Chicago Booth Magazine Changing the Way the Game is Played

Javaid, with his background in advanced analytics, turned to hard data for the answer. He scoured sales figures from past seasons. This led him to institute one minor but crucial change to the design of the concession stands: he redirected the queues so that they snake around stanchions, like an airport security line, at 75 points around the stadium. By the midpoint of the season, the move had increased sales by one extra transaction per line per minute since—amounting to an extra $20,000 per game, or $200,000 over the course of each season’s home games, including the preseason.

It might sound like loose change—beer money—for a franchise that collected $427 million in revenue in 2014, the first season in Levi’s Stadium, in the heart of the fans and fortunes of Silicon Valley. Javaid, however, understands that the small improvements add up, and his is a top-five revenue-producing team. “In the NFL, it’s not tremendously difficult to make money,” he said. “You can do B+ work by not really doing much, and you can still be profitable. But in order to get to the next level, you need to find those little wins.”

This season, he and his five-person business ops team found those victories by introducing not only the stanchions but also more beverage-only lines, and by repositioning the retail kiosks, increasing the efficiency of mobile ordering, and pushing the stadium opening time back 30 minutes. That last modification alone, brought about after Javaid’s team noticed that most fans lingered in the parking lots pregame anyway, created what he describes as a “six-figure impact” in what the 49ers saved on staffing.

Though no one’s going to argue that the redirection of a concessions line makes for high Hollywood drama, Javaid is essentially applying Moneyball-style stats to the business of football, with hard data driving industry decisions. “Without data,” he said, “I think it’s really difficult to find those $20k, $40k, $50k opportunities that, when multiplied over the number of events, generate millions of dollars in aggregate to the bottom line. That’s giving our ownership a little more financial flexibility so we can put the best product on the field.”

We’ve all seen the Brad Pitt movie at this point. We know that baseball managers consult computer printouts with statistics and graphs before every decision, and that basketball players can study the pick-and-roll defensive tendencies of every opponent they’ve ever faced at any moment in their careers.

But football hasn’t changed much in the nearly 50 years since the first Super Bowl (1967), and the NFL has seemed lukewarm about embracing analytics. This is, after all, a sport that measures its first downs with the analog precision of a chain-link cord extended between two people on the sidelines. There is not even a clear way to measure the productivity of its most important player. Would that be determined by QBR or passer rating?

Compared with the other major North American sports, football appears to be in its analytics infancy. That makes the sport an appealing enterprise for Booth professors John Huizinga, Kevin M. Murphy, PhD ’06, and Tobias J. Moskowitz—the three men who created Booth’s pioneering Sports Analytics course, now in its third year—and alumni such as Javaid and Ken Kovash, ’06.

Huizinga, who’s Walter David “Bud” Fackler Distinguished Service Professor of Economics, and Murphy, who’s George J. Stigler Distinguished Service Professor of Economics, are offering the course this spring. This time they’re co-teaching it without colleague Moskowitz, Fama Family Professor of Finance, who is on leave.

For the NFL, it turns out, the winning analytics equation is a lot longer than once thought.

“In baseball, every franchise has a team of people devoted to analytics efforts,” said Kovash, VP of player personnel for the Cleveland Browns. “In the NFL, that’s far from the case.” The Browns, in fact, typically don’t make Kovash available for interviews and don’t even use the word “analytics” to describe his work with data. “We talk about it in terms of being progressive,” he said. “Trying to find competitive advantages and trying to have more and better information than the 31 other teams. The opportunity to really contribute to something at its early stages is really, really attractive and exciting.”

The average NFL team is currently worth more than $2 billion. That money, which has approximately doubled since 2011, could be considered an investment in the future of the team, whether it’s used for signing players or coaches, enhancing facilities or equipment, food or trainers, toward future performance. Competitive balance has given the league more parity than ever. Television ratings for week one of the 2015–16 season hit an all-time high.

On a fan level, the rabid success of daily fantasy football (see “Data and Daily Fantasy Football”) confirms an appetite for inventive DIY data crunching even at home and on a 24-hour loop. At the pro level, standing out requires creativity, and more and more it’s the analytics experts in the organization making that happen.

In 2012, Javaid was a member of a Booth student team that beat out 22 other business schools to win the NFL’s first case-study competition, run through the MIT Sloan Sports Analytics Conference. The challenge was to determine how franchises should go about conceiving the “right size” of a new-build stadium, including a methodology for allocating seating arrangements and suites and creating a pricing structure. Javaid’s team fulfilled the brief and then went further, examining why some existing NFL stadiums were not ideally designed to maximize revenue. They reasoned that the Superdome in New Orleans, for instance, would be optimized with 11,000 fewer seats at higher average prices. And they suggested that the Philadelphia Eagles were undervaluing their average ticket prices by $26, inadvertently leaving money on the table.

Javaid said that the project helped him get in the door for an interview with the 49ers, who hired him in March 2013—just as they were designing their own new stadium. That same spring, Moskowitz, Murphy, and Huizinga launched Booth’s first Sports Analytics course. Moskowitz is an award-winning financial economist and a consultant for hedge fund AQR Capital Management. But he is perhaps best known among sports fans for Scorecasting, which he coauthored with sportswriter L. Jon Wertheim. The 2011 book defied conventional sports wisdom through quantitative reasoning. Among Moskowitz’s chief arguments: football teams should not punt the ball on fourth down nearly as often as they do.

His explanation relied on statistics that either nobody had bothered to look for, or nobody had bothered to connect and publish. It seemed reminiscent of the early work of Bill James, the onetime security guard at a Kansas pork-and-beans factory who began publishing annual baseball abstracts in the 1970s, which finally caught the attention of the Oakland A’s general manager, Billy Beane, in the late 1990s, and whose story became the basis for Moneyball. More than a decade later, baseball games are now atomized behind the scenes beyond a point of recognition. Every pitch, at any level, can produce dozens of metrics for teams of stat gurus to ingest.

“Now every ball ever hit by any player anywhere, in the minors or majors, is tracked seven ways to Sunday,” said Girsch, today the team’s assistant general manager, with World Series rings from 2006 and 2011. “I think as an industry we’ve done a pretty good job squeezing the majority of what’s going to be squeezed out of that sort of analysis.”

Let’s just say that colored pencils aren’t necessary anymore. Statcast, the terabyte-per-game camera capture system set up by MLB in each of its stadiums last season, has blown all previous statistics-generating operations out of the water. Every MLB franchise has a team of five or six people like Girsch hunting down the numbers behind what’s left. “Whatever you call old school today is unrecognizable from what old school was 10 years ago,” he said. “It’s changing fast.”

Football too moves forward, albeit at its own pace. A decade ago, Kovash, newly graduated from Booth, was determined to work in his favorite sport. But football teams weren’t interested. He essentially had to launch his own venture just to get the attention of the NFL—to demonstrate that analytics have a place in the game.

He began a project with University of Chicago’s Steven Levitt (coauthor of Freakonomics) that applied economic game theory to play calling in the NFL as a way to assess coaches’ decision making. He traced data back five years and found a trove of interesting patterns. “The whole goal of that was to get me a job with a team,” Kovash said. “That process took a lot longer than I ideally hoped.”

Despite Kovash’s initiative and enthusiasm, NFL teams still seemed resistant to hiring a guy with a background in finance, economics, and marketing, but not in scouting. Kovash then went to San Francisco to do analytics work for Mozilla, the software giant. Finally, in 2010, the Dallas Cowboys called. They wanted him to come build one of the league’s first analytics departments.

Kovash was elated. Then reality hit him.

“It was essentially like starting from scratch,” he said. Much of the first few months simply entailed convincing coaches and scouts that his role was necessary. His tasks were not especially intensive. The ways teams quantified the values of players or assessed game situations had not changed much in decades. Much of Kovash’s job entailed “sitting down with [then] head coach Wade Phillips, or spending a million hours with the scouts and just learning a lot and understanding, ‘OK, does information exist in the world that can help answer these questions?’” Often it does, but somebody like Kovash needed to know where to look. “You can’t gain the competitive advantages you are seeking unless you collaborate with the coaches and scouts that have the football expertise to help you find the questions and the answers.”