|
BGonline.org Forums
Counterexample?
Posted By: eXtreme Gammon In Response To: Counterexample? (Chuck Bower)
Date: Tuesday, 5 February 2013, at 4:26 p.m.
Chuck: I don't know if this is how MWC errors are tallied by XG, however, so maybe XG's method wouldn't give > 100%?
Errors in equity or MWC are simply added and can be greater than 1 or 100% respectivly in a DMP.
The only time I renormalize MWC is on the line "Luck base rating" in the match summary. I apply the following formula
VR := Result - DeltaCostLuck; VR := 1 / (1 + exp((0.5 - VR) * 4)); This insure that a player cannot be more than 100% (or less than 0%) favorite. The slope is 1 around 0.5 so for most practical uses the value reported is very close to the difference of luck MWC.
|
BGonline.org Forums is maintained by Stick with WebBBS 5.12.