The evolution of sports analytics: A Q&A with UNC Charlotte expert Michael Schuckers
The field of sports analytics continues to become increasingly important across all sports. From the court to the turf, UNC Charlotte students are getting hands-on experience and helping teams make winning decisions.
Michael Schuckers, a professor in the School of Data Science, has extensive experience in sports analytics, consulting for teams in the National Hockey League and Major League Baseball. He answered questions about the role of sports analytics in baseball, hockey and football.
How has the sports analytics field evolved in the last 10 to 20 years?
There are a couple of ways that sports analytics has changed in the “Moneyball” era. First and foremost, sports analytics has become a more widely used tool across most of the major sports. Second, there has been a revolution in the data that is available to teams. Most of this new data is in the form of player location data, which tracks the coordinates of all players and the ball/puck ten times or more per second. Having this data provides much more gradual information about player actions. Third, AI and all the tools/computational power needed for that have allowed us to build better and better predictive models with all of these new features.
How has data collection evolved in Major League Baseball? What are the main metrics analysts are looking at?
As with other sports, data analysts for teams have more and more data available to them. Baseball was among the first to have data on player locations captured through cameras at the ballpark, and baseball has had pitch-tracking data, velocity, movement, and location – collected 60 times per second since 2008. More recently, they have improved the quality of pitch tracking and added data collected on swings by the batter such as the swing path and the attack angle.
How has data collection evolved in the National Hockey League? What are the main metrics analysts are looking at?
Like the evolution of data collection in other sports, we have seen the NHL progress from box-score counts to more sophisticated measures. Much of what people study in hockey is based on shot attempts, since as Wayne Gretzky said, “You miss 100% of the shots you don’t take.” We have gone from counting shots to weighting them by the probability of their leading to a goal, often called expected goals (xG). As with other sports, hockey now has tracking data: the location of players on the ice and the puck many times per second. However, these data are not available outside of team employees and so the innovation that has happened in American football has been siloed to NHL teams in hockey.
How has data collection evolved in the National Football League? What are the main metrics analysts are looking at?
The traditional metrics for football such as yards gained and percentage of passes completed were augmented starting around 2014 with tracking data: the location of the players and the ball at ten times per second. Beginning around 2018, teams gained access to data from all 32 teams. This has revolutionized the analysis of players and teams as we can know so much more about who is involved and their impact on each play. Likewise starting in 2018, the NFL created a yearly data analysis competition called the Big Data Bowl (BDB) under the leadership of Mike Lopez, who is now senior director of data and analytics at the National Football League. The BDB has improved the quality of analysis in football. Budding young analysts have access to interesting data and teams have better ways to evaluate whom to hire. The last number I saw was that 75 BDB participants have been hired in sports analytics by professional teams. I’m teaching an advanced football analytics course this semester as part of our new sports analytics major at Charlotte and our students will compete in the 2027 BDB as their final project in the course.
What types of internships or hands-on experiences are UNC Charlotte students getting in the sports analytics field?
We have a variety of experiential learning opportunities for students as part of our sports analytics major and certificate programs. Each semester, my colleague John Tobias runs an internal internship program where Charlotte students work with Charlotte varsity teams on analytics projects. Last spring, I ran a remote internship course with an NHL team where they built data-based models to help that team. That team would like us to repeat that again next semester. Experiential learning opportunities like these are fantastic for students so they can take what they are learning in the classroom and apply it to practical settings.
Written by: Andrew James and Michael Schuckers