Does the NFL have a quality-of-play problem? Over the first two weeks of the season, it's fair to say that the league hasn't exactly delivered much in the way of exciting action. After a thrilling season opener between the Chiefs and Patriots, national contests such as Seahawks-Packers, Giants-Cowboys and Texans-Bengals have been marred by dismal offensive line play. We've had eight games decided by more than 20 points through two weeks, double last year's total, and that's without getting to the Lions and Giants on Monday Night Football.
I have to admit that I'm skeptical of most quality-of-play arguments surrounding the game. Most are anecdotal and products of a small sample, pointing to a bad play, a terrible quarterback or an ugly week of games. Often, they turn into a case of whack-a-mole. Once one surefire sign of subpar play pops up and is proved to be false, another entirely different argument leaps ahead. There's no definition of what actually represents bad football, and as such, both fans and professional analysts often start with the premise that the game has taken a turn for the worse and work backward to try to prove their point.
I'm going to suggest that those arguments don't have a strong case through two weeks in 2017. Furthermore, if you really believe that there's a quality-of-play issue built around offensive problems, there's a quarterback sitting in the free-agent market who would be an upgrade for a handful of teams and a high-quality backup in case of injury for just about every other team. You can't simultaneously make an argument saying that the league is suffering from a dearth of quarterbacks and also suggest that Colin Kaepernick doesn't deserve a meaningful role in the NFL. But more on him in a minute.
Is the football bad now?
It's true that games haven't been particularly close so far. Through the first 30 games of the 2017 season, the average contest has been decided by 12.9 points, and two-touchdown games usually aren't very interesting. Through the first two weeks of the 2016 season, the average margin of victory was just 8.9 points.
Of those two, though, the 8.9-point figure is more of an outlier than this year's mark. Since the league moved to 32 teams and its current schedule format in 2002, the average margin of victory over the first two weeks of each season has been 11.3 points per game, closer to this year's margin than last year's. In 2011, for example, the average margin of victory was an identical 12.9 points per contest. In 2003, the number was all the way up to 14.8.
Games are closer during the modern era of football than they were during whatever era you might want to classify as the glory days of football. Games during the first two weeks of seasons in the 1970s were decided by an average of 12.0 points per contest. That dropped to 11.5 during the 1980s, but it was back up to 11.9 during the '90s and hung around at 11.5 during the first decade of the 21st century. Since 2010, games in Weeks 1 and 2 have been decided by an average of 10.9 points per contest. Even with 2017's slate of relative blowouts, we're living in an era of more competitive games.
The problems so far in 2017 have been concentrated on offense, and while it might be reasonable to interpret that as defenses doing a better job of employing personnel, rotating pass-rushers and employing smarter coverage concepts, the argument instead seems to be that offenses -- and quarterbacks, in particular -- are playing worse than ever.
Scoring through two weeks is down. Through those 30 games, teams have produced an average of just 20.3 points, down nearly 11 percent from last year's two-week average of 22.5 points. It's reasonable to assume that poor quarterback and offensive line play is keeping teams from scoring points in bunches the way they have in years past.
Right? Well, again, not exactly. Teams have been scoring at record rates in recent years, which seems at odds with the idea that there are trash quarterbacks and offensive lines around the league. The average team produced 22.3 points per game through the first two weeks of the season between 2010 and 2017, the highest rate for any decade since the AFL-NFL merger of 1970 by more than a full point. Teams in the 1980s averaged 21.1 points per contest, while teams from the '70s were all the way down at 19.1.
This isn't a two-week fluke, either. The seven most recent seasons are also the seven highest-scoring seasons since the 1970 merger, with teams hitting a peak at 23.4 points per game in 2013. Teams averaged 22.8 points in 2016. If the offensive freeze struck and NFL teams averaged 20.3 points per game for the remainder of the season, it would rank 35th among the 48 post-merger NFL seasons, exclusively ahead of seasons from the '70s and '90s, which are being held up as a contrast to the disastrous, sloppy offensive play of today's NFL.
While there are certainly sloppy plays, the evidence that the NFL is a messier league than ever before just isn't supported by reality. As I wrote last year, turnover rates have dropped dramatically from previous eras to historic lows, even after accounting for the move to shorter passes. Some have blamed the proliferation of games on short weeks thanks to the expansion of the Thursday night schedule, but when I looked into the issue in 2013, turnovers and drops didn't occur more frequently on Thursday than they did in games played on Sunday or Monday. That hasn't changed, either; since the beginning of the 2013 season, 2.1 percent of the snaps on Thursday games have resulted in turnovers, better than the Sunday rate of 2.3 percent. Meanwhile, 3.8 percent of Thursday passes are dropped, which is below the 4.0 percent drop rate with full rest on Sundays.
It wouldn't be fair to suggest that these are foolproof measures of performance. There are things for which the numbers can't account. If these statistics did back up the idea that the quality of play in the NFL has sunk -- if the turnover rate was up on Thursday night or games were historically high blowouts -- they would be the cornerstones of those arguments. It's telling that the complaints about the lack of scoring and close games didn't come up when we were having these quality-of-play discussions last year. We can never refute the idea that quality of play is down, and I think there are certainly issues with offensive lines around the league, but there's certainly not a foolproof case that NFL offenses are playing worse than they did in days gone by.
The quarterback problem
The quote you'll hear if you watch the NFL closely for more than a week or two is that there simply aren't 32 starting quarterbacks to go around to play for 32 teams at any one time. It has always been a frustratingly naive quote, and there are a few reasons why it's an inaccurate cliché. For one, by definition, you can't have 32 above-average starting quarterbacks in a league with 32 starting quarterbacks. More important, the league doesn't operate in a way to get the best passers opportunities:
NFL teams aren't very good at scouting and sifting through quarterbacks. The Cowboys-Broncos game we saw on Sunday is a great reminder of how organizations that spend millions of dollars on both identifying and acquiring quarterback talent are led more by luck and circumstance than anything else. The Cowboys, who had former undrafted free agent Tony Romo as their starter for a decade, tried to trade up and grab Paxton Lynch in the first round of the 2016 draft. They were beaten to the punch by the Broncos for Lynch, with the Raiders subsequently pipping Jerry Jones & Co. to fallback plan Connor Cook. The Cowboys had to settle for Dak Prescott, who looks like every bit of a franchise quarterback, even if he struggled against the league's best pass defense in Week 2.
Meanwhile, the Broncos haven't gotten much out of Lynch. Their third-stringer is Brock Osweiler, whom Denver attempted to re-sign after the 2015 season with a reported three-year, $45 million offer, only for Osweiler to turn them down and sign with the Texans for an ill-fated campaign. The Broncos then traded up to grab Lynch, who has been subpar and lost consecutive training camp battles to former seventh-round pick Trevor Siemian, who has been on the roster all along and certainly appears to be the best quarterback of the three.
Both Prescott and Siemian would have lingered on the bench if it weren't for circumstances unexpectedly breaking their way. Prescott might be a backup stuck behind a franchise quarterback in a different city if the Cowboys had been able to nab Cook, and he only got his chance because Romo (and Kellen Moore) suffered an injury during the preseason. Had the Broncos re-signed Osweiler, he might have kept up his previous level of competence and Denver might have cut Siemian in favor of a veteran backup.
It would be foolish to assume that there aren't at least one or two more useful starters like Prescott and Siemian lurking on the bottom of rosters elsewhere around the league, waiting for a chance. There probably aren't enough starters to get to 32 useful players, but the 32 players who would represent the 32 best options in football aren't the 32 guys currently getting starting reps.
NFL teams often look for one prototypical quarterback archetype at the expense of other styles of play. What is true is that there aren't 32 pocket passers who are 6-foot-4 with incredible arm strength, which is the model teams try to target to an almost comical degree. They place a remarkable emphasis on height and a quarterback's ability to "make all the throws," which is how Brandon Weeden was drafted ahead of Russell Wilson in 2012, and why the 6-foot-6 Mike Glennon racked up $18.5 million in guarantees from the Bears this offseason despite having exhibited modest aptitude for the game in Tampa Bay.
Players who look like the quarterbacks that general managers dream about late at night also get second and third chances to prove themselves as viable passers. Sometimes, it works out: Sam Bradford was mostly a mess with the Rams and wasn't much better during his year in Philadelphia, but he held up behind a dismal offensive line with the Vikings last season and looked great in the season opener before missing Week 2 with a knee injury. Meanwhile, Blake Bortles has never been good and is entering his fourth season as a starter, while mediocrities like Glennon and Blaine Gabbert have racked up the opportunities to prove that they're not viable quarterbacks.
This isn't anything new: You might remember Doug Flutie spending most of his career in the CFL before being forced to split reps in Buffalo with another prototypical quarterback in Rob Johnson. What also isn't new is the idea that there are more starting jobs than useful starting quarterbacks. Look around the league and you'll see guys like Bortles, Glennon, Jacoby Brissett and Josh McCown starting games. Veterans Carson Palmer and Eli Manning appear to be fading, while rookie starters DeShone Kizer and Deshaun Watson are works in progress.
There isn't the sort of leaguewide depth you would hope for, but again, this isn't anything new. Consider the guys who started 20 years ago in Week 2 of the 1997 season. Sure, there were unquestioned franchise quarterbacks such as Brett Favre and Drew Bledsoe to go along with seemingly ageless future Hall of Famers Dan Marino and John Elway, just as there's unquestioned top-end talent among the best quarterbacks in football today.
What you forget is that there were all kinds of anonymously mediocre-or-worse quarterbacks starting for teams. There were rarely effective veterans such as Erik Kramer, Dave Brown and Scott Mitchell taking serious reps. Draft busts Heath Shuler and Jim Druckenmiller started games, with the latter making his first and only professional start after being taken in the first round of that year's draft. Third-stringer Steve Matthews also made his lone pro start as a replacement for an injured Mark Brunell in Jacksonville, while professional tall men such as Kent Graham and Todd Collins dutifully threw passes off line for bad teams.
The only reason anybody would be nostalgic for the quarterback play of 20 years ago, relative to now, would be that they remember the stars of 1997 as compared to the full breadth of NFL quarterback play today. (And even that's generous: Chris Chandler and Trent Dilfer made it to the Pro Bowl in 1997, suggesting that they were two of the top eight quarterbacks in the league.) Indeed, as Jason Lisk of The Big Lead noted last year in a must-read story, people have been complaining about the decline in both the quality of NFL play and the abilities of quarterbacks for 25 years now.
One possible solution ...
If you don't agree with all of that and believe that there's still a historically significant dearth of quarterbacks and problem with quality of play in 2017, you're entitled to your opinion. If that's true, though, it's close to impossible to make an argument against Colin Kaepernick starting for a handful of NFL teams right now, let alone being on an NFL roster. For whatever flaws Kaepernick has as a quarterback, a look at history suggests that there hasn't been a single quarterback in the post-merger modern NFL to play like Kaepernick did in 2016 without getting another shot at a job afterward.
Let's leave aside Kaepernick's abilities as a runner -- which are still significant, given that he led all quarterbacks in rushing DYAR and was fourth in rushing DVOA last season -- and ignore his two-plus excellent seasons as a quarterback under Jim Harbaugh, both of which would be treated as reasons to shell out millions on any other quarterback's résumé. Instead, let's simply look at what Kaepernick did last season under Chip Kelly on a frustratingly bad 49ers team while throwing to Jeremy Kerley, Quinton Patton and Garrett Celek as his most targeted receivers. (They've combined for four catches through two weeks in 2017.)
The best simple metric for quarterbacks using raw stats is adjusted net yards per attempt (ANY/A), which is essentially a supercharged passer rating with better coefficients and sacks incorporated. Pro-Football-Reference.com provides index statistics that allow you to compare a quarterback's numbers to players from his era and scales them to a 100-point system through standard deviations above and below the mean. Kaepernick's 2016 ANY/A+, the index statistic for ANY/A, was a 97. He was one-fifth of one standard deviation below the mean as compared to every other 2016 quarterback, finishing 23rd among 31 qualifying quarterbacks in the category.
It's rare for a young quarterback who gets regular reps in the NFL in any given season to finish his career without getting at least another shot or two. How rare? I went looking for quarterbacks who had not yet hit their 30s and threw 200 passes or more in a season without ever playing again. Kaepernick will turn 30 in November, so 2016 was his age-29 season.
There are only 14 other quarterbacks besides Kaepernick to do that in NFL history, not including other passers from 2016 who were hired elsewhere (such as Osweiler and Matt Barkley) or Teddy Bridgewater, whose future is uncertain after a catastrophic knee injury. Kaepernick's 97 ANY/A+ was third best among the group, but even if he'd played worse in 2016, it shouldn't have mattered. Every one of them had an excuse for ending their career without getting another shot at an NFL job that wouldn't apply for Kaepernick.
Six of the quarterbacks suffered career-ending or career-shortening injuries, including Neil Lomax (arthritic hip), Steve Ramsey (double ankle surgery), Tim Couch (rotator cuff), Heath Shuler (repeated foot injuries) and Gary Marangi (shoulder). Pat Haden suffered a knee injury and retired while rehabbing the injury to take a broadcasting job. Johnny Manziel and JaMarcus Russell left the league amid substance-abuse issues. Kaepernick would not fit into either of these categories.
The other six quarterbacks -- each with a fraction of Kaepernick's résumé and a less impressive final season as a regular quarterback -- were given one or more opportunities to sign with an NFL team but never made it back onto the field for a regular-season pass. Joey Harrington, Mike McMahon, Cade McNown, John Skelton, Craig Whelihan and Randy Wright each went to training camp with one or more teams without making it onto the field. This was also true of some of the injured players. Consider that Marangi went 0-7 as a starter and set a still-standing league record for lowest career completion percentage. The Packers still tried to trade for him, only to be rebuffed when Marangi failed a physical. The Browns signed Marangi anyway.
The other arguments about Kaepernick's on-field abilities are flimsy. He isn't a great pocket passer, but he has posted a totally reasonable in-pocket passer rating of 87.6 over the past three seasons, better than plenty of other backup types who racked up millions of dollars in contracts this offseason. Teams might want a quarterback who knows their system having gone through camp, but in a league in which Jacoby Brissett can go from Patriots third-stringer to Colts starter in eight days, it's hard to believe Kaepernick can't even make his way onto a roster whatsoever.
The 49ers were 4-20 in Kaepernick's final 24 starts, which is a totally arbitrary end point but admittedly not good. They also allowed an average of 27.5 points per game in those starts. It's fair to wonder why a bad record would keep Kaepernick out of the league, while Josh McCown, who was 2-22 in his most recent 24 starts before this season, signed a one-year, $6 million deal with the Jets. The best argument is that McCown would somehow be a useful mentor to Bryce Petty and Christian Hackenberg, which ignores the reality that McCown's list of pupils hasn't exactly been impressive.
There's no on-field precedent for a quarterback as active as Kaepernick to linger in free agency without a job after serving as a regular quarterback the previous season, let alone one as effective. The reasons Kaepernick is unemployed right now have to be off-field, then, and they're also filled with holes. The word "distraction" is tossed around as some vague catchall, but Kelly told Adam Schefter that Kaepernick and his decision to protest was "zero distraction" to the team last year.
If the idea is that Kaepernick will continue to draw attention for protesting the national anthem, that also doesn't hold up. It was reported in March that Kaepernick would stand for the anthem in the future. Furthermore, players around the league have continued to kneel or sit during the anthem without repercussions or any notable public outcry in 2017, including Eric Reid, Marshawn Lynch and Michael Bennett, the latter of whom was the subject of a pregame rally before Sunday's 49ers-Seahawks game. The rally also showed support for Kaepernick, once a Seattle arch-nemesis.
The other anecdotal arguments haven't held up. Kaepernick has continued to make steady donations to charity to hit the $1 million donation target he set last year. There were questions about whether Kaepernick wanted to play, but firsthand reports on Sunday revealed that Kaepernick wants to play football and is in shape to suit up.
Kaepernick himself denied suggestions that his salary demands were preventing him from finding a new team, although it's hard to believe that a league which offers Jay Cutler $10 million a year to come out of retirement after being benched and cut by the Bears couldn't justify paying Kaepernick millions. Concerns about any financial hit after a Kaepernick signing don't make sense given that the vast majority of each team's income comes from fixed television rights, while Kaepernick still ranked 39th in jersey sales in August despite being a free agent.
It's true that you can find current players like LeSean McCoy and Joe Thomas who disagree about Kaepernick and suggest that he would be a distraction and isn't good enough to overcome whatever attention he might bring to a team. (McCoy also suggested that Michael Vick was 10 times the quarterback Kaepernick once was, which doesn't exactly lend credence to his case.) It's also true that both Tom Brady and Aaron Rodgers have come out in recent weeks and expressed their surprise that Kaepernick isn't employed, and they're pretty qualified to talk about modern quarterback play.
There just isn't a strong enough case for teams to pass on Kaepernick, given the desperate public outcries for useful quarterbacks and the nature of his protests. Even if you disagree with his stance, it's bizarre to contrast his peaceful political dissent as a crime that should keep him off rosters in a league in which even marginal players embroiled with confirmed or alleged incidences of domestic and/or sexual assault can sustain careers.
Watching so many NFL teams willfully make ignorant choices and then complain about the lack of quarterback options while leaving a clearly qualified candidate on the sidelines makes it seem like the quality of decision-making in the modern NFL is far worse than the quality of play.