Kelley began to wonder how one might replicate that fighting spirit in the classroom. He soon imagined a computer application that would use students’ day-to-day results to match them up with comparably skilled contestants in head-to-head academic competition—in everything from classroom pickup games to bleacher-filling, live-broadcast amphitheater tournaments.

In September 2012, Kelley called Steve Dunbar, director of the American Mathematics Competitions, or AMC, an elite program sponsored by the nonprofit Mathematical Association of America, with the idea of a competition based on AMC problems. The competition, founded in 1950, enrolls about 400,000 students, but it still uses pencil and paper and can take weeks to score. Dunbar had actually been searching for a way to bring AMC into the 21st century, and as soon as Kelley described his vision, Dunbar knew that this was what he’d been looking for. In two months, Kelley had a prototype. In five months, he and Dunbar had selected 16 high schools to field-test the software. By February 2013, the first trials began.

To those who blanch at Coleman’s vision of making academics a spectator sport, Kelley says the focus of Arete, as with the rowers’ fitness test, is on helping students achieve “personal best” milestones, a strategy that most schools rarely use. “Once kids see they’re getting better, it just perpetuates improvement,” he said. When I met Kelley, he was working on a tool that would allow spectators to view Arete matches live online. He said he hoped that would “bring enough glory to the math department, or enough glory to the math students, that everybody else says, ‘I’d like to try this, too.’”

In September 2013, after the pilot testing, 468 schools showed up for the beginning of the first Arete fall competition, and Kelley soon had 10,000 kids on the platform weekly. By November, he had arranged the highest-scoring 384 teams into six 64-team brackets. Two weeks before Christmas, the Final Four teams in each of the six divisions fought for their division’s title. In the highest division, TJ actually made it to the Final Four, but was outscored by the Academy for the Advancement of Science and Technology in Hackensack, New Jersey. Hackensack lost in the finals to San Jose’s Harker School. The following September, nearly 600 schools and 15,000 students showed up to play, paying a modest fee of between $120 and $195 per school, for access to the platform for the entire season.

In October 2015, Kelley received a grant of nearly $150,000 from the National Science Foundation to further develop his project. Soon, students will be able to arrange matches on their own. What’s more, hundreds of thousands of 6th- to 12th-grade students will be able to compete simultaneously in a challenge that decides a national and eventually a worldwide champion. After implementing that feature, Kelley wants to expand the same tournament model to other school subjects and grades.

AMC’s Dunbar hopes that Arete will ultimately bring high-level math to a larger audience—the traditional AMC is focused on just the top 10 percent of students in the top 10 percent of schools. “One of the things that I do, one of the things that gets me up and here into the office every day, is that I want to get more good math in front of more kids, more often, in as many ways as I possibly can,” he said. International competitions pitting our best students against the best in the world could be thrilling. “If you look at the top level of competition, the United States is as strong as any other country in the world,” he said. “It would be good and it would be competitive. It would be exciting.”