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

BGonline.org Forums

Paper on EMG inconsistencies

Posted By: David Rockwell
Date: Tuesday, 27 November 2007, at 3:54 a.m.

In Response To: Paper on EMG inconsistencies (Jeremy Bagai)

I have read and enjoyed your article. I particularly liked your articulation, "A backgammon position is not really what's on the board—it is all the legal continuations from what's on the board. Two positions may look different, but if they present the exact same set of options, they are the same position." EMG clearly isn't working in the table shown in your post.

I think you've also identified the key issue, "...getting numerical feedback from Snowie and gnubg is how most players advance their game." I've checked the error in an opening 6-2 played 24-16 at two different scores using Snowie 3 ply, 0-0 /65 & 60-60 / 65. EMG shows an error of .033 & .034 while MWC show .06% and .25%. As you point out in the article, MWC says more about when the error took place within the match rather than the quality of the error. I want to get feedback that gives the same result for the same error unless ME considerations legitimately affect the quality of the error. EMG looks like the better of the two choices here.

Are there significant problems with EMG other than cube decisions with overage? I would be surprised if EMG did not work reasonably well for checker play. I would not be surprised if EMG failed to compare cube and checker errors well. Before finding a solution, it would be good to get our arms around the extent of the problem. (First do no harm - I've always thought that was a lame approach to medicine but good advice for software development.)

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.