The Jacksonville Jaguars have been one of this season’s surprise teams at 4-3, but they have been anything but consistent. Jacksonville opened the season with a dominant win in Houston and then lost to Tennessee by three touchdowns at home … before destroying Baltimore in London. They followed that with an overtime loss to the Jets and an impressive win in Pittsburgh, a Super Bowl contender. After losing by 10 at home to the upstart Rams, the Jaguars decimated the Colts in Indianapolis, delivering a shutout.
What will the Jaguars do on Sunday against Cincinnati? Your guess is as good as mine, because the NFL has been such an up-and-down league in recent years that it borders on chaos.
At least, that is the perception of today’s NFL. To test whether that’s really the case, I used Football Outsiders’ DVOA system (that’s Defense-adjusted Value Over Average, which is explained here).1 That system breaks down every single NFL play and determines its value when compared to a league baseline based on situation (i.e., down and distance, field position, score).
Among NFL teams, Jacksonville has the highest statistical variance (36.6 percent) in weekly DVOA performance this season, meaning that no team’s performance varies from week to week as much as Jacksonville’s. In fact, the Jaguars’ variance is the second-highest in the Football Outsiders database going back to 19862 (the 2005 49ers were at 36.7 percent). So the Jaguars are an extreme case of inconsistency. They are also the exception.
Just last season, the New York Giants set the benchmark for the lowest variance in our database, at 2.5 percent, meaning that their performance almost never wavered from its average. When looking at all 32 teams, the average DVOA variance for the 2016 season was 11.7 percent, the lowest of any year since 1986, which suggests that teams were never more predictable week in and week out. So far this season — through Week 8 — the average DVOA variance is 11.9 percent, which would be the second-lowest season. Of the five lowest seasons in variance since 1986,3 four of them have come since 2013.
By this measure, one could argue that teams have actually been more consistent on a weekly basis than at any other point in the last 30 years. That might sound crazy, but crazy is the NFL’s first language.
Every NFL season seems strange at the time
As we enter the midpoint of the season, you may feel as though little in 2017 has stayed on script. A rookie (Deshaun Watson) who wasn’t even the Texans’ Week 1 starter was rewriting the NFL record book until he tore his ACL in practice. The Rams and Jaguars, perennial doormats, have the best scoring differential on a per-game basis. The Giants (1-6) and Raiders (3-5) each already have more losses than they did all of last season, when both made the playoffs. Miami has a minus 60-point differential, the worst ever for a 4-3 team.
I could go on and on, but this feels like the strangest NFL season since … well, since 2016. That’s the one in which two rookies helped Dallas to a No. 1 seed, the Lions set a record with eight fourth-quarter comebacks, and the Falcons blew a 28-3 lead in the Super Bowl after Matt Ryan had one of the greatest passing seasons ever.
The fact is “crazy” is what the NFL is all about. Take the 2003 season, for example, when teams had the highest average DVOA variance since 1986, at 17.2 percent. That season oozed drama: The Buccaneers and Raiders fell apart after appearing in the Super Bowl in 2002; the Peyton Manning-led Colts erased a 21-point deficit in the final four minutes against Tampa Bay in Week 5; the Ravens beat Seattle 44-41 in overtime after a late 17-point comeback in Week 12; Brett Favre threw four touchdowns in Oakland the night after his father passed away in Week 16; and in the final week of the season, the Vikings were eliminated from the playoffs when Josh McCown hit Nate Poole in the end zone on fourth-and-25 for the 3-12 Cardinals.
The NFL thrives on improbable events. Recency bias, especially in the social-media era, makes us want to believe that the craziest season is the most current one.
Where has all the variance gone?
Chaos is the perception, but what’s the reality? By digging deeper into the DVOA numbers, I found that variance has been trending downward in the passing game — suggesting that teams are becoming more predictable in how often and how well they will take to the air. It currently stands at 16.9 percent, and if it stays at that level, 2017 will have the lowest variance since 1989.4 It would also be the third year in a row to set a new low. Meanwhile, the running game — which has traditionally been fairly consistent — has also seen its variance (5.9 percent) reach its lowest level since 1989 this season.
The low variance we are seeing in 2017 can be somewhat chalked up to only being halfway through the season. As the season progresses, injuries are bound to shake up more teams’ performance — losing an Aaron Rodgers or an Odell Beckham Jr. has an impact on the consistency in an offense’s performance.
However, the decrease in variance in the passing game is the most significant as the NFL continues to be a passing league. This change can be explained by one word: innovation. It’s the innovation the NFL experienced with the record-setting 2007 Patriots and the lack of innovation we have seen in the years since.
Let’s start with how the Patriots changed the game. For as much as head coach Bill Belichick is known as a defensive genius, his innovation on offense with Tom Brady at quarterback has arguably been the most impressive part of New England’s nearly two-decade run. The Patriots were already ahead of the curve with the “dink-and-dunk” offense under offensive coordinator Charlie Weis, but 2007 took things to unseen levels.
That year was about far more than just a collection of talent (Randy Moss, Wes Welker and Donte’ Stallworth) that Belichick added around Brady to break offensive records and challenge a 19-0 season. These Patriots introduced the NFL to the shotgun-spread offense. New England was the first team in Football Outsiders’ database to use shotgun on at least 50.0 percent of its plays (50.4 percent to be exact). The next season, Kansas City and Arizona leaned on the formation just as much, with the Cardinals making it to the Super Bowl behind Kurt Warner. Currently, 23 offenses in 2017 are using the shotgun at least 50 percent of the time.
The shotgun is important, because the average shotgun play since 2007 produces a DVOA that is 8.7 percentage points higher than non-shotgun plays. In other words, the shotgun produces better offensive efficiency, and most of the league has caught up to this fact. The league-wide shotgun usage was just 7.3 percent in 1996. It climbed to 27.2 percent in 2007, hit 40.7 percent after the lockout in 2011, and there was another jump in 2013 to 58.7 percent, which is about where every season since stands. The 2017 shotgun rate is currently 58.5 percent.
Back in the day, only the NFL offenses with great quarterbacks liked to put the ball in the air a lot, but that has changed. Practically every offense in this era calls more pass plays than run plays in a season. Throwing for 4,000 yards and completing 60 percent of your passes used to be rare air reserved for only the best passers. Before 2007, no more than a handful of quarterbacks hit those benchmarks in a season. In 2016, 13 quarterbacks hit those numbers — that’s the most ever.
This excessive use of the shotgun and short, quick passes has wreaked havoc on a defense’s ability to produce havoc-wreaking plays of their own. In 2016, the sack rate (5.8 percent) was the lowest in NFL history — or at least in seasons in which sacks were officially counted. You can count on a minimum of 1,500 screen passes a season from the 32 offenses, and those are nearly impossible for the defense to stop from being completions. The 2016 interception rate (2.3 percent) was also the lowest in NFL history, and the last six seasons5 feature the six lowest interception rates in NFL history. In some cases, quarterbacks seem afraid to throw passes downfield for fear of being benched because of picks — just ask DeShone Kizer about his job security in Cleveland.
Despite trying to emulate a record-setting offense like the Patriots, most teams simply cannot replicate the level of talent New England had. Randy Moss was the greatest vertical threat of his era, and the Patriots redefined the role of a slot receiver with Welker. They used his short-area quickness and ability to read coverages on the same page with Brady to perfection for years.
A lot of offenses are looking for “their Welker” now. The Patriots still adore that role more than anyone, evidenced by the Welker clones they’ve used like Julian Edelman and Danny Amendola. The rest of the league has caught on — which you see in the usage patterns of Randall Cobb (Packers), Golden Tate (Lions), Doug Baldwin (Seahawks), Cole Beasley (Cowboys), Jarvis Landry (Dolphins), Sterling Shepard (Giants) and Cooper Kupp (Rams).
And the rise of these receivers has led to pass defense conformity as well. To combat the threat in the slot, many defenses are using an additional cornerback in their base defense — and thus decreasing the use of the traditional 3-4 or 4-3 set. According to Football Outsiders’s 2016 game charting, defenses were in nickel for 57 percent of plays compared with 30 percent for a 3-4/4-3 base.
The Patriots also changed the way running backs were used as specialists in the passing game. While Kevin Faulk isn’t even the best receiving back with that surname in NFL history, his steady use in the passing game for the Patriots paved the way for others to follow, including Danny Woodhead, Shane Vereen and James White. Back in the day, usually only the great passing offenses (Don Coryell’s Chargers, Bill Walsh’s 49ers, Dan Marino’s Dolphins) threw a lot to the running backs. Today, we see even some of the worst offenses lean on a receiving back. Woodhead (Ravens) and Vereen (Giants) play roles similar to those they had in New England, although for struggling offenses. Chris Thompson (Redskins), Theo Riddick (Lions) and Duke Johnson (Browns) are other examples of committed receiving backs this season. The Panthers even used the No. 8 pick in the draft on receiving back Christian McCaffrey, who currently ranks third in the NFL with 49 receptions. He has just 49 carries, so his 1.0 carries-to-catches ratio is way out of line with the average ratio (8.4) of the last 35 first-round running backs. McCaffrey’s role given his draft status just wouldn’t have made sense in an older era, when a running back’s pass-catching ability was an afterthought.
Now that almost every offense has tried to copy New England, we’re in a period in which no one really wants to be different. If you try something different, it is usually met with mockery. Take the Wildcat, for example. Miami used the formation — in which a running back receives a direct snap — in 2008 to shock the Patriots in an upset win, and then continued to sprinkle it in for one playoff season. In 2017, the Wildcat is basically an internet meme — it’s most notable appearance this season was when Jay Cutler used it to express his general disinterest in being a Dolphin.
John Fox is viewed as an ultra-conservative coach, but he tried to do things differently with Tim Tebow and his version of the read-option on the 2011 Broncos. That even led to a playoff win, but the Broncos couldn’t help but dump Tebow for Peyton Manning’s conventional passing perfection in 2012. Elsewhere, the craze over the read-option offense peaked in 2012, with teams like Carolina, San Francisco, Seattle and Washington adopting the offense that was on loan from college football. But none of those offenses ever used it 200 times in a season. Five years later, the two most prolific examples of that brief movement — Colin Kaepernick and Robert Griffin III — are not even on any NFL roster.
And remember when Chip Kelly was going to revolutionize the NFL with his hyper-fast offense? That lasted for about a half a season in 2013. His offenses sure snapped the ball quickly, but a steady decline in results and eventual conformity to coaching norms made Kelly a failure in the pros.
Fox has tried something different in Chicago this year: ignore the passing game. The Bears have won games in which the quarterbacks did not even have 100 net passing yards. Rookie Mitchell Trubisky had four completions on seven throws against Carolina in a Week 7 win. The Bears seem intent to just hand the ball off 50 times a game and hope the defense scores a return touchdown (or two). But this is not a sustainable strategy, and the Bears have devoted significant resources to the quarterback position, including the high-priced signing of Mike Glennon.
Chicago’s approach might be more appreciated if the offensive production and salary management were more practical, but no one said sustaining success in the NFL was meant to be easy. It’s just getting harder to stay ahead of the curve when every team has a basic understanding of what it takes to succeed, but certain limitations with the roster or coaching continue to hold certain teams back.
Passing hasn’t gotten easier so much as it has gotten smarter and safer. That’s not even a nod to the rules that aim to protect skill players from vicious hits, but a remark on the risk-averse nature of so many passing offenses today. Why hold the ball to get sacked or take a shot to get intercepted when the 2-yard checkdown on third-and-12 is always open and no one has to put their job at risk.
True craziness may just have to come in the form of a coach who eschews the punting unit outside of obvious situations and plays four-down football from the opening drive. If even just a few teams copied that strategy, then variance would go back up, and there would be even more interesting stats and decisions to cover on a weekly basis.
You can believe 2017 is more unpredictable than usual, but statistically, that’s likely not going to be the case. Also, the Patriots (6-2) are tied for the best record in the AFC behind Brady and Belichick, and the Browns are 0-8 and don’t have a quarterback of the future on their roster. See, some things in the NFL are as consistent as ever.
Check out our latest NFL predictions.