Friday, September 21, 2018

The Greatest Week In Quarterback History


There have been a lot of changes in the NFL over the past 20 years, but none have been as consistent or as talked about as the flourishing of the passing game. It’s gone beyond cliché at this point to say that the NFL is a passing league, as it has been accepted by all but the most stubborn coaches out there. Statistical milestones will continue to fall left and right, offenses will look completely unstoppable, and you will get as sick of hearing about it as I am of writing about it.

But even in this modern era, last weekend was a particularly noteworthy outburst of passing productivity across the league. You may have seen the statistics covered already by others across the internet, but the best job was done by Chase Stuart at Football Perspective, who wrote about the record breaking week and included this chart that makes it clear just how much of an outlier these games were.


I’m generally not a fan of passer rating as a statistic, but its ubiquity makes it convenient for analysis like this. When the system was developed decades ago, it was scaled so that game performance roughly matched the letter grade system used in most schools. A score in the 60s was bad, like getting a D in a class. The 70s were average, like a C (back when Cs were the average). A score in the 80s was a solid B, and any passer rating above 90 was like getting an A.

With this in mind, the chart above becomes even more eye opening. On average across the entire league last weekend, quarterbacks performed at a level higher than an A+. Even the steadily rising trend cannot explain this, and the numbers in week one don’t come anywhere close to matching week two.

There’s been a lot of words spilled about this already, speculation about the cause of this stunning performance. The new rules around hitting the quarterback have gotten the most attention, as have the crackdown on aggressive hits farther down the field. There have been a lot of new quarterbacks enter the league over the past two seasons, and there are very few franchises without at least some long term plan. The league feels flush with quarterback talent in a way I can’t remember happening before, and this explosion of passing efficiency may just be a sign of things to come.

It's worth keeping an eye on, but I don’t have much more to say about it now. Instead I want to go in a different direction, looking at the individual performances that drove last weekend’s spike. Who played out of their minds to lead us to this, and is it reasonable to believe they can keep it up?

My favorite efficiency metric for quarterbacks is Adjusted Net Yards per Attempt, or ANY/A. I’ve gone through this before, but just in case you haven’t memorized every obscure statistic I’ve tossed out there, here’s how this is calculated. It starts with Yards per Attempt, then adds 20 yards for each touchdown pass and subtracts 45 for each interception. It then takes away yards lost to sacks, and divides by the total number of pass attempts plus sacks.

There are some things missing that I would like to see added—scrambles should be counted just like sacks, though it would be hard to separate these from designed runs, and I wouldn’t mind seeing some reward for first downs like is given for touchdowns. But of the common metrics ANY/A is the one that best combines clarity and relevance.

ANY/A is a good measure of efficiency, but efficiency isn’t everything. If a quarterback throws for 100 yards on 10 passes that’s great, but are we really going to say we’d take that over someone who throws for 300 yards on 35 passes? Volume matters as well, which is why my favorite way of looking at an individual game as a whole is using what I will call Surplus Adjusted Net Yards (SANY).

The logic and mathematics behind this are both simple. Basically you ask what a quarterback would produce on an average throw, then look at how many yards above or below that the quarterback is. This is then multiplied by the number of attempts to give the total surplus yards produced by a quarterback. So if a quarterback throws 20 passes with an ANY/A of 10 and the average is 7, he produces 3 more yards on average per pass attempt, which is multiplied by 20 to give him 60 more yards than what you’d expect from an average quarterback.

I calculated the SANY for each quarterback that played this weekend, using the average ANY/A of the entire league going back to the start of the 2017 season. I then compared each quarterback’s performance back through his career prior to last weekend, seeing where the game ranks against all previous performances (using the ANY/A for the season in which the game took place). I did this to produce three different percentiles: a quarterback against his own career, against Tom Brady’s career (the peak option), and against Joe Flacco’s career (the most average quarterback who came to mind).

The results are summarized below, broken into tiers to show how each quarterback fared in this past weekend’s games.

The Elite Performers 
By the Brady percentile score, five different quarterbacks were higher than the 80th percentile. To give that some context, you would expect Brady to have a game like this one out of every five times, and it would take him more than a season and a half to put up five such games.

You can get an idea of just how impressive these performances were by looking at how each of these quarterbacks performed against their own careers. Ben Roethlisberger is a future Hall of Famer, and even in a losing effort on Sunday he was better than 86 percent of games he’s played over 15 years in the league. Kirk Cousins just signed a fully guaranteed free agent contract, then went out and played one of the best games of his career against a division rival.

The other three quarterbacks are not ones you would expect to see this high. Blake Bortles has been the worst starting quarterback in the NFL for several years, but against New England on Sunday he was stellar, throwing the ball 45 times with impressive efficiency numbers in leading the Jaguars to a surprise victory over New England. And Ryan Fitzpatrick at the age of 35 just had the third best game of his career, bettered only by a 6 TD and zero INT game with the Texans in 2014 and his opening week performance against the Saints.

(Here’s something fun I encountered while putting this data together. I only collected individual game performances for quarterbacks who are currently active, so I don’t have data for retired players. But in my 3900 game sample, Fitzpatrick has both the third best and single worst game by SANY. In fact, his zero TD/6 INT performance with the Jets in 2016 had an SANY of -360, more than 80 points below the second worst performance.)

And then of course there is Pat Mahomes. Of all the quarterbacks this past weekend, Mahomes was the only one to set a career high in SANY, and since this was only his third game that doesn’t mean much. However, this performance was better than all but three games by Brady, and better than anything Flacco has done his entire career. I have no idea if Mahomes can keep this up, but it is hard to imagine a better start to a career.

Very Good Performances
This next group is made up of the other eight quarterbacks who eclipsed the 50th percentile compared to Brady’s career. That gives us a total of 13 such players, which is an astounding number when you consider the marker we are measuring them against. This is about what you would expect if you took every team in the league not in California or Texas and replaced their starting quarterback with a top five passer of all time.

Unsurprisingly, a number of quarterbacks on this list would count this past weekend’s game among the best of their career. Well established veterans Philip Rivers, Matt Ryan, and Andy Dalton all did better than the 80th percentile, and Jared Goff came close as well. Every one of these quarterbacks—nearly half the league—performed in the top quartile of Flacco’s career.

Still Pretty Good
At the beginning of this I used Joe Flacco as a barometer for average quarterback play. We can debate this plenty (and God knows the internet has spent plenty of time debating Flacco’s status as an elite quarterback), but I think it’s reasonable to say that through ten years he’s been consistently somewhere in the middle of the pack. Which makes it extraordinary that a total of 22 quarterbacks did better than his median performance.

The names in this group are fascinating for several reasons. It isn’t surprising to spot good games from middle of the pack quarterbacks like Tyrod Taylor, Ryan Tannehill, and Nick Foles (owner of the single best SANY performance among active quarterbacks FYI), and while it’s fun to notice that Blaine Gabbert had one of the best games of his career and still finished among the bottom half of quarterbacks last weekend, three names definitely stand out on this list.

Tom Brady, Drew Brees, and Aaron Rodgers have all established their places as top ten quarterbacks of all time, and all three had rather poor performances this past weekend. Brady and Brees were just below the median levels for their careers, while a hobbled Rodgers squeaked above the lower quartile of his.

This is what truly defined the excellent statistics of this past weekend. On a given week you would expect 16 quarterbacks to do better than the median level for their careers, and we didn’t see anything that shocking this weekend with 17 that hit their own personal upper half. The difference was mostly the distribution. The mediocre quarterbacks had excellent games, and the quarterbacks who struggled were some of the all time greats, for whom even a bad game is statistically impressive.

I don’t know where the passing game will trend over the rest of the season, but I have a feeling last weekend was mostly an outlier. Things will settle into a more normal pattern, and the average level of quarterback play will drift back to where it’s been for the past few seasons. But week two was definitely fun to watch, and I wouldn’t hate it if quarterback excellence was the new trend moving forward.

The Rest
In a way it’s reassuring to know that even in a week defined by great performances, there will always be a few to remind us just how insanely difficult this position is to play.

No comments:

Post a Comment