As a  master’s degree candidate in the Masters of Applied Data Science program at the University of Michigan School of Information, I had to create a final Capstone project which highlighted some of the learning in the program. I was fortunate to work with two outstanding team mates, Anthony Giove and Avinash Reddy. We created BaseballML.com, a site with dashboards and tools for new ways to analyze the game of baseball.

There is a lot of work done on sports analytics, but we wanted to contribute something new. There are some interesting things on the site such as some great visualizations and machine learning models, including …

Player Explorer Dashboard

The dashboard allow you to explore stats about individual players including an interactive tool for exploring baseball historian and statistician Bill James’ player similarity scores and Hall of Fame (HOF) Worthiness explorer to estimating the likelihood if induction to Cooperstown at different points in a  player’s career.

Strike Zone Explorer Dashboard

This dashboard features some great visualization for understanding strike zones in actual game conditions. It includes Batting Average Profiles for hitters and pitchers to visualize player success based on pitch location, an Umpire Strike Zone Visualizer for seeing strike zones based on actual umpire calls, and, a “What if I pitch here?” tool for pitching outcome prediction to analyze the likelihood of throwing a strike under different conditions.

Team Similarity Tool

This tool started as an idea to expand on player similarity scoring to better understand team rankings and possibly predict team performance.  There’s still work to do on this tool, but we’ve made an interesting start.

There are a couple of other nice things on BaseballML.com including a data download tool where you can take advantage of our web scraping and data collection work to access our database of over seven million pitches for your own analysis work.

Our first Baseball Machine Learning Blog explains the data science behind our work. Its a very long blog, but will have some interesting content for baseball fans and data junkies.

Thanks again to Anthony and Avinash for being great team mates!