“They’re calling out our plays.”
It’s more innocuous of an observation than it seems at face value. There isn’t an offense in the league that doesn’t occasionally get its plays called out by the defense—and for a team like Baltimore, whose offense was the primary subject of intensive offseason study for defensive coordinators, it was an expected obstacle. The “wait until they get film on Lamar” argument, often made by detractors during his 2019 MVP season, is largely overestimated in its impact—but does have a kernel of truth at its core: it is easier to stop an offense for which you have several months, not one week, to prepare.
But there’s no avoiding the harsh reality of the Ravens’ offense, which is both disappointing and still pretty solid in 2020. Expectations ran high after Jackson’s MVP season, in which Baltimore paced the league in both passing and running game DVOA and generated a positive EPA/play on designed runs over a full season—something most offenses never accomplish. This year, the Ravens remain a top-five rushing offense by DVOA—still quite good, but missing the head and shoulders advantage they held over fellow rushing attacks in 2019.
It’s the passing game that has lagged far behind. In every major category of a passing game’s efficiency, the Ravens have endured significant hits: EPA/play has gone from second to 19th per RBSDM.com; DVOA has gone from first to 23rd; Jackson’s ANY/A has fallen all the way from 8.19 to 6.20. These dropoffs, of course, cannot be completely attributed to an opposing defense correctly identifying tendencies and play-calls on the field—especially when you consider Jackson’s play-action numbers. Still one of the heaviest play-action passing teams in the league, the Ravens have actually improved the play-action passing game in 2020 off of their already astounding 2019 numbers: off of play-action, Jackson is completing 65.8% of his passes (64.5% in 2019) for 8.0 yards/attempt (7.5 in 2019). If opposing defenses have sniffed out Baltimore’s play-action tendencies, it hasn’t been reflected at all in their success against it.
No, the failure of Baltimore’s passing game is in their standard dropback offense. Still a good 65% of their passing game as a whole, Jackson's 6.7 yards/attempt and 61.4% completion percentage without play-action are markedly lower than 2019’s 8.0 yards/attempt and 66.9% completion percentage under the same context.
The values of play-action have long been lauded and investigated, and with a running game as effective as Baltimore’s folded into the mix, it is not surprising that Greg Roman’s passing game remains one of the league’s best. But even without a run fake attached to a passing play, the opportunities to stretch second-level defenders, script easy throws with motion, or move safeties for deep shots all remain accessible to good passing game designers. This is where Roman has failed the passing offense in Baltimore. Their designs are both unimaginative and unintuitive, and critically, do little to stress defenses in coverage.
Take the pick-six against the Steelers as an example.
The Ravens push running back J.K. Dobbins out wide into an empty set to give Jackson a key on whether the defense is in man or zone coverage. The Steelers are in zone coverage here, as cornerback Joe Haden stays wide over Dobbins. But the Ravens’ passing design here has almost no teeth against zone coverage: with two quick, in-breaking routes layered underneath another in-breaking route, the Ravens deposit all three receivers to the middle of the field, where they can all be accounted for by only two zone droppers for the Steelers—let alone the flat defenders and deep middle safety lurking around this poorly spaced combination. Jackson comes off his first read and works to rookie wide receiver James Proche, who has worked too much toward the middle of the field and linebacker Robert Spillane, condensing the space into which Jackson could deliver an accurate throw. The ball is picked.
There was not a single defender put into any state of conflict here, and unfortunately, that is not an isolated sensation for the Ravens’ passing designs. When working their dropback passing game, Jackson is infrequently given progression reads, as most of his options see their routes break at similar times. Accordingly, he has become more prone to a quick scramble when his first read is covered, dropping his eyes instead of getting to anything on the backside—there is usually nothing available anyway. Against man coverage, if tight end Mark Andrews isn’t uncovering, the rest of the Baltimore pass-catching corps often lacks the chops to win on breaking routes.
Despite Jackson's poor production on dropback passes, his isolated play remains about the same. Per Sports Info Solutions, Jackson's 72.9% rate of on-target passes on traditional dropbacks (32nd in the NFL) is only barely lower than his 74.0% on-target rate (27th) from last season. He isn’t less accurate—but everything else around him is worse. The passes are going for fewer yards and completions, as we’ve seen, but his touchdown rate has nearly halved (9.3% to 4.9%) while his sack rate has nearly doubled. This sensation is closely linked to that identified in Steven Ruiz’s piece on the Baltimore Ravens offense for For The Win, in which he noted that Jackson’s success from empty sets has gone from first to worst in the last season, with blitzes and pressure playing a significant role in his steep decline. Empty sets not only inherently exclude play-action, which allows defenses to blitz more aggressively, but they offer route combinations with which opposing defenses are more familiar, but without the elite wide receiver talent of teams built differently than Baltimore.
Unfortunately, this sensation is not new to the Roman offensive approach. The Ravens’ 2019 offense has often been compared to the 2012 San Francisco 49ers offense, and appropriately so: that team was fifth in both rushing and passing DVOA, Roman called the plays, and once star running quarterback Colin Kaepernick took the reins from Alex Smith, he deployed a play-action heavy passing attack that utilized pistol formations with heavy personnel to open up the QB run game. That playbook was almost copy-pasted into the Ravens’ base offense last year—and the issues that followed it San Francisco are following it to Baltimore.
From 2012 to 2013, Kaepernick experienced a similar and significant dropoff in his dropback passing efficiency, going from 63.8% completion rate and 8.5 yards/attempt to 57.2% completion rate and 7.3 yards/attempt. It is almost precisely the hit that Jackson took from 2019 into 2020, and it is accompanied by a similarly buoyant effect of the play-action game, in which Kaepernick not only remained effective but improved from 2012 to 2013.
What is truly concerning isn’t what happened to Roman’s offense from 2012 to 2013—it’s what happened in 2014. Coming off a still successful but slightly concerning offensive season in which he was criticized for a pass-happy approach and poor red zone play-calling, Roman stripped the offense down to its base parts in 2014. He called it a “clean the garage” offseason, in which they would go back to their “basic elements.” This philosophy shift, accompanied by a commitment to wide receiver talent in the acquisition of Stevie Johnson and Brandon Lloyd, and the drafting of Bruce Ellington, meant an abandonment of the play-action passing game. The Niners went from leading the league to average in the league, and with that shift, careened into a more traditional spread offense that no longer took full advantage of Kaepernick’s running ability.
Through 2012 and 2013, the Niners used multi-back sets (22 and 21 personnel) on a majority of their snaps, more than doubling the league average at the time—no team put three receivers on the field less frequently. In 2014, the Niners suddenly based out of 11 personnel with 40% of their total snaps, running out of that grouping only 22% of the time, per Football Outsiders. The offense lost its true character, and with it, Roman lost his job.
Again, you can see the foreshadowing of this trap in the Ravens’ current trend. Baltimore led the league in 22 personnel snaps in 2019 (15%) and was fifth in 13 personnel snaps (6%). This season, while still pacing the league in 22 personnel, the trade of Hayden Hurst has all but eliminated the Ravens’ three-tight-end offense (two snaps). The accompanying uptick comes in 11 personnel, which has risen from 46% last year to 51% this year, and with it comes only a 41% successful play rate on passing plays. This was the Ravens’ least effective personnel grouping for throwing the football—not only are they deploying 11 personnel more frequently, but they’re also throwing it from 11 personnel more frequently as well.
Again, this shift was foretold by investment. The Ravens drafted two wide receivers in 2019 with significant capital: Marquise Brown as the first wideout off the board at No. 25, and Miles Boykin as the 13th at No. 93 overall; this past draft they added another third-round receiver in Devin Duvernay at No. 92 overall. With Boykin as the slowest receiver selected with a 4.42s 40-yard dash, the Ravens were clearly oriented on team speed and deep passing, which head coach John Harbaugh listed as their priority on offense during the offseason.
As these picks have struggled to fully pan out, the Ravens’ accompanying neglect of other positions has glared. In trading away Hurst and electing not to add another tight end, blocking tight end Nick Boyle has stepped into the TE2 role and failed to produce in the passing game as he has previously. Marshall Yanda retired and the Ravens filled his shoes with third-round rookie Tyre Phillips, a college tackle who has struggled mightily in his first season. Late-round and undrafted rookies Bradley Bozeman and Patrick Mekari both found starting gigs through last season and went unexposed in the scope of a dominant offense, but this season have struggled when asked to play beside returned center Matt Skura, who does not look fully like himself coming off of a knee injury in 2019. Yes, the Ravens invested in the future of their running game in second-round pick Dobbins—but for as long as they insist on splitting his carries with Gus Edwards and Mark Ingram, they won’t see the added value to the running attack that a top-flight guard would have offered.
Of course, there’s only so much you can predict with injuries and retirements; so much faith you can put into young players and your scouting department’s projections thereof. What you can see clearly, however, is not only how the Roman offense of Niners and Kaepernick lore quickly tailed off, but how poor drafting accelerated that process. Through 2012 and into 2014, the Niners missed on A.J. Jenkins, Quinton Patton, and Bruce Ellington at wide receiver; LaMichael James, Marcus Lattimore, and, to an extent, Carlos Hyde at running back; Marcus Martin, Joe Looney, and Brandon Thomas on the interior offensive line. The offense shifted away from what worked, but it also lacked the quality young talent to shift toward, and the move was accordingly aimless and hopeless.
The fall of the Niners under Harbaugh and Roman had far more contributing factors than a stale dropback passing game and a shift to spread personnel. There was friction with the front office, a far worse draft resume than the one in Baltimore, and the quarterback, while talented, was not as good a player as Jackson is now. Baltimore has far more encouraging signs even in its struggling offense than San Francisco did: the elite play of Dobbins, the on-target throws from Jackson, and the aggressive fourth-down decision-making of Harbaugh all contribute to that sunnier outlook.
But just like defenses know the Ravens’ next play, we can also look to Roman’s offense in San Francisco and see a potential future. One in which a struggling passing game continues to turn to empty sets and shaky wide receiver talent to galvanize a passing game that must instead swing the other way, using heavy personnel and play-action to create easier throws. One which Roman simplifies instead of complicates, looking to assist his quarterback and instead demystifying his offense for opposing defenses.
Defenses have seen this act before; we have as well. Jackson could always play better—most quarterbacks could always play better—but the Ravens’ poor passing game is a symptom of a larger ill at the levels of the coordinator and personnel department, and the necessary salves must come from those departments accordingly.