Pass Pressure vs. Coverage Ability: A Statistical Analysis

[thebandit27]

As some of you know, every off-season I put together one mega-post that endeavors to delve deeply into previous season’s analytical data in order to provoke discussion. This year, I decided to focus on an age-old football question:

 

Is it more important to be able to pressure the QB or cover pass-catchers?

 

And boy is this one a doozy. It’s very, very long (hey, I’ve been in need of distractions lately), so read at your own risk.

 

While we all have our opinions (and I have my own), this post won’t include any. It is simply an unemotional analysis of the data at hand. To aid me in this undertaking, I relied on the following advanced statistics from Football Outsiders, which examines the performance of every team’s defense in all plays on which they provided pressure versus those on which they did not:

 

http://www.footballoutsiders.com/stat-analysis/2013/2012-pressure-plays-defense

 

Read the article for the details, then come back for some exhausting number-crunching (go ahead, I’ll wait).

 

Glad you’re back; let’s get to it…

 

Teams heavily reliant on DBs for pass defense (defined by at least an 8-spot improvement, which denotes 25%):

 

Team (# spots improved)

STL (8)

NYJ (24)

ATL (12)

WAS (9)

SF (8)

NO (8)

TB (14)

 

Teams heavily reliant on pass rush for pass defense (again, defined by at least an 8-spot improvement, which denotes 25%):

 

Team (# spots improved)

DEN (10)

NYG (10)

MIN (14)

DAL (10)

OAK (8)

BUF (23)

NE (13)

CAR (8)

CLE (19)

 

Teams that are good either way (defined as a top 8 ranking under both scenarios, so as to represent top 25% of teams in both scenarios):

 

Team (rank w/ pressure, rank w/o pressure)

ARI (2,3)

HOU (8,6)

CHI (1,1)

SEA (6,2)

 

Teams that can’t get it right either way (defined as a bottom 8 ranking under both scenarios, so as to represent bottom 25% of teams in both scenarios):

 

Team (rank w/ pressure, rank w/o pressure)

KC (24,26)

JAX (27,28)

DET (28,25)

PHI (32,32)

 

Now, some analysis…

 

1) I’ve noticed that there isn’t a strong correlation between # of plays producing pressure and a high ranking in both categories, nor is the converse true (as you’ll notice teams like Detroit and Philadelphia didn’t struggle to produce pressure at all). This is a significant finding, to the point that simply generating a pass rush isn’t enough; a team simply has to be good on the back end as well in order to have a successful pass defense.

 

2) There is, however, a very strong correlation between raking high in both categories and a high ranking in overall defense, as all of Arizona, Houston, Chicago, and Seattle ranked in the top 12 in the NFL in yards/game, yards/play, and points/game allowed (the one exception is Ari, which ranked a bit lower in points/game allowed). The same can be said for the converse, as all of Kansas City, Jacksonville, Detroit, and Philadelphia ranked in the bottom 7 in the NFL in the same 3 categories (save for Phi & Det ranking a bit higher in yards/game allowed).

 

So what’s the point? Well, there’s 2 of them I’d like to focus on:

 

1) For the hometown crowd: As evidenced by the numbers, Buffalo’s pass rush needs to be more consistent. Given that the team ranked 23^rd in applying pressure and had the 7^th-best pass defense when applying pressure, that one is obvious.

 

2) Answering the question posed at the onset of the topic: is it more important to apply pressure or to be able to pass defend with DBs when the pass rush fails? Here’s what I’ve noticed:

 

The teams that ranked in the top 12 in both yards/play and points/game allowed (which, in statistical terms, represent the most effective defensive units on both a per-play and per-game basis) in 2012 were as follows:

 

Team (YPPA rank, PPGA rank)

SEA (6,1)

DEN (1,4)

PIT (2,6)

SF (3,2)

CHI (4,3)

CIN (5,8)

HOU (8,9T)

BAL (10,12T)

 

You’ll notice that 3 out of the 4 teams that ranked highly both in pressure and non-pressure creating scenarios are on the list, so in order to normalize the data to answer the question at hand, we have to eliminate them, and shorten the list to the following teams:

 

Team (YPPA rank, PPGA rank)

DEN (1,4)

PIT (2,6)

SF (3,2)

CIN (5,8)

BAL (10,12T)

 

If you’ll notice, only one of these teams, Denver, fielded a defense that was measurably more effective under pressure-creating situations than in non-pressure situations. For the others, the difference was either negligible (in Pittsburgh’s case), or measurable in the other direction (more effective in non-pressure situations).

 

If we’re going purely by the numbers, it appears that it’s more important to have a defense that can cover well in the event that pressure isn’t created than it is to focus on creating pressure. However, there could be a number of variables that skew this analysis. Therefore, I’m now opening the floor to discussion. Please feel free to provide feedback as to the topic at hand, as well as how it relates to our home-town team.

 

Thank you for reading.

[MDH]

What about turnovers? Is there a correlation to pressure or coverage and having more turnovers?

[uncle flap]

Nicely done. Thanks for the effort.

 

I think it goes without saying that pressure helps the coverage. If you've got good coverage, you may not need the pressure.

 

There are a few variables that don't seem to be included in the analysis. I imagine down and distance play a big role, as well as the schemes and the tendencies of each team on both sides of the ball.

 

I think the conclusion is correct that good coverage trumps all, but I'm guessing that if you *know* you've got good coverage, you don't need to generate as much pressure.

 

The DCs philosophy likely plays a big role in this too. If he is confident in his DBs, he may feel more comfortable sending extra blitzers and leaving his guys in single coverage. He may just as likely feel comfortable not emphasizing attacking the QB, because he doesn't think his secondary needs the boost from pressure, and lets the DBs and LBs sit back in zones.

 

I find it interesting there is no mention about man vs zone in the article, nor any correlation to run defense. If a team is stout against the run, the odds increase in their favor defending the longer passes required to convert 3rd downs.

 

As far as implications regarding the Bills, I'd say I'm on board with Pettine's philosophy of pressuring (as well as creating the threat and/or illusion of pressure) will help the secondary. Coverage trumps all but I assume pressure helps a weak secondary more than it would a secondary that is strong on its own. Perhaps Wannstedt's biggest downfall is that he thought his players were talented enough to simply win each individual battle. But like a house of cards, it only takes one screw up to dismantle his entire scheme.

 

An attacking defense should help out some of the DBs to an extent, but I'll certainly agree with the stats here: Pressure is not a substitute for coverage.

Edited August 1, 2013 by uncle flap

[thebandit27]

What about turnovers? Is there a correlation to pressure or coverage and having more turnovers?

 

Great question...I'll have to add that in when I get home.

 

Nicely done. Thanks for the effort.

 

I think it goes without saying that pressure helps the coverage. If you've got good coverage, you may not need the pressure.

 

There are a few variables that don't seem to be included in the analysis. I imagine down and distance play a big role, as well as the schemes and the tendencies of each team on both sides of the ball.

 

I think the conclusion is correct that good coverage trumps all, but I'm guessing that if you *know* you've got good coverage, you don't need to generate as much pressure.

 

The DCs philosophy likely plays a big role in this too. If he is confident in his DBs, he may feel more comfortable sending extra blitzers and leaving his guys in single coverage. He may just as likely feel comfortable not emphasizing attacking the QB, because he doesn't think his secondary needs the boost from pressure, and lets the DBs and LBs sit back in zones.

 

I find it interesting there is no mention about man vs zone in the article, nor any correlation to run defense. If a team is stout against the run, the odds increase in their favor defending the longer passes required to convert 3rd downs.

 

As far as implications regarding the Bills, I'd say I'm on board with Pettine's philosophy of pressuring (as well as creating the threat and/or illusion of pressure) will help the secondary. Coverage trumps all but I assume pressure helps a weak secondary more than it would a secondary that is strong on its own. Perhaps Wannstedt's biggest downfall is that he thought his players were talented enough to simply win each individual battle. But like a house of cards, it only takes one screw up to dismantle his entire scheme.

 

An attacking defense should help out some of the DBs to an extent, but I'll certainly agree with the stats here: Pressure is not a substitute for coverage.

 

Very good points.

 

One other variable that wasn't included was whether pressure was generated via a standard rush scheme o a blitz. Obviously, the extra rushers would compromise men in coverage, so it would be interesting to see how that bore out.

[thebandit27]

What about turnovers? Is there a correlation to pressure or coverage and having more turnovers?

 

Okay, getting back to this...

 

The top 8 teams in takeaways for 2013 were (most at the top):

 

Chicago

New England

NYG

Washington

Atlanta

Arizona

Cincinnati

Houston and Cleveland (tie)

 

Only Cinci, Altanta, and NYG ranked in the top 9 (since there are 9 teams listed above) in pressure plays created. Conversely, only 3 of these teams ranked highly in pass defense on plays that were not pressure plays (Chi, Ariz, and Houston).

 

So it appears that TOs are more a reflection of either (a) being good in both pressure and non-pressure scenarios, as is the case for Chi, Ariz, and Houston (which IMO is a bit of a "duh" statement), or (b) being one of the teams that sees a marked improvement under one of the two scenarios, as is the case for NE, NYG, Atl, Was, and Cleveland.

[bowery4]

So, you're saying we're screwed again

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Just kidding, great job!

[ddaryl]

I guess my next question is which teams did better in run defense ?

 

http://www.footballo...m/stats/teamdef

 

 

(***Having some trouble with the info on the link I posted as they use 2011 weighted data vs 2012 non weighted data, and I am not 100% sure what is what***)

 

Only 1 teams in the top 10 were not in the top half (16) of the rushing D

 

HOWEVER, the top 10 Pass D teams were in non sequential order the top 10 D teams which supports your point well

 

I figure your are trying to make an argument on what works best for a team defense, a pass rush or a top D Backfield, but how does it affect Run D if a team is heavy on the DL/LB or DB's. We can debate all day on what is better against the pass, but if you are vulnerable to the run then you are still vulnerable and are still going to lose games. Case in point Buffalo

 

Buffalo's run D was so horrible last year that teams didn't have to pass on us, and some teams ran up the score so quickly on us they just sat back and ate up the clock. We were rated #31 in run D

 

We have solid DB's and Solid DL IMO, what we didn't have was Solid LB's and a D scheme that worked consistently...

 

IMO the teams that are balanced and capable of generating some consistent pressure, have decent DB's with an ALL-Pro mixed in and can stop the run which requires a stout DL and LB core are the ones that went far. That thought is basically as obvious as it gets.

 

if you go heavy on the DB you are probably getting burned on the run game, if you go heavy on the DL or relying heavy on the blitz you expose the DB's and then other teams eventually break a long play or 3

 

 

 

 

 

I still stand by the old saying that the game is won and loss in the trenches

 

 

my thoughts were a little scattered and I am still digesting your thoughts

Edited August 2, 2013 by ddaryl

[Dorkington]

Very interesting read, thanks for this!