The FiveThirtyEight film “The Cowboys and the Indian,” which debuted last week, tells the story of A. Salam Qureishi, who brought computerized player analytics to the Dallas Cowboys in the 1960s. At the time, few other pro sports front offices used advanced statistics to make decisions about player acquisition and game management.
By 2001, that wasn’t the case in baseball, but the NFL hadn’t progressed much since Qureishi’s days. That year, Paraag Marathe joined the San Francisco 49ers as a one-man team with goals similar to Qureishi’s: improve player acquisition.
Marathe, like Qureishi, is of Indian descent. Unlike Qureishi, he knew a lot about football before working in the NFL. Marathe grew up in the Bay Area town of Saratoga, California, as a big fan of the 49ers and other Bay Area teams. He worked for the 49ers first as a consultant, on a three-month stint from the consulting firm Bain and Co. Then San Francisco executive Bill Walsh offered him a full-time job. “It was a no-brainer for me,” Marathe said in a telephone interview last week.
Today, Marathe, 37, is one of the elder statesmen of NFL analytics. He oversees it for the 49ers as team president. He sees more of his competitors using similar tools, looking for every competitive advantage they can find. (We spoke last week amid a late-season slump that has eliminated the 49ers from the playoffs after three straight conference-championship-game appearances.) In the following transcript of our interview — lightly edited for brevity and clarity — Marathe explains why it remains easy to hire talented analysts, why communication is more important than statistical rigor and why plenty of good work is still being done outside the league.
San Francisco 49ers President Paraag Marathe speaks at the annual Bay Area college football media day at Levi’s Stadium on July 30 in Santa Clara, Calif.
Alex Washburn / AP
Carl Bialik: Did you know the history of analytics in the NFL before you joined the league?
Paraag Marathe: I definitely did. When I came into the league in 2001, analytics was certainly more prevalent in baseball. It was just starting to become prevalent in basketball. The NFL was sort of the latest adopter. You see it a lot more now. Unlike baseball, where it’s all around player evaluation, the NFL is more complicated. It’s much more of a team sport, with much more covariance between positions. Is a running back’s success due to his ability to break away, or his line’s ability to run-block, or his quarterback’s ability to pass, which makes the run easier?
But the NFL also has two other areas where analytics plays a big role. The first is game management: How you manage the clock, when to go on fourth down, the run/pass play selection, those sorts of things. The second is the salary cap. With the advent of the salary cap in 1994, and where I made my mark with the 49ers and the NFL, is managing the salary cap much more analytically, similar to how a portfolio manager would manage a stock portfolio, managing risk.
CB: I’ve read that you’ve applied analytics to fans. How does that work and what have you learned about what they want?
PM: Oh yeah, yeah, absolutely. And it’s not just what they want, as in, what are their desires when they come to stadium, in terms of what they want to consume — content, beverages, coming closer to the game — but also the lifetime value of a fan. When you capture a fan’s loyalty, someone who becomes a fan at an early age, they will stay there. There’s a lot of loyalty. For professional teams 20 years ago, maybe as recently as 15 years ago, the vice president of marketing was all around what’s the cheerleader uniform and what’s the rallying cry for team. Now it’s all around what’s the content for the website, what’s the lifetime value of a fan, and so on. It’s much, much more analytical.
CB: How did you first connect with Bill Walsh?
PM: I was working at a consulting company, Bain and Co., on a bunch of sports-related projects. Bill Walsh and Terry Donahue were looking at drafts — not the players themselves, but draft slots. Is there a better algorithm, a better way to do the draft chart? It was a three-month project. I was the junior guy on the team. We sort of hit it off, and they asked me to come on board full-time.
CB: How often, while you’ve been on the job, have people asked you if you’ve played football? How did you answer?
PM: They either asked me, or they just assumed I didn’t. If they did ask me, I didn’t play college football, and I barely played high school football. I played baseball growing up, mostly. If you’re asking, did I feel like an outsider from the beginning when I started, I certainly did, but times have changed, and you earn respect with the work you do.
CB: What was the status of analytics at the 49ers when you arrived?
PM: It was a one-man show. To be fair, I never really did that much. It wasn’t so much on evaluating player talent on the field. It was a lot on the salary cap and how to be more efficient on managing the cap.
CB: How about now?
PM: We’ve got four or five folks, whether helping scouts better evaluate players, helping coaches, as well as the salary cap.
CB: Has the whole organization bought into analytics?
PM: Yeah, I’d like to think so. It’s definitely more accepted around the league. The Ravens just hired a head of their analytics department in 2012. You see it all across the board now. Clubs are trying to look for any competitive advantage they can. It’s not just, spend $1 more than the next team, it’s, what’s every competitive advantage you can squeeze out of this product?
CB: Is it tough to find good people, with so many teams hiring?
PM: No, I wouldn’t say that. There are so many good analysts across traditional industries, and sports is still such a sexy field, that there is no shortage of good talent. There are a lot of people who want to work in sports. Just go to the MIT Sloan Sports Analytics Conference every year.
CB: How important is it to be able to communicate the findings to people who aren’t technical?
PM: At the end of the day, it boils down to this. The information is only as good as it is to the person receiving it. I’ll take a C+ piece of analysis communicated perfectly over an A+ piece of analysis that’s not communicated well. Only a small portion of the work is the analytics itself. The rest is putting it in a practical format so the salary-cap person and the coach can appreciate it and use it. Instead of trying to go overboard with analytics, focus on the practical: Focus on the things that have the highest impact on your organization.
CB: Because of the importance of retaining a competitive advantage, do you generally not disclose specifically what you’re looking at, and what you’re finding?
PM: Generally speaking, we don’t really talk about a lot of those things. But it’s not just analytics. In nutrition, sleep studies, and psychological aspects, people are looking for advantages every place they can.
CB: Can you detect the spread of analytics in the league from how hard it is to get certain players in the draft, or from tactics of opponents?
PM: It’s mostly through conversations. I’ve been in the league now 14 years, and just having conversations with people in every level, I’m starting to see changes. Not starting to see — there’s been a lot of changes.
CB: Is analytics work being done within teams better than the work done outside it?
PM: Yes and no. That’s a tough question to answer. There are only 32 teams and there are seven billion people. There’s a lot of stuff that’s not even published that’s probably really good. The difference between what’s happening with teams and what’s happening in the ether, is what’s out there is pretty theoretical, whereas what teams do has to be pretty actionable. The most actionable things are being done in clubs, but I’m sure there are some excellent things being done out there.
CB: Which analytics publications do you read?
PM: Football Outsiders, Pro Football Focus, different blogs, even you guys do some awesome stuff. All the stuff that’s done is really good.
CB: You said in 2005 that even at 45 or 50, you’ll never be a football guy. Do you still feel that way?
PM: That was almost 10 years ago. What I meant is, not having played or coached the game, it’s just different. I don’t understand the nuances of the Xs and Os, nor do I try to, in terms of schemes and things like that. There’s no point in me trying really hard to be average at something. It’s important to focus on the things I know I can do well, like manage the salary cap. I won’t be a coach or GM, nor do I aspire to be.
CB: Does the NFL support analytics sufficiently? For instance, you’ve criticized the rule barring computers from the coaches’ booth.
PM: They’re still getting better. There’s the rule against laptops, even calculators. It’s difficult for an offensive coordinator to even capture simple things like average yards per play on a drive, or how successful a certain play has been, in terms of even crunching it in Excel. They have to do it on a notepad. Things like that are frustrating. I wish they would react a little faster to technology changes. They’re getting there. Now they’re allowing tablets on the sidelines, so you don’t have to have the binder full of photos of plays.