| |
BGonline.org Forums
Luck Adjusted Result
Posted By: MaX In Response To: Luck Adjusted Result (Christian Munk-Christensen)
Date: Monday, 10 August 2009, at 11:27 a.m.
One thing to keep in mind: lar computations are done (by default) 0ply. Doing them 2ply is possible, but a bit longfor my taste. This, combined to the fact that for a neural net computing the luck in a consistent way is more difficult than computing the "simple" equity of a move, may explain deviations from te expected results.
It still is, imo, a useful piece of info and it makes sense conceptually.
BTW, I'm not sure what people mean saying "luck estimation is unbiased" ...
MaX.
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.