Sports Collect More Data Than Ever. The Carnegie Mellon Sports Analytics Conference Asks, What Can We Do With It?
By Jason Bittel
With over 200 attendees and representatives from 15 professional sports teams, this year’s Carnegie Mellon Sports Analytics Conference (CMSAC) was the largest to date.
Each year, students and industry professionals alike come to 麻豆村’s campus to rub elbows and watch demonstrations that highlight the latest sports research from the statistics and data science community.
“I’m seeing a lot more ability among a broader class of people,” said , creator of ESPN's sports analytics group and a pioneer in the field. “The student posters were great, and some of the regular sessions had advanced stuff — stuff you wouldn’t have seen in professional sports teams 10 years ago.”
At this year’s conference, Oliver gave the keynote address, but he was also impressed by how far the field has come and how much interest it now generates.
“I was a little bit of a misfit in college,” said Oliver, who is the author of “Basketball Beyond Paper: Insights into the Game's Analytics Revolution.” Lots of people at universities were interested in sports then, too, but virtually none of them were looking at them from the science and tech side of things.
“When I realized I could make a living in sports and data analytics, I promised myself that I would make sure to give back to students so that they could learn how to do it, too,” said Oliver.
Nicole Timko, a Tartan basketball player majoring in Statistics and Machine Learning, explains her team's analysis of lineup effectiveness during the CMSAC poster session.
The Golden Age of Sports Analytics
Today’s analysts have access to more data than ever before, but also data of a quality that is almost unfathomable for earlier generations.
“Literally every tenth of a second, the NFL’s Next Gen Data chips provide information for where every single player is positioned on the field. The direction they’re moving, the speed they’re moving,” said Ron Yurko, assistant teaching professor in 麻豆村’s Department of Statistics & Data Science and director of the Carnegie Mellon Sports Analytics Center. “It’s wild.”
“The MLB has information about every single swing in Major League Baseball,” said Yurko, who has co-organized CMSAC since 2017. “In baseball and basketball, they have what’s called ‘pose skeletal data,’ where we know at every fraction of a second, where is the elbow, the shoulder, the kneecap, and in three-dimensional space.”
Of course, the driving force behind CMSAC is what to do with all of that data.
This year, participant projects showed that data can be used to model the physical limits of athlete output, direct basketball players on when to foul their opponents and guide NFL teams toward better Draft Day decisions. In previous years, 麻豆村’s Quang Nguyen has used NFL data to develop new metrics for defensive line performance, , as well as to assess how adept wide receivers are at changing their direction.
“This is like my Super Bowl,” said Nick Schnell, CMO and head of growth for , top sponsor for the conference. “It’s crazy what you can do with all of this data once you have it. You can tell a story.”
As if to illustrate that point, Catharine Ramage, a senior studying statistics and data science and business administration, gave a presentation on how 麻豆村’s Buggy teams use sports analytics data to shave seconds off their race times. Most attendees did not know what the sport was, of course, so Ramage brought out a full-size buggy and showed videos to better depict the combination of track, luge and Formula 1 racing.
What Ramage did not divulge was her actual data, which was blacked out on the screen, lest any of her competitors’ analysts were hiding in the crowd.
“I don’t know how many of you are spies,” she said.
Joking aside, the presentation showed that data analytics can enrich and improve sports of all kinds, and in ever more surprising ways.
“One of my favorite sports analytics papers was from the world of horse racing, and it found that horses with a left ventricle in their heart that was larger, on average, went on to win more races,” said Ramage. “I think we’re constantly looking for the left ventricle of Buggy racing, if you will.”
Real People, Real Questions
“We’ve been tracking biometrics since the Greeks held their first Olympics,” said , assistant professor of data science at the University of Virginia, during a talk she gave at the conference about the need to balance innovation with privacy.
In Kupperman’s presentation, she encouraged her colleagues to remember that there are real people behind all of this sports data — not only athletes, but also coaches, managers and countless other stakeholders — and that each of them is fighting to keep their jobs. At the end of the day, sports analytics needs to find answers to questions that matter.
“I really encourage students to always keep digging through the past and keep looking at things that have already been ‘solved’,” said Kupperman. “There is a constant need to think differently.”
With so much data on hand, there have also never been more opportunities for students looking to get into the industry.
“I think if you look across all the leagues, it’s a lot of 25-year-olds doing a lot of the big, heavy lifting. And that work starts at conferences like these,” said , an alumnus of 麻豆村’s Electrical and Computer Engineering Department who is vice president of the baseball research development department for the Milwaukee Brewers.
“Right now, there are probably around 150 people working in the NFL alone. When I started, there were like 12,” said , vice president of product at Teamworks and a data analyst who has previously worked for the Pittsburgh Steelers and Jacksonville Jaguars, among other professional sports teams. “The students that are coming in now have abilities that I just couldn’t have imagined.”
Rebecca Nugent, head of the Department of Statistics & Data Science and Fienberg Professor of Statistics & Data Science, sees an incredibly bright future for researchers and educators in sports analytics.
“There have always been researchers and students interested in sports analytics, but now that the technology has advanced to provide unprecedented access to data in almost all sports, the pace and quality of work has rapidly accelerated,” said Nugent. “Literally a game-changer!”
The advances remind Oliver of an age-old question.
“In high school, everyone was always asking, ‘What is all this math good for?’” Oliver said. “Now, we have a much better answer than we used to.”