What if we had a new way to determine pass-game quality in the NFL that would adjust for opposing defense quality?
It could provide important context and eliminate overpraising of a quarterback — and his aerial attack — against a cake schedule or over-criticism of others who faced a challenging slate of defenses.
Introducing: Actual Passing vs. Expectation (APEX)
This new metric measures how every NFL pass game performs relative to how those defenses perform against the rest of their opponents.
The idea
with APEX (just like DIP, the defensive version of this metric) is to provide a cleaner, more fair snapshot of each pass game’s effectiveness by accounting for the quality of the collection of defenses it has faced throughout the season.
APEX adds vital context to EPA per dropback by calculating the difference in each pass game’s performance against a specific defense compared to its baseline against other pass games.
For example, if a quarterback plays an elite-level defense allowing -0.09 EPA per dropback and manages an EPA per dropback of +0.02, it would actually be viewed as a “successful” outing for that pass game because it registered a higher EPA per dropback than what was expected against that defense entering the game.
On the surface, a pass game’s EPA per dropback of +0.02 EPA per dropback would be universally viewed as a blah outing for an aerial attack unit. But not with APEX.
The opposite would be true if a pass game only managed a “low” EPA per dropback of, say, +0.13 against a defense that otherwise was surrendering an EPA per dropback much further into the positives.
And now that we’re seven full weeks into the NFL season, somewhat of a baseline for each defense has been established, so we can measure how pass games are performing relative to expectation.
Because APEX is measuring EPA per dropback and not EPA per individual defensive matchup when looking at pass games collective APEX throughout a season, the formula here is pretty simple:
APEX = (Total Actual EPA – Total Expected EPA) / Total Qualifying Dropbacks
Like all offensive EPA-based statistics, a higher (positive) 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%.
The calculation here is per dropback, so scrambles and sacks are counted as well. When determining the EPA per dropback allowed baseline for a defense, the pass game’s specific performance against that defense was excluded. While designed runs are excluded, scrambles and sacks (of course) count. Spikes and kneeldowns don’t.
Yes, a lot of explainers, I know. Kind of necessary for the introductory article, an all of this is similar to (but the inverse of) DIP.
Here is APEX through Week 7:
Some quick takeaways:
- The Bills sit at the top of these rankings, which isn’t incredibly surprising, but does speak to the legitimacy of Buffalo’s offense through six games. It’s been outstanding relative to competition, both easy and challenging.
- Miami’s offense is actually not performing brutally compared to the defenses they’ve faced. Being 18th in APEX doesn’t exactly align with the team’s 1-5 record.
- While C.J Stroud has had a fine start to a critical 2026 campaign, Houston’s APEX is 14th, compared to being 9th in regular EPA per dropback through seven weeks.
- Jacksonville’s 7th-place spot in APEX is two spots lower than their regular EPA per dropback.
- The most sizable jump from EPA per dropback to APEX belongs to the Jacoby Brissett Cardinals. They’re 26th in EPA per dropback and 21st in APEX. Games against the Seahawks, 49ers, and Broncos have hurt their regular pass game performance.
