The bracketology package is aimed at creating an easy to use interface for analyzing historical NCAA mens basketball tournament data & accelerating the process of creating and evaluting bracket prediction algorithms.
Documentation is hosted by Read The Docs
Get the package from PyPi:
pip install bracketology
Create a Bracket object
from bracketology import Bracket, Game, Team
b19 = Bracket(2019)
Check out the teams that made it that year
b19.regions
# or
b19.round1
# or
b19.result.get('first')
Simulate a result
from bracketology.simulators import upset_prob
b19.score(upset_prob(0.15))
This is for the actual package files; what gets uploaded to PyPi
and downloaded when you pip install bracketology.
This folder contains the files to create the Read The Docs documentation. The structure was automatically generated with Sphinx-quickstart
The file index.rst contains the main info for the documentation.
Some Jupyter notebooks for the original data cleaning, exploratory analysis, and (eventually) some baseline modeling.
These are used for uploading the package to PyPi.
This project is still in it's nascent stages. There is plenty of room to contribute! Feel free to branch or fork this repo and PR for contributing, or just have it for personal use.
At the moment, there is not strict coding standards. But any PR's should be comprehensible and have comments where neccessary.
If you are passionate about this topic and have large scale feature ideas, or any ideas on how to better achieve the goals of this package, feel free to reach out! I would love to work with you to make this better for everyone.
Maybe one day we will get a perfect bracket.
Message me on LinkedIn
or Email: [email protected]
Big ticket feature in progress
- adding regular season data for each team each year
- more robust error handling
- adding more simulator functions to go with each package
- TBD