The top of the NFL, for the most part, stayed put this week in our EPA power rankings. No team moved into or fell out of the top-5. But just outside of that, there was quite a bit of chaos and movement with these teams, headlined by the resurgence of a powerhouse NFC team of last season. The Cowboys, on the other hand, continue to slip. As always, these rankings are based on expected points added (EPA), adjusted for strength of opponent.
Expected Points, the foundation of many analytical arguments, uses data from previous NFL seasons to determine how many points a team is likely to come away with on a given play based on down, distance, time remaining, and field position. The difference in expected points at the start of a play and expected points at the end is referred to as expected points added, or EPA.
Dallas slid two more spots this week after dropping eight spots last time. They’re now third in the division by this measure and entering into the realm of a New York team, though mercifully still nowhere near as bad as the Jets. Still, Dallas is going to have to get a pretty miraculous performance from Ben DiNucci (or whichever quarterback is starting next week) to avoid dropping to the worst adjusted EPA differential in the East.
The Bucs remain atop the list after a dominant win over a Raiders team that, despite having a statement win over the Chiefs just a week ago, has never really been more than average in terms of their EPA differential.
Then we get some real movement in our rankings, with our biggest risers in two California teams. The Chargers are suddenly in the top-10 as Justin Herbert continues to shine, and the 49ers appear to be back to their 2019 selves after handing Bill Belichick the worst home loss of his Patriots career.
These two teams have something else in common too, aside from a good move up in the power ranks. They both feature offenses that throw the ball at or behind the line of scrimmage on over a quarter of their dropbacks. EPA/play is a stable and relatively predictive measure of quarterback success, but it doesn’t take into account how many screen passes or jet sweep toss plays different teams run. For example, here’s a Jimmy Garoppolo “throw” that went for +1.3 EPA, despite Garoppolo targeting Deebo Samuel 5 yards behind the line of scrimmage:
Now, it’s not fair to punish quarterbacks for this by say, using the air EPA on the play of -1.2, but if we’re trying to evaluate quarterback talent it’s not necessarily fair to include these either. There’s also no easy way to define which throws came on plays like this and which were checkdowns or throwaways. But we can get a decent approximation with simple air yards, the distance the pass is thrown relative to the line of scrimmage. By separating throws at or behind the line of scrimmage from throws beyond it, we can add more even more context to each quarterback’s play through this year.
Herbert and Garoppolo trail only Russell Wilson in EPA/play on passes at or behind the line of scrimmage, but they also run these plays nearly twice as often as Wilson (~26% for Herbert/Garoppolo, 15% for Wilson). So while Herbert and Garoppolo rank 11th and 8th, respectively, in fumble adjusted EPA/play (per rbsdm.com), they look much more average when asked to throw beyond the line of scrimmage. Herbert only drops to 13th, but Garoppolo goes all the way down to 17th in the NFL in EPA/play when just looking at plays with positive air yards.
Aaron Rodgers is far ahead of the pack in EPA/play on passes beyond the line of scrimmage, but he’s also asked to do it less often than anyone else. Rodgers currently throws a pass at or behind the LOS on nearly 31% of his dropbacks, the only QB above 30% in the NFL. It’s actually quite scary to think what might happen if he keeps up this EPA/play and the Packers decide to air it out a bit more often.
Meanwhile in Dallas, there’s a pretty clear distinction between the offense with Dak Prescott and the offense with Andy Dalton. Dalton is second to Rodgers in the rate of passes behind the LOS (27.1%) while Dak ranked 14th (22.6%). There’s also the, well, more obvious difference in their EPA/play. Dalton has had the least success on short passes and the third-least success on longer passes. He hasn’t played two full games yet, so be cautious making complete judgements here, but these signs almost point to the idea that DiNucci can’t really be much worse, right? That’s my shot at hope.
The optimism isn’t lasting long this time, however, after looking at the adjusted EPA predictions for Week 8. These predictions like Dallas a bit more than the Vegas spread of 7.5, but I’m guessing the further we get from Dak the more in line these predictions will be with the rest of the crowd. Last week we went 10-4 with the adjusted EPA predictions, let’s see how we do this time.
[vertical-gallery id=656807][lawrence-newsletter]