The Number Everyone Keeps Repeating
What if the number that convinced everyone college football is dying is just the wrong number?
FBS broadcasts on Nielsen-rated networks averaged 1.7 million viewers through the first four weeks of this season, down 11 percent from the same stretch of 2025, a figure Front Office Sports reported last month that has since been recycled into a verdict: the sport’s audience is collapsing. Only two games cracked 10 million viewers in that span, Texas-Ohio State’s 14 million on ABC and Ole Miss-LSU’s 10.9 million, compared to four games clearing that mark in Week 1 alone a year ago. Oklahoma-Michigan drew 8.2 million on Fox. Wisconsin-Notre Dame pulled 8.1 million on NBC. Solid audiences, just smaller than the ones everyone got used to measuring the sport against.
Here’s the part that didn’t travel nearly as far. Fox Sports’ own president of insights and analytics, Michael Mulvihill, posted that through September, total time spent watching college football across every outlet is up 0.2 percent compared to a year ago.
https://twitter.com/mulvihill79/status/2105420564945084489
Nobody built a headline around that, even though Fox has every incentive to spin its own numbers the gloomier direction if the data supported it. The audience supposedly fleeing college football spent marginally more total time with the sport than it did last year. That single fact is the whole case for this piece, and it means the 11 percent figure is measuring something much narrower than “demand.”
Is College Football Viewership Actually Down in 2026?
Average per-game viewership for college football is down 11% in 2026, but Fox Sports’ own data shows total time spent watching the sport is up 0.2% year over year. The audience didn’t shrink. It spread across far more broadcast windows and streaming platforms than in past seasons.
Those two numbers describe two different layers of the same market, and only one of them tells you whether people still want to watch college football.
Where the Extra Games Went
Start with the playoff. College football went through the playoff’s expansion to 12 teams, up from four teams and three total games as recently as 2023 — eight new playoff broadcasts that didn’t exist a few seasons ago. Layer on top of that a new Big Ten media rights deal worth roughly $1.1 billion a year, split across three partners. Fox alone is carrying 137 college football games in 2026, including 89 Big Ten games. CBS airs another 14 to 15 Big Ten games. NBC has 14 to 15. The Big Ten Network carries dozens more on top of that. That’s before ESPN, TNT, USA, Peacock, Paramount+, and The CW, nine traditional networks splitting the sport, plus a stack of standalone apps (ESPN Unlimited, Fox One, Peacock, Paramount+, Disney+) that didn’t exist in anything like this form five years ago.
Translate that into plain terms: the sport multiplied the number of rooms its viewers could be sitting in at any given moment. A fan who had three windows on a Saturday afternoon a few years ago now has nine, plus four separate apps competing for the same four hours. Spread an unchanged audience across three times as many simultaneous broadcasts and the number you’d expect to fall first is average viewership per window. That’s exactly, and only, what fell.
The Honest Counterargument
There’s a real objection to this, and it deserves more than a dismissal. Brobible points out that total viewing time is partly a function of available inventory, not just audience size. Unlike the NFL, which plays a fixed slate of games every year, college football added more hours of programming in 2026. Mulvihill’s 0.2 percent figure is measured against a much larger pool of total hours than it was twelve months ago, so a modestly engaged audience spread thin across that pool could, in theory, still produce a flat-to-up number that looks healthier than the underlying engagement actually is.
Grant the point. It still cuts against the decline story rather than for it. If the audience were genuinely shrinking, it would have to do more than hold steady on the games it already watched last year — it would have to absorb dozens of brand-new broadcast windows, the extra playoff rounds, and the additional Big Ten inventory without the total falling. A genuinely eroding fan base wouldn’t have enough people left to fill that much new space — it would retreat, not spread out to cover more of it. The fact that total time held roughly flat while the sport handed fans several times the live-game inventory it carried five years ago is itself a sign of real demand, not an illusion created by a bigger denominator. I ran the comparison against the NFL’s own number for the same stretch, down 0.7 percent on a fixed game count, and the gap only makes college football’s flat-to-up figure look sturdier.
What Would Actually Prove the Sport Is Shrinking
None of this means the ratings conversation this fall is manufactured from nothing. Awful Announcing has reported a real appetite for a decline narrative, tied to off-field drama that has nothing to do with how many channels are carrying games: congressional hearings, the proposed Protect College Sports Act, another churn cycle through the transfer portal and the sport’s NIL spending gap, and a Big Ten-SEC realignment standoff that has made scheduling feel adversarial even to casual fans. A weaker marquee schedule through September didn’t help either; four openers clearing 10 million viewers a year ago versus two this year means fewer appointment-television moments to anchor a Saturday around.
Those are legitimate threats to the sport’s audience. They’re just a different threat than the one the 11 percent stat gets dragged in to support. The number that would actually prove college football is shrinking is Mulvihill’s own number going negative, total time spent watching the sport falling even as the inventory keeps expanding. Until that happens, the programming decisions worth making are about how thin the audience has been sliced across too many windows, not about chasing bigger games or cutting inventory to chase a crisis that the sport’s own data says hasn’t arrived yet.
