Well, the model failed us on Day 1 with the Coyotes keeping it close for half the game and then giving up 5 goals in the span of a few minutes. Onto the next one!


My NBA Winner model is showing strong predictions for 65% greater rates, and after going through all of the correct prediction percentages from the test data, correct results STRONGLY correlate to prediction probability which is the sign of a strong model. It is predicting at nearly a 75% correct rate for probabilities of 70% and greater, as well as a 67% rate for probabilities between 65-70%. Here are tonight's predictions:

The Bucks have a moneyline of -120 for implied odds of nearly 55%. Our model today has them winning 65% of the time which, as mentioned, has been true to results from the test data used. My Moneyline model is using 4 years of data over a 12-game rolling average, including things like Two-pointers, Three-pointers, and Free Throws Made, as well as Effective Field Goal % and historical Moneyline results. I've been anxious to put this model to work since NBA Predict was released recently and LOVE the difference in implied odds to probability. For these reasons, locking in Milwaukee Bucks ML -120 BOL!