Who are college football’s biggest fair-weathered fans?

Everyone knows that college football fans attend games less during losing seasons (for similar results for NFL, click here). But, let’s dive a little deeper into this effect, and see which teams and conferences are the biggest “fair-weathered” fans–if not for pure, unbridled interest, then to have potential ammunition for your next debate with a rival team fan.

Overall Effect

For the specific methodology, scroll to the bottom of the page.

Season attendance and win percentage were accumulated for each FBS team from 2007-2016 (courtesy of NCAA). On average, one loss equals a 507 person decrease in attendance (per game). This relationship is (gasp!) not too surprising, but it does help us better understand the exact size of the effect.

Losses and Attendance


A potentially more interesting question: which teams’ attendances are hurt most (and least) by losses? Results are provided in the table below.


Wins and Attendance- Individual Teams

A potentially even more interesting question: are there any factors that predict a team’s position on the list?

There are a wide range of teams whose attendances all are strongly affected by losses (e.g., Pitt, Louisville, Memphis, Purdue, Oklahoma St.), but, interestingly, they do not appear to have much in common. Data-wise there were no significant commonalities between teams high on the list, at least in the factors in my data set (results not shown here–they’re boring)…but more research may reveal some interesting insights.


Last, I examined the effect by conference.



Attendance and win percentage were accumulated from 2007-2016 (courtesy of NCAA)

To calculate attendance decreases per loss, I conducted an OLS linear regression for each team.

Stadium changes for each team were handled on a team-by-team basis. For example:

  • From 2007-2015: Stadium had a capacity of 50,000 and an average attendance of 50,000
  • In 2016: Stadium had a capacity of 55,000 and an average attendance of 55,000

In this case, I changed the 2016 stadium attendance (55,000) to the previous high (50,000). In more complicated cases, some subjectivity was required (please comment for details for a specific teams).











  1. Pitt, Louisville, and Memphis are in actual cities, with other sources of entertainment should the teams become less-desirable to watch.


  2. Northwestern is highly skewed due to the number of fans of other Big Ten programs that live in Chicago. For example, when Michigan plays at NU, the stadium is sold out and at least half the fans are rooting for Michigan. So my guess is that it is away fans of stronger programs that, on average, win against NU that are causing their attendance to go up. Not the strength of the NU fanbase (which definitely rides the bandwagon).


  3. One factor that is hard to take into account is that some teams (like Pittsburgh) have a minimal consistent fan base, yet are in a locale where it is easy (and much cheaper) for visiting team fans to come to the game and get tickets. So you may get a sell out for a big name game thanks to opposing fans, then crappy attendance the rest of the year.


  4. Sounds terrible and probably is but the man has a point about PITT. They have lots and lots of visiting fans that attend games that simply are not going to show up for subsequent games. Example, the Clemson giant win on the road. Terrible turnout for the next game (Duke). How does that graph explain that? The answer is bad weather and late season holidays


  5. Thanks for putting this together, but, there are so many factors (ranking of opponent, weather, consecutive home games, opponent’s fanbase, etc, etc) that go into this there’s really nothing you can take from it.


  6. Something to consider (make a few separate charts) would be if the next game was a BYE or away game, meaning a 2 week span as opposed to a 1 week span between home games. Also, consider a Thursday game/ night game/ and who the opponent is.
    For example: Clemson in 2016 lost 1 game (Pitt) then was on the road next week and had a night game the following week. Clemson in 2015 lost at a neutral sight for the national championship. The 1st game of the next season was away and the 1st home game was week 2 against a sun belt team (where attendance is already lower than other games).

    Interesting numbers none the less, but I feel like there is a lot more than what meets the eye. The article is a bit simple in retrospect.


    • Hi Jordan,

      Would such factors become moot with a sample size covering 9 seasons? For example, do you think any impact you might see from a Clemson loss followed by a road game and then a night game in one season could be balanced out by a game schedule in a different season that temporarily inflated attendance (perhaps a higher-than-average number of weekend games at ideal times)? I wonder if the impact of something like scheduling would neutralize with a large sample size because teams are bound to have schedules that both help and hurt their attendance over the years. Curious to hear your thoughts!


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s