cjf4 / football_analysis Goto Github PK

My goal for this project is to fit a classification model to NFL plays to derive a metric called “Win Probability”, and an additive metric “Win Probability Added”. My hypothesis is that WPA will be a better indicator of player performance than traditional player metrics.

Data was obtained using the nfldb postgres database and associated python package. The data collection scripts require a connection to an installed local copy of the nfldb databse. Directions for installation can be found here.

Once downloaded, raw data was queried via raw SQL and transformed to an appropriate format for classification model fitting.