A better probability model for the racetrack using Beyer speed numbers

The dominant model in research related to the racetrack is that a horse's probability of winning a race is equal to the fraction of the win pool bet on that horse when adjusted for a favorite long-shot bias. A new model is developed using Beyer speed numbers to estimate the probability of a horse winning a race. An SPRT like test is developed to determine which of these two models is better. Although developed for the racetrack, this SPRT like test can be utilized whenever there are two models for assigning probabilities and the better model needs to be selected.