Posted August 30, 2013

Map matters: Facebook map shows geographic fandom of preseason Top 25

Important Map Matters
College fan map

(via Facebook)

In a neat little project, here’s a map of college football fans broken down by region, using data from the number of “likes” a team has on Facebook. Maps and data are fun, and so is football. It’s always important to be playing this while talking about cartography-related endeavors.

Obviously, this map is a bit misleading since it only features schools in the AP Poll Top 25, so it looks like West Virginia is ALL OHIO STATE fans, despite the fact that we know a lot of West Virginia roots for WVU and Penn State and Pitt (and Marshall!).

Here are some other observations:

• Speaking of Ohio State, how is it reaching into Alaska? I’m from Ohio, and there is no way I’m moving to Alaska, ever. I’m sure it’s nice, but I’ve had my fill of cold. I imagine moving to Alaska is like asking everyone in a room to punch you in the stomach, then going into another room and asking everyone to punch you in the face. At some point, it’s just too much.

• The Gators are Florida’s team. Which is a bit sad, in a way, but I guess Ryan Lochte is a much bigger star than we all originally thought.

• What is going on in that one county in California? Why are there Texas fans there?

• Nebraska’s reach extends a lot further than I thought it did.

• I’m a little surprised Notre Dame doesn’t have a bigger portion of the map than it does. There are a lot of Irish fans all over the U.S. I figured there would be gold sprinkles all over the map.

What else do y’all see? Any giant realizations or insights into #maps? Leave them in comments.

9 comments
splooz
splooz

Clearly, this fool has never been to Alaska.

Isle_of_Orleans
Isle_of_Orleans

I think this is interesting because of its limitations. For instance, you'd never otherwise know that folks in Mississippi prefer LSU to Alabama unless Ole MIss and Miss State were eliminated from the running. I hope they keep putting this out over the course of the season; it'd be interesting to see how it changes as schools  move into and out of the top 25.

Here's something a little more rigorous:

http://commoncensus.org/sports_map.php?sport=5

JEPD
JEPD

One COUNTY for Clemson.

vieaa
vieaa

I see justification for the Big 10 expansion. Nebraska has a big draw, and Maryland+Rutgers could do really well if they get competitive.

John NoLastName
John NoLastName

I personally believe that US Americans, like such as, don't have maps like this and the Iraq. Like such as.

Ron aka Buddy
Ron aka Buddy

The first problem is using crap like Facebook.

KJKSLC
KJKSLC

This is a horrible map, with skewed data based on bad sampling and small sample sizes.

First, because it only takes into account the top 25 teams, you're way overestimating the reach of some schools.  Second, because it represents only the most favorite team per county, it way underrepresents others.  Third, because it's county-by-county, large swaths of area might be represented by a very few votes.  Fourth, since it uses facebook likes, that's a very selective and not indicative sample.

This shows why you're surprised at Nebraska's reach and why you don't see more Notre Dame.  Let's use Iowa as an example.  For school loyalties, the state is pretty evenly divided between Iowa and Iowa State, with probably a small edge to Iowa.  But neither of those teams is in the Top 25. Nebraska fans are a distant third, perhaps 10% of the population (but higher in Western Iowa).  But because there's not a a concentration of fans of any other school, those 10% are the highest bloc and it looks like Iowa is 75% Husker fans.  Similarly, among my friends, maybe only 5% were Notre Dame fans.  That is still a lot of fans statewide, but it won't overcome the 10% who are Nebraska fans.  I suspect that's true throughout the country -- 5%-10% of all college football fans in a given region are Notre Dame fans (certainly down from a few decades ago) but none would get over the hump to be the "leading vote getter" in any particular locale.

Your comment about the county in California demonstrates the bias and sample size problems in the survey.  That's Madera County -- not particularly small at about 150,000, but also not particularly big in California standards and not close to any major college football center.  All you would need would be a couple of businesses that hire UT graduates, and a population that is otherwise largely divided between Stanford and USC, and boom! a crazy anomaly.  


Sorry to be a wet blanket, but I'm just not sure this chart tells us much.

howieloader
howieloader

@KJKSLC Exactly. This is what happens when people who have no idea about statistical analysis get involved in things they're not qualified to handle...

NeMoeImya
NeMoeImya

@KJKSLC I was going to reply, but you basically already did it for me.  This map is borderline useless because of the data set it uses.


It would have been very interesting if you included, say, all of the BCS conference schools.  That might actually tell you something about the size/influence of relative fanbases, but not this.