Writers of Pro Football Prospectus 2008

19 Sep 2011

Geography of College Football Fandom

Nate Silver on the New York Times college football blog explores the geographic allegiances of college football fans, a data-rich perspective for the current and shifting conference landscape.

Posted by: Brian Fremeau on 19 Sep 2011

15 comments, Last at 23 Sep 2011, 1:27pm by Aaron Brooks Good Twin


by Guido Merkens :: Mon, 09/19/2011 - 4:16pm

This is an extremely informative article, but it misses one key point with relevance to conference expansion: obviously television money is driving expansion, and simply counting the number of fans of a particular team in a particular market ignores "unwilling" subscribers.

Let's take Rutgers as an example. Rutgers has shown interest in joining the Big Ten, and although it only brings 600,000 of its own fans in the New York market, according to these numbers 20% of the college football fans in New York like Rutgers. Thus, there may be enough Rutgers fans in the New York market to force Time Warner to put the BTN on its standard cable package, or at least the general sports tier. If they do this, then not only will the 600,000 Rutgers fans in NYC pay for the BTN, but 2.4 million fans of other teams will pay for the BTN as part of a general sports package.

In this way, although recognizing that some less populous areas like Alabama or Omaha have a higher concentration of fans of one team, adding a school in a populous market like NYC, even if they have to share the market, can be more valuable because you rope in a number of subscribers who don't care about the new team but will nevertheless pay for the sports package.

by zlionsfan :: Tue, 09/20/2011 - 9:06am

I think that is intentional. There are cases where footprint expansion is more valuable than fan adoption (like Syracuse and Rutgers), but it's a reasonable approximation otherwise. The SEC's desire to add Texas A&M still makes sense, and after all, the Big Ten made a lot of noise about considering pretty much anyone, but they ended up adding a relatively small market with a rabid fan base.

After all, it's no different than the rest of Silver's stuff or the work FO does. It's meant to be a guide ... I don't think he means to suggest that Navy would be more valuable to the ACC than Wake Forest, for example. It just gives us another factor to consider as we play the Realignment Projection Game. (Any fantasy leagues available for that yet?)

by mm (not verified) :: Mon, 09/19/2011 - 4:18pm

Interesting reading.

I'd like more description of how he "adjusted" the CommonCensus info. While I like looking at that site once or twice a year, it certainly has been affected by people posting the weblink on certain schools fansites over time.

by Josh :: Tue, 09/20/2011 - 12:02am

This article is complete garbage.

by andrew :: Tue, 09/20/2011 - 8:21am

THat map won't work for me.

by Southern Philly :: Tue, 09/20/2011 - 10:58am

This is useless. The methodology is strange to put it mildly.

"One way to estimate the regional variances is to look at Google search traffic. For instance, according to Google Insights for Search, the term “college football” is searched for about 5 times as often in Birmingham, Alabama as it is in New York City, relative to overall search traffic."

That's a really bizarre way to utilize search traffic. People who follow college football already know where to look for the information they want, they're not going to google "college football."

The map he links to says, right off the bat:

"Important: this map is highly inaccurate and should be understood only as a demonstration, and not as any kind of reliable data yet"

And yet he uses it as a starting point.

The results are totally off. Georgia Tech ahead of UGA in fandom puts the entire thing into question. If that wasn't bad enough, he has Clemson as the 10th most popular program in the nation.

by Bad Doctor :: Tue, 09/20/2011 - 11:38am

Totally agree on the Google map ... and yet it looks quite reasonable, doesn't it? Biggest loyalties in the SEC and Big 12 areas, followed by Texas and the Big Ten, with the worst scores in the always non-sports crazy Pacific coast and decidedly NFL-heavy Amtrak corridor. I think Texas looks a bit low, but other than that it looks believable.

I would think a better method would be to gauge all "football" searches, then depress the number by all "nfl" searches. But maybe using the simple "college football" tracks the activity of very casual fans or non-fans who are looking for gifts for loved ones, which ends up being a fair measurement in its own right.

by Southern Philly :: Tue, 09/20/2011 - 12:16pm

"Totally agree on the Google map ... and yet it looks quite reasonable, doesn't it?"

Not from my direct experiences, Georgia and Alabama in particular. And the color schemes are messed up. UGA and FSU are the same color so it's tough to tell what's supposed to be what near the border. The marker for Toledo is in Illinois, the marker for Miami of Ohio is in northeast PA, the marker for Temple is in central Missouri, Northwestern's is in Wyoming... it's bad.

"But maybe using the simple "college football" tracks the activity of very casual fans or non-fans who are looking for gifts for loved ones, which ends up being a fair measurement in its own right."

It's a terrible search parameter, especially when you're trying to map out the actual fans, not the casual fans. Nobody that follows a sport searches for only "college football" or "baseball" or whatever. How many times have you ever searched just "football?" I'm guessing never.

by DisplacedPackerFan :: Tue, 09/20/2011 - 2:47pm

The way the google metrics work if you searched for "college football scores" it would count for this. As long as college and football are in the search used it would count. "leading rushers college football" "college football stadium sizes" "college football injuries" "tickets college football" all count. It's not looking for exact matches, it's more or less just a grep or regex parse for those keywords within any specific search.

That being said I'm not saying it's a great metric, but as the article mentioned without a specific study, it's a related data set. How strong the correlation is, I'm not sure, but it could be tested if you get some more specific data. Since revenue data is also used the die hard fans who don't hit google are likely counted because they probably went to a game or bought something so they are getting counted that way. I would have liked to see TV data integrated as well. But the thought process he used to try and get something useful out of the data at hand isn't completely asinine.

As to the Georgia / Georgia Tech thing. I spend most of my time in the midwest so I'm not really in that market. However, I actually know several GT fans, I don't know any Georgia fans. They are all older (35+) and I recall GT getting national recognition many years back. Since college football fans are loyal, I could believe that there was some surge in fan base 20-30 years ago that would still impact current fan stats. I suspect that Notre Dame's fan base is shrinking as again I know very few young ND fans, but I know a lot of 50+ ND fans.

Again though I wouldn't really have a problem with a better study showing UGA with more fans than GT, I'm just saying I can believe the other way.

by JonFrum :: Tue, 09/20/2011 - 6:21pm

Don't believe the other way. I lived in Athens for five years. Much of the Georgia business/professional/legal establishment is 'a bulldog.' Georgia is the state school. Georgia Tech is a local school that represents the Atlanta area. Uga the dog is more famous in Georgia than any GT player or coach. Not close.

by Aaron Brooks Go... :: Fri, 09/23/2011 - 1:27pm

Even John Heisman?

Density seems to be a function of proximity, school size, and relocation pressure.

The Megalopolis (Richmond-Boston) doesn't have strong loyalties, because many people aren't from there, and Maryland and BC are basically the only Div-IA schools worth mentioning.

Alabama has a huge density because they own their local market, are the biggest school around, and people never move in or out, and thus there's no dispersal of the fan base. A school like GT, which is primarily a grad school, widely-disperses its graduates, who unlike Georgia grads, can read.

by Southern Philly :: Tue, 09/20/2011 - 11:59pm

"As to the Georgia / Georgia Tech thing. I spend most of my time in the midwest so I'm not really in that market. However, I actually know several GT fans, I don't know any Georgia fans."

I live in Atlanta, the most pro-GT part of the state. And even in the city, it's UGA by a landslide. To say that GT has 500k more fans than UGA is just plain incorrect.

by DisplacedPackerFan :: Wed, 09/21/2011 - 9:58am

Fair enough. I was just pointing out the fan counts were national, and that from my outside perspective it matched as I had seen GT fans in WI, MO, MN, IL, ND while not seeing any Georgia fans. As mentioned it wouldn't take much to convince me otherwise.

by Alexander :: Tue, 09/20/2011 - 11:52am

I have never googled "college football" nor have i googled "Pitt Football" or anything similar. However, I am a Pitt fan who watches all the games.

The New York Slimes strikes again.

by Jetspete :: Tue, 09/20/2011 - 2:44pm

i hate the ny times, but this is a tremendous article. thanks for the link