When it comes to sports and statistics, there’s a lot more at play than the box score. Charlie Rohlf (CDM MS ’12) is one of many individuals working at the intersection of professional athletics and big data. As a product manager at STATS LLC in Chicago, Rohlf oversees SportVU, a sophisticated tracking system that provides critical information for professional soccer and basketball teams, including all 30 NBA teams.
“SportVU is a camera-based computer vision system designed to detect the movements of the players and the ball during a basketball or soccer game,” Rohlf says. The product generates an ‘X’ and a ‘Y’ coordinate for every player in the game 25 times per second, locating him on the court or field. SportVU also tracks the coordinates of the ball. “By the end of the game, we have more than a million data points for the movements of the players and the ball,” Rohlf explains. “Then we have to make sense of all that data.”
Software engineers apply a variety of data science techniques to pull out useful information. Using algorithms, they can detect shots, passes, dribbles and so on. Rohlf actually started as a software engineer when he first joined STATS in 2012, and that background, as well as his master’s in computer science degree from DePaul, continues to help him succeed in his current role. Not only can he explain to a coach, scout or trainer what the data mean, but also he can translate the team’s needs into terms the software engineers understand.
This fluency in both worlds started when Rohlf was an undergraduate at Duke University. He worked as a manager on the basketball team for four years under legendary coach Mike Krzyzewski. When Rohlf accepted the position, he was an engineering major, but during sophomore year, he switched to computer science. His interest and knowledge in technology soon proved to be an asset for the team.
“At the time, we were transitioning from recording games on VHS tapes to doing everything digitally using video-editing software,” Rohlf says. “As I advanced in my degree, I also ended up helping with some programming and scripts to calculate our scouting statistics more efficiently.”
Rohlf’s passion for basketball and statistics continued when he enrolled at DePaul a few years later. DePaul appealed to him in part because the program’s flexibility allowed Rohlf to continue teaching computer classes and coaching high school girls’ basketball at Nazareth Academy in La Grange Park, Ill. For his thesis, Rohlf ran algorithms on footage from the academy’s games, writing code to identify the basketball in the video.
It was a logical extension of his CDM courses in computer vision, which Rohlf describes as “teaching the computer to see.” He found a mentor for this type of work in Professor Daniela Stan Raicu. “I owe a tremendous debt of gratitude to her,” he says, noting that he took all of her courses related to computer vision. “I learned so much, and I loved every minute of it.”
Although he’s no longer Raicu’s student, Rohlf has returned to DePaul as a guest lecturer in her data-mining courses. “I’m always impressed with the caliber of the students and the quality of their questions,” he says. Rohlf jokes that he doesn’t have to be particularly interesting himself because the subject matter tends to catch people’s attention. “Even if you’re not a sports fan, it’s easy to engage with this topic,” he says. “As someone who loves sports in every form, I’m lucky to have a career where I get to think about the logic and mathematics side of it every day.”
Shortly before this issue went to press, Rohlf accepted a new position as senior director for basketball technology and analytics at the NBA.