What’s more valuable: a five-yard gain on 1st-and-10 from your 20 yard line or a five yard gain on 1st-and-10 from your opponents’ 20 yard line? On the surface, you could argue that they’re the same – 5 yards is 5 yards. But what if I told you there was a different way to look at their value?
This is where the concept of Expected Points (EP) and Expected Points Added (EPA) come into play. If you’re unfamiliar, Expected Points is a metric that captures the value of a given down, yards to first down, and yard line combination (also check out my post further explaining this topic and the EPA Calculator API). The metric is generated by taking the average of the “next score” for each of these combinations. Generally speaking, the closer you get to your opponents’ end zone, the higher the EP! Here’s a chart that illustrates this point:
EPA is the change in EP from the beginning of a play and the resulting situation after the play has concluded. For example, if the initial Expected Points value was 0.5 and the play resulted in an updated Expected Points of 0.65, the Expected Points Added is +0.15.
For the remainder of this post, I will focus on EPA and how it can be used to evaluate teams! EP and EPA are also great metrics that help evaluate going for a 4th down conversion, but we’ll save that for a future post.
For the 2019 season, I took play-by-play data (NFL Savant) and applied EP values to generate EPA for each play! I then aggregated EPA for each team and compared the differential between the EPA gained by a team’s offense and the EPA given up by a team’s defense. As you can imagine, a higher total differential means that a team both did a good job with gaining EPA and preventing EPA. The opposite could be said about teams that did poorly in this exercise. Looking at the graph below, it’s not a surprise to see teams like the Baltimore Ravens and San Francisco 49ers at the top of the list. Interestingly, the Super Bowl Champion Kansas City Chiefs are in the middle of the pack, which can be attributed to their defense’s lack of ability at preventing EPA.
Seeing what the differential “Power Rankings” are, here are the offensive and defensive EPA rankings:
(A negative, or smaller number, is desired)
With the charts provided and an understanding of what EPA is, you may be wondering how it compares to traditional metrics, like overall offensive or defensive ratings (which is based off total yards gained or given up), or newer metrics like DVOA (measures a team’s efficiency by comparing success on every single play to a league average based on situation and opponent).
As you can see, EPA ranking does not translate quite completely for the team’s offense or defensive ranking. But it does line up in some cases with either the team ranking or their DVOA/Offensive DVOA ranking.
Long story short, I believe that EPA is a great tool in the football analysis toolbox. It doesn’t explain everything, but it does provide value – especially when making in-game strategic decisions. If there is a 2020 NFL season this year, I will dive into examples where EPA can help push a team to go for a 4th down conversion or play it safe and punt the ball (boring…).
Thanks again and as always, leave a comment or question! AND PLEASE CHECK OUT THE EPA CALCULATOR FOR YOURSELF!