Predicting NBA Game Winners

Overview

This project is based on a lecture by Robert Layton given at PyCon Australia, 2015. The goal is to build predictive models that determines whether the home team will win an NBA regular season basketball game, then evaluate the how well the models perform. The data used is from the 2012-13, 2013-14, and 2014-15 NBA seasons.

Python Libraries Used

Numpy
Pandas
sci-kit learn

Models Used

Decision Trees
Random Forest

What's In This Repository?

Datasets (folder)

• 2012-2013 Regular Season Standings.csv
• NBA Regular Season Results 2013-2014.csv
• 2013-2014 Regular Season Standings.csv
• NBA Regular Season Results 2014-2015.csv

Predicting NBA Game Winners.ipynb (Jupyter Notebook)

Additional Resources

Recording of lecture: https://youtu.be/k7hSD_-gWMw
The dataset for this project was retrieved from: www.basketball-reference.com