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

BGonline.org Forums

Improved Cube Handling in Races

Posted By: Axel Reichert
Date: Sunday, 29 June 2014, at 7:46 p.m.

In Response To: Improved Cube Handling in Races (JM2004)

Thanks for your feedback.

The mere fact that you fitted your formula to a small database (all 3003 bear-off positions with at most 8 checkers) did not necessarily mean that it could not be applied to positions with more checkers or requiring some bear-in. Having read your article, it was simply unknown to me whether your method perhaps would work reasonably accurate or not.

Since I Liked the approach of your paper very much, but wanted a method as general as possible, I decided to test it on Tom Keith's much larger database. This is perfectly legitimate and the fact is clearly mentioned on page 20 in my article. And the outcome of this test definitely represents some new knowledge: Applied on these more than 50000 positions, the average EPC error of my method is 1.04, whereas the average EPC error of your method is 1.70. Your method is very accurate for the small database (3003 positions, average EPC error 0.48), for which my method has an average EPC error of 1.05. But again, I wanted a general method, not something for a rather small subset.

I did not restate the fact that you tested for the small database only, since I recommended to read your article in full anyway.

Hopefully this clarifies things. When extending my article (I got a lot of feedback and interesting suggestions that I think are worth pursuing), I will rework the relevant paragraph. Please be patient in the meantime.

Best regards

Axel

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.