Stat Analysis

Advanced analytics on player and team performance

2017 Defensive Personnel Analysis

by Bryan Knowles

Earlier this week, we looked at what personnel groupings were most popular on offense. Now it's time to turn around and look at things from a defensive standpoint. In a league where three wideouts has become not just the most popular formation but the default formation, how have defenses reacted?

Just as three-wide formations have become the offensive default, nickel defenses are now the NFL's primary defensive formation. Nickel first became more prevalent than base defenses in 2012, and became a majority of all plays in 2015. There's a reason the AP All-Pro team added an extra "defensive back" position in 2016; slot corners are now more likely to see the field than your seventh guy in your front seven.

Base defenses remain more effective on the whole, though that is misleading. Nickel packages are by a significant margin better at handling 11 personnel (0.8% DVOA for nickel, 8.4% DVOA for base sets), so it makes sense that that would become the most played defense. Base personnel are still better at handling things like two-tight end sets, resulting in better overall DVOA. Nickel is the best defense for defending the most popular offense, so it has become the most popular scheme


Defensive Personnel Groupings
Personnel 2016 Pct 2017 Pct Difference
2017 DVOA
Nickel 57.4% 52.3% -5.1%
1.1%
Base 29.7% 33.1% +3.4% -3.6%
Dime+ 11.0% 12.9% +1.9% 5.7%
Goal Line 1.2% 1.0% -0.2% -1.5%
Big 0.7% 0.7% 0.0% -0.9%

A couple quick notes here:

  • We no longer separate 3-4 and 4-3 fronts in our stats. In all honesty, the distinction is becoming more and more meaningless in the modern NFL; the difference between a 4-3 defensive end and a 3-4 outside linebacker is more or less whether or not they have their hand in the dirt at the snap. Hybrid defensive schemes are the name of the game now, and trying to cram 2018 defensive strategy into a 1980s framework is less than useful.
  • For the record, however, 55 percent of base snaps were 4-3 defenses, and 44 percent of them were 3-4 fronts. That doesn't add up to 100 because there were a handful of 2-5s, 5-2s, and 1-6s that popped up here and there. Carolina was the only defense to never stray from their front; they were a 4-3 defense and only went to a three-man line on 26 snaps all season long, all of them in nickel defenses. At the other extreme was New England, who freely flipped between 3-4 and 4-3 defenses when they could be bothered to be in base defense at all; more on that in a moment.
  • "Dime+" includes any package with more than five defensive backs. That includes all your dime packages, as well as the 314 snaps in quarters and six snaps with eight defensive backs on the field, all Hail Mary defenses in the last 14 seconds of halves. Half-dollar defense? Sacagawea? Paper currency?
  • "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 those defenses were used on the 1-yard line, but that wasn't a literal necessity; the Texans actually used it once in the other team's red zone.

As one might expect, the rise in nickel on defense is directly related to the rise in 11 personnel on offense. There's a 0.95 correlation between the two; it's a fairly obvious decision to put extra cornerbacks on the field when your opponents bring out the extra wideouts. When the frequency of 11 personnel dipped this season, so did the frequency of nickel.

That doesn't mean that every 11 set is matched with a nickel defense, of course. Last season, 74.2 percent of 11 sets were met with nickel. Dime+ packages were used 18.6 percent of the time, while base defenses actually ticked up by half a percentage point to 7.2 percent. Base defense use against 11 had been holding steady at about 6.5 percent for the previous three seasons. My hunch is that this a charting anomaly; that the rise of the hybrid linebacker/safety is being handled slightly differently by different charters. That's something to keep an eye on in future seasons.

On offense, every team used the same personnel grouping the most, with 11 being predominant everywhere. That was not true on defense, however; while nickel sets were the most predominant personnel groupings for 23 teams, there was variety throughout the league. There's really no way to talk about this without giving a massive data dump, so here's all 32 teams, with the percentage of time they spent in their base, nickel, and dime+ packages.


Defensive Personnel Frequency
Team Base Rk Nickel Rk Dime+ Rk
CLE 66% 1 31% 29 1% 27
LAR 58% 2 37% 25 2% 23
CAR 50% 3 49% 20 0% 32
IND 42% 4 44% 23 11% 14
DEN 41% 5 27% 32 30% 5
PIT 38% 6 36% 26 24% 7
HOU 38% 7 34% 28 24% 8
NYJ 37% 8 53% 18 8% 16
MIA 36% 9 60% 14 1% 25
SF 36% 10 45% 22 15% 10
TB 36% 11 62% 11 0% 30
OAK 35% 12 57% 16 7% 17
WAS 34% 13 63% 10 2% 24
JAX 33% 14 66% 6 0% 29
BUF 33% 15 66% 5 0% 31
BAL 30% 16 40% 24 28% 6
Team Base Rk Nickel Rk Dime+ Rk
CHI 30% 17 64% 9 5% 20
CIN 30% 18 67% 4 1% 26
TEN 30% 19 64% 8 5% 19
DET 29% 20 52% 19 15% 11
NYG 29% 21 60% 15 11% 13
SEA 29% 22 68% 3 3% 21
KC 27% 23 29% 30 42% 3
ARI 27% 24 65% 7 7% 18
PHI 27% 25 60% 13 10% 15
LAC 26% 26 27% 31 46% 1
DAL 25% 27 55% 17 20% 9
ATL 24% 28 72% 2 2% 22
NO 23% 29 61% 12 13% 12
GB 23% 30 34% 27 42% 2
MIN 21% 31 77% 1 0% 28
NE 15% 32 48% 21 37% 4
Avg 33% -- 52% -- 13% --

Using the phrase "base defense" to describe New England's front seven is sort of awkwardly inaccurate. Not only did they freely flip between three- and four-man fronts, they rarely even had a front seven to speak of. We've seen teams use base defense less frequently than this -- New England did so in 2016, as did Green Bay -- but the 2017 Patriots stand out as the only team to go nickel-dime-base, in descending order.

Joining the Patriots as teams that laugh at the phrase "base defense" were the Packers, Chiefs, and Chargers. Their base defenses were, in actuality, dime packages. In general, we're talking about three safeties and three corners, rather than going with four cornerbacks -- Green Bay called it their "Nitro" package -- but it's something that's growing in popularity around the league. When you're weak at linebacker depth, going with an extra safety like Morgan Burnett or Adrian Phillips in the box makes a lot of sense. The hybrid safety/linebacker position is only growing in popularity, which makes nickel formations all the more enticing.

Cleveland was the polar opposite to New England, in so many ways. The Browns stuck in their base defense 66 percent of the time, which insane in the modern NFL. You might think that this is what happens when you are constantly facing opponents running out the clock on your way to 0-16. Except that the Browns were doing this from the very start of the game! In the first half, the Browns were in base defense 66 percent of the time. When the score was tied or the Browns actually had a lead, the Browns were in base defense... the same number, 66 percent of the time.

Cleveland was the only team in the last four seasons to even hit 60 percent. This may go part of the way to explaining why they were fourth in rushing defense and 26th in passing defense; if you never go into nickel or dime packages, you're going to have trouble stopping receivers in the modern NFL! Cleveland had 245 snaps where their base defense was matched up against 11 personnel; only the Panthers and Rams joined them above 100. To put it another way, roughly one-sixth of all 11-versus-base snaps in the NFL were played by the Browns defense. At least they were unique!

The Rams were the only other team to use their base defense more than half the time, though Carolina missed out on joining them by just four plays. Carolina only had 10 plays not in either base formation or nickel, and nine of those were at their own 1- or 2-yard line. As mentioned earlier, every single one of Carolina's base sets was a 4-3 front, so no team in the league had as little defensive variety as Carolina. That sounds like my kind of defense, by which I mean the kind of defense I would literally design because I am neither good nor creative at designing defenses. For all that, Carolina ranked seventh in defensive DVOA, because when you have Luke Kuechly, Thomas Davis, and Shaq Thompson as your linebackers, you can get a lot of versatility without actually having to sub anyone.

Minnesota, Atlanta, Seattle, and Cincinnati were the least balanced defenses in the league, sitting in nickel more than two-thirds of the time. Contrast that with Houston, Baltimore, and Pittsburgh, who didn't use any defensive personnel package more than 40 percent of the time, freely flipping between base, nickel, and dime packages depending on what personnel the offense trotted out.

It's interesting that even with the homogenization we've seen in personnel packages on the offensive side of the ball, there are still so many different ways that defenses choose to handle it. It's not like there's an obvious right or wrong way to do it, either -- the Vikings, Ravens, and Rams were about as far from one another as you can possibly get, yet all three were in the top six defenses by DVOA.

It doesn't seem to matter whether you stick to what you're best at or substitute to match every single variation offenses put out there. Either can work -- as long as you have the talent to back it up.

Comments

21 comments, Last at 27 Jun 2018, 10:58pm

1 Re: 2017 Defensive Personnel Analysis

by Bryan Knowles // Jun 21, 2018 - 6:35pm

The Browns were all about that base, 'bout that base (no nickel)
All about that base, 'bout that base (no nickel)
All about that base, 'bout that base (no nickel)
All about that base, 'bout that base, base.

Yeah it's pretty clear, they'll stay in Cover-2
They simply do not care what that offense do
They make their LBs and slot receivers race
And use all the 4-3 in all the wrong places.

Yeah, my momma, she told me, make sure you cover the screen
'Cause defense don't matter when you go oh-and-sixteen
It don't matter if modern D tactics proclaim we're wrong
So if you want modern football then go ahead and move along.

Login or Register to post comments

2 Re: 2017 Defensive Personnel Analysis

by Vincent Verhei // Jun 21, 2018 - 8:54pm

You're fired. Clean out your virtual office.

Login or Register to post comments

3 Re: 2017 Defensive Personnel Analysis

by Bright Blue Shorts // Jun 22, 2018 - 4:10am

Lucky the Browns can offset their lack of defensive ingenuity by having a head coach who is an offensive genius ...

Login or Register to post comments

13 Re: 2017 Defensive Personnel Analysis

by Theo // Jun 24, 2018 - 11:58am

This is the base of me
And you’re never gonna ever run a yard on me, no!
This is the base of me
And you’re never gonna ever run a yard on me, no!
Throw your screens and your outs, throw your bombs and your posts
But I'm not gonna care cause,
This is the base of me
And you’re never gonna ever run a yard on me, no!

Login or Register to post comments

15 Re: 2017 Defensive Personnel Analysis

by TheIdealGrassLaw // Jun 24, 2018 - 1:38pm

Never gonna give base up,
Never gonna stop a first down,
Never gonna go around, and sack you
Never gonna make you try
Never gonna tip a pass high
Never coming from behind, won't pull through

Login or Register to post comments

20 Re: 2017 Defensive Personnel Analysis

by Broncos2015 ---- formerly Ninjalectual // Jun 27, 2018 - 8:20pm

No songs from Ace of Base?

Login or Register to post comments

21 Re: 2017 Defensive Personnel Analysis

by TheIdealGrassLaw // Jun 27, 2018 - 10:58pm

I'm sure there's a video somewhere on youtube of Ernie Adams "Seeing the sign."

Login or Register to post comments

4 Re: 2017 Defensive Personnel Analysis

by jtr // Jun 22, 2018 - 8:12am

Yeah, so it turns out leaving your 4-3 out there against modern 3-wide offenses is a good way to go 0-16. Kinda saw that coming. I wonder if that's by (flawed) design, or a matter of the team not being organized enough to match up their substitutions with the offenses. Or maybe they just don't give a damn, that would explain a lot too.

For Carolina, a lot of the reasoning behind the Shaq Thompson pick a few years ago was so they could keep flooding the field with linebackers even against modern offenses. All of the pre-draft chatter about Thompson was hand-wringing about how he's not quite a linebacker and not quite a safety, but what do you know, linebacker-safety hybrids are a great thing to have around in the modern NFL. It also helps that Carolina is primarily a zone defense team, so when you put Thompson in the slot, he's not really "covering" a WR in the conventional sense, that's just where he lines up before he drops into his zone.

Login or Register to post comments

5 Re: 2017 Defensive Personnel Analysis

by ssereb // Jun 22, 2018 - 1:15pm

I think the fact that only one cornerback from last season, Brian Boddy-Calhoun is still on Cleveland's roster goes a long way toward explaining why they were in a base defense so often. After they traded Joe Haden their most accomplished cornerback was...Jamar Taylor, I guess? They probably felt it was better to have their top linebackers on the field than the Couldn't Beat Out Jamar Taylor All-Stars.

Login or Register to post comments

6 Re: 2017 Defensive Personnel Analysis

by Aaron Brooks Good Twin // Jun 22, 2018 - 2:27pm

I have to imagine a lack of depth of talent plays in here.

You might as well play a good LB against 11 than a terrible CB. At least the LB can eventually tackle them.

Login or Register to post comments

7 Re: 2017 Defensive Personnel Analysis

by Bryan Knowles // Jun 22, 2018 - 3:22pm

You'd think that, wouldn't you?

But no.

Browns defensive DVOA, 11 personnel versus base set: 17.1%
Browns defensive DVOA, 11 personnel versus nickel: 10.0%

And if you prefer just their passing numbers:

Browns defensive DVOA, 11 personnel versus base set: 36.6%
Browns defensive DVOA, 11 personnel versus nickel: 13.0%

Cleveland's base set WAS better than their nickel set overall, so I see what you're saying. But the numbers indicate that, at least in Cleveland's case, it's better to put out worse players who can actually matchup with your opponents than better players in the wrong situations.

Login or Register to post comments

10 Re: 2017 Defensive Personnel Analysis

by jtr // Jun 23, 2018 - 8:40am

Do you guys have game by game charting stats for this? I'm curious as to whether these base-vs-11 snaps cluster into a few games or if they're spread across the season. If it's just a few games, then I guess it's just bad gameplanning; maybe they wanted a linebacker to match up with Pittsburgh's very physical slot receiver Juju Smith-Schuster for instance. If it's across a bunch of games, then I suspect it's a matter of general disorganization. It's not a trivial feat of organization to identify opposing personnel, get the correct 11 guys on the field, rotate the defensive line as needed, and communicate a formation and play call in the 10 seconds or so before the offense gets lined up. Most teams can do that reliably, but then again most teams win at least one damn game per season.

Login or Register to post comments

11 Re: 2017 Defensive Personnel Analysis

by Bright Blue Shorts // Jun 23, 2018 - 10:19am

I have told previously of my one experience of watching the Hue Jackson Browns.

- Confusion on the opening kickoff return with two guys not sure who was taking it.

- Calling timeout before their first offensive play because they had 12 men on the field.

And this was week 10 of the season so not like it should have been teething troubles.

Login or Register to post comments

14 Re: 2017 Defensive Personnel Analysis

by Theo // Jun 24, 2018 - 12:13pm

This very much sums it all up.
The Browns have had a GREAT and I mean stellar offseason - they added a real QB in Tyrod Taylor, a good runningback in Hyde, a real NFL receiver in Landry, Flash Gordon is back for a full season (since 2013, one might hope), they added a free safety in Randall and drafted Mayfield to play in 2019 or something happens with Taylor and drafted Ward who the vets already voted into the starting lineup.
But this could all be outdone by the stupidity of Hue and/or Gregg Williams.

Login or Register to post comments

16 Re: 2017 Defensive Personnel Analysis

by jtr // Jun 25, 2018 - 8:52am

I had forgotten that Gregg Williams was the defensive coordinator in Cleveland. It's weird that he was so reluctant to match up with the offense this past season, because he was such a big user of dime packages in LA. He was one of the early adopters of the dimebacker trend that's all over the NFL right now. It's weird to see the man who made Mark Barron an every-down "linebacker" go to another team and stick a bunch of linebackers on the field against slot receivers.

Login or Register to post comments

17 Re: 2017 Defensive Personnel Analysis

by Theo // Jun 25, 2018 - 11:56am

I wonder what personnel was ran against Cleveland - maybe they faced a lot less 11 personnel than NFL teams.

Login or Register to post comments

18 Re: 2017 Defensive Personnel Analysis

by Bryan Knowles // Jun 25, 2018 - 2:03pm

Depends on your definition of "a lot"

Cleveland faced 11 personnel about 51% of the time, third-lowest in the league. Is that a "lot less" than the league average of 59%? I could see that argument...

In one-score games, the league average was 56%; the Browns faced it 52%. This matters because the Browns were often losing, and so other teams would be looking to run out the clock on them. Is that still a "lot less"? I'm less sure.

Login or Register to post comments

19 Re: 2017 Defensive Personnel Analysis

by Theo // Jun 26, 2018 - 2:55am

I don't know how those one-score games were built.
I think some of the Browns' one score games were actually multiple score games, that were just one score games because the other team took the foot of the peddal or played their backups on offense (wk17). But I could be wrong.

Login or Register to post comments

12 Re: 2017 Defensive Personnel Analysis

by Bryan Knowles // Jun 23, 2018 - 2:30pm

They're pretty spread out, with a very slight bias towards doing it more at the end of the season.

Overall, they were in base against 11 47.5% of the time. 10 of their games they were somewhere between 40 and 60% -- and they got up to 65% against the Packers. I thought that last bit was because Green Bay likes to have a receiver as a running back and thus teams in general might be willing to use more base, but no; teams didn't go base v 11 versus the Packers more than anyone else (Arizona, in fact, was the team people did that the most against).

Of the five remaining games, the Minnesota and Chicago games were at 39%, so that's not substantially different from their window; that's just falling one play short. Jacksonville saw them at 31%, tiny by Cleveland standards (but again, remember, the league average was 7.2%!)

That leaves two other games. Weeks 2 and 3 against Baltimore and Indianapolis, where the Browns were at 21.7% and 12.9%, respectively. First guess why it was low there was Myles Garrett not playing, but he didn't play in Week 1 or Week 4, either. Jamie Collins suffered a concussion in Week 2 and missed Week 3, so that might explain it somewhat...except he missed Weeks 4 and 5 as well, and the Browns were back up to 56.8% and 60.0% in Weeks 4 and 5.

I think you're right in that it's bad gameplanning, I just don't think those bad plans were limited to a handful of games <_<

Login or Register to post comments

8 Re: 2017 Defensive Personnel Analysis

by Raiderfan // Jun 22, 2018 - 5:15pm

Your comment only makes sense if you left off minus signs. Did you, or am I just confused?

Login or Register to post comments

9 Re: 2017 Defensive Personnel Analysis

by Bryan Knowles // Jun 22, 2018 - 6:03pm

Remember, lower defensive DVOA is better!

EDIT: Unless you mean the last paragraph, which I could see being confusing.

Cleveland's base set had a better (lower) DVOA in general, looking at every play. However, when they faced 11 personnel, Cleveland's base set was worse than their nickel set. Yet, they still continued to avoid using their nickel more than any other team in that situation.

Login or Register to post comments