| |
BGonline.org Forums
Questions about MET
Posted By: Matt Cohn-Geier In Response To: Questions about MET (adambulldog)
Date: Wednesday, 17 September 2008, at 6:55 a.m.
1. Is the g11 better than the Kazaross MET? I loaded the Kaz one as it seemed to be the more recent.
The Kazaross MET only works for 2-point matches, I believe, but could be wrong here (maybe goes up to 5 point matches?). The g11 goes up to 11 point matches.
2. Do you really believe in the g11 (or other MET) to the level of precision that is presented? It seems to me that gammon rates probably vary among players (even among players of similar overall skill) and certainly vary between human players and computers.
Do you believe in the results of any rollout? g11 isn't perfect, but it is our best guess. I don't memorize it to 4 decimal places and try to use it in match play, but it certainly is more accurate than Zadeh. And, considering it was rolled out by GNU, it is probably the closest for 0-ply/2-ply's actual gammon rates and so on.
2a. The game in question was between a Danish advanced player and me. I'm not convinced that the difference of 39.7% ME for 3-away, 2-away given in the Zadeh MET and the 40.1% presented in the Kaz MET actually represents a meaningful difference given the players involved.
Maybe true. GNU will also vary its checker play based on the MET, which will sometimes yield a different result.
I'm pretty sure I win fewer gammons than does gnu or snowie.
Doesn't the advanced player also save less gammons? If you're significantly better than he is, you should win more games and gammons than GNU or Snowie, right?
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.