This artifact is the final report of my project for MCS 100: The Mathematics of Sports. My partner and I decided to use predictive modeling to try to determine which NBA player should win the Most Valuable Player (MVP) award that season.

Along with Beating the Bubble, this is one of my first independently-led projects. I was really proud of the way we carried this project, especially because the small amount of data (there is only one MVP every season) made it very challenging. From a science communication perspective, I thought that we adopted the correct structure for a research paper. I think the structure of this paper is quite strong, and it is easy to follow our progression through the project, what we did and why we did it.

I am also very happy with our use of plots and tables. I think that we did a great job of exploring our data before doing anything, which helped us set a clear outline of what we wanted to for the rest of the quarter. Moreover, I think that the plots that we have are a great way to explain to our audience what our data was like as well as why the research question we are tackling is hard to answer.

Moving forward, I learned that it is sometimes better to use styling for a specific conference to build stronger ethos in a research paper. In this case, we could have considered using the MIT Sloan Sports Analytics Conference report templates. In addition, I think we could have explored our results a little more. I wish we’d had the chance to show some visuals about where our models made mistakes, rather than just a large table of results. For instance, we could have decided to plot all players on a points per game versus assists per game chart, highlighting the season’s MVP and our prediction.

This project was completed in collaboration with Gonzalo Majluf.