NFL Prediction Model

To predict the match-winners of a gameweek of the National Football League based on game stats parsed from NFL.com's live GameCenter using the nflgame API - API used in order to get the in-game stats.

Things done

• Built a model using RandomForest Classifier and utilized statistics like Rushing Yards per game, Passing Yards per game, etc. after applying Feature Selection.
• Wrote ETL(Extract,Transform,Load) job to collect stats for the home and away teams for the last 10 years using python nflgameapi.

This project works on python 2.7 as the API first worked only for python 2.7 but this would be updated to python 3 in some time.

The outcome and the predictions for the gameweeks are all posted on my Blog on Medium. Do check it out.

If the notebook is not rendering then use the link given below in order to view the ipynb file.

• nbviewer