouldn't care whether a bot on some ply goes haywire for something in a nearly impossible position and for esoterica. What we do care about is how well bots play overall and in the 99% of all positions that are rather usual.

Of course bots need to know how to roll primes home, but it is MUCH more important that they be able to do it from there own outerboard ..ie 6 thru 11 pts or 5 thru 10 pts. Positions with snakes from something like the 13 to 18 pts might occur once a year (likely less) in Champ level ABT play since people with a clue won't hit enough blots even if you try to get all of your checkers sent back.

MUCH more important for the bot to be able to play well after later hitting from a backgame like this for example. Here Black had a nice backgame but just rolled 44 last turn.

is Player 2

score: 0
pip: 215
Money session
Jacoby Beaver
pip: 85
score: 0

is Player 1
XGID=-bcCdCDCB------------b-bb-:1:-1:1:52:0:0:3:0:10
to play 52

eXtreme Gammon Version: 1.21

Messages In This Thread

 

Post Response

Your Name:
Your E-Mail Address:
Subject:
Message:

If necessary, enter your password below:

Password:

 

 

[ View Thread ] [ Post Response ] [ Return to Index ] [ Read Prev Msg ] [ Read Next Msg ]

BGonline.org Forums is maintained by Stick with WebBBS 5.12.