Matt Ryan's blistering hot season may be one for the record books, but he has more than just himself and his teammates to thank. Ryan and all recent quarterbacks have been riding on a wave of inflation in passing statistics. Although Ryan's 2016 rivals Kurt Warner's 2000 "Greatest Show on Turf" season, the two seasons aren't easily compared -- Warner's passing yards were much harder to come by 16 years ago than they would be today.
It's no secret that the passing numbers in the NFL keep climbing. Analysts have been calling the NFL a passing league for the past few years, but the truth is it has been so for two generations, and there's no end in sight for passing's ascendancy.
Passing's nadir came in the mid-1970s, just prior to rule changes that opened things up for offenses. In 1977, teams averaged 142 yards on 25 attempts per game. After the rules were altered in 1978 to prevent contact with receivers downfield and to allow linemen to block with extended arms, those numbers jumped to 159 yards on 26 attempts and have steadily climbed to where they are today. In 2016, teams averaged 242 yards on 36 attempts per game. All the while, rushing averages have hovered near 4 yards per carry throughout the entire period.
Virtually every measure of passing success continues to climb in recent seasons. Yards per attempt -- the stat most telling of passing effectiveness -- climbed from 6.8 in 2001 to 7.2 this past season. Completion percentage went from 59 percent to 63 percent and interceptions per pass attempt fell from 3.4 percent to 2.3 percent over the same period.
A stark discontinuity like the one between '78 and '79 leaves little doubt as to the reason for the increase in passing stats in the years immediately following, but why did the numbers continue to grow in virtually every other season, and why do they continue to grow today? There are many possible explanations, so let's dig into several plausible theories for why passing stats continue to climb in recent years, and see which ones hold water. A deep dive into the stats can lend some insight and narrow down the theories as to what's causing the rise. Stats can't prove any of this is correct, but whatever theories we have must be consistent with the facts.
(Note: All numbers are for the regular season, and the seasons chosen for comparison were selected on the earliest year in our data for each stat cited.)
Theory: Teams are just passing more often
Drew Brees led the league this season with 673 pass attempts. Fifteen years ago, the leader in attempts was Jon Kitna with 581. Although it is undoubtedly true that teams pass more often -- pass attempts have climbed by 12 percent since the 2002 expansion -- that trend doesn't explain why passing efficiency has improved over the same period. In fact, with offenses passing more often, defenses can be geared more often to defend against the pass, so we might logically expect to see efficiency decline rather than improve as much as it has.
Verdict: Undoubtedly true, but it's an effect rather than a cause of the explosion of passing numbers.
Theory: QBs are making better decisions
In 2001, first-year starting QB Tom Brady was picked off on 2.9 percent of his throws. This past season, rookie Dak Prescott was picked off on only 0.9 percent of his, and Brady on only 0.5 percent. These numbers reflect of a broader trend that has seen the league-wide interception rate plummet over the past 15 years. In 2001, it was 3.4 percent, and in 2016 it was 2.3 percent, lower by a factor of one-third. For historical comparison, in 1977, the year prior to the passing rule changes, the interception rate was 5.7 percent.
If interception rate is an indicator of how well QBs make decisions, then yes, QBs are improving on that front.
It's true that the recent improvements in these numbers are partially due to a greater reliance on screen passes, which are relatively hard to pick off, but that can't come close to accounting for a one-third drop in the interception rate.
Looking at how QBs handle pressure can tell us more about their decisions. When the QB is pressured (which is typically on downfield plays and rarely on wide receiver screens) all the major passing efficiency metrics are improving significantly. Since 2009, yards per attempt has increased from 4.3 to 5.7. If sack yards are factored in, yards per attempt has climbed from 2.0 net yards per dropback to 3.3. Off-target percentage has dropped from near 35 percent to 28 percent, and interception rate has dropped from 4.3 percent to 2.9 percent. Those are stunning changes in just a six-year span.
Verdict: True, and under pressure in particular.
Theory: Receivers are getting better at catching
In 2016, Jordy Nelson led the league with 0.7 percent drops on his targets. In 2006, Lee Evans led the league with a 1.5 percent drop rate. Looking at the entire league, drop percentage has fallen from 5.4 percent in 2009 to 3.3 percent this season -- a reduction by well over one-third. Drop percentage may not tell the whole story, but it is representative of overall receiver ability to catch. Some of this reduction is likely a byproduct of a reduced average depth of throws and the increase in screens, but only a small amount.
Verdict: True, and likely part of the answer.
Theory: Receivers are getting more yards after catch (YAC)
Donald Driver led all wide receivers in 2006 with 483 yards after catch. This past season, Jarvis Landry of the Dolphins had 639 yards after catch to lead NFL wideouts. Our YAC numbers go back only through 2006, and there is quite a bit of year-to-year variation, but the trend is increasing. In '06 receivers generated 4.8 YAC per reception, and this season they generated 5.1.
Receivers are getting more YAC per reception in recent years than in years past. If we use a little math, that converts into a 2.9 and 3.2 YAC per attempt difference between 2006 and 2016, an improvement of 0.3 YPA, which would explain a significant amount of the increase in overall yards per attempt in recent years. But that increase is due to the fact teams are throwing more screens, which tend to have a lot of YAC, and some of which are largely free yards behind the line of scrimmage. The numbers on regular downfield passes have held steady at 2.5 yards per attempt on average.
So YAC is a big part of the picture, but its overall effect is muted.
Verdict: True, but on balance does not explain the increase in passing success.
Theory: QBs are getting more accurate
Sam Bradford was the king of completion percentage this season with a 71.6 percent rate, one of 10 passers with rates above 65 percent. In 2001, only Kurt Warner and Rich Gannon completed more than 65 percent of their throws. Although completion percentage has increased, that doesn't necessarily mean QBs are more precise with passes. The average pass has become shorter, and shorter passes are easier to complete. In 2006, the average pass attempt traveled 8.7 yards downfield. That number has steadily fallen since, and this season it's 8.1 yards. Our adjusted completion percentage stat, which accounts for depth of each throw and ignores spikes and throwaways, has improved slightly since 2006, the first year it was tracked, going from 66.6 percent then to 68.3 in 2016.
At the same time, however, our off-target percentage stat, which measures over- and underthrows on targeted passes, has crept up from 17 percent to 19 percent since 2006. No matter how we measure it, the 18 percent increase in total yards cannot be explained by a small improvement in accuracy. Completion percentage is a minor contributing factor in bigger overall passing numbers, but the increase in completion percentage is more than explained by the reduction in receiver drops, which itself is related to depth of throws.
Verdict: Unclear, but if true can account for only a small part of the answer.
Theory: Offensive schemes and strategies are changing
As discussed, offenses are throwing many more screen passes than they did just a few years ago. In 2006, the first year we tracked pass types, offenses threw 750 screens compared with 1,151 this season -- a 53 percent increase. And that's down from a peak of 1,359 screens in 2015. The number of screens exploded between 2006 and 2012. One reason the number of screen passes has dropped is that their effectiveness has decreased steadily. Offenses gained an average of 6.7 yards per attempt in 2006, but only 5.6 yards per attempt in 2016. The typical screen pass actually lowers the overall yards per attempt average in the league, but one important statistical aspect of screens is that they usually provide more YAC than the typical downfield pass, because the first several yards are usually behind the line of scrimmage and easily gained by a receiver.
Play-action passes have become slightly more common since 2006, but are well short of their 2012 peak. Play-action passes remain an underused part of the passing game in the NFL. They are significantly more effective on average than downfield passes and screens, improving from a mean YPA of 6.7 in '06 to 7.7 this season, with about 0.6 yards of that gain coming from YAC. Interceptions and sacks are also less common when using play-action.
Verdict: True, and explains improvements in sack rates and interception rates, but not overall efficiency.
Theory: Penalties are helping passing offenses
Defensive penalty yards on dropbacks have increased about 2.5 percent per year since 2002. This total went from 2,676 in 2002, to a high of 4,711 in 2013 to 3,822 in 2016 (the lowest figure since 2012, but still a 30 percent increase over the 2002 total). But these extra yards can't be a cause of the passing stat increases, because they aren't considered passing yards. The official passing yards totals don't include penalties on pass interference or unnecessary roughness, but penalties on defense can mask otherwise negative plays on offense. Unfortunately, those are statistics that we never get to see, because they aren't officially kept.
Further, the increased safety, or perhaps the perception of increased safety, provided to offensive players by a heightened focus on rules that protect them would plausibly allow them to make plays in what might otherwise be dangerous circumstances. Unfortunately, that's not something easily measured, but would be consistent with the improvement we see by QBs when under pressure.
Verdict: Not directly responsible, but indirect effects may be part of the answer.
Theory: Offensive lines are improving
Here, the evidence is mixed. Quarterbacks are feeling more pressure, but are getting more time in the pocket. The pressure rate on all dropbacks has increased from 24 percent in 2009 to 27 percent in 2016, while at the same time blitzing occurs slightly less often. This combination of factors makes the QB's job more difficult. In the interests of full disclosure, it's hard to be sure whether the pressure rate is truly increasing or our game charters' criteria has evolved over time. But another indicator suggests QBs are under more pressure. The NFL's QB hits stat has increased in terms of a per dropback rate, from 16 percent in 2009 to 19 percent in 2016. Sacks per dropback have held relatively steady over that period at about 7.5 percent, so the picture isn't very clear.
Our own offensive line metrics suggest defenses are gaining a slight upper hand in control of the line of scrimmage on pass plays. The line control percentage in 2011 was 50 percent for offensive lines, and in 2016 it was slightly lower at 48 percent.
We can use time-in-pocket metrics to assess how much time offensive lines are giving their quarterbacks to throw downfield. The way time-in-pocket is commonly reported, as an average, is very misleading, however. The time-in-pocket reported for a quick three-step drop will be very short, but it tells us almost nothing about how long the QB would have had. It can tell us only that he had at least that long. That's why we need to look at time-in-pocket in terms of something known as a "survival function," which measures the probability something (your smartphone, your laptop hard drive, or a QB in the pocket) will last for various lengths of time.
When we look at time-in-pocket in those terms, QBs appear to be getting more time to throw. In 2011, a QB could expect a 50 percent chance of surviving at least 5.5 seconds, and in 2016 he could expect a 50 percent chance of surviving at least 5.8 seconds. Another way to look at it is that the 50 percent chance the QB had of surviving those 5.5 seconds in 2011 is now closer to 60 percent. (5.5 seconds may seem like an extremely long time in the pocket, mainly because QBs tend to throw by 3.5 seconds. But keep in mind that doesn't mean the QB would have been sacked at 3.501 seconds. And if a QB did regularly hold on to the ball for 5.5 seconds, he would have a 50 percent sack rate!)
Verdict: Unclear, and if true, not likely to be to a significant degree.
Theory: Secondaries are getting worse
This would certainly be consistent with most of the other factors already discussed, such as the drop in interception rate. Passes defended have dropped slightly, from 17 percent in 2009 to 15 percent in 2016. On the other hand, YAC on downfield throws has been consistent at 2.5 yards per play, so defenders are making tackles just as well. And defenses are improving against the screen. What may be happening is that all the screen passes are forcing defenses out of where they would prefer to be for downfield attempts, making those plays easier for the QB.
Verdict: Possible, but unclear.
Here's what we know: Passing totals and passing efficiency numbers have continued to improve in recent years. Receivers appear to be getting better at catching. Defenses are getting more pressure on QBs, and doing it without blitzing as often, but QBs are getting smarter and making better decisions, particularly under pressure. Pass protection appears to be losing ground to the pass rush, but not enough to overcome. Screen passes have dramatically increased, and have had a direct impact on interception and sack rates, but not on overall efficiency. Secondaries are being stretched horizontally by screens, but are adapting and getting better.
The theory that may best fit the evidence is that passing offenses have simply improved. Regardless of what some pundits claim and despite the practice restrictions in the current collective bargaining agreement, QB play has been improving significantly for years, particularly with regard to decision-making. The training and development of the passing game from high school up through the NFL is consistent with what we see in the numbers. Changes in the rules, and a greater emphasis on protecting players, have amplified this improvement. Completion percentage is up, but it's not the QBs' throwing precision that has improved. It's their cognitive abilities and development that have improved. In addition, defenses have been stretched by changes in scheme and have been forced to adapt to a more horizontal passing attack, and to a lesser degree more play-action.
So if it's true that QBs are simply better now than they were, and that's the primary cause of the recent increase in passing numbers, then maybe Matt Ryan's 2016 season is comparable to Kurt Warner's 2000 after all. Maybe we shouldn't be discounting today's stellar passing numbers as much as we tend to do. Maybe we should just accept the greatness we've witnessed for what it is.