Just bet their top score to show because of very high in the money%. Their top win bets end up being over bet with a negative ROI, but as show bets they have a positive ROI.
How many derby winners your system pick, since inception, for you to be making such bold claims? Why not say par instead of class, which makes more sense in horse racing language
Speed figures are incomplete power ratings. I actually published my power rating method free through links in my videos. SPY options are the much better gambling game anyway.
You should have a calibration method for each regression of point figures. Otherwise, your guesstimate of speed figures will be beyond of upcoming reality.