Mike Lopez, associate editor of the Journal of Quantitative Analysis in Sports, is a senior director of football data and analytics at the National Football League and a former assistant statistics professor at Skidmore College. In 2020, he was honored with the American Statistical Association’s Statistics in Sports Significant Contributor Award.
When were you first interested in statistics applications in sports?
Each time we used to stay at hotels as a kid, I would open USA Today and look at two things. First was the sports page, and second was the infographic they would always show in the corner on the front cover. So, sports and statistics have always been two of my passions—I just had to wait a little while until there were careers in the intersection of those fields.
What are your professional duties as a sports statistician?
Our job is to use data to enhance the game of football, and that covers anything that relates to the on-field tendencies of players, coaches, and teams. We have metrics relating to game excitement, fairness and equity, player health and safety, officiating, and pace of play. We also play a role in innovation regarding the future of the game.
What path did you take to become a sports statistician?
I never had the goal of being a sports statistician. Instead, once I realized my passion for analyzing data, my focus was more generally on building skills that could translate to careers in analyzing or teaching statistics. That path was a mathematics major at Bates College, high-school teacher (and assistant football coach!), six years of grad school at the University of Massachusetts-Amherst and Brown, and then four years at Skidmore College as an assistant professor. None of those roles were explicitly designed to end in a sports statistician role, but between the public speaking, subject-specific expertise in football, ability to teach and work with other researchers, technical/coding skills, and, most importantly, evolution of the sports world, the stars aligned.
Tell us about your typical day at work.
We have a football analytics team of seven folks right now, and so most of my time involves ensuring the short- and long-term success of that group. We are a mix of both junior analysts and successful data scientists covering a variety of disciplines within the game. I will also spend a good amount of time coding—working to build new metrics, improve old ones, or create reports or presentations based off our work. In season, we are responsible for reporting on the season as it progresses, and we work with the NFL’s Competition Committee each offseason when it comes to rules changes and ways to enhance the game.
What skills and academic training (e.g., college courses) are valuable to sports statisticians?
Here are a couple avenues for folks to think about:
- Most statistics or data science courses will in some way aid a career in sports data. One of the first ways I got started in the field was by taking the methods I learned that week in grad school and seeing if there was a corresponding way to answer a sports question with them. Between the Poisson (goals and penalties) and binomial (win/loss, made/missed) distributions, hierarchical modeling (players as random intercepts), spatial statistics (heat maps), random forests (win probability models), and survival analysis (time until an injury), sports is an excellent sandbox to play in while improving your statistics acumen. Even better, most sports data sets are free and can be found online.
- Volunteer with a local team or sports organization. My time helping coach high-school football was invaluable to what I do now at the league office. Knowing what, for example, coaches are looking for, how to make a scouting report, or how officials call the game can improve the questions you are asking with data. Additionally, improving skills in how to communicate with folks who are not data experts is something we are constantly striving for.
Are there specific data-management software or statistical tools you find especially helpful in your work?
Our Football Data and Analytics team is a mix of R and Python, and most of my coding is done in R. Aggregating to the NFL, I’m guessing it’s about 60 percent R, 30 percent Python, and 10 percent other languages (e.g., Stata, SAS, etc.).
For tools, nothing beats a powerful visualization. Even the most complex of our models are almost always distilled into something easier to interpret.
What are first steps you would suggest for entering the sports industry as a statistician?
One aspect of the field that has changed over the years is that, more and more, folks joining teams have already built a portfolio with examples of work. In most cases, this is a good thing, as it puts less emphasis on what degree you have or where you went to school and more focus on the skills you can bring to a team.
At the NFL league office, for example, our group puts on a data science competition each year called the Big Data Bowl, in which the goal is explicitly to find talented folks who can handle the league’s player tracking data. In four years of the competition, more than 30 participants have gone onto jobs with NFL teams or companies that analyze NFL data.
Are there particular websites for the interested student to visit to learn about working in sports data analytics?
Most job posts are on Teamworks. But more than any place to learn about the field, I would encourage folks starting out to think about interesting questions in whatever sport they are interested in. What is the question that needs to be addressed? What problem are you trying to solve? What other work has been done in the field?
One of the fun parts about working in sports is that because it is such a passionate area of public discourse, there is a lot that can be learned just by listening and problem-solving. Additionally, if you are working in a domain in which you are passionate, it can be fun to solve interesting questions.
Do you have other general advice for high-school or college students who are interested in a career as a sports statistician?
Though it is certainly a great time to work on a team in sports analytics, more broadly, there are only about 130 professional sports teams in the big four North American leagues. That is not a ton of jobs each year, and even if you do land one, it may come with a demanding set of hours and in a less-than-ideal location.
Outside of teams, though, there are several jobs in sports data that may also prove exciting while requiring a more reasonable workload. These could be jobs with vendors of league data (TruMedia, Telemetry, Zelus Analytics), the league offices, companies involved in sports science or sports performance, or even some college teams. Regardless, having skills that can help a professional sports team win games will always translate outside of sports, too.
Do you have a favorite sport or team?
There are 32 great NFL teams! Outside of the NFL, I grew up outside Boston, so the Bruins, Celtics, and Red Sox are most likely to capture my attention.
Leave a Reply