What if we had a new way to determine defensive quality in the NFL that would adjust for opposing quarterback quality?
It could provide a new answer on just how porous the Buffalo Bills defense really been through six games this season.
Introducing: Defensive Impact on Passing (DIP)
This new metric measures how every NFL defense impacts quarterback performance relative to how those quarterbacks perform against the rest of their opponents.
The idea with DIP is to provide a cleaner, more fair snapshot
of each defense’s effectiveness by accounting for the quality of the collection of quarterbacks it has faced throughout the season.
DIP adds vital context to EPA per dropback by calculating the difference in each quarterback’s performance against a specific defense compared to his baseline against other defenses.
For example, if a defense plays an elite-level quarterback averaging +0.348 EPA per dropback and allows an EPA per dropback of +0.212, it would actually be viewed as a “successful” outing for that defense because it surrendered a lower EPA per dropback than what was expected from that quarterback entering the game.
On the surface, allowing +0.212 EPA per dropback would be universally viewed as a bad outing for a defensive unit. But not with DIP.
The opposite would be true if a defense allowed a “low” EPA per dropback of +0.08 against a quarterback who otherwise managed an EPA per dropback well into the negatives.
And now that we’re six full weeks into the NFL season, somewhat of a baseline for each quarterback has been established, so we can measure how defenses are performing relative to expectation
Because DIP is measuring EPA per dropback and not EPA per individual quarterback matchup when looking at defenses collective DIP throughout a season, the formula here is pretty simple:
DIP = Total Actual EPA – Total Expected EPA / Total Qualifying DropBacks
Like all defensive EPA-based statistics, a lower (negative) number is better, and I’ve adjusted for garbage time, so the “qualifying” stipulation in the denominator equates to those dropbacks when the the win probability is between 10% and 90%.
Although I decided on the name DIP, the calculation is per dropback, so scrambles and sacks are counted as well. When determining the quarterback EPA per dropback baseline for a defense, his performance against that specific defense was excluded. Also, for a quarterback’s baseline to be counted, he needed to have at least 30 total dropbacks against other defenses to date this season.
Yes, a lot of explainers, I know. Kind of necessary for the introductory article.
Oh, and lastly, before I show you chart through Week 6, like all defensive EPA statistics, a negative number is better.
Some quick takeaways:
- Unsurprisingly, the Seahawks defense is still incredibly stingy.
- The Patriots defense has been sneaky good. Before their Week 7 game against the Bears, they were 14th in EPA per dropback allowed in non-garbage time but were in the No. 3 spot in DIP through Week 6.
- Strong showing for the Jaguars in DIP too, as they went from No. 17 in EPA per dropback rankings yet are seventh in this metric. Coordinator Anthony Campanille has done a fine job, and he has the horses, particularly in his front seven.
- The Panthers defense only has 64.2% of their defensive dropbacks faced actually counted in DIP because they faced Cooper Rush in Week 2, Shedeur Sanders for part of Week 3, and Tanner McKee in Week 6, and none of those quarterbacks have at least 30 dropbacks against other defenses to compare to.
- Philadelphia’s defense has been better than probably given credit for despite facing a relatively strong set of quarterbacks in its first six weeks of the season, including Jayden Daniels, Matthew Stafford, Trevor Lawrence, and Bryce Young.
- The Texans defense is still not to its 2025 level, but it’s playing better than the raw figures would suggest. Houston is 21st in DIP (as opposed to 26th in regular EPA per dropback).
- The most sizable discrepancy between regular EPA per dropback allowed and DIP belongs to the Saints. They’re 22nd in regular EPA per dropback and 9th in this new metric.
- Teams with a lower DIP than their actual EPA per dropback allowed through Week 6 include the 49ers (7th to 16th), Browns (5th to 15th), Ravens (11th to 19th), and Giants (12th to 17th).
- The Bills get a bump from 25th in EPA per dropback to 23rd in DIP. They have seen a rather daunting run of quarterbacks to begin the year including C.J. Stroud, Goff, Maye, Stafford and Kirk. And it’s not as if things are getting easier when Buffalo returns from its bye in Week 8 to welcome the Ravens to Orchard Park. While the end of November does feature a clash with Patrick Mahomes, they do have games against Malik Willis and Geno Smith sandwiched between those battles with former MVP quarterbacks.
