Advanced Stats 101: EPA and the resurgence of the Packers’ offense

The Green Bay Packers offense has officially roared to life after another impressive showing on Sunday against the Detroit Lions.

You’ve likely noticed Pro Football Focus has Aaron Rodgers as their top graded player in the league and everyone is abuzz about Aaron Jones’ incredible performance on the ground and in the passing game. It’s a great time to be a Packers fan even this early in the season and we’re going to do something a little uncommon in this week’s look at an advanced statistic: we’re going to use it to praise Green Bay instead of being a buzzkill.

That’s right, analytics say nice things too! You wouldn’t know it by how some dismiss advanced stats, but like we said last week in our crash course on DVOA, they really aren’t all that bad and can even be complimentary of a team instead of critical. To do that, this week we will be looking at Expected Points Added or EPA. No, this has nothing to do with the Environmental Protection Agency.

What is EPA? The acronym stands for “Expected Points Added” and simply put, it measures the value of individual plays, expressed in points. It’s important to know that by points we aren’t talking about specific touchdowns, field goals or safeties.

To calculate the “expected points,” a lot of factors are considered. Down, distance and field position are the main ingredients in calculating EPA, which can be expressed as either a positive or negative number.

What play results can affect EPA, either positively or negatively? Well, a five-yard gain on 2nd and 3 is a much better result than the same gain on 2nd and 10. The former moved the chains while the latter sets up a third down with a moderate distance to go. Therefore the EPA on the first scenario would be higher than the second.

The stat of course grades plays that move the chains higher than ones that don’t. That brings us to the Packers and an area where they’ve greatly improved from last season.

Fans will recall that Green Bay struggled mightily on third down last year with a 37.6% conversion rate, ranking 19th in the league and the lowest of any playoff team. There was relative success on first and second down last year but those downs are not as weighted in grading as highly as third down, often a “do or die” down where the drive is either extended or ended.

Third down is also quite often a passing down, which is where the EPA grade on a quarterback comes into play. ESPN had Rodgers ranked 13th in passing EPA for 2019. This was a common statistic cited that pointed to where Rodgers’ play was suffering, even if it caused many fans to roll their eyes.

Through two games in 2020, Rodgers is currently ranked third in EPA and the Packers as a team have improved in third down conversion percentage at 52.17%, good for sixth in the league.

Coincidence? Not at all.

When weighing all aspects of the offense, Ben Baldwin of The Athletic says Green Bay currently ranks number one overall in EPA. If you’ve watched the first two games of the season, this shouldn’t surprise you. The Packers offense has been a model of efficiency, early red zone struggles against the Vikings aside. Rodgers has also been attacking the middle of the field more, which has increased his yards per attempt (Y/A) to 8.16, versus 7.03 a year ago.

Imagine that! Film and an advanced stats line up!

EPA can be applied to the defense as well and of course it’s the inverse of how it is calculated on offense. Judging by what we have seen of the Packers defense so far, it’s been very much a mixed bag. Opportune turnovers have come but there have also been times the defense has been suspect.

In last week’s discussion, we learned DVOA was not always the best measure of garbage time, and EPA isn’t either. It bases the values of a down/distance/field position situation on neutral game situations — namely, when the point differential is 10 points or fewer in the first and third quarters. How much points one would practically expect from an offense on 1st-and-10 from the 50 might be different in those situations than when a team is down three scores early in the fourth quarter, but EPA treats every play, even late in games, as if it were happening early on.

That means the 24 point hemorrhage that the Packers’ defense gave up in the fourth quarter against the Vikings counts just as much as it would have if they gave up those points early in the game.

Last season the Packers ranked 17th in overall EPA. Through the first two games of this season, they’re actually worse at 20th. Besides the late-game struggles in week one, the other big reason for the fall is that they’ve gotten worse on third down (16th in EPA on third down this year thus far compared to 9th last year). The Packers are of course struggling at times to get off the field if you have watched any games this year. Look at that! The film and stats line up again!

To summarize, EPA is a great tool for measuring efficiency, particularly at quarterback and on defense. As you can see, Rodgers is looking much better and the defense is looking shaky. In other words, it might be time to party like it’s 2011 again.

Hopefully this makes EPA look a little less scary and should the Packers hit a speed bump down the road, this should be one of the first stats you look at to explain why.