Most college students who get invited to the National Football League’s Scouting Combine — where top prospects showcase their athletic talents ahead of the NFL draft — have blazing 40-yard dash times, brand deals and a lifetime of football behind them. Eric Steinberg had none of that when he went to the 2025 Combine.
The Emory College senior has never played a down of organized football. Emory doesn’t even have a football field or team. Nonetheless, he and his teammates were named finalists in the NFL Big Data Bowl, the league’s premier data science competition held in conjunction with the Combine each February.
“I joked with my friends that my project would get me drafted,” says Steinberg, majoring in quantitative theory and methods (QTM). “I never thought I’d actually be boarding a flight to Indiana.”
The Big Data Bowl is a Super Bowl of sorts for the sports analytics community. Hosted annually by the NFL, it invites participants to tackle a themed football data challenge. Anyone — including students, coaches and professional analysts and data scientists — can enter the competition as an individual or team.
This year’s theme focused on using pre-snap data to generate actionable predictions about post-snap plays. Participants typically spend several months developing and submitting their analyses.
Only five teams from more than 7,000 submissions were invited to present their work at this year’s NFL Scouting Combine in Indianapolis. Finalists received $12,500 and presented their findings to a packed audience of team analysts, general managers and league officials.
Steinberg’s team was one of the five.
He and his teammates — Lindsay Fleishman of the University of Georgia, Lucca Ferraz of Rice University and Daniel Barcelona Soriano of the University of California, Davis — examined how a tight end’s alignment and movement could signal whether they would block or run a route. Their project, TEndencIQ, used NFL tracking data to train a model that offered practical insights for defensive coaches.
“We built a gradient-boosted model to predict tight end behavior using only what’s visible before the snap,” says Steinberg. “It relied on the same positional and motion cues a defender might use in real time.”
Clifford Carrubba, chair of Emory’s Department of Quantitative Theory and Methods, praised Steinberg’s achievement.
“It is really exciting to have students be able to go out and compete in national events and be successful and show what they can do on real projects in the real world beyond Emory and the classroom,” Carrubba says.
Steinberg is equally proud of his Emory and QTM education. He says his performance in the NFL competition demonstrates Emory can excel in unexpected arenas.
“The QTM department really gave me the tools to take on something like this,” he says. “What I learned in class directly helped me take on the Big Data Bowl.”
Running an unconventional route
Steinberg’s passion for football didn’t follow a conventional route. Growing up in San Francisco, home to one of the NFL’s most storied franchises, he was at best a casual observer.
But in college, he became unexpectedly obsessed with learning the game. Not just the highlights, but the detailed layers of strategy behind every play.
“Honestly, it started with me going down a YouTube rabbit hole of 2000s NFL lore breakdowns,” he says. “A year or two later, I’m up at 1 a.m. watching (popular football analyst) Brett Kollmann break down split-zone blocking from a two-tight end set against Cover 3.”
That kind of detail, he says, was what pulled him in. “I realized almost everything in sports comes down to a mix of probabilities, outcomes and measurable patterns,” Steinberg says. “It’s all numbers at the end of the day, and I love numbers.”
While the Big Data Bowl is open to anyone, Steinberg applied to its selective Mentorship Program, which pairs participants with data analysts from NFL teams and the league office. The program offers guidance and feedback throughout the project, along with insight into how analytics are used at the professional level.
The mentorship proved invaluable — and led Steinberg to connect with his eventual teammates by scouting out fellow participants whose skill sets complemented his own. Despite working across three time zones and through two study abroad programs, the group met regularly throughout the project. That commitment, Steinberg says, was key to developing a focused and well-executed analysis.
“He and his teammates were always asking questions, taking the initiative to set up meetings and messaging each other,” says Ally Blake, a data scientist for the NFL and one of Steinberg’s mentors. “They used each of their strengths to bring the best version of the project to the competition.”
The team members met in person for the first time in Indianapolis at the Scouting Combine in February. In addition to presenting their findings, Steinberg and his compatriots hobnobbed with head coaches, hall-of-famers and potential draft picks.
Nate Silverblatt, a QTM visiting instructor and teacher of a sports analytics class in which Steinberg is a student, says he isn’t surprised Steinberg finished well in the Bowl.
“He’s been a really impressive presence in the class based on his questions during lectures,” Silverblatt says. “He asks thoughtful, specific questions that show he has a strong grasp of the material and a real curiosity about how these concepts lay out in professional sports settings.”
Although a student team from NYU was declared the overall winner of the competition, Steinberg is proud of what he accomplished. With graduation around the corner, he’s ready to turn his passion for sports analytics into a career.
“Getting paid to watch and analyze football?” he says. “Sounds like a dream job to me.”