The Bills’ fanbase has reached a level of optimism that we haven’t seen in years, and given how the last two years have transpired, I understand why. The front office and coaching staff have done a great job of adding talent, specifically cornerstone talent, on both sides of the ball, and they have tailored their schemes and philosophies to match their personnel rather than trying to force their personnel into their schemes.
Fans saw this firsthand in 2018 as franchise quarterback Josh Allen got his first taste of NFL action. While his overall numbers were pedestrian, the change the offense underwent from his first six weeks of the season to the last six was glaring. As we all know, there was much more youth and speed added into the lineup post-Allen’s injury, but no one has really gone in-depth into why the team made those changes or how they changed how Allen was defended by the Bills’ opponents. That is, until now. So with the help of SportsInfo Solutions (SIS), I want to break down how teams defended Allen and the Bills’ offense differently pre- and post-injury, and show you why they made those changes and what it means going forward.
One-High vs. Two-High Safety Looks
After scouring the statistics, one of the more interesting trends that I noticed was how teams treated Allen’s passing acumen. Surprisingly, teams primarily played single-high coverages against Allen. For the sake of this article, single-high coverages are Cover 1 and Cover 3, whereas two-high looks are Cover 2, Cover 2 man, Tampa 2, Cover 4, and Cover 6. Omitted for this section were Cover 0, screens, prevent, and combo coverages.
In 2018, Allen saw single-high looks on 76% of his dropbacks. He completed 51.5% of his passes vs. single-high looks for 1,429 yards, 9 TDs, and 5 interceptions. He also had a much higher touchdown percentage (4.4%) when facing one safety vs. two (1.4%).
Breaking it down even further revealed that Allen faced a lot more single-high looks in weeks 1-6, facing that coverage shell 79.3% of the time vs. 20.7% of two-high. That’s why OC Daboll made things as easy as he could for the rookie, like giving him packaged run-pass plays. Here he sees a stacked box and a single-high look, so he checks to the pass play in the call and gets an explosive ball to Andre Holmes. Packaged plays were something the staff implemented and carried in the offense throughout the season.
But the offense as a whole struggled, including Allen, who only mustered up two touchdowns via the passing game vs. those single-high looks in the first six weeks and none against two-high looks.
One of those touchdowns came against the Vikings in week three.
Man vs. Zone Coverage
Now let’s dive a little deeper into the coverages by examining man (including Cover 0) vs. zone coverages. Defenses were not scared of the offensive weapons that the Bills had outside, so they appeared to play a lot of single-high looks, as mentioned earlier, and man coverage. With receivers like Kelvin Benjamin and Andre Holmes playing prominent roles in the passing game, teams elected to stop the run and play man coverage to the tune of 62.9% of Allen’s dropbacks. Which includes 55% of his dropbacks vs. Cover 1 in weeks 1-6. So Allen was throwing into tight man coverage to receivers who generally lacked the ability to separate. You can see why Brandon Beane and Sean McDermott decided to add some speed into the lineup as Allen returned from injury, even admitting “we thought we were lacking in the speed department on offense. That’s been an emphasis the last couple of weeks.”
The minor tweaks in philosophy and personnel changed defenses’ coverage approach. Defensive coordinators not only played less single-high coverages (-7.1%), but they played 17% less man coverage against Allen and the Bills’ offense in the second half of the season.
Allen’s completion percentage dropped against both man and zone coverages in the second stanza but saw an increase in yards per attempt vs. man coverage.
More importantly, Allen increased his touchdown percentage across the board once he returned to the lineup, including this strike versus the Dolphins. With defenses playing a little more two-high AND zone looks in the back half of last season, such as this Cover 6 look, Allen was able to increase his touchdown percentage.
Once Allen returned, the offense became a lot more explosive with more speed in the lineup, which allowed Allen to uncork his big arm on passes over 20 yards 40 times, compared to the 23 in weeks 1-6, which was an increase of 5.6% (Per Pro Football Focus). Allen’s deep pass percentage went from 16.5% to 22.1%, which led the NFL from weeks 12-17. His accuracy percentage on passes over 20-yards went up from 30.4% to 40%, and his touchdowns jumped from one to six. His arm started to become a major factor.
No QB topped Josh Allen's average depth of target of 11.48 last season pic.twitter.com/8yLDxdqaTJ
— PFF Fantasy Football (@PFF_Fantasy) March 9, 2019
Robert Foster and Zay Jones were the recipients of most of these deep touchdowns. Here’s one of Foster’s touchdowns vs. Cover 1.
Allen’s Athleticism Took Over
Another peripheral result of the changes was that Allen took fewer sacks. On the year, Allen took 28 sacks, 21 of which were vs. man coverage. Fifteen of those sacks were in weeks 1-6. That’s a drop from 16.7% sack percentage to 7.7% (as seen in the chart above). Per SIS, Allen broke eight tackles/sacks in the pocket, which was third behind Patrick Mahomes (11) and Baker Mayfield (9), who both had many more dropbacks than Allen.
The game appeared to slow down a bit, and Allen became a little more decisive. This includes his decision to just pull the ball down and take off, something that he did a lot of. Per PFF, Allen finished with 89 rushing attempts for 613 yards and eight touchdowns. Forty-seven of his rushing attempts, 513 yards, and five touchdowns were off of QB scrambles. Only 22 rushes were designed QB runs, per PFF, while the rest were QB sneaks, kneels, or other. Here, the Jaguars blitz and play man coverage on the back end. Allen feels the pressure, so he wastes no time tucking and running.
Daboll embraced Allen’s athleticism and let the wild stallion run free. In weeks 1-6, Allen rushed 35 times for 155 yards, which is 4.4 yards per attempt, including 86 yards after contact and three touchdowns. From the Jaguars game on, Allen ran the rock 54 times for 476 yards, which is 8.8 yards per attempt, five touchdowns, a whopping 248 yards after contact on 12 broken tackles, and converted a 1st down on 53.7% of his carries. The offense truly took on the identity of their rookie quarterback through the air and on the ground.
What does it all mean?
While it took General Manager Brandon Beane and Brian Daboll some time to completely understand how Allen’s play at Wyoming would translate to the NFL, they appeared to have carried those adjustments into the 2019 plans.
Per PFF, Allen was under pressure on 43.4% of his dropbacks in 2018; that was the second most behind Houston’s Deshaun Watson. Eleven of Allen’s 28 sacks surrendered occurred in 2.5 seconds or less, which was the fifth-most among all QBs. Nine out of those 11 sacks under 2.5 seconds took place in the first six weeks of the season. So the free agent pickups and drafting of guys like Mitch Morse, Ty Nsekhe, Quinton Spain, Spencer Long, and Cody Ford, just to name a few, should help relieve some of the duress Allen was under last year.
The upgrades along the offensive line and running back talent should jumpstart the run game if defenses want to play as many two-high looks as they did to end the 2018 season. Sadly, the Bills’ rushing yards per attempt dropped from 3.4 in the first six weeks to 3.0 from weeks 12-17. This is likely due to the shuffling of offensive linemen, specifically adding Wyatt Teller from week 10 on and going from Russell Bodine at center for most of the season to Ryan Groy — perhaps the worst run blocker of them all — from week 12 on. John Miller was another hog who was in and out of the lineup late last season. Overall, the offensive line was one of the worst units in recent memory.
But as is often the case, the biggest key was giving Allen the correct weapons. The big, contested catch guys don’t mesh well with his game. Getting receivers that separate with their physical abilities, specifically their speed or quickness, helps him more. So adding guys like Cole Beasley and John Brown to the WR room, along with an athletic tight end like Dawson Knox, makes perfect sense if teams are going to continue to defend Allen with single-high and/or man coverage looks. At the very least, their ability to separate and get deep will help Allen continue to lead in deep pass percentage as he did in 2018, attempting a whopping 19.7% of his passes over 20 yards. That’s 2.9% more than the next QB.
The Firebaugh native showed improvement as the year went on in 2018, but very little of it showed in the stat box. Now that the staff understands how to use him, he should be in much better situations, situations that he has likely seen before, but this time with a few more weapons that will help him be more successful.