This is one of those analytics terms that gets thrown around on NFL Twitter like everyone is supposed to already know what it means.
We had to look it up ourselves a couple of years ago when we kept seeing people argue about it in quote tweets so here is the plain English version.
The short version
EPA stands for Expected Points Added. It is a way of measuring whether a play in a football game actually made things better or worse for the offense.
Every situation on a football field has a historical expected point value based on the down, the distance, and where the ball is on the field.
When a play happens the situation changes and so does the expected point value. The difference between the before and after number is the EPA.
A 25 yard completion on second and eight from midfield has a positive EPA because it put the offense in a much better spot.
A sack on that same play has a negative EPA because it made things worse. That is really all there is to it.
Why you keep seeing it on social media
EPA blew up because it is really good at exposing the difference between quarterbacks who actually help their team win games and quarterbacks who just pile up stats in garbage time.
We have all watched a guy throw for 350 yards in a game his team lost by 20 and thought those numbers felt hollow.
EPA is the stat that proves it. His yards were real but most of them came in situations where they did not change the outcome and his EPA reflects that.
Does it matter for betting?
We started paying attention to EPA a couple of seasons ago and it has been one of the more useful stats for spotting NFL teams that are playing better or worse than their win loss record suggests.
A team sitting at 3-4 with strong EPA numbers on both sides of the ball is probably due for some positive regression and that is where you can find value before the sportsbooks adjust their lines.
