David Sparks

David Sparks received his PhD in 2012.  He currently serves as the Director of Basketball Analytics for the Boston Celtics.  

 

The event that I remember most fondly was a small conference at the University of Michigan in honor of George Rabinowitz, where I presented a paper with John Aldrich. I met Philip Converse, Keith Poole, and many other scholars doing work on scaling, ideology, and polarization. It was such a great experience to exchange ideas with people I admired so much and whose work I had been citing -- I felt very connected to the discipline.

 

The thing I find myself missing the most is, oddly enough, our department's pick-up basketball games. Our games were such a relaxed respite, and it was a great way to get to know people.

 

I was heavily influenced by my involvement with Political Institutions and Public Choice (PIPC) and the Ralph Bunche Summer Institute. The PIPC program, which let me work closely with Dave Rohde and John and other graduate students, was a great environment in which to work. I got to learn from other students, like Mike Brady, Jacob Montgomery, and Brendan Nyhan, how to develop research and begin building a career. I am grateful to Paula McClain and Scott de Marchi for letting me work with the Bunche students; each summer was a great learning experience for me.

 

My advisor, John Aldrich, was also influential, particularly in my research and professionalization. When I first came to Duke, I had a very nebulous sense of my research interests. Over the course of the program, my focus changed substantially, shifting toward projects on classical scaling, polarization, and partisanship, in which John and I were both interested. I also really appreciate John's mentorship, his work ethic, and the good-natured way he interacts with everyone.

 

I think the Empirical Implications of Theoretical Models (EITM) approach to integrating theory and empirics has served me well. Even when I am not explicitly developing formal models, I approach my statistical work from the perspective of a social scientist, always seeking to test our theoretical expectations, rather than solely mining data for patterns.

 

I also prospered from my involvement with the Social Science Research Institute and the Program for Advanced Research in the Social Sciences (PARISS) program, which, along with my coursework, exposed me to a wide variety of methodological approaches. At Duke, I was never forced to adopt one particular tool at the expense of others, and as such I am now able to apply a wide variety of techniques and adopt new ones relatively quickly.

I currently serve as the Director of Basketball Analytics for the Boston Celtics. From the beginning of my time at Duke, as I was introduced to statistical computing and various methods, I would reinforce my learning by finding applications in sports. As I developed those skills, I got the opportunity to work for the Celtics remotely as a statistical intern. By the time I finished my PhD, my relationship with the team had progressed to the point that I was offered a full-time position.

My advice to undergraduate and graduate students is similar: take classes that cover difficult material, and be bold and experimental in the challenges you pursue. As a student at Duke, you are being taught by people who are not only pushing the boundaries of their field, but who put lots of time and effort into helping you learn from their expertise. Outside of school, and to a lesser extent post-prelims, autodidacticism seems to be the rule. I certainly wish that I had taken more statistics and computer science courses under the guidance of people whose job it was to ensure that I learned those skills.

 

Also, Duke as an institution really wants its students to do impressive things. This means that there are resources available if you want to try something, and there is little downside risk relative to that which you might encounter after college. It is difficult to generate good ideas, but if you have some, Duke is a great place to explore them.