2018 Defensive Personnel Analysis
by Bryan Knowles
Last week, we looked at which personnel groupings were most popular on offense in 2018. Now, it's time to flip that around and look at how defenses have responded. But first, we must begin with a question, courtesy of our head honcho Aaron Schatz.
It just occurred to me. At what point do we change the FRONT SEVEN section of the FO Almanac unit comments to FRONT SIX? Would be more accurate in the current NFL.
— Aaron Schatz (@FO_ASchatz) June 24, 2019
The answer is "now." We can change that right now, as the front seven rarely plays together any more.
"Base" defense is, at this point, a term that needs replacing; it's simply no longer even remotely accurate. Nobody played base on the majority of their snaps in 2018; only two teams (the Broncos and Rams) even had it as plurality of their snaps. This isn't because plays are being equally distributed amongst formations, either; 25 teams spent more than half their defensive snaps in nickel defenses, and one (the Chargers) even spent more than half the time in dime packages. "Base" defense in the modern NFL is a five-man secondary; your nickel corner is a far more valuable player than your third linebacker or fourth lineman.
This has been going on for years at this point, of course, but we reiterate it this year because we've reached a milestone. For the first time since we began tracking personnel in 2011, base defenses were used on fewer than 25 percent of all snaps. And, for the first time, nickel defenses were used on more than 60 percent of all snaps. It's a trend that's been going on for years, but it's always interesting to see these whole-number milestones go past.
|Defensive Personnel Groupings|
|Personnel||2017 Pct||2018 Pct||Difference||2018 DVOA|
A couple quick notes here:
- For the record, 55 percent of base snaps were in 4-3; 43 percent were in 3-4. That doesn't add up to 100 because we saw 92 snaps of a 2-5-4 (mostly from Cleveland) and 49 snaps of a 5-2-4 (mostly from Miami). Even when teams are using base defenses, they're beginning to use them in less conventional ways.
- We group 4-3 and 3-4 defenses together to look at base sets as a whole. Teams do generally still stick to one or the other, however. 11 teams primarily ran a 3-4, most prominently Denver, the only team with zero 4-3 snaps. 15 teams primarily ran a 4-3, including Indianapolis, the only team with zero 3-4 snaps. That leaves seven teams which had neither front hit 75 percent of their base snaps: Arizona, Atlanta, Cleveland, Houston, New England, and both the Giants and Jets.
- "Dime+" includes any package with more than five defensive backs. That includes all your dime packages, as well as the 346 snaps in quarters and three snaps with eight defensive backs on the field, all Hail Mary defenses in the last 11 seconds of halves. The responses to eight defensive backs? An interception by Jared Goff, a throwaway by Cam Newton and a meaningless 3-yard screen pass by Mitchell Trubisky.
- "Big" defenses are 4-4-3 or 3-5-3 lineups, while "Goal Line" includes all other personnel groups with fewer than four defensive backs. More than half of those defenses were used on the 1-yard line, but that wasn't a literal necessity; the Cardinals got marked with a goal line defense on third-and-15 from their opponent's 44-yard line.
You'll note that defenses still had a better DVOA in base than in nickel. This was true in 2017, as well, but that could be explained by nickel being the defense of choice against 11 personnel. 11 personnel was the most successful offensive personal grouping, and nickel performed better than base defenses did against it. So, even with a worse DVOA, it was still the better defense; it was being used more successfully in tougher situations. This has generally been the case throughout the 2010s.
That was not the case in 2018, however. While nickel was still the predominant answer to 11 personnel, being used 77 percent of the time to base's five percent, it performed worse overall. Nickel vs. 11 had a defensive DVOA of 3.4%; base put up -8.1%.
|Defensive Personnel versus Offensive Personnel|
|Defense||% Used||Def DVOA||% Used||Def DVOA||% Used||Def DVOA|
So, what's going on here?
Whenever you're faced with a one-year reversal in a stat, especially one that is as counterintuitive as less cornerbacks being better against more wide receivers, the first guess should always be that it's a one-year blip; that weird splits happen and normality will return in 2019. That may be true, and we'll keep an eye out for that next year. Base defenses are used so rarely against nickel that small sample sizes could be wreaking havoc -- 1,006 base-versus-11 snaps are a lot, but there are over 15,000 nickel-versus-11 snaps to look at. It's 300 fewer snaps than a year ago, so that sample size is only shrinking.
However, there are more substantive reasons why base could have ended up better than nickel. Those base-versus-11 snaps were more likely to happen earlier in the game, with the score close. They were more likely to be runs, and the pass attempts that did occur tended to be shorter. Opposing offenses gained fewer yards against base than they did against nickel, despite the DVOA difference. It's possible that what we're seeing isn't "base defenses are more effective against nickel," but "conservative play calling is less effective than bold play calling." That wouldn't explain why base defense was better in 2018 specifically, but it could be a general explanation for why the numbers might be seem counterintuitive; it could be situational rather than strategic.
Another potential explanation is that teams that can't cover 11 with their base defenses might simply not do it. That was not the case in 2017, when Cleveland had one-sixth of all base-versus-11 snaps and was terrible at it, as they were at most everything in that 0-16 season. In 2017, eight teams used base-versus-11 at least 10 percent of the time; five had a positive defensive DVOA and only one was lower than -6.0%. (Remember, positive DVOA means bad defnse.) In 2018, six teams used base-versus-11 at least 10 percent of the time; only three had a positive DVOA and two of them (Denver and Pittsburgh) were in the negative double-digits. Cleveland led the league in base-versus-11 snaps in 2017; they had a 17.1% DVOA doing so. Denver led the league in 2018; they had a -14.1% DVOA doing so. It's a biased sample -- if teams that are bad at doing something stop doing that thing, the split will look better.
Personally, I'd place my money on "weird, small sample size split," with a side of "the Hue Jackson Browns were terrible," but it's something we'll keep an eye on going forward.
Now, onto the team level stats, which as always requires a big honkin' table.
|Defensive Personnel Frequency|
Denver clocked in at just under 45 percent, but still led the league in base usage. That 45 percent mark is the lowest number for a league leader we have seen; last year, three teams were up over 50 percent. The Broncos will probably change significantly going forward with Vic Fangio calling the shots; you can see the Bears were nearly the exact opposite of the Broncos in terms of base-nickel splits.
That defensive musical chairs game continues, as ex-Broncos head coach Vance Joseph is heading down to Arizona, who finished dead-last in base defense a year ago. No team stuck in one personnel grouping as much as the Cardinals did last year, but we're likely to see a ton more of seven men in the box as Joseph tries to help Arizona's defense bounce back. Ex-Cardinals coach Steve Wilks moves to Cleveland, so they should see their nickel percentage rise dramatically in 2019. He replaces ex-Browns defensive coordinator Gregg Williams, who heads to the Jets, so you'd expect New York to use more base as they adopt Williams' scheme. That means ex-Jets coach Todd Bowles will take his scheme to Tampa Bay, so you'd expect the Buccos to run out more defensive backs in 2019. You might need a corkboard and some colored string to keep track of all this, but these defenses will look mostly schematically the same in 2019, just all wearing different jerseys.
There are six ways to order base, nickel, and dime set frequencies, and all six were represented last season. Twenty-three teams went nickel-base-dime, as that's now the league default. There's some variety in a group that large, of course.
One group tends to stick in nickel whenever possible, epitomized by the Cardinals, Saints and Bears. These are teams that tried to find ways for their fifth defensive back to pitch in in the box, rather than bringing in extra big bodies on a regular basis. They make some of the fewest personnel switches in the game; the Cardinals took that to the extreme by being in nickel a whopping 84 percent of the time. As a result, the Cardinals had three of the fifteen defensive backs with the most plays on running attempts in the league with Antoine Bethea, Budda Baker and Tre Boston.
There's another group, represented perhaps the most by the Colts, Texans and Seahawks, who swap defensive backs in and out on nearly a play-by-play basis to try to get themselves in the best possible matchups. These teams may use nickel the most, but it's much closer to an even split between the three personnel groupings; it requires more specialists who can fill specific roles as opposed to the previous category's need for more versatile all-around players.
And then you have the "brother, can you spare a dime+" group, with the Jaguars, Dolphins, and most extremely Panthers representing that subset. Most of those zeroes in the dime+ column are actually just fractions of a percentage point, but not Carolina. The Panthers only ran five plays outside of their base and nickel packages; all goal line plays on the 1-yard line. They play the Rams Week 1; you may remember from last week that the Rams had the least offensive personnel variation in the league. We may set a record for seeing the same basic personnel groupings over and over again right off the bat when Ron Rivera and Sean McVay lock horns.
While nickel-base-dime is the predominant defensive structure out there, it can be more interesting to look at the exceptions. Nickel-dime-base has been the Patriots way for quite some time, as the Patriots routinely spend among the most time in the league with five or six defensive backs out on the field. Belichick was ahead of the curve on this, trotting out five- and six-back groups back at the turn of the millennium to shut down the Greatest Show on Turf Rams; he may still be ahead of the curve as his teams use base defenses less and less. That Belichick way also has gone to Detroit (Matt Patricia, ex-Belichick defensive coordinator) and Philadelphia (Jim Schwartz, ex-Belichick assistant in Cleveland). This year, Baltimore also joins the nickel-dime-base crew; a wide departure from 2017's numbers now that Don Martindale has taken control. The other ordering used by multiple teams was the dime-nickel-base grouping used by Green Bay and Pittsburgh, generally involving three safeties to make their dime packages a little bigger and stand up more to regular play.
That leaves three odd ducks out there. We've already mentioned the Broncos; they basically ignored nickel packages entirely and either went with their base defense or dime package depending on the situation. As that will change in 2019, that might leave the Rams standing alone as the last team to have their base defense be their primary unit. Wade Phillips has been doing this for a long time, he's very successful at it, and doesn't need to change just because some crazy college kids are coming up with their wacky spread offenses and whatnot, thank you very much.
And then you have the Chargers, who played 64 percent of their snaps in dime+, blowing everyone else out of the water. They led the league in 2018, too, but at just 46 percent. Part of this was injury-related; the Chargers finished with the third-most adjusted games lost at linebacker and the tenth-most AGL on the defensive line. For most of the season, their best healthy defenders happened to be defensive backs, so you might as well cram them all onto the field at the same time. This peaked during the playoffs, where health issues meant that they basically ran nothing but dime all the time, shutting down Baltimore and then getting destroyed up front by New England.
It's not just injury-related, though. Gus Bradley has always liked fast defenses, though it was mostly nickel-related in Jacksonville. The Chargers' versatile defensive backs make it more viable, too. Derwin James can line up anywhere except nose tackle and have success, so Los Angeles can get different looks without changing personnel. Adrian Phillips plays linebacker when needed, which was early and often thanks to their injury issues. If more teams had versatile players like that, you'd see more dime packages around the league. That's why they're the one dime-base-nickel team; rather than take James or Phillips off for a linebacker, they simply moved one or both of them up when they wanted six men up front, only subbing people out when they wanted to go beyond that. You don't need a nickel package if your dime defenders can play multiple roles.
It's doubtful the Chargers will be up over 60 percent this season, as health should bounce back and a new talent injection into the front seven should keep things more balanced, but speed and versatility are the general watchwords of the modern edge of defensive football, and the Chargers look to remain at the forefront.
One final note to keep in mind. When we looked at offensive personnel, we saw the league homogenizing, with most teams not only using 11 personnel most frequently, but being most effective when using it. Defensive personnel groupings are much different. The top quartile of defenses last year came from all of these different groupings:
- the Bears were an all-nickel team;
- the Bills, Jaguars and Vikings were no-dime teams;
- the Ravens were dime-nickel-basers;
- the Broncos used base more than any other team;
- the Texans were one of the most balanced defenses in the league in terms of personnel;
- and the Chargers ran all-dime, all the time.
Defensive philosophy, at least in terms of what players you're running out onto the field, isn't as foundational to defensive success. Any system can work, as long as you have the players to run it.