It is basically decided at this point that the NFL will have few or no fans in the stands barring an absolute miracle for the 2020 season. The Green Bay Packers confirmed that there will be no fans at their first two games, and in my opinion, it’s unlikely we ever see a volume of fans at any stadium this year that reach a meaningful number. Even the Packers talked about 10-12 thousand at max, but that is, at best, 15% capacity.
To give you an idea of what 10-12 thousand looks like, that’s a little under the listed capacity for the University of Wisconsin-Whitewater’s football stadium, Perkins Stadium, which is 13,500. I attended UWW for 1.5 years and let me tell you, the noise experience of Perkins is a massive gap to the noise experience at Lambeau or other NFL venues.
With a quieter environment, the sounds from the field will be heard clearly by everyone, especially those within the box. I expect this to lead to an environment of higher offense. The reason for this is primarily that communication of checks and audibles will be much easier in this environment. Teams will also not have to use silent counts, which eliminate the snap count advantage that the offense holds over the defense. These will be pretty uniform across the NFL.
There is one aspect of this that will not be uniform, and that is the free play. The free play is an art form and it is not evenly distributed. We all know that Aaron Rodgers excels at this, and he even mentioned the possibility for more free plays on his Zoom call with the media earlier this week. The numbers I will be using in this article come from NFL Operations, which has data on this since 2006. Note, the data is from November of 2019, so it is not fully up-to-date, but it is the most recent available data tracking free plays. The EPA and total dropback numbers are from the NFL’s publicly available play-by-play data.
Head and shoulders above all others
Rodgers is in a complete league of his own when it comes to free plays. He has thrown for nearly 2,000 air yards, virtually all of that coming since 2010. The only player within 1,000 yards of him is Ben Roethlisberger, and no one is within 900 air yards. Aaron is the Babe Ruth of this very specific skill.
On a per-season basis, Aaron has averaged about 225 air yards on free plays, and he takes full advantage of those throws, averaging about 25 air yards per attempt. Not only does he have an insanely high volume of these plays, but he also presses the advantage these opportunities present when he gets them.
For the 2020 season, I would expect these to go up across the entire league. Even quarterbacks with rather meek hard counts will probably generate an extra jump or two, but I hypothesize that Rodgers should be able to increase his volume substantially given his elite vocal chords.
How much would more free plays improve Rodgers’ overall efficiency?
There is no way to really predict what free plays will look like in 2020, but my goal here is to play out one scenario to show how big (or small) the impact of this could be. Ultimately, Rodgers averages about ten free plays per season. I think it is fair to assume that the vast majority of these happen at home, where the crowd is quieter when the Packers are on offense. This is admittedly a rather lazy exercise in mathematics, but if we double that number of plays — because every game is essentially a home game from a crowd noise perspective in 2020 — we end up with 20 free plays. If we assume Rodgers continues to utilize free plays in a similar matter, with 25 air yards averaged, how do these few plays improve his efficiency?
The first thing to note is that the variance on this is going to be pretty wild. We’ll be dealing with a very small sample size of throws deep down the field, where variance is already quite high. We also don’t know exactly which plays that will be replaced, so for the sake of simplicity, we’ll be replacing Aaron’s average play from last season. There are a few different ways to try and add these pieces in, but I am going to stick with the simplest way because I think it still tells a quite accurate story.
Last year, on attempts that traveled at least 20 yards down field, Rodgers averaged 0.29 EPA per play. On plays in which the opposing defense committed a penalty, Rodgers produced 2.55 EPA per play. Neither of these numbers is perfect though. The latter includes all sorts of in-play penalties like pass interference, holding, and illegal contact. The pass interference one is particularly problematic for the topic of this study since it can cost a defense dozens of yards.
Unfortunately, in the NFL PBP data, I was unable to remove that from the datasets in an appropriate way, specifically for Aaron Rodgers. What I was able to do is isolate the entire league and on every play where there was offsides, encroachment, or too many men on the field (the primary calls that lead to free plays), the NFL average EPA was .72 per play. It is important to note that this was only plays in which the penalty was enforced, not when it was called and declined. As good as play by play data has become, it still has blind spots.
For the entire year in 2019, Rodgers averaged 0.066 EPA per pass. So what happens if we double the number of free plays for Rodgers? Like I said, there’s no easy way to do this, and this is mostly meant to be a directional hypothetical. Rodgers had 608 dropbacks last year. If we assume we can replace ten of his average dropbacks with ten free plays, how would we do that?
Firstly, there is a baseline minimum that he can do. It is impossible for Rodgers to do worse that 0.72 EPA per play. That is the league average for an offsides/encroachment/12 men on the field penalty. What Rodgers does beyond that is in his control. He averaged .29 EPA per play on throws more than 20 yards down the field in 2019, but that includes plays in which the pass was not completed. If we only include passes that traveled 20 yards and were completed, Rodgers EPA’ per play leaps to 2.53. How often does Rodgers complete those passes though? According to Sports Info Solutions, Rodgers has completed passes of at least 20 air yards at a 33.2% rate over the past five seasons.
We didn’t say there would be no math
So now it is time for some math. Not only is it time for math, but it’s time for incredibly irresponsible math. Taking the 608 dropbacks and removing ten at the average rate of .066 and replacing them with different values is necessary. In order to figure out those values, more math is involved, albeit not too difficult. There is a 33.2% chance that Rodgers gets a play that is typically valued at 2.53 EPA per play, making its average value on a random free play .84. The remaining probability of 66.8% is at 0.72, making its average value on any random free play 0.48. The total value that Rodgers averages on a free play would then be 1.32 EPA per play. So what happens to Rodgers when we add ten 1.32 EPA plays into his seasonal dropbacks? It moves him from 0.066 EPA per play up to 0.086. That is a not-insignificant jump. But although it moves Rodgers up a few spots in EPA per play, it still keeps him leaps and bounds from the elite passers (0.2 or better).
There isn’t a precise way to forecast this out because the variance at play here is incredibly large. We’re talking about ten-ish plays per year. It’s possible that Rodgers connects on a greater than normal number of these passes, which could drive the average EPA up close to 2 or 2.5 — but even that only moves his EPA per play to just over .1.
Long-story short, I don’t expect a few extra free plays to make a huge difference for Rodgers this season. The biggest improvements in efficiency will likely have to come from middle of the field usage increases and better efficiency on play-action. Any marginal gain helps though, and this is one I think Aaron will look to exploit.