Databall
Betting on the NBA with data
Overview:
DataBall is a project that combines data science and sports by attempting to predict NBA winners against the spread. The project uses statistics pulled from the NBA stats website, point spreads, and over/under lines from covers.com. All predictions are made using Python and the scikit-learn machine learning library.
Features:
- Scraping data: The project includes a Scrapy project that can scrape point spreads and over/under lines from covers.com.
- Support functions: The project includes a Python module called "databall" that provides support functions for tasks such as collecting stats to a SQLite database, simulating seasons, and customizing plots.
- Documentation: The "docs" folder contains the necessary code to build the GitHub Pages site for this project.
- Notebooks: The "notebooks" folder contains Jupyter notebooks of all analyses.
- Report: The "report" folder contains LaTeX files for the report and slides.
- Jekyll
Jekyll is a static site generator written in Ruby that allows you to create simple, fast, and secure websites without the need for a database.
- GitHub Pages
Gridsome is a Vue.js-based static site generator that makes it easy to build fast and flexible websites and applications by leveraging modern web technologies like GraphQL, Webpack, and hot reloading
