This model uses a Linear Regression AI model to determine which factors are the most predictive for who will win football games by a certain margin (i.e. covering the spread). Historically, this model predicts at above a 60% rate which is better than most professional bettors.

NFL Prediction Model

The model below uses machine learning to predict a binary outcome. The model determines whether or not a football team should cover the spread with Predicted Home Cover as the dependent variable. In this example, we used 20 different independent variables including many defensive and offensive statistical categories. The model’s accuracy was over 64% when moving the Prediction Score slider on the top left to at least 60 (i.e. higher precision). In
the example above, the ROI on a high velocity product was about 4x, which would be a great return on a business investment.

The Do-it-yourself (DIY) NFL model gives you the chance to build your own NFL model and increase your Return on Investment (ROI). You can move the various sliders to help you find the perfect model.

DIY NFL Model

My Services

Transforming Data into Insight

Starting from scratch, we compile complex stats from numerous sources. We use a variety of skills and tools to create advanced analytics that lead to actionable insight.

A Passion for Sports

Since growing up as the grandson of the SEC commissioner, my passion for sports started at a very early age. I have since combined that with my professional analytics experience.

Actionable Insights

Receive easy-to-follow advice and predictions for your bets.