| |
BGonline.org Forums
Match Equity Formula Reviewed and Revised
Posted By: Rick Janowski In Response To: Match Equity Formula Reviewed and Revised (David Rockwell)
Date: Wednesday, 19 June 2013, at 1:45 a.m.
That’s an interesting question, David. It hadn’t occurred to me when carrying out this analysis that there were any noticeable discontinuities between the 15 point and 25 point match zones. However, I thought I would check against a 25-point match MET derived purely mathematically, i.e. where implicitly there should be no discontinuity because there is no joint. I used mec26 for this purpose (the full mathematical version, not the one incorporating Gnu rollouts).
The results appear to compare very favourably. Both METS would have max errors of 2.2% when the formula parameters are derived for the purpose of minimising errors in a 25-point match. When the objective is to minimise errors in a 15 point match, mec26 has a maximum error in this zone of 1.1% (cf.0.9% for R-K) and has a maximum error in a 25-point match of 2.5% (cf. 0.028 for R-K). For the mec26 MET, errors do not exceed 1.5%. up to and including a 22-point match (compared to 19 point match for R-K). After that both METs display fairly rapid increase in error, with the same type of pattern with maximum errors occurring at near 15 away for either the match leader or trailer vs 19-25 away.
It looks to me like the 15-25 point match extrapolation process you and Neil carried out worked very effectively. A job very well done.
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.