One particular defensive position has the greatest impact on NFL run defense
Which defensive position has the greatest impact on run defense in the NFL?
This question arose in my mind when thinking about the New York Jets‘ defensive roster and its distinct strengths and weaknesses in the run game.
New York has some elite run-stoppers at cornerback and safety but is lacking in this category at the defensive tackle and linebacker positions. Can the Jets be a good run-defending team with this kind of talent distribution?
Of course, basic logic would lead us to believe that defensive tackles and linebackers are far more important in the run game than cornerbacks and safeties, making it difficult for the Jets to succeed at stopping the run with such a roster. But I wanted to get a concrete answer on the exact value that each defensive position group carries in the run game.
Which position group is the most important in succeeding at stopping the run? Which positions are not all that important? What is the disparity in value between the most important positions and the least important ones?
I crunched the numbers to answer these questions.
Methodology
My plan was to find the correlation between positional run-defense performance and team run-defense performance in the 2021 NFL regular season. Which individual position’s success had the greatest correlation with team success?
To quantify team success, I used Football Outsiders’ rush defense DVOA, an efficiency metric that calculates the quality of a team’s run defense by accounting for factors like down, distance, game situation, opponent quality, and more.
To quantify the run-defending success of each position, I took the Pro Football Focus run-defense grade (a metric that grades an individual player’s performance against the run through the analysis of their film) of every defensive player in the league and then calculated the cumulative PFF run-defense grades for all five position groups on each of the league’s 32 teams.
In other words, I calculated the cumulative PFF run-defense grade for these five units on every team:
- Interior defensive linemen (IDL)
- Edge rushers (EDGE)
- Cornerbacks (CB)
- Linebackers (LB)
- Safeties (S)
After calculating the run-defending strength of every positional unit in the league, I then went position-by-position finding the league-wide correlation between rush defense DVOA and cumulative PFF run-defense grade.
In other words: What is the average correlation between a team’s rush defense DVOA and the PFF run-defense grade accumulated by its IDL unit? What is the correlation between rush defense DVOA and PFF run-defense grade by the EDGE unit? And so on.
The results revealed that one position stood above the rest when it came to importance in the run game.
The results
Here are the correlation coefficients (r=) for the relationship between each position’s individual run defense (cumulative PFF run-defense grade) and team run defense quality (rush defense DVOA):
- LB: r= -0.614
- IDL: r= -0.433
- CB: r= -0.348
- S: r= -0.305
- EDGE: r= -0.273
For reference, a perfect negative correlation is -1.000 and zero correlation is 0.000. Keep in mind that a lower DVOA is better (less production allowed), so that’s why these correlations are negative (i.e. as run defense grades increase, DVOA goes down).
With a fairly strong correlation coefficient of -0.614, it is the linebacker position whose individual success correlated the most with team success – and by a wide margin, too. The difference between LB and the second-ranked IDL position was greater than the difference between IDL and the fifth-ranked EDGE position.
LB and IDL coming in as the two most correlative positions is no surprise, but I was certainly surprised by the wide disparity between them. It is a blaring signal that the linebacker position is the sport’s most important when it comes to run defense.
I was also taken aback by the cornerback and safety positions outpacing the edge rusher position. Perhaps defensive backs have a greater hand in run support than we might think. After all, while they don’t get involved often, their role tends to be extremely important when the ball does find them. They are often the difference between a nice 15-yard chunk and a back-breaking 80-yard touchdown.
This is good news for the Jets. Their newfound run-stopping talent in the secondary (in the form of D.J. Reed, Jordan Whitehead, and possibly Sauce Gardner pending the speed of his development) might actually boost their run defense quite significantly rather than being merely a nice tack-on to the resumes of those players.
But the main takeaway of this research is undoubtedly the monumental importance of the linebacker position. It’s here where teams’ fates in the run game are decided.
If your LB unit is good against the run, your run defense will probably be good. If your LB unit is bad against the run, your run defense will probably be bad. It’s that simple.
Among the top-9 teams in PFF run-defense grade at the LB position, 8 of them finished top-11 in overall rush defense DVOA:
- 1. Seattle Seahawks (72.9 grade) – Finished 8th in rush defense
- 2. Indianapolis Colts (69.7 grade) – Finished 3rd in rush defense
- 3. New Orleans Saints (67.3 grade) – Finished 1st in rush defense
- 4. New England Patriots (66.7 grade) – Finished 10th in rush defense
- 5. Buffalo Bills (63.9 grade) – Finished 11th in rush defense
- 6. Green Bay Packers (63.1 grade) – Finished 28th in rush defense
- 7. Las Vegas Raiders (62.6 grade) – Finished 9th in rush defense
- 8. San Francisco 49ers (62.0 grade) – Finished 2nd in rush defense
- 9. Baltimore Ravens (60.5 grade) – Finished 4th in rush defense
Among the bottom-9 teams in PFF run-defense grade at the LB position, all of them finished bottom-14 in overall rush defense DVOA:
- 24. Detroit Lions (46.1 grade) – Finished 31st in rush defense
- 25. Jacksonville Jaguars (44.8 grade) – Finished 19th in rush defense
- 26. Minnesota Vikings (43.0 grade) – Finished 25th in rush defense
- 27. Houston Texans (42.1 grade) – Finished 22nd in rush defense
- 28. New York Giants (41.7 grade) – Finished 32nd in rush defense
- 29. Pittsburgh Steelers (41.2 grade) – Finished 27th in rush defense
- 30. New York Jets (40.0 grade) – Finished 26th in rush defense
- 31. Atlanta Falcons (39.3 grade) – Finished 29th in rush defense
- 32. Chicago Bears (32.7 grade) – Finished 24th in rush defense
These numbers illustrate why the correlation coefficient at LB was so strong.
It’s important to remember that correlation doesn’t always equal causation. With that being said, all of this data still makes for an interesting addition to the discussion about positional value in the defensive run game.
Regardless of what happens at other positions, the fate of the Jets’ run defense rests in the hands of the LB unit: C.J. Mosley, Quincy Williams, Jamien Sherwood, Hamsah Nasirildeen, Marcell Harris, Del’Shawn Phillips, DQ Thomas, and Javin White.
Can Mosley stay effective into his thirties? Can the Jets’ coaching staff get at least one of their many young-and-athletic linebackers to break out?
If the answers to these questions are “yes”, the Jets had better hope that they find out quickly. Two of last year’s top-five rushing offenses await New York within the first two weeks of the season: the third-ranked Ravens and the fourth-ranked Browns.
Great read, Michael. Thanks! If not Kwon Alexander, let’s hope a quality veteran LB gets cut loose and the Jets pick him up.
Would be interested to see stats on 1) Yards before contact, and 2) percentage of 15+yd plays.
I just did this for 15+ yard runs. There’s very little correlation for every position except S, where the correlation is actually pretty strong. Lines up with what you’d expect. Safeties are responsible for preventing those breakaway runs.
If the ball carrier gets past the lineman consistently and the linebacker is cleaning up the plays than the lineman is not doing his job. Time to get a player that will clog the middle and draw double teams! Can’t wait! Just extend THIS season!
Ah yes, but what about the idea of the IDL “keeping the linebackers clean?” When evaluating a play, how does PFF consider a player’s job on that play if he’s got a guard in his face as opposed to not? That is to say, if a hypothetical IDL stands up a double team and a LB blows up a run, does that LB get the same grade if the IDL lets the guard slip off him and the guard forces that LB to try and bring down the RB with one arm? These correlations are based on a lot of faith of PFF scoring, which is the tool you have to use, but I think some of the findings might be susceptible to the same dynamic that makes a LB the leading tackler on every team, and not an IDL.
Right this is definitely the primary flaw that this analysis could have. What if the LBs on good run-defending teams just look better (and thus grade better) because the DL in front of them keeps them clean and makes their jobs easier? There’s certainly credence to this idea. Which is why correlation and causation don’t necessarily go hand in hand here.
The point of a grading system like PFF is to account for these variables – grading a player as an individual regardless of how good or bad his teammates are – but I’m sure there is some bias toward players who put in better positions to succeed. Naturally, they’ll screw up less often and make more plays. If the grading system is doing its job, it’ll raise the LB’s expectations in situations where he is set up well and lower them in situations where he is put in a horrible spot, but I honestly don’t know for sure how much they account for that.