Odds are, living in the US, you’ve seen a National Football League (NFL) game on TV at some point. It’s almost an eventuality, given the NFL’s complete dominance of the airwaves—in 2015, NFL games accounted for 46 of the top 50 most-watched television programs, and all of the top 25. Of those 46 games, 19 had more than 25 million viewers. The only thing that came close was the first Grand Old Party (GOP) debate last fall—what a beautiful, distant memory, when I didn’t check the FiveThirtyEight election forecast every hour.

This year is a different story. After seven weeks of game action, ratings are down across the board by about 11 percent. Given the sheer gap between the NFL’s viewership and that of literally everything else on TV, you could be forgiven for thinking that this isn’t a huge deal; the NFL could hemorrhage viewers and still lap contenders like the National Basketball Association (NBA), Game of Thrones, and whatever else people watch nowadays. But for a league that perennially deals with negative press over its deplorably haphazard response to problems of domestic violence by players, any sign of slipping dominance is likely inspiring panic in the hallways of the league’s New York headquarters.

Because the NFL is such a reliable ratings juggernaut, this sudden decline has already been extensively discussed: among the most commonly named culprits are the presidential election’s domination of the media landscape, backlash to a recent wave of political protests by NFL players, oversaturation of the TV landscape, competition from a particularly newsworthy baseball playoff, a decline in quality of play, and a lack of recognizable stars relative to previous years.

(Why would I write about such a talked-about topic? Obviously because I think everyone else is wrong and I’m right. In all seriousness, although some of the potential factors causing the NFL’s current ratings decline make sense, I’m less confident in some of these factors having the impact people think they do.)

For starters: I don’t think the election is killing the NFL’s ratings. While two of the three presidential debates scheduled for this fall did compete against NFL games, two instances of extremely low ratings (the first debate drew roughly 100 million viewers, while the NFL’s counterprogramming drew only 8 million) shouldn’t be enough to drive the across-the-board decline we’ve seen so far this fall. Furthermore, other games that didn’t have to compete against the debates drew low (though slightly better) numbers as well. And, although we know that NFL fans have a higher voter turnout index than average, I find it hard to believe that even an election as attention-grabbing as this one is pulling away statistically significant numbers of fans from the TV on Sundays, Mondays, and Thursdays. Of course, the election could be impacting NFL ratings in a more obscure way than by simply drawing away viewers—we’ll find out after election day, but I suspect the NFL’s problem is more complex than “viewers are distracted.”

Meanwhile, while this year’s World Series was more exciting than usual—or so I’ve heard, given that I’d rather watch paint dry than baseball—I don’t think baseball is capable of stealing too much of the NFL audience. While the ratings for the World Series beat those for Sunday Night Football once this year, that extrapolation doesn’t work for the entire baseball playoffs and the entire football season so far. For starters, it was the World Series for only one week—assuming that just because baseball’s championship series beat football in viewership once translates to what’s been happening on a much larger scale for eight weeks is more than a little analytically unsound. Beyond that, the viewership calculus for the NFL regular season and the MLB playoffs are fundamentally different.

The NFL’s regular season core viewership is much larger and much more geographically distributed than that of the baseball’s playoff viewership: while people living in Cleveland may prefer to watch their championship-contending baseball team than the eternal circus of breathtaking incompetence that is the Browns, that logic likely doesn’t apply to major football markets like New England, New York, and Texas that don’t have any stake in the World Series. Furthermore, while baseball schedules playoff games almost every night of the week, football only takes place on Sundays, Mondays, and Thursdays. Because of this, conflicts of interest for dedicated viewers of both sports are negligible except in rare cases, especially given that ratings numbers factor in same-day DVR viewings or internet streams, which should account for the normally-dedicated fans that do choose baseball over football.

While I personally don’t think that the recent wave of political protests by NFL players is adversely impacting ratings, I’m a little more cautious in claiming that this potential cause of decline is categorically untrue. There is some evidence that people who would normally watch the NFL have been boycotting games due to the league’s non-response to the protests: a Rasmussen poll from early October found that roughly one third of respondents said that they were “less likely” to watch games because of the protests. However, I would take this with a grain of salt—it’s only one poll, and there’s other evidence throwing its findings into doubt.

For example, while searches (all data on this is from Google Trends) for Colin Kaepernick has spiked this fall, along with other prominent players involved in his protest, none of them appear among the most-searched players in the last few weeks, and apart from a spike in search volume at the time his protest was first publicized in August, Kaepernick’s search volume has fallen back to being roughly equal to or lower than the search volumes for other well-known quarterbacks like Tony Romo, Ben Roethlisberger, Cam Newton, and Aaron Rodgers. Similarly, the geographic breakdown of searches on him shows that interest is mostly limited to California and Nevada—i.e., the areas where the most 49ers fans are. This pattern of search interest in a player being concentrated to their team’s home region holds true for every other quarterback (QB), even Brady. It seems that although the protest made Kaepernick more famous than before, it hasn’t driven any particularly long-lasting or notable interest relative to that sample of other QBs.

Furthermore, although the NFL hasn’t taken any official steps to punish the players involved in the recent wave of protest, TV networks have, on their own, attempted to placate conservative viewers by cutting away from players who kneel during the National Anthem. Finally, the actual viewership numbers for particular games don’t seem to bear out the idea that protests by viewers against Kaepernick are hurting the NFL’s TV numbers. Games involving either Kaepernick or other prominent protesters—like New England Patriots tight end Martellus Bennett—don’t draw particularly lower viewership than other games. Viewership, instead, seems to be correlated with the presence of popular teams—the Cowboys, the Packers, the Patriots—to teams known to be good this year, or with major stars. While the theory that the election is killing football ratings has some basis in empirical evidence, there’s very little out there supporting the idea that protest is hurting the NFL.

What is hurting the NFL’s ratings, then? I believe that a combination of decreased quality of play and a relative paucity of attention-grabbing stars, paired with the NFL’s oversaturation of the TV market—between Sunday-morning games played in London, and Sunday, Monday, and Thursday night games, the NFL now broadcasts a significant number of its games outside its normal Sunday-afternoon programming bloc—seems to be causing a general decline in viewership, especially among younger fans.

I’ve already written about a relative decline in the quality of quarterback play this year, but in addition to a sinking passing game, the league has become increasingly reliant on younger players, for many of the same structural reasons that good quarterbacks have become scarce; essentially, the economics of the league make it prohibitively expensive to hang onto middling players and try to develop them as they age. So instead, teams are taking on more and more young, inexperienced players out of college who may be unfamiliar with many aspects of the NFL game, leading to an overall decline in quality. While these factors have been working for several years now, it’s possible that the decline in the quality of the NFL’s on-field product has finally gotten to the point where viewers are starting to get bored. It also doesn’t help that referees have become more active than ever—no one wants to watch endless official reviews, or see what seems like every other play get called back on a penalty.

Meanwhile, the league is at somewhat of a low point in terms of superstar players—the ones who sell jerseys all around the country and draw even unaffiliated TV viewers to tune in to their games. Tom Brady was suspended for the first four weeks of the year (since he’s returned, the ratings for Patriots games have skyrocketed); Peyton Manning, Marshawn Lynch, and Calvin Johnson all retired last year; Tony Romo, the starting QB for the Dallas Cowboys, the most popular team in the country, is injured for an indefinite period of time; JJ Watt, the most dominant defender in the league, is out for the season. Meanwhile, other stars like Cam Newton and Aaron Rodgers are going through significant patches of subpar play. While top-level talent is overflowing at positions like wide receiver and along the defensive line, there will inevitably be an adjustment period as new players become famous and fans search for their new favorites.

It doesn’t help that the NFL has begun to take an extremely hard line against outbursts of joy and individuality among its players, further limiting its ability to develop popular, bankable stars that will draw viewers to games they wouldn’t ordinarily have a rooting interest in. “Excessive celebration” penalties (yes, I know that’s a ridiculous concept for a league that sells what amounts to a supersized playground game) are seemingly being doled out more than ever, and stars like Antonio Brown of the Pittsburgh Steelers are being fined for doing cute, endearing stuff like wearing cleats with pictures of their kids’ faces on them. For some reason, the NFL has decided that the best way to sell its ever-shrinking stock of engaging, watchable players is to make it seem as if they have no personalities and never have fun playing football. This is because the NFL is ruled by a cabal of ignoramuses.

Finally, there’s the issue of oversaturation. I’ve written about how NFL creep in the past—essentially, the creeping overproliferation of the NFL’s product, particularly playing games after short weeks of rest or after very long plane trips, as well as the year-round media coverage of relatively unimportant parts of the football landscape like offseason minicamps—harms the game by putting undue strain on players and by cheapening the overall product and the discourse around it, which in turn makes us, the viewing public, less likely to want to care in the first place. This by itself isn’t a massive problem. It genuinely sucks that Thursday night games are bad for the health of players, but people still watch the games; it’s undeniable that treating the NFL as a year-round fascination worsens the quality of media coverage, but people (myself included) still read endless pre-draft coverage in the otherwise football-less months of March and April. In a perfect world, the NFL offers a compelling-enough product to survive oversaturation.

However, as I just spent four paragraphs pointing out, this isn’t a perfect world. The NFL is dealing with crises of game quality and player marketability, both basically self-inflicted, and the sport’s current oversaturation worsens both those problems. People might be happy to wake up early and watch a game in London when recognizable stars are involved, but if the game’s a dud, they might give up on the concept—similarly, ratings for games in standalone timeslots (Sunday, Monday, and Thursday nights) suffer when unaffiliated viewers have no real reason to tune in. There’s a very real danger that the league has overshot its environmental carrying capacity, so to speak, and is due for a correction.

Unfortunately for the NFL, there’s no easy fix here. The quick solution would be to drastically scale back on excessive celebration penalties and reduce the role of officials in the game, and hope that the return of injured stars later this winter and next year is enough to stop the bleeding. Much more significant structural changes would be necessary to address issues with quality of play, and scaling back NFL creep is pretty much out of the question for a league that hates to admit defeat. There, is, luckily, the possibility that I and all the other commentators freaking out about sinking ratings are wrong—maybe the election and World Series will wrap up and everything will go back to normal. Maybe all the injured stars will come back better than ever, and all the underprepared young players will leap forward into competence just in time for the playoffs. But if there is a fundamental and intrinsic problem with the NFL that’s causing low ratings, I don’t trust the league to fix it.