The Great American "Pizza" Map

Having spent considerable time over the past months analyzing the distribution of a number of strangely juxtaposed social phenomenon (ranging from religion to bars), today's map should come as no surprise to anyone. After all, isn't the comparison between the number of user generated references to "pizza", "guns" or "strip club" an obvious one to make? Perhaps not, but we're doing it anyway.

To our knowledge this is the first time that this particular set of fractures within American society have been mapped. And it is likely to be the last time as well, so don't blink.

The Great American "Pizza" Map

The map reveals that America seems to be largely pro-pizza with the red dots of pizza spreading across the landscape (the Midwest and Northeast seem particularly well endowed) like a rich and robust tomato sauce. Perhaps there is something after all that everyone can agree upon. (Although the unity behind pizza would likely break down over a fight between "New York" and "Chicago" styles.)

Moreover, there are a number of clusters -- represented by green dots – with more references to guns than pizza. We're not quite sure what life there is like (as we live in a particularly large swath of sauce) but it does give one pause for thought. Are these patches of green olives? Peppers? Jalapeños? Pesto? In any cases, these clusters are predominately rural areas – where apparently there are more guns than pizza.

Lastly one sees a few scattered locations where strip clubs are more popular (we're looking at you Las Vegas) than either guns or pizza. Much like the token scattering of pineapple that one is forced to endure on pizza (there's always one person who suggests it!), strip clubs represents a decidedly marginal activity in the aggregation of human experience illustrated in this map.

In any case, this post has made us hungry so we're off to grab a slice.

Informal versus OFFICIAL Fun & Vacation

In some of our previous explorations of user-generated representations of place, we looked at the "funnest" places in North America. Although Toronto was by far the "funnest", when normalized by population, Cape Cod Bay looked like the best place to go for a good time (because, you know, nobody actually lives in the bay). In order to get a better grasp on where the fun is and from whence it comes, we compared our previous data on user-generated Google Maps placemarks mentioning the word "fun" with listings in the Google Maps directory that mentioned "fun". In short, how do informal notions of fun (user-generated) compare to OFFICIAL fun (directory listings).

Locations of Informal (user-generated) and Official (directory listing) Fun

Perhaps unsurprisingly, Toronto again appears prominent because of the prevalence of informal, as opposed to official, "fun". Toronto is largely an anomaly among urban areas in North America as most cities are decidedly tilted towards fun of the official type variety. Clearly we need to do some fieldwork in Toronto!

Likewise a considerable area in the rural western U.S. also displays a favoritism towards user-generated/informal fun. Upon further examination, many of the areas displayed in orange above correlate to the locations of US National Parks and National Forests. Because there would be few, if any, directory listings in these protected areas (as opposed to urban areas, which would have a much larger directory), user-generated placemarks are more prevalent than those generated automatically using sources like the Yellow Pages.

Also of interest is the high correlation in the differences between user-generated and directory content for "fun" and the differences between user-generated and directory content for "vacation" (see below). Here again wide swaths of the western US have more user-generated vacation reference than directory content, despite the general trend across the U.S. and Canada being the opposite. One site that shows up prominently as a cluster for user generated fun and vacation is Wall, South Dakota, home to the famous Wall Drug. Many a weary traveler driving across the country on I-90 have sought a few hours of refugee/distraction here. And apparently many have chosen to document it as well.

Locations of Informal (user-generated) and Official (directory listing) Vacation

So even if only in the relative prevalence of user-generated representations of places that are both fun and good for vacationing (don't they go so well together?), rural areas have found their place in the American cyberscape.

The Beer Belly of America

At FloatingSheep, we're willing to search for and analyze almost anything that falls within the realm of human experience. Sometimes this is mundane (pizza) and sometimes it is contentious (abortion) but most of the time it falls somewhere in between. Such as, where can I get a drink?


Total Number of Bars


We were quite surprised, however, when we did a simple comparison between grocery stores and bars to discover a remarkable geographically phenomenon. We had expected that grocery stores would outnumber bars and for most parts of North America that is the case. But we could also clearly see the "beer belly of America" peeking out through the "t-shirt of data".



Starting in Illinois, the beer belly expands up into Wisconsin and first spreads westward through Iowa/Minnesota and then engulfs Nebraska, and the Dakotas before petering out (like a pair of love handles) in Wyoming and Montana.

The clustering was so apparent that we wanted to check how it compared to the "official" data on this activity. So we gathered 2007 Census Country Business Pattern on the number of establishments listed in NACIS code 722410 (Drinking places (alcoholic beverages)) and divided by Census estimates for state population totals for 2009 and found remarkable correspondence with our data.

On average there are 1.52 bars for every 10,000 people in the U.S. but the states that make up the beer belly of America are highly skewed from this average.

Rank	State	Bars per 10,000 Population
1	North Dakota	6.54
2	Montana	6.34
3	Wisconsin	5.88
4	South Dakota	4.73
5	Iowa	3.73
6	Nebraska	3.68
7	Wyoming	3.4

Another slice of the Google data which shows the relative number of bars in a location further confirms this concentration. So looks like Wisconsin is your best bet.


Specialization in Bars

Google's Geographies of Religion

“Religion is probably, after sex, the second oldest resource which human beings have available to them for blowing their minds”
Susan Sontag

Following up on the earlier discussion of the user-created geographies of religion, the following maps simultaneously display all four religious references (Allah, Buddha, Hindu, Jesus) in order to visualise distinct religious cyberscapes. Below we see the data on a global scale. This map clearly mirrors many of the expected religious geographies of the offline world: references to Allah being most prominent in the Middle East, references to Buddha being most prominent in East Asia, references to Hindu being most prominent on the Indian subcontinent and references to Jesus being prominent in Europe and much of the Western Hemisphere.

Interestingly, there are are no large-scale homogeneities in the data and this reflects the sometimes scattered nature of religious practice in the world. Looking at the below map of user-created religious references in Europe, it can been seen are a significant number of places (e.g. parts of Switzerland, Germany, the UK) in which there are more references to Buddha than any other religious terms. Likewise there are parts of Belgium and France with a dominant number of references to Allah, and parts of the UK with a dominant number of reference to Hindu. (The cluster of Hindu references on the Estonian islands of Saaremaa and Hiiumaa is tied to a village named Hindu rather than religious practice). Also of note is the transition of religion as one moves eastward and southward with references to Allah becoming more prevalent in Muslim North Africa and Turkey. However, one can also see how this is far from monolithic with references to Jesus also sprinkled throughout this region as well as strong clusters in Israel/Palestine as well as within Armenia.

In Asia a similar amount of diversity can be seen. The United Arab Emirates is a particularly interesting example. While officially a Muslim country, Indians make up the largest demographic presence and the dominance of references to Hindu (rather than Allah) is likely a reflection of this fact. Likewise the Malay Peninsula and the Indonesian archipelago (particularly the island of Java) illustrate the complexity of religious practice in this region. References to Buddha, Allah and Hindu are all in evidence on Java. Other examples include the predominately Buddhist nation of Sri Lanka with some Hindu areas to the North and the difference between Pakistan (more Allah) and India (more Hindu).



Finally, it is informative to include one additional map with this set. Here we included placemarks that reference the word ("sex"), a popular and international used term with very different connotations than the religious keywords used earlier. The purpose of including this term is to compare user interest in religion to user interest in sex. If (as some say) the three topics to avoid in polite conversation are religion, sex and politics it seems only right that this Internet blog (the antithesis to polite conversation if there every was one) takes on the question. Sadly, the inclusion of politics will have to wait until another day.

In Asia there are very few places where there are more references to sex than Allah, Buddha, Hindu or Jesus.

Constrasting this is Western Europe (especially the UK and Scandinavia) and North America (especially the East and West Coasts) there are more references to sex than any of the four religious terms that we searched for. The distribution in the U.S. mirrors our early maps of the virtual bible belt and church-bowling-firearms-strip clubs. There are, however, exceptions such as the Iberian countries of Spain and Portugal which continue to show more references to Jesus.

So it would seem that Susan Sontag's observation has some merit, at least in the European and North American context.

Metropolitan Level Maps of Cyberscapes

Over the next weeks we will begin to post a range of metropolitan level maps of user generated Google placemarks. These maps illustrate one of the "cyberscapes" of these cities, i.e., the cloud of geo-coded data in cyberspace which provides an additional layer for human interaction.

Cyberscapes, consist of multiple layers, e.g., Google placemarks, Wikipedia articles, geotagged Flickr, Twitter Tweets, etc., but the metro maps are limited to user generated placemarks.

The examples of New York's and Capetown's cyberscapes below illustrate how these cyberscapes vary over location and topic. The white box in each map indicates the area for which we have data. Stay tuned....

All User Generated Placemarks
New York City, January 2009


User Generated Placemarks Referencing Crime
New York City, January 2009


All User Generated Placemarks
Capetown, South Africa, January 2009



Upcoming Cites: Baghdad, Beijing, Buenas Aires, Capetown, Chicago, Dublin, Hong Kong, Jerusalem, Lagos, Lexington (KY), London, Los Angeles, Mexico City, Mumbai, New Orleans, New York, Pyongyang, Rio de Janeiro, San Francisco, Tallinn, and Washington D.C.