|
Re: Amit's total results |
Amit; No, if you make predictions for different quakes I have to compute the probability of each quake separately. Then the score reflects the different odds and combines them into a single result. The score uses the natural log of the probabilities. If you're interested, I can send you a write-up of the method, a bit too long to post here. Roger Follow Ups: ● Re: Amit's total results - Amit 18:17:08 - 12/16/2012 (80834) (1) ● Re: Amit's total results - Roger Hunter 18:46:25 - 12/16/2012 (80835) (1) ● Re: Amit's total results: CANCEL - Roger Hunter 19:13:44 - 12/16/2012 (80836) (1) ● Re: Amit's total results: CANCEL - Skywise 19:23:26 - 12/16/2012 (80837) (1) ● Re: Amit's total results: CANCEL - Roger Hunter 21:45:11 - 12/16/2012 (80838) (0) ● Re: Amit's total results - Skywise 15:55:04 - 12/16/2012 (80826) (1) ● Rules for evaluation - Roger Hunter 16:08:31 - 12/16/2012 (80827) (0) |
|