Analytics in sports: There’s a stat for that

For which batters should the New York Mets put on the infield shift?

Paul Sabin has a stat for that.

Why is Lebron James still playing near his peak at an age when most other NBA greats have already retired?

Sabin has a stat for that, too.

Whereas most sports fans see athletes, X’s and O’s when watching a game, Sabin – an ESPN sports data scientist who received his Ph.D. in statistics from Virginia Tech in 2019 – is all about the numbers. He will be sharing his analytical insights at 1 p.m. Wednesday, Feb. 16 as the keynote speaker in “Nerds and Jocks Unite – What's Next in the Sports Analytics Revolution?” part of the Northern Virginia Technology Council Data Science Speaker Series in partnership with Virginia Tech. Registration is required for the online event.

“In some sense, sports analytics has become a loaded term,” said Sabin, “but it’s really the idea of deriving insights from data and using data to make better decisions when it comes to scouting, drafting, in-game decisions like going for it on the fourth down or trying for a 2-point conversion, and so forth.”

Analytics has become so sophisticated that in today’s NBA game, player-tracking technology allows teams to gather 25 data points per second for each player on the basketball court.

“What’s next is not just using player tracking data for game strategy, but advancing frontiers in sports science and injury prevention,” Sabin said. “I think we’ll see huge strides in that, and players like Tom Brady playing until [the age of ] 44, or Serena Williams, or Lebron James playing not too far from his peak at an age when Michael Jordan was wrapping up his career.”

At ESPN, Sabin is one of just four data scientists. “Our motto is to serve sports fans anytime, anywhere, so we’re always trying to create good content and storytelling,” he said. “It’s our job to educate fans, to provide content and context for games.”

While not quite to the level of ESPN’s Monday Night Football broadcast with Peyton and Eli Manning, Sabin has produced and hosted alternative ESPN broadcasts that focus more on data, statistics, even gambling odds, for a few NFL and MLB games – a trend likely to continue and grow

“There’s no other sports media network that has a team like us that is as ingrained in production and broadcasts,” he said.

Of course, much of the nation’s fascination with analytics can be traced to the 2011 movie “Moneyball” starring Brad Pitt as the Oakland A’s general manager who was among the first to adopt a data first strategy over eye-based scouting. 

At the time of the movie’s premier, Sabin, a native of Montgomery County, Maryland, was an undergraduate statistics student at Brigham Young University (BYU). He and his group of friends “all chose sports projects when we had a choice to do research,” he said. “I don’t think I would have been able to have gotten a Ph.D. if I couldn’t have worked with sports data. The subject became a lot more tangible when I could apply it to sports.”

Today, those friends have all found careers in sports: one is director of analytics for the Sacramento Kings; another worked for the Cleveland Cavaliers and Detroit Pistons.

Ready for graduate school in 2014, Sabin desired to move back east, and a BYU professor connected him with David Higdon, head of Virginia Tech’s Department of Statistics, part of the College of Science.

While Higdon’s research focused on more traditional applications such as a company’s customer orders and supply chain issues, the professor has become a huge cheerleader of Sabin’s career.

“When I started in statistics, people just didn’t collect data in this way [for sports],” said Higdon. “Sports analytics was only just starting then, and no one was having a career.”

Higdon is working to ensure Virginia Tech keeps up with this demand. He’s approved one-time classes like “Baseball Statistics” as well as independent study projects for stats students who helped the Hokie softball team design better batting orders. 

“So far it’s been a one-off thing, and the next step for us in our department is how to make this more permanent,” Higdon said. “I think we’re going to try to crack that nut over this next year. Certainly, we want to make hires with people who have expertise in sports analytics. Of course, competing for them right now is difficult because there aren’t a lot of them out there, which is why it’s great that we’re able to bring in Paul. [The College of Science] has a degree called Computational Modeling and Data Analytics, so I’m going to talk to the head of that program and see how do we make this more permanent with Paul.”

Through Higdon, Sabin was invited by Virginia Tech statistics professor Thomas Woteki, who is director of the Academy of Data Science, to speak to the Northern Virginia Technology Council. 

“I got my Ph.D. about 100 years ago,” Woteki laughed. “Nobody ever envisioned a statistician having a career in sports. If I had foreseen that I would probably be a very wealthy man.”

Following Sabin’s keynote, there will be a panel discussion featuring fellow Hokies Zac Robertson ‘13, who is director of analytics and research for the Miami Heat; Ken Pomeroy '95, creator and college basketball analyst of KenPom.com; and Tanya Coutray ‘92, ‘94, analytics strategy leader at Amazon Web Services.

“Analytics affects every part of sports business from evaluating strategy during a game to the back end with social media engagement and the return on investment from your business partners,” Woteki said. “Everyone understands sports as a fan or spectator, but that same impact is pervasive in almost every line of business. We’re all providers of information these days whether we intend to be or not because of the devices we carry around. That’s how Facebook makes money by providing this information to advertisers. It’s the way business is done.”

Woteki is especially excited about the timing of “Nerds and Jocks Unite” because of its proximity to two of the biggest sports events of the year: the NFL Super Bowl and March Madness.

Still, what makes sports exciting isn’t the increasing reliance on data and statistics, but the unquantifiable human element.