This project aims to create a comprehensive suite of NFL betting models that leverage machine learning and statistical techniques. The main objectives are:
- Effective Data Pipeline: To fetch, clean, and aggregate play-by-play NFL data.
- Statistical and Machine Learning Models: To predict game outcomes for both spread and total betting.
- User Interface: To display predictions and insights through an intuitive UI.
Data Loading: Successfully loaded play-by-play data.
Week-by-Week Data: Currently perfecting the creation of week-by-week statistics.
Rolling Averages: Creating rolling averages from the week-by-week data.
Betting Models: Development of various betting models including:
- Linear Regression with Various Quantile Loss Functions
- Neural Networks with Various Quantile Loss Functions
- Potentially other model types, like tree-based models. UI Development: User interface for model interaction.
Python Pandas for data manipulation nfl_data_py for fetching NFL data Planned: PyTorch for machine learning models
Instructions to set up and run the project will be coming soon.
Contributing Please read CONTRIBUTING.md for details on our code of conduct, and the process for submitting pull requests once the project gets up and running.