| |
BGonline.org Forums
MET update
Posted By: NJ In Response To: MET update (David Rockwell)
Date: Wednesday, 23 September 2009, at 8:17 a.m.
For Backgammon NJ, I created my own MET. I essentially just did a cubeful evaluation of the opening position at each match score. The neural net gave the gammon chance of the opening position as 13.77% per side, or 27.54% overall. Backgammon chance was 0.57% per side. Cube efficiency was set to 0.67. The resulting MET somewhat resembled mec26.
I'm interested in how you arrived at a 21.7% take point for the leader at -2, -13. Using my met and a 0.67 cube efficiency, my program says the take point should be 13.9% at -2, -13. That is a big difference.
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.