A reinforcement learning approach to developing an optimal policy for NFL offensive playcalling
The following project is based upon the idea of implementing various reinforcement learning algorithms in order to develop an optimal policy for calling plays. The policy development is based upon the Kaggle Detailed NFL Play-by-Play dataset.
- Python >= 3.5 (type hinting)
- Pandas
- Kaggle API
In order to setup this project, it's necessary to clone the repository and download the dataset from Kaggle
git clone https://github.com/jacobeturpin/nfl-offensive-playcalling-optimization.git
cd nfl-offensive-playcalling-optimization
kaggle datasets download maxhorowitz/nflplaybyplay2009to2016 -f "nfl-play-by-play.csv" -p ./data/ --unzip