By As an update to 4.0 users, we tested the 0.25 alpha setting again over the weekend for the exponential model compared to the 0.5 setting at four tracks, specifically, Belmont, Churchill, Gulfstream and Santa Anita, using data for five recent cards at each. The study considered all races, including maidens. The model had positive ROI results with its top selections at Belmont (+15.1), at Gulfstream (+30%), and negative at Santa Anita (-5.5%) and Churchill (-33%). Separately, on Sat., June 1, the program had positive ROI% at Belmont (+2.0%), Gulfstream (+4.0) and Churchill (+25%), while negative at Arlington (-30%).
According to this latest data, it appears that the exponential model does better with the 0.25 alpha setting than at 0.5. For more information on the results, see the bottom of the HW 4.0 Studies page.
Pro-Handicap Analytics 