| |
BGonline.org Forums
a new type of match play--Net Score Gammon
Posted By: Bob Koca In Response To: a new type of match play--Net Score Gammon (Phil Simborg)
Date: Sunday, 12 August 2018, at 7:20 p.m.
"and learn how it is done."
First decide what is the winning margin (e.g. 7). For regular backgammon suppose you have a 21 point MET. It could be used for any smaller match length but possibly annoyingly there would need to be a different MET for each match length. What is "nice" though is that there is less to know. For a regular 7 point match a MET has 55 entries but for your variation there are only 7 (and one of them is trivial 50% for net score of 0). I am assuming the Crawford rule is in effect for any score of one less than the goal and in the following discussion that you are in a goal of 7.
As a first step just play a bunch of matches between equal opponents. I would start though at different scores (we already know 50% for net score of 0). If the score ever reaches 0 you should settle that as a 50% win. Along the way keep track of every score reached. For example if the scoring goes +4, +2, +3, + 4, -4, + 7 then it is like you had 5 different trials that got settled.
For the score of +4 count 2 wins and 1 loss. for the scores of +2, + 3 count 1 win for each.
Once you have the MET it is pretty much the same logic as you know already. For example when should oppoenent accect a race when opponent is +4. If drop is at +3. The MET value for +3 is value from taking. If accept are either at +2 or +6. The loss L is MET value at +3 - MET value at +2 and the gain G is MET value at +6 - MET value at +3. The takepoint is L/(G+L). That takepoint ignores cube value and bonus wins which could be taken into account as they are in regular backgammon.
Once you get an MET which is close enough you could modify a program such as GNU to play the matches for you and the increased sample size would get you a more accurate MET.
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.