| |
BGonline.org Forums
Adjusted match equity
Posted By: Fabrice Liardet In Response To: Adjusted match equity (Dimitar Draganov)
Date: Thursday, 18 November 2010, at 2:52 p.m.
OK, here is how I did it for Crawford or post-Crawford scores.
Start with the probability of winning a single game (1-point match), say 60% for instance (I know that is a bit extreme, but I want simple figures). The probability for the favorite to win two in a row is 36%, while it is 25% for equal opponents. For three in a row, it is 21.6% instead of 12.5%. Based on that observation, one can get an approximation by interpolating logarithmically to get the "60% favorite" match equity starting from the regular match equity.
Formally, it means that one solves the two equations MET50 = 0.5n and MET60 = 0.6n for MET60. The solution is MET60 = MET50log2(5/3), where log2(5/3) ~ 0.737.
For instance, if the favourite trails -7,-1 Crawford the regular MET says 11.6%, and my formula says that his MWC is in fact 0.1160.737 = 20.44%.
Whether the above can be somehow generalized to other scores and/or rendered easily computable, I don't know.
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.