How StkUltra uses
Statistical Probability
to produce the odds line

We can think about the chances of a horse winning a race by first considering a perfect race. In this race, each horse has exactly the same qualifications as the other. We put 10 such horses in the starting gate, and BANG, let the doors pop open.

There are an infinite number of things that can and do happen between the gate and the wire that makes one of these horses the lucky winner. Each horse had one chance in ten, for equal odds of 9 to 1. (not 10 to 1 because you get your $2 bet back when you win.)

StkUltra correctly shows this, you can test it by setting all the variable weights to zero as in the figure to the right. This gives each horse an equal score for the StkUltra calculation. The StkUltra selection process settles point-score ties by favoring the lower post position, so in this situation the A, B, C, and D selection goes to posts 1, 2, 3, and 4.

The total point score of all horses in a race divided by the point score of a single horse in the race gives the point-score odds for that horse. The odds-spread function of StkUltra pushes low odds lower and high odds higher while maintaining the correct overall statistical probability. (The sum of the reciprocals of the odds line must equal 1)

The raw point-score odds will seldom be accurate.
We can force them to be accurate by spreading the odds with the odds-spread function of StkUltra.

The odds-spread analysis report above shows results that a correctly set odds line would produce. Predicted wins are calculated based upon the fact that correctly set odds will produce wins at a rate of (horses-at-those-odds / odds-number + 1). The analysis compares predicted wins to actual wins and accumulates the difference across the first 20 odds numbers.

Variable #18 under the Odd column in the image above is the odds-spread variable. The bigger that number the greater the odds are spread between the lowest odds and highest odds. The number is user controlled. The training method #3 provides the analysis shown above wherein the spread may be changed like any other variable, simply by entering it's number and responding to the Change-it-to- prompt with the desired number.