I just noticed this thread, so some of what I interject here may be redundant.
I have been tracking StatFox and Sagarin most of the season. Unfortunately, I experienced a "crash" around Xmas and lost valuable data and formulas. I began tracking again on Jan 28th, 2010.
The results are games in which there was a disparity of 3.99 or more using StatFox and 3.84 using Sagarin Predictor, vs. their respective predictions.
StatFox: 176 Plays: 90 Wins, 86 Loses, 51.1%
Sagarin Predictor: 367 Plays: 194 Wins, 173 Loses, 52.9%
The most accurate was Pomeroy. I'm sure you're aware he uses adj. efficiency ratings and pace figures to determine outcome. Once again, using a difference of 3.99 from the posted line (usually Betjam), his record in the same span of time is:
Pomeroy: 97 Plays: 60 Wins, 37 Loses, 61.9%
I've built a program that uses most of what Pomeroy does, but I use a range disparity of 3.49 to 4.99 in increments of .5 pt. to rate selections. The 4.99 and up produces 68% winners, but only 37 bets in the time frame. The 3.49 pt range produced 59% winners on 102 bets. As always, the trick is to optimize the number of plays and NOT decrease the winning percentage.
Once again, if this is redundant sorry. Just inserting some info I've collected this season.....hoping it helps save you some time.
Good luck, guys, with building a revised tool. I'll check back periodically to see how it goes.