One of the settings in the program that can affect results is the alpha setting, which determines the degree of exponential smoothing in the model for simulation. The alpha setting is found on the User Settings sheet toward the bottom. With a 0.25 alpha, the greatest weight is placed on the most recent final speed figures in determining the means and standard deviations for the simulation, while 0.5 alpha takes into account more final speed figures than just the must recent in doing the same. A case where it might be advantageous to change the setting from 0.25 to 0.5 is when a good horse had an off performance in his most recent race, where he or she might be expected to bounce back. The 0.5 alpha would not discount that off-performance as much as the 0.25 alpha would. Sometimes this leads to winning results where such horses find themselves in the top three selections, where otherwise the market is heavily discounting them. This happened in the G1 Alfred Vanderbilt race at Saratoga mentioned on the Disclaimer page, where Imperial Hint, who won the race, had odds of 5-1 because of recent sub-par form in terms of final speed figures, but as it turns out, also had the highest winning percentage in the field, adjusted for stakes, at 37.6%, reflecting that historically Imperial Hint was a winning horse, even more so than the heavy favorite, Mitole, in that race. By being aware of the difference between the 0.25 and 0.5 alpha, a user can make the handicapping adjustment in the model that can lead to more successful results in certain cases. Or if he prefers, a user can choose the 0.5 alpha setting all the time to angle for such horses who would be discounted or ignored by the market because of recent sub-par performance, but because of that, offer higher odds and have greater winning payouts.
Pro-Handicap Analytics 