Jeffrey Lurie has specific ideas on how to run his football team. Each of his last four head coach hires including current head coach Nick Sirianni—most recently the Indianapolis Colts offensive coordinator— rose through the ranks as offensive assistant coaches. The organization is unusually aggressive at acquiring quarterbacks; the surprise usage of the fifty-third overall pick on Jalen Hurts less than a year after Carson Wentz received a $128 million contract comes to mind. According to Overthecap, the Eagles ranked third in spending on offensive players at the beginning of the 2020 season. It’s no secret that Lurie believes building the offense—particularly the passing offense—is the best path to establishing a Super Bowl contender.
This is a logical philosophy in my view. Analytics research has shown that offensive efficiency is more stable year-to-year than defensive efficiency. It also shows passing offense efficiency is more stable year-to-year than rushing offense efficiency. It makes sense to search for a head coach with expertise in offensive play-calling and quarterback development. This approach mitigates the risk that comes when a head coach who delegates those functions to an offensive coordinator who is likely to become a head coach himself for another team.
Like outgoing head coach Doug Pederson, Sirianni had a stint in the NFL as a quarterbacks coach. In Sirianni’s case, he got that opportunity with the Chargers between 2014 and 2015 to oversee Philip Rivers when Frank Reich (current Colts head coach) was their offensive coordinator. Reich, of course, left the Eagles after a successful stint as the offensive coordinator for the Super Bowl-winning team and he brought Sirianni with him to the Colts as his own offensive coordinator. Sirianni also has experience as a receivers coach, ranging from Division II program IUP (2006-08), the Kansas City Chiefs (2012), and the Chargers (2016-17). With Lurie’s emphasis on the passing game, it’s easy to see how a resume like Sirianni’s interested him.
No one knows for sure what the Eagles offense will look like under Sirianni or how his approach might differ from Pederson’s. Another open question is the influence of offensive coordinator Shane Steichen (formerly the same position with the Los Angeles Chargers in the second half of 2019 through 2020). There’s no certainty that the Sirianni regime’s plan will even be an improvement. There are clues which can help us make an educated guess though.
Sirianni will likely use Shotgun (and run the ball from Shotgun) at a similar rate to Pederson
Reich had been Pederson’s own offensive coordinator, so it is natural for there to be similarities between the Colts and Eagles offenses the last three seasons. The most glaring one to me is their similar usage rate of Shotgun formations. Neither the Eagles nor the Colts had been using them at the near-exclusive clip of the Ravens (97%) or the Cardinals (92%) in 2020. But they do fall squarely into that second-tier of teams: the Eagles ranked 6th (80%) and the Colts ranked 9th (75%). Only the 2019 Colts ranked outside the top 10 in the three years since Reich departed for Indianapolis (they ranked 14th, so still above league-average).
The natural question to ask might be, “why does it matter how often an offense aligns in Shotgun?” The answer is that Shotgun formations facilitate some things better than Under Center formations do. Quarterbacks typically prefer the Shotgun in the dropback passing game because they can see more of the field from a five-yard depth than they can under center. This superior vision then allows for more Spread formations with three or more receivers detached from the box. And while option football originally evolved through Under Center offenses like the Split Back Veer and Wishbone, Shotgun formations can package the option threat with modern run schemes. For example, the Ravens use modern zone, power, and counter schemes in their option run plays involving Lamar Jackson to attack the perimeter.
All NFL offenses pass more out of the Shotgun than they run, even the Ravens. The choice to favor the Shotgun then, for Reich and Pederson, is predominantly to enhance their passing games. With that said, both the Colts and Eagles ranked in the top 10 for run percentage on Shotgun plays each season between 2018 and 2020. Steichen’s Chargers offense ranked just lower in 2020 at 11th. Neither the Colts nor Eagles extensively used their quarterbacks as a run threat in Ravens’ fashion (outside of Hurts’ four starts last year under Pederson). But that doesn’t mean that option football was absent. It just existed in a different form.
According to NFL Network correspondent James Palmer, the Eagles ranked 3rd in usage of run-pass options (14.4% of all plays) from 2018-2019 and the Colts ranked 5th (10.7%). The Eagles famously led the NFL in RPO usage in the 2017 Super Bowl season with Pederson and Reich both on staff. The five-yard depth from Shotgun allows the quarterback to make a more precise pre-snap or post-snap read on a defender, so he knows whether to hand the ball off or make a throw to the vacated area.
In short, RPOs are cheat-codes, especially when the read defender chooses to defend the pass. PFF contributor Seth Galina wrote, “the design of a solo running play might not have enough answers to create the required yardage, but there are built-in answers whenever RPO is attached. Over the last three NFL seasons, the average run play gained -0.108 expected points added (EPA) per play, while runs without an RPO attached gained a lowly -0.117 EPA per play. When teams attached an RPO, that figure rose to 0.070.”
The Eagles’ first-round draft pick this year, Alabama receiver Devonta Smith, has plenty of experience as a target on RPOs. According to SEC StatCat, 40 of Smith’s 118 catches (and 44 of his 151 targets) last season were part of RPOs. Smith’s absurd ability to pick up yards-after-catch (or YAC) makes him an ideal target on “replace” slant routes or digs which put linebackers in conflict between their zone coverage pass responsibilities and their run fits. And if the defense tries to play man? Exploit the defenders who have turned their back by letting Hurts run.
It’s fair to expect the Eagles’ new offense to remain near the top of the league in Shotgun usage and among the most balanced (in terms of run-pass) when in Shotgun formations. Hurts’ presence may even dictate a higher usage rate of Shotgun than either the Eagles or Colts offenses achieved the last three seasons. The modern quarterback run game offers a new dimension which fits perfectly in Sirianni’s likely inclination to lean heavily on the Shotgun.
Sirianni’s usage of Under Center formations will likely be different than Pederson’s
Down-and-distance splits offer some more granular perspective on formational tendencies. The Eagles had just a slightly higher percentage of snaps in Shotgun than the Colts did on 1st-and-10, 2nd-and-Long, 2nd-and-Medium, and 3rd-and-Short. Even with Steichen’s Chargers ranking closer to the middle in under center usage, it seems likely the Eagles will remain a Shotgun-heavy team in most situations.
The one notable exception appeared in 2nd-and-Short; the Eagles were in Shotgun on 71% of those plays, compared to 48% for the Colts. The Colts’ went without disguise on their under center plays in this situation. They literally ran the ball on 100% of 2nd-and-Short plays in which they were under center. They were still run-heavy from shotgun, ranking 4th with a run percentage of 59% on 2nd-and-short. This is a striking difference with the Eagles who ranked 23rd in run percentage from Shotgun and 30th from under center.
| Shotgun Percentage (2020 Regular Season) | PHI | IND | LAC |
| 1st and 10 | 73% (7) | 68% (10) | 49% (18) |
| 2nd and Long (8-10) | 86% (6) | 82% (12) | 77% (17) |
| 2nd and Medium (4-7) | 76% (9) | 69% (11) | 76% (8) |
| 2nd and Short (<=3) | 71% (5) | 48% (15) | 29% (28) |
| 3rd and Short (<=3) | 79% (11) | 77% (13) | 72% (16) |
Although they’re a small part of the offense, the under-center formations are where I expect some significant differences in Sirianni’s tendencies compared to Pederson’s. The Colts ranked third in run percentage on under center plays last season, while the Eagles ranked second-to-last. 2nd-and-short plays are certainly a driver of this divergence. This disparity basically holds up in the previous two seasons; the Eagles ranked no higher than 22nd in under center run percentage, while the Colts ranked 3rd in 2019 and 15th in 2018.
There are some interesting differences between the Colts’ and Eagles’ run schemes the last three seasons which add some context. In short, the Colts’ run game was a lot more schematically diverse than the Inside Zone-centric Eagles. For instance, Ryan Weisman found the Colts called nearly twice as many Outside Zone plays as the Eagles last season.
Outside Zone is the base run play of teams like the Vikings, Titans, Rams, and Browns who lead the league in under center usage. It’s a downhill play that often excels with tight line splits and under center alignment. The Colts’ higher run rate from under center is likely driven by their Outside Zone usage, along with a higher usage of the Man scheme (also known as Duo or “power without the puller”). The Duo scheme looks very similar to Inside Zone, but there are some key differences. The most relevant one for the Eagles’ personnel is that Duo is run to the side of the tight end; Goedert may have his shortcomings as a receiver, but he’s an excellent run-blocker.
Steichen’s Chargers offenses featured Outside Zone even more frequently than Reich and Sirianni did. There was also a high dose of Duo in the Chargers run game, like there was for the Colts. The key to making both Outside Zone and Duo work is the double-teams along the offensive line. The return of four experienced starters and the development of Goedert as a run-blocking tight end makes a shift to more Outside Zone and Man from under center more feasible.
In fact, the 2017 Eagles’ run scheme usage resembles the 2018-20 Colts’ more than the 2018-2020 Eagles’. Not surprisingly, they ranked 14th in rushing percentage from under center in 2017. The Eagles were open about Reich’s particular influence on the run game in his time as the Eagles offensive coordinator. The drop in Outside Zone between 2018 and 2020 surely reflects his departure.
Sirianni might not be as analytics-oriented as Pederson on the play-action pass and passing on early downs
Conversely, these under center tendencies hint at an incremental drop for the Eagles in play-action passes under Sirianni. Football Outsiders showed that under center play-action passes produce a target depth of 12.0 yards compared to 9.2 for shotgun play-action passes on 1st-and-10 (the disparities are similar in all other down-and-distance categories). Under center play-fakes also net a higher yards per attempt on 1st-and-10 and 2nd-and-5 to 10 (traditional running downs), relative to shotgun play-fakes. Play-action passes (especially those from under center formations) are a cheat code. Reducing their usage is a real concern I have with the incoming coaching staff.
I don’t have the data to totally prove this hypothesis, but Pederson’s relatively pass-heavy tendencies from under center are likely driven by an effort to maximize play-action. The Pederson-era Eagles acted like they were on a mission to prove Ben Baldwin’s conclusion that play-action passing efficiency is independent of run volume. The Eagles even dialed up 21 play-action passes (a blistering 49% of total passing attempts) in the Super Bowl victory against the Patriots. They ranked 4th in percentage of play-action passes to total passing attempts in the 2017 season. This is clearly Andy Reid’s influence on Pederson. The Chiefs ranked third on play-action percentage on 1st downs in 2020, despite also being the pass-heaviest team in competitive first downs. Despite conventional wisdom, it is not necessary to be a run-heavy team to succeed with play-action.
When Reich left, there was a clear divergence between the Eagles and Colts in this aspect of their offenses. The Eagles proceeded to rank 3rd, 6th, and 17th in play-action percentage from 2018 to 2020 in that order, while the Colts ranked 19th, 10th, and 30th. For what it’s worth, Steichen’s Chargers offenses ranked 27th and 16th in play-action percentage the two seasons he was a play-caller. Reich certainly had his fingerprints on aspects of the Super Bowl offense like the run game, but his post-Super Bowl tendencies suggest we should credit the play-action passing game to Pederson at this point.
Is a dropoff in play-action under the new coaching staff reason for concern? Maybe a little. Play-action passing consistently outperforms both non-play action passing and rushing in terms of efficiency by a wide margin. To be clear, the delta between a Pederson and a Reich offense amounts to between 2 to 3 play-action attempts per game. It’s not like the Sirianni Eagles will become the Pittsburgh Steelers who ranked dead-last in play-action percentage and called play-action almost half as frequently as the next-lowest team in 2020. But this delta also is not trivial. Over time, those 2 to 3 play-action attempts per game would accumulate to several explosive plays across the season and may swing a game or two.
Another aspect of the Super Bowl offense where Pederson appears to have final say on was early down tendencies. FiveThirtyEight shows that passing on early downs raises average EPA, while rushing lowers it. The Eagles ranked fifth in passing percentage (53%) on 1st-and-10 in 2017, before accounting for how often their win-probability was high. They remained in the top-half of the league each of the seasons after Reich left. In Reich’s first season with the Colts, they actually passed more often on first down (58%, 4th-most) than the Eagles did (55%, 9th-most). That changed the following two seasons without Andrew Luck, as their rates were 48% in 2019 and 47% in 2020. There are many parts of being a play-caller which aren’t so easily reductive to analytics. But it’s fair to say that, for the parts which are, Pederson was more willing to push easy buttons like more play-action and passing on early downs than Sirianni’s old boss.
Sirianni is likely to use shorter passing concepts (specifically more shallow crossers) than Pederson
The Colts had three very different starting quarterbacks in Sirianni’s three seasons in Indianapolis: Andrew Luck, Jacoby Brissett, and Philip Rivers. But one thing all three had in common was they threw a lot of crossing routes to Colts receivers. According to Ben Solak, “[i]n three seasons, Colts’ passers were among the league’s top quarterbacks in gross targets on shallow routes: Luck with 53 such targets in 2018 (2nd), Brissett with 35 in 2019 (4th), and Rivers with 46 (2nd).” Solak also refers to a clip of Sirianni describing the foundational ideas to the Colts offense: “getting ‘the ball to our players within 5 yards of the line of scrimmage, running.’”
This was not achieved by the Eagles offense in the Pederson-Wentz era. Wentz attempted just 16 shallow routes in both 2018 and 2019, only to see that number fall to 8(!) in 2020. Some of this was specific to Wentz; Solak notes that even with Reich in Philadelphia in 2017, Wentz threw 16 shallow routes that season. Even with Wentz gone, the Eagles’ new quarterback (Hurts) shows an even more glaring evasiveness to throwing at the middle of the field. Hurts ranked second-to-last in 2020 on percentage of attempts to the middle of the field among qualified passers and dead-last in success rate on these passes.
If there is one aspect to Hurts’ game that Sirianni must correct, this is it. The building block passing concepts of Sirianni’s offense will not function if Hurts remains averse to the middle of the field. In my view, it’s such a major leap to think this habit will change adequately enough. I argued in the spring that the Eagles should not forgo the opportunity to draft a quarterback prospect for precisely this reason, among others. They obviously disagreed with me. There happened to be one quarterback prospect, Justin Fields (Ohio State), available at the time of their pick who excelled at throwing in the middle of the field.
The fact that I think the Eagles made a mistake in assessing the quarterback position is true, regardless of whether Pederson or Sirianni were the head coach. For one thing, many of the concepts Sirianni is likely to bring to Philadelphia are ones Pederson has been using as well. Emblematic of the West Coast offense influence on Reich, the Drive concept is a throwback to Bill Walsh, Joe Montana, and Jerry Rice. It forms a triangle read by pairing a shallow crosser by an outside receiver with a deep dig by the inside receiver and a checkdown to the running back. The Air Raid influence on both coaches is best captured by the Mesh concept, characterized by two intersecting shallow crossers from opposite sides of the field. This was the play the Eagles shredded the Patriots on four times in the Super Bowl for 81 yards and three third/fourth down conversions, according to The Athletic. Crucially, these are all passing concepts which threaten the middle of the field.
The high volume of crossing routes in the Colts’ offense shows how Sirianni might compensate for the marginal loss in efficiency from fewer play-action attempts. Andrew Rogan (recently hired by the Baltimore Ravens’ analytics staff) used PFF and NFLFastR data to show that targets on crossing routes have higher average EPA than all other targets. This was one finding that led him to believe the high efficiency on play-action passes may, in part, be a product of omitted variable bias. The plain English way of saying that is this: because so many play-action passes feature crossing routes, it might be crossing routes which drive efficiency on these plays (not play-action). If this is true, then a drop in play-action volume from Pederson to Sirianni may be less of a big deal than it first seemed.
Rogan doesn’t think there is no boost in efficiency from the actual play-fake by the quarterback. Quite the contrary. In fact, the basic point he makes in the post is that a “successful running attack leads to play actions being more effective. The key unseen to this finding really is depth of target.” Using statistical techniques, Rogan found there is a significant relationship between efficiencies for rushes and play-action passes with a depth of target less than 10 yards (but not passes with higher depth of target). That relationship extends to crossing route targets specifically as well. The idea is that linebackers are most sensitive to play-fakes (an observation backed by Hermsmeyer), so the space linebackers vacate is the most exploitable. That space happens to be the same area that receivers occupy on shallow crossers. It may not be necessary to run the ball often, but it helps to run the ball well.
This is where the Colts’ formula starts to make more sense. They were one of just two teams (the other being the Ravens) to generate positive average EPA on run plays in the second half over the last 3 seasons. This is remarkable because they also rank in the middle of the pack in average rushing EPA in the first half. In other words, the Colts offense is the one true example in the NFL of “body blows” over a large sample period. The offensive line of Anthony Castonzo, Quenton Nelson, Ryan Kelly, Mark Glowinski, and Braden Smith genuinely seemed to inflict such punishing physicality that the Colts became better at running the ball with each additional attempt.
To be clear, the very exceptional nature of that Colts offensive line means that this probably isn’t a reasonable benchmark for the Eagles. Jason Kelce, Brandon Brooks, and Lane Johnson have all been terrific run-blockers throughout their careers, but the Colts’ achievement is one that not even the 2017 Eagles line managed. Even so, an adequate run game and a higher dose of crossing routes is a schematic pairing which makes sense.
Sirianni will use three receiver sets more often than Pederson
Rushing EPA is correlated with lighter boxes and, in turn, lighter offensive personnel groupings (see charts from FiveThirtyEight). Personnel groupings are an area where there has been another divergence between Reich and Pederson. The Colts used 11 personnel (1 running back, 1 tight end, 3 receivers) at a clip of 72% (6th), 62% (16th), and 69% (9th) of snaps over the last three seasons (the Chargers ranked 6th at 71% in 2020). Under Pederson, the Eagles used it just 56% (27th), 41% (29th), and 59% (18th) of snaps.
The modern consensus is that simply being in 11 personnel achieves greater run efficiency. I happen to agree with that idea. But I think the causality works in reverse too. PFF contributor Eric Eager found that interior defensive linemen who defend the run well allow defenses to allocate more resources to the passing game. My thinking is the same mechanism should apply to the offensive linemen who excel at run-blocking. This is an example of what Football Outsiders writer Zach Binney calls a DAG (short for directed acyclic graph). To simplify his idea, the offense only partly influences its own rushing success with personnel. The defense has a say in it as well by choosing to be plus-one in the box or not. If the offense is particularly excellent at running the ball from 11 personnel, the defense may opt to be plus-one in the box and create advantageous spacing in the passing game.
Anthony Reinhard found the box count tendencies deployed against each offense through Week 13 last year (sorted by probability of passing play). The Eagles happened to face lighter boxes than league-average in most situations, while the Colts and Chargers both essentially faced league-average box counts. Neither is necessarily better than the other in a vacuum. The Chiefs and Bills were clear outliers in facing low box counts, precisely because Patrick Mahomes and Josh Allen rightfully inflicted the fear of God in defensive coordinators. The Titans and Browns were similar outliers in facing high box counts due to their prolific rushing attacks.
Some important context: Reinhard doesn’t distinguish between personnel groupings in his analysis. The Eagles used 12 personnel more often than the Colts and Chargers in 2020, yet they somehow faced lower—not higher—box counts on average. This captures how Pederson became a victim of his own past success with two tight end sets. According to Sheil Kapadia, the Eagles passed on 57% of snaps from 12 personnel in 2017 (second-highest behind the Chiefs), compared to the league-average of 45%. This was a small dropoff from the 60% passing rate from 11 personnel that season. This was a big driver of the Eagles’ Super Bowl run. The 27 snaps that Zach Ertz and Trey Burton played together in the playoffs resulted in 8.9 yards per play (9.7 yards per pass attempt) with 15 first downs and 4 touchdowns.
This included the Super Bowl-winning touchdown pass to Ertz whose play-design Reich was heavily involved in:
[Receivers coach Mike] Groh and Reich agreed that a bunch play would definitely be a smart inclusion—and all three men thought splitting out Ertz, as he had against Atlanta, would be smart, as well as the unique motion to the bunch side by the back. This would be a variation of the bunch they’d never run … and, of course, a variation the Patriots hadn’t seen. Groh mined the cluster-route concept, Reich added the running-back motion, Reich and Groh liked splitting out Ertz, Groh brought it to Pederson, and Pederson liked it all.
Sharp Football Analysis found that, in 2018, the Eagles had an identical 61%-39% pass-run split in 12 personnel sets and the overall offense. But while that formula achieved astonishing success in 2017 and 2018, Steven Ruiz showed how it cratered by 2019. The Eagles ranked 31st in average passing EPA from 12 personnel that season and ranked 11th in average rushing EPA in that personnel group (less impressive when the lower EPA on runs versus passes is considered).
Ruiz referenced Seahawks linebacker K.J. Wright who claimed his teammates were calling out the Eagles plays before the snap; Wright specifically mentioned in the Philadelphia Inquirer seeing tells on a Pederson favorite from the Reid playbook (the tight end middle screen). Ruiz then added his own analysis: “It’s not hard to figure out when an RPO is coming (when the back is lined up away from the tight end side in gun formations on early downs) or when the Eagles are about to run their trusted Mesh play (when the receiver to the tight end side and the slot receiver are in tight splits).”
So did the Eagles upgrade at head coach?
Play-designs had clearly stagnated after Reich departed Philadelphia. But I don’t think all of that falls on Pederson’s shoulders. The Philadelphia Inquirer reported that Wentz checked out of calls at the line of scrimmage frequently (even just to spite his head coach). Judging by their actions this offseason, the Eagles front office and Wentz both seemed to believe Reich held the answers for rekindling their success in the 2017 Super Bowl season. The last few seasons have provided some strong hints for where Pederson and Reich each distinctly deserve credit for earning the Lombardi Trophy. I emphasize that they both contributed.
The conventional wisdom may be that Reich was the brain trust behind the Super Bowl. I’m deeply skeptical. Pederson had (mostly) done well at pushing easy buttons like passing on early downs, calling lots of play-action passes, and going for it on 4th down. That’s not something Reich did in Indianapolis. There are areas where Sirianni may improve over Pederson, like more emphasis on crossing routes and lighter personnel groups. But because Sirianni has no experience as a play-caller, it’s hard to take a shot in the dark that he’ll outperform his Super Bowl-winning predecessor. That is a wager I won’t make.
Liberty Gridiron 