When asked how much analytics informs the New England Patriots‘ decision-making in key situations, Bill Belichick answered simply, “less than zero.” This prompted several snide responses from the Twitter machine, such as:
Obviously “less than zero” is a mathematical impossibility, and it’s hard to imagine that the famously thorough Belichick is utterly unfamiliar with the data on this subject. The New England coach has been known to be coy with the media before. But it’s still worth asking: how much analytics do the Patriots use?
Belichick has been pretty consistent through the years in pooh-poohing the value of analytics and their role in the Patriots organization. On the one hand, he’s been known to protect some of New England’s practices with a secrecy worthy of the intelligence industry. But we don’t have examples of him use analytics to justify a decision, as Baltimore Ravens head coach John Harbaugh did the other day, for instance. Nor have any of his past coaches or other employees shed much light on the role of numbers in the Patriots’ decisions.
When people talk about analytics in the New England organization, they often point to one man: Director of Football Research Ernie Adams. Between his title, his bespectacled appearance, and his history as a bond trader on Wall Street, it’s no wonder people assume Adams is a statistical genius. Adams and Belichick open themselves up to speculation with how closely they guard what Adams does on a day-to-day basis. When the duo worked for the Cleveland Browns, team owner Art Modell once promised,“I’ll pay anyone here $10,000 if they can tell me what Ernie Adams does.”
That said, we do have some information on Adams’ role, and what we know doesn’t necessarily have a heavy statistical component. In the NFL Films documentary Do Your Job about the 2014 season, Adams revealed that he runs the scout team, watching film to understand what New England opponents are likely to run and coaching the practice squad to run those plays. Julian Edelman remarked, “I ask him questions on defenses, because he legitimately has [photographic] memory, so I’ll be like, ‘Ernie, when they’re in the red area and they’re down four points, what do we expect this team to do?’ And he’ll rattle off: this, this, this.”
Charting and cataloging an opponent’s plays on film is an analytical process, but it’s not “analytics” in the sense that most people think of it. It’s really old school advanced scouting, not too dissimilar from a role in Pro Personnel (and Adams did serve as Director of Pro Personnel for the New York Giants in the 1980s). Randy Dean, who played for Northwestern and the Giants when Adams was on staff, described him as “one of the premier resources for anything in film.”
Adams does stay versed on analytics—he once called Rutgers statistician Harold Sackrowitz to evaluate New England’s two-point conversion chart—but he appears to stay versed on everything, reading nearly every football book that comes out. He knows playbooks back and forth, serves as the gameday rulebook expert to help Belichick know when to challenge, has a tremendous knowledge of NFL history, and yes, dabbles in numbers. But from the information publicly available—and there isn’t a lot—it looks like analytics is more of a minor element of his role than a focus.
The analytics community would certainly like to claim Belichick as one of their own, given New England’s tremendous success. The Sloan Sports Analytics Conference at MIT once granted him a lifetime achievement award. There are certainly aspects of New England’s approach that align with analytical thought. Belichick famously went for it—and failed—on a key 4th-and-2 against the Indianapolis Colts in 2009. As de facto general manager, Belichick frequently trades down, an efficiency that economists Cade Massey and Richard Thaler documented in their seminal paper, “The Loser’s Curse.”
But there are many facets of Belichick’s job, as both head coach and head of personnel, and he doesn’t always align with analytics. He used a first-round pick on running back Sony Michel in 2018, for instance, in defiance of emerging consensus that “running backs don’t matter.” The Patriots have gotten more conservative on fourth downs in recent seasons, ranking 19th in 4th-and-2 or less attempts in 2018. And even where they follow the guidelines of analytics, it’s not clear that they’re being influenced by them. They were trading back even before Massey and Thaler’s paper came out in 2005, suggesting they came to their own conclusions about the merits of trading back, rather than following analytics research.
The Patriots rarely do things that are analytically stupid, but that doesn’t mean they’re on the cutting edge of numerical analysis, either. My read is that Belichick, Adams, and the rest of the Patriots machine are very smart people who have put a lot of time and thought into what makes for winning football. It should be no surprise that the result of that effort aligns with analytics in many ways; that’s to the credit of analytics, not the detriment, even if New England rarely explicitly uses analytical findings.
The root of the Patriots’ success appears to be based on film study. Belichick’s father Steve wrote Football Scouting Methods, a revolutionary text on how to watch football with an eye toward how to exploit weaknesses and combat strengths. Adams and Belichick initially bonded at prep school because Adams had read the elder Belichick’s book. Continuing that legacy—painstaking tape scouting, not analytics—is the true bedrock of New England’s success.