Strife, second-guessing and a polarizing figure: Inside the Eagles’ analytics conflict
When Nick Sirianni interviewed with the Eagles at Jeffrey Lurie’s mansion in West Palm Beach, Fla., most of the usuals were there: general manager Howie Roseman, president Don Smolenski, other key members of the front office, and, of course, the owner himself.
Each had a role in the exhaustive process of finding the team’s next head coach, with several from the group responsible for a specific line of questioning. The topics ran the gamut, from scheme to personnel, coaching staff to coaching philosophy, and more.
In the area of analytics, its representative was the only interrogator, however, who wasn’t actually employed by Lurie: his only son, Julian.
Five years earlier, Julian Lurie was among those in the room when Doug Pederson and other candidates interviewed with the Eagles. But he was present mostly as an observer. This time around, the 25-year old was given an active purpose in the realm in which he was most familiar and his father had long been a proponent.
The Eagles have long been ahead of the curve on analytics with an owner who has increasingly championed its use. Lurie may not have won his first Super Bowl without Pederson’s numbers-based aggressiveness. But as the team’s fortunes declined and the coach rebuked chief strategist Alec Halaby and his analytics department, the owner felt compelled to increase his presence, sources close to the situation said.
There were other significant issues that led to Pederson’s firing, but at the core was Lurie strengthening his grip on his team. The questions now are how involved will he remain with Sirianni as his second straight first-time head coach, and how much of it will revolve around analytics?
The Inquirer spent months talking to a dozen team sources, past and present, as well as other league sources familiar with the inner workings of the Eagles to cast a light on the importance of analytics in the organization and to Lurie, the internal conflicts over its usage, and how it may be used going forward.
The sources, in most cases, requested anonymity because they were unauthorized to speak publicly on the subject, feared retribution, or felt they couldn’t speak freely if named.
Pederson to Sirianni
When Julian Lurie presented to Sirianni and other prospective coaches a report based on analytics and asked for an assessment of the data, according to sources who were either present or had direct knowledge of the Eagles’ interview process, it wasn’t viewed as an extreme reach.
The Harvard alum had already completed a two-year rotational internship with the NFL from 2017-19 and his inclusion was just the latest step as the elder Lurie grooms his son for a senior position and eventually to succeed him as chairman and CEO.
“He’s paying his dues and learning the business,” an Eagles source said.
But that he would do so at the expense of Halaby, the Eagles’ vice president of football operations and strategy, was peculiar. Leadership did select a committee that wasn’t made up entirely of senior-level staffers as it wanted a diverse set of viewpoints.
Halaby, though, has long been a polarizing figure in the organization. He was Roseman’s hire and over his 14 years with the Eagles has become a trusted adviser to Jeffrey Lurie.
But on the coaching and scouting side, he has had his detractors. While there might be an element of “football guy” prejudice, the Eagles’ last two coaches — Pederson and Chip Kelly — both felt a need, when given the clout, to move Halaby out of the football wing of the NovaCare Complex.
Pederson had gotten so frustrated with Halaby that for nearly a year they didn’t communicate, multiple team sources said. The coach did embrace analytics and used the statistics in both game planning and game management. In 2017, his aggressive play-calling on fourth down and two-point conversations played a role in winning a championship.
But Pederson had come to rely more on his assistants, particularly Ryan Paganetti, for his numbers. This created a natural tension with the analytics department. Some coaches welcomed Halaby’s analysis and worked closely with him, but there were also many football-side staffers who deemed his department’s reports either esoteric or radical.
“You would just get a report dropped on your desk, and they never backed it up with why they felt that way,” one Eagles staffer said. “The results would be black and white without wiggle room. Sometimes they weren’t even entirely data-driven. They felt more like opinion pieces.”
Halaby, through a team spokesperson, declined to be interviewed for this story. Pederson, who is currently unemployed but planning on a return to the NFL next year, declined a request, as well.
Lurie has been credited by some with spearheading the Eagles’ use of analytics. He created an environment that allowed for his coaching and personnel departments to push boundaries, even beyond the parameters already set by the analytics community, if they were willing. Pederson initially benefited from this freedom.
“We are pretty obsessed with it,” Lurie said during the 2019 offseason, “and we’re always looking to figure out how we can be much better at it. ... We have to be the best we can be, and it’s not just about mining lots of data. We will collect a lot of data and it’s what you do with it. I think that’s the key.”
As the Eagles regressed following 2017, and the conflict between Pederson and Halaby lingered, Lurie tried to bridge the gap. But the coach came to view his boss’ espousal of analytics during their weekly Tuesday meetings as merely a means to second-guess him, sources close to the situation said.
“I think Doug was always trying to be accommodating and respectful,” one source said. “I think he did a fairly good job of smoothing things over with Jeffrey. But at some point it reached a breaking point where Jeffrey thought that wasn’t good enough.
“He doubled down and said to Doug, ‘You’re not using this information enough.’”
Lurie, too, declined to be interviewed for this story. While Halaby reported to Roseman, some Eagles staffers considered the analytics-preoccupied Lurie to have a closer bond with the 34-year-old. Still, when the owner conducted his search for Pederson’s replacement, his son was the front man for analytics, not Halaby.
Halaby did provide the analysis for prospective coaches to review, but Julian Lurie was considered a natural liaison because of his relationship with Halaby and his predilection toward analytics.
It’s unclear how Sirianni handled that part of the interview. When Lurie introduced his new coach in January, he never once said the word “analytics” during his 20-minute opening statement. He mostly focused on the 40-year-old’s love of football, emotional intelligence, and teaching ability.
Sirianni never mentioned analytics, either, and only recently was asked how much he planned on using it as a first-time head coach.
“I think it’s just a piece of the puzzle,” Sirianni said last month. He added: “I think a wise man avoids all extremes. You don’t want to be like, ‘I’m just putting all my eggs in this basket or just putting all my eggs in this basket.’ You use all the pieces available to you to help put your team in the best position.”
The NFL has been slow to adopt analytics, but there have been more teams, coaches, and general managers using advanced statistics in recent years. Sirianni has previously been open to using them, but as head coach several sources who worked with him said they expect him to lean heavily on traditional tactics.
“He will not be driven by analytics,” one of the sources said.
How that may sit with Lurie remains to be seen. There is the possibility Sirianni evolves or is more accepting of Halaby. Offensive-minded coaches tend to be more aggressive, which is one reason why Lurie has favored that side of the ball with his last four hires.
But even Pederson, who finished fourth in Football Outsiders’ rankings of the most aggressive coaches in the last 38 years, wasn’t utilizing analytics enough to Lurie’s liking. The issues between the two at the end were manifold. You don’t finish 4-11-1 because of just one thing.
Sirianni said that Lurie’s lone message thus far has been one of full-pledged support. Pederson similarly said the same publicly. And in some ways Lurie has been an exemplar boss. But in pushing Halaby’s conclusions, the owner crossed a threshold with his former coach.
“Instead of using it in a constructive way, he was using it in a critiquing way,” a source familiar with the Eagles’ inner workings said of Lurie. “If a coach perceives that an owner is using somebody else’s information to undermine, there’s [a] 100 percent chance you have a division in the building and it’s consequential.
“Jeffrey never really understood the consequences of embracing and using Alec’s information in that way. If he does the same to Nick over the next two-three years — and they’ll have a honeymoon period just like everybody does — it’s going to result in the same kind of friction.”
Rise of analytics
In the mid-1990s, the Eagles created the first known analytics department when then-team president Joe Banner hired three MIT students as contractors. Lurie supported the initiative and over time advocated for its implementation, believing the smartest teams use data in concert with traditional scouting.
The Eagles looked for non-obvious ways to create a competitive advantage in a league that was designed to negate competitive advantage. Analytics was just one approach, although the initial studies were just advanced models of the tendency reports low-level assistants had been quantifying for years.
The department grew in sophistication and over time as Lurie’s interest grew. While his understanding of conventional game-planning increased the more he was around football, analytics was an easier entryway into the game because it was numbers-based.
“Jeffrey gets a lot of his data from the analytics department,” a source who has worked closely with Lurie said, “and it influences a lot of his opinions on football.”
The analytics community was way ahead of most in disputing the establish-the-run theory and this deduction, for example, has long fostered Lurie’s belief that offenses should favor throwing early in games and on first or second down to get ahead.
“What’s the right way to say this?” he said in 2018 when asked about the establish-the-run theory. “It’s just not a truthful way of reporting based on all the information we now have. OK? That’s sort of a nice way to say it.”
When it came to fourth down, the analytics community had been calling on coaches to be more aggressive years before a few caught up. The Eagles were among the more passive teams on fourth down when Andy Reid and Kelly were at the helm. But Pederson, with a nudge from Lurie, was more receptive to the idea.
“A lot of teams — ours is one — where it’s all [planned] is in the offseason, and done with mathematics,” Lurie said of going for it on fourth downs in 2017. “It’s not based on any form of instinct. If it’s going to be 50/50, 48/52, then a coach is going to have their instinctual predilection, right?
“But what we found is, there’s been so many decisions over time that are too conservative for the odds of maximizing your chance to win at the opportunity. I mean, you’ve seen certain coaches that are deemed more aggressive because the math leads them there. That’s all it is.”
Halaby first worked for the Eagles as an intern in 2007 and 2009. He graduated from Harvard with an English degree, but minored in economics and became immersed in analytics after a brief stint at Football Outsiders and from doing his own research.
Unlike Lurie and Roseman, Halaby actually played football and was a quarterback for his high school team in Madison, Wisconsin. He was hired full-time as an analyst in 2010 and used various methods in scouting, but analytics was his vehicle into the NFL and would become his domain with the Eagles.
He was promoted to special assistant to the GM two years later and despite his relative inexperience was someone Roseman publicly and privately tabbed as a potential future GM. The one time the Eagles allowed Halaby to talk with reporters, he stressed that analytics was just one element of a “full-information approach.”
“I think there’s not a holy grail or a magic bullet or a single play or a single number that’s gonna answer the question,” he said in 2013. “It’s more of trying to get a 360-degree idea, get all the angles.”
Halaby was open to learning more about film study and other traditional forms of scouting. And certain coaches and scouts in turn took his information and applied it to their trade. At the very least, they considered it.
But there were others, some from the old guard, who would literally crumple up an analytics report and toss it in the trash in full view of others, sources close to the Eagles said. Jerry Azzinaro, Kelly’s defensive line coach, once derisively referred to Halaby as that “Harvard guy” during a coaches clinic when he argued passionately against the onslaught of numbers-based reasoning.
“Some coaches really got fed up with Alec,” a source who worked in the Eagles front office said. “But it was more because they thought he was producing ammunition for them to be second-guessed, not because they really hated him.”
Still, early into Kelly’s tenure, he had Halaby removed from the coach’s box during games. And when the coach won a power struggle and was placed in charge of personnel, he made sure Halaby was moved out of football operations along with Roseman.
Kelly, like Pederson, wasn’t adverse to analytics. But the head coaches and other staffers sometimes had a difficult time relating to Halaby and how his information was packaged.
“There are two important parts to being good at analytics. One is just the analysis part. The second is knowing what are the right questions I can figure out that are actionable,” a source close to the Eagles said. “Alec is very good at the first part. The framing — he should be the leader of what are the questions we can ask, what are the areas we can make a difference in — he’s not good at that.
“It needs to come from a collaboration or somebody else.”
Lurie, though, was often enamored by the data. The Eagles’ investment in analytics is just one example of how they’ve been ahead of many teams in embracing innovation. Kelly brought with him a trailblazing sports science department, and Lurie was predicting as early as 2016 that player-tracking devices would revolutionize athletics.
But in terms of the draft, there hasn’t been as much of a technological breakthrough. Film study, player interaction and shoe-leather investigating are still considered the best ways to predict a prospect’s success.
Still, the Eagles and other teams fail more times than not in personnel decisions. Lurie acknowledged as much when asked in 2017 how analytics factored into the draft.
“All you’re trying to do really is beat the odds,” Lurie said. “Anyone who studies the NFL draft knows the odds of hitting on draft picks is not high, whatever round you’re in. … So analytics, I think increasingly around sports, is very important, but it’s a tool.
“It took a lot more than analytics to decide that we were going to select Carson [Wentz].”
If the Eagles had followed Halaby’s advice, they may not have drafted Wentz. In one pre-2016 draft analysis, he had Paxton Lynch rated as his top quarterback, according to personnel department sources. The report was later a source of ridicule among the anti-Halaby faction, especially after Lynch was a bust.
It should be noted that John Elway and the Broncos drafted Lynch in the first round, and Wentz, after a disastrous 2020 season and subsequent trade to the Colts, is now viewed through a different lens.
Three years later, though, one of the driving forces behind the JJ Arcega-Whiteside selection was the math behind the wide receiver’s catch radius. Lurie latched onto this idea, as reported earlier by The Inquirer, and was able to influence Pederson and Roseman into going along.
Some scouts were appalled. They later argued that by focusing too much on one trait, the analytics failed to see the forest from the trees: that Arcega-Whiteside may have had a large sample of catch radius plays because getting separation — even against Pac-12 competition — could be a struggle.
The personnel department received Halaby’s reports, but because he reported to Roseman, and in essence Lurie, there wasn’t a clear delineation between departments when it came to evaluations. The secrecy and perceived subterfuge led some staffers to jokingly refer to the five-person analytics team as the Eagles’ “deep state.”
Halaby, meanwhile, continues to question aides as to whether coaches are playing to Arcega-Whiteside’s strengths, two team sources said. The third-year receiver made the 53-man roster last week, but just barely.
Part 2 of the series continues with a behind-the-scenes look at Pederson’s usage of analytics, the rift between the coach and Halaby and how Sirianni may implement the use of numbers with Lurie over his shoulder.