| |
BGonline.org Forums
Some history
Posted By: Chuck Bower In Response To: Supremo MET in the Real World (jdg)
Date: Wednesday, 16 December 2009, at 2:36 a.m.
Kit outlined how his now famous MET was built in the Mar-Apr 1992 issue of Inside Backgammon vol. 2 #2, p. 4. If you can't get hold of the article, here is a synopsis.
Circa 1980 Kit wrote a computer program to generate a match equity table based upon some assumptions he doesn't detail in this article. He combined his table with two independent ones generated by Kleinman and Robertie and published a consensus table in Backgammon Times in 1981.
For this 1992 article he compared the previously published table's results with ~1000 matches provided by Hal Heinrich. He then modified his program to account for these empirical data. In particular he mentions Post Crawford gammon rate of 21%. I believe that number was used throughout his program in generating the new table.
On one page in the article he shows comparisons of the original table's MEq values and the database results, with number of actual matches tallied for each score. The largest number of matches for any entry was 207, followed by 206 -- those being the only scores where more than 200 matches were used. For Crawford scores, four entries had between 65 and 79 matches tallied and three scores had between 23 and 42 matches investigated.
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.