About Predictor

My name is Joey Bloom. I graduated from Cornell College in 2018 with a degree in business analytics and I currently work in marketing as a solutions architect in the Chicagoland area. The summer I graduated, I started working on my first betting model and called it "NFL Predictor."

With some trial and error, I created a process to efficiently develop machine learning models for betting, and this website was created to share the predictions.

The models are trained on data wrangled together from 2010 - 2017. The model was evaluated using data from 2018 - current season. Data from nflfastR and nflgamedata.com was used to create the training dataset.

2 machine learning algorithm techniques were used -
- Generalized Linear Model (GLM)
- Random Forest (RF)

For each algorithm, a base model and a tuned model (“T” added to the end of the name) were developed. The base and tuned models both share the same attributes, but different algorithms may have a separate set of attributes.

These models are not guaranteed to produce any profits.

I am also active on twitter under my account @data_bears. Please feel free to reach out to me there if you have any questions.