November 12, 2014

The (Rust) Belt of Basic-ness? Mapping the Pumpkin Spice Latte

As fall gives way to winter, we're all left clinging to the best vestiges of the passing season: the changing leaves, college football, temperatures above freezing and, for many of the most basic amongst us, the pumpkin spice latte. Debuted by Starbucks in 2004, and featuring no actual pumpkin content, the pumpkin spice latte has become a staple of fall, with Ugg boots and yoga pants-wearing women practically crawling out of the woodwork to get their hands on the thing. And while Starbucks touts that over 29,000 tweets have mentioned #pumpkinspice since 2012, we suspected there was much more to the story of the pumpkin spice latte [1]. Despite the fervor, we noticed that there's been no definitive tracking of the geographical expansion of the pumpkin spice latte as it seeks to colonize the world of regular, everyday people drinking plain ol' coffee.

Searching only for the latest manifestation of the pumpkin spice phenomena, we collected all geotagged tweets in the continental United States for September and October 2014 with references to either "pumpkin spice" or "#psl", yielding a total of 19,537 tweets. But rather than simply mapping the basic distribution of these tweets, we've instead normalized this data by tweets referencing "coffee" during the same period. Using a 25% sample of all of these coffee-related tweets -- totaling 42,696 tweets -- aggregated to hexagonal cells, we calculated the odds ratio at the lower bound of the 95% confidence interval in order to provide a bit more context and account for any number of biases within the data. Using this measure, we've identified those places with greater-than-usual numbers of pumpkin spice latte tweets relative to those tweets referencing coffee (orange), and vice versa (purple), as seen in the map below.

References to Pumpkin Spice Lattes relative to References to Coffee [2]

Based on our binary classification, it's evident that the vast majority of the country has stuck with their preference for coffee, even during the PSL's peak season. But given that our interest is in mapping the prevalence of the PSL in particular, we want to pay closer attention to the smattering of orange hexagons in the map. While there are no definitive clusters of PSL-related tweeting, if you squint your eyes you can just barely visualize a belt of pro-PSL places stretching from St. Louis up to Chicago, and from Cincinnati up to Toledo and Detroit, and from Cleveland to Pittsburgh, what we've termed "the basic belt". While this belt roughly corresponds to the vernacular region of the Rust Belt, Ohio in particular sits at the center of this pumpkin spice-loving portion of the country, representing the buckle on our belt [3]. Given this clustering of PSLs, we suspect that the Buckeye State might well be on its way to becoming the Pumpkin Spice State.
----
[1] Well, actually, Renee Kaufmann had this hunch. All credit for the idea behind this post goes to her.
[2] Sorry about the Web Mercator projection, y'all.
[3] Can one still wear a belt with yoga pants?

October 27, 2014

Geographies of Grits

Throughout the history of this blog, we’ve mapped any number of geographically-specific social phenomenon. But often times, we’ve been drawn to mapping things associated with the American South, whether because it’s arguably the most distinctive cultural region in the United States or because all of us have lived on its outskirts for some time or another, we’re not quite sure. But if you had to choose just one region to map using social media data, the South is probably a good place to start.

Continuing this persistent obsession, we decided to map one of the South’s most prominent culinary traditions: grits. As such, we collected all geotagged tweets in the United States from June 2012 to September 2014 mentioning the strange (read: awesome) ground-corn porridge-like dish, totaling around 64,000 tweets. Keeping in mind that geotagged tweets still represent only around 2-3% of all tweets, this figure represents a breakfast table conversation of several thousand tweets per day, and highlights the ability of this kind of social media data to provide insight into a particular cultural phenomenon that is relatively more difficult (though certainly still possible) to measure through more conventional means.

The map below represents a normalized visualization of grits-related tweeting throughout the continental United States. Using a grid of hexagonal cells, the number of grits tweets were normalized using an odds ratio by a random sample of tweets from the same time period. In this measure, a value of 1 signifies that there are exactly as many grits tweets in a given location as one would expect according to the baseline measures of tweeting, with values greater than 1 indicating that there is a greater predominance of grits tweets than one would otherwise expect. In effect, this analysis cuts out the potential for these maps to simply reproduce maps of population density, honing in on the actual phenomena at hand.

Mapping the 'Grits Belt'

Indeed, here you can see that while the South in general demonstrates a general preference for grits over the rest of the country, it is actually a relatively small number of coastal localities in the low country that have the strongest connection to grits through social media. While New Orleans represents something of an outlier in the far corner of the South, there is also a consistent band of concentrated grits tweeting stretching from just north of Charleston, South Carolina down through Beaufort (though seemingly skipping over Hilton Head, a popular tourist destination that might be understood as relationally disconnected from much of the rest of the distinctly southern culture surrounding it) and Savannah, Georgia, all the way to Brunswick.

In general, this map demonstrates the general potential of this kind of method to locate geographically-specific cultural practices in space, as well as the notion that these kinds of maps can reinforce the persistent connectedness between virtual representations of the world and people’s everyday lives and material practices. But there is more that we can do with this data by putting it into relation with other datasets. The map below does just that, by comparing our existing dataset of tweets mentioning ‘grits’ with all geotagged tweets during the same time period that mention ‘oats’. We again employ the odds ratio measure, but rather than comparing using a baseline population of tweets, we use the oats-related tweets to normalize our values. In this analysis, values less than 1 signify a preference for oats, while values greater than 1 represent a tendency towards grits. Not only does this comparison continue to affirm the identification of a ‘grits belt’ in the South, but it also highlights other areas of the country – an ‘oats oval’ stretching from the Northeast to the Midwest – that stand in stark contrast to the southeast in terms of digital porridge discourse.

Grits vs. Oats: The Emergence of the 'Oats Oval'

Thus a key avenue for analysis of digital social datasets is examining the relationships between individual users or individual messages. It is also possible to identify relationships between places, based on visits or tweets made by the same person in these different places. While we’ve already identified the South as the key locus of grits-related tweeting in the United States, it’s important to not simply ignore all of the other data points available to us that are just not quite as spatially clustered. Indeed, given the strong connections between the cultural practice of grits preparation and consumption and the vernacular region of the South, we might hypothesize that even those people tweeting about grits outside of the South are likely to have some kind of connection to the South, perhaps as a kind of diasporic community now living in other parts of the country, or even just traveling for a short period of time.

Relational 'Gritspace'

To examine this relationship, we begin by looking for users in our original dataset that have tweeted about grits more than once – yielding a total of 8,958 users – then drawing a line from the tweet locations in chronological order. The resulting map below clearly shows that there is a strong relational connection with the South for those who tweet about grits from other places, even for cities like Los Angeles that are quite distant in absolute space, as well as in terms of cultural identity. Indeed, the gravity of grits appears quite strong, as of the users tweeting about grits from outside the South, approximately 55% of these also sent tweets from inside the cluster identified in the first map in this post. So even for those grits-obsessed tweeters outside of the South, the pull of porridge remains strong… and, we would expect, even stronger when you throw a bit of cheese and jalapenos in, too.

-----
If you’re interested in reading more about the methods used to make these maps, and about the utility of mapping geotagged social media data more generally, you can check out the following pre-publication version of a forthcoming book chapter from which this work was drawn:

Poorthuis, A., M. Zook, T. Shelton, M. Graham and M. Stephens. Forthcoming. "Using Geotagged Digital Social Data in Geographic Research". In Key Methods in Geography, eds. N. Clifford, S. French, M. Cope and T. Gillespie. London: Sage.
Abstract: This chapter outlines how one might utilize the massive amounts of web-based, geographically-referenced digital social data for geographical research. Because much of these data are user-generated and produced through social media platforms, we also focus on the pitfalls associated with such sources and the benefits of a mixed methods approach to these data. Not only can digital social data be mapped for visual analysis, it is also useful to use a range of quantitative methods to understand relationships between different subsets of the data. In addition, closer, systematic readings via qualitative methods of social data provides insights of particular people’s perceptions and experiences of the world around them. Thus, while making maps is often the starting point for geographers working with this kind of research, it is rarely the end point.

September 17, 2014

Hashtags and Haggis: Mapping the Scottish Referendum

The past weeks have been quite eventful in Scotland as a monumental election unfolds. Everyone wants to know, which way will the Scots vote? While we here at Floatingsheep certainly don't have the answer or power to predict the referendum, we thought it might be interesting to see the geographic dimension of how Scots (and the rest of the world) are tweeting about a fundamentally geographic decision [1].

We pulled data from DOLLY from the last month and a half for a number of hashtags and terms that we thought might be helpful in taking the pulse of Twitter discussion around the independence referendum. Most obviously, we collected the hashtags #VoteYes and #YesBecause due to their association with the pro-independence movement, and the hashtag #NoThanks because of its association with anti-pro-independence sentiment [2].

We started by comparing the prevalence of 'no' (i.e., pro-union) hashtags versus 'yes' (i.e., pro-independence) hashtags the global level. In the map below, orange indicates a greater prevalence of 'yes' tweets and purple indicates that there are more 'no' tweets. Perhaps the most interesting thing here is that we can see the United Kingdom swing towards a 'yes' vote, which has, for the most part, appeared to be the underdog in more conventional polling leading up to the referendum. Then again, most of Western Europe, along with Thailand and Australia, also have a general preference for 'yes' tweets. Oddly enough, the United States is the staunchest defender of the union, based solely on it's massive preference for 'no' tweets. Strange for a country that yearly celebrates its breaking away from Mother England

Comparing 'Yes' vs. 'No' Tweets at the Global Scale

Looking closer at the UK, we can see that much of Scotland has a roughly equal number of tweets in support of both the 'yes' and 'no' positions -- reflecting the contentious and hotly-contested nature of this referendum. But the Central Belt in particular -- where a lot of actual votes will be coming from, as it is the most densely populated part of the nation -- swings heavily towards 'yes'. The English, on the other hand, seem very much inclined towards pro-union or anti-separation tweeting.

Comparing 'Yes' vs. 'No' Tweets in the United Kingdom

To take an alternative look at support for the different positions, we mapped the percentage of each of the three hashtags that originates in each of the administrative sub-regions of both Scotland and the UK as a whole. The Highlands and parts of the Central Belt again show up as strong bastions of 'yes' votes.

Percentage of Referendum-Related Tweets from Different Regions

But seeing as we're interested in doing more than just mapping distributions, the next question is how are we to put all of this into context? The only proper place to start is, of course, with the Queen. The map below illustrates those places which also tend to have higher-than-normal levels of tweeting about the Queen (in orange) and those places that are tweeting less about the Queen than might usually be expected (in purple), based on a baseline measure of tweeting activity. Sadly, the whole country seems to be ignoring her. Apart from Glasgow, that is. In the interests of not upsetting an 88 year-old lady, we have chosen not to explore these tweets in any more detail.

Tweets referencing "Queen"

Building on this, we also explored the geography of references (using the same method described above) to something inherent in most people's definitions of Britishness: tea and crumpets

We see an all-around tea-depression; hardly anywhere is particularly pro-tea at the moment, truly a shocking state of affairs. The British are clearly not being their usual selves, and for their sake we're glad the referendum will be over soon, regardless of the outcome. Scotland, in particular, has average tea counts that are low by historical standards.

Tweets referencing "tea and crumpets"

This analysis would, of course, all be meaningless unless we mapped the geographies of a range of uniquely Scottish phenomena: haggis [3], kilts and Nessie. Still using the same method as above, the map below shows without a shadow of a doubt that Scotland is destined to become its own nation.

Tweets referencing "haggis", "kilts" or "Nessie" 

The Scots are tweeting about these topics at a greater-than-usual rate, while their southern neighbors remain distinctly uninterested. If ever there were an indication that these nations are divided by more than just a line on a map, we see that manifested in the topic of people's Twitter conversations. In short, the Scottish referendum is not just simply about "yes" or "no" but seemingly touches on much more fundamental questions of ovis-based cuisine, men's wear and mythological creatures.

So even if the 'no' votes win out in and the Kingdom remains united, the geographies of haggis related tweeting (along with a few other things) has revealed that these are two very different nations, indeed.

UPDATE (9/18/14 @ 12:45pm):
We've added another map to our analysis below, which shows the relative prevalence of #VoteYes and #NoThanks tweets throughout Great Britain, at the level of administrative sub-regions, rather than the hexagons used above. This map makes for a stark contrast between the English (and Welsh) and the Scottish... while there are a few areas of Scotland that show relative parity between 'yes' and 'no' tweets, most of the nation demonstrates a relatively strong prevalence for 'yes', while much of England demonstrates at least a slight preference for 'no'. 



--------------
[1] In case you don't know what Twitter, is we refer you to the Scots Wikipedia page on the subject, which states: "Twitter is an online social networkin service an microbloggin service that enables its uisers tae send an read text-based messages o up tae 140 characters, kent as 'tweets'".
[2] Perhaps we could have simplified this phrasing, but then we would have lost the chance to type "anti-pro-independence", which is a lot of fun. Anti-pro-independence. Anti-pro-independence.
[3] Normally the Floatingsheep collective avoids conversation about sheep heart, liver, and lungs that are boiled in a sheep stomach. But we made an exception this time.

September 10, 2014

Mapping #RussiaInvadedUkraine

From snarky exchanges between official Canadian and Russian Twitter accounts, conflicting representations of Crimea in Google Maps and OpenStreetMap, and a recent piece by Peter Pomerantsev in The Atlantic on how Vladimir Putin is revolutionizing information warfare, the ongoing conflict in the Ukraine has been widely reflected online. One particular manifestation of this conflict on social media was the #RussiaInvadedUkraine hashtag, which emerged at the end of August as Russian troops appeared in Eastern Ukraine. The hashtag has served as a social media rallying point for supporters of Ukraine, with the New York Times reporting that in the first day of its existence, over 500,000 tweets using the hashtag were sent.

Wondering what the spatial distribution of this hashtag looked like across Europe, we fired up DOLLY and collected all geotagged tweets containing the hashtag sent from European countries between August 27th and September 7th, 2014, resulting in approximately 4,500 tweets. To control the effect of single, very active, individuals sending many tweets -- and to better represent aggregate rather than individual actions -- we only included the first five tweets from any single user, resulting in a total of about 2,100 geotagged tweets. These tweets were aggregated to the country level and then normalized by the total number of tweets sent during this same time period, resulting in a location quotient for each country. The location quotient indicates the relative prevalence of tweets containing the #RussiaInvadedUkraine hashtag compared to the overall level of Twitter activity during this same time. Values greater than one indicate that people in a given country contributed a greater number of tweets about this topic than would be expected based on usual tweeting levels, with values less than one meaning that they were underrepresented in tweets using this hashtag than one might usually expect.


The map above illustrates a strong concentration of the #RussiaInvadedUkraine hashtag in countries that are nearer to the Russian border. In short, a classic example of a distance decay function, in which distance from a phenomenon is inversely related to attention or presence of a given phenomenon. In general, most countries within Eastern Europe -- including Russia itself -- show a higher level of Twitter activity around this hashtag, with some exceptions such as Moldova, Slovakia and Romania. In particular, however, the Ukraine and its neighbor Belarus show an extremely high level of activity around this issue, with the Ukraine alone contributing roughly 48x more of the tweets using this hashtag than it did to the baseline sample used for normalization. Conversely, as one moves westward, the level of participation in this social media meme drops considerably. While Germany, which is both geographically and relationally more proximate to Eastern Europe has a location quotient of just 0.81, the Netherlands and Italy have scores of 0.25 and 0.22, with the UK and France having extremely low location quotient values of just 0.05.


Of course, Twitter is not an unproblematic representation of the population, and tweets containing this hashtag can express a range of sentiments from both sides of the conflict [1]. Quite clearly, the interest and official response from western states (and their militaries) is not tied to the level of popular participation in social media activism. Instead, as we showed in the case of tweeting related to Ferguson, Missouri and the protests around the shooting death of an unarmed black teenager at the hands of a police officer, geography matters when it comes to directing our attention to news and current events, with people more directly connected to these issues having a much greater level of interest and concern [2]... even when it involves the invasion of military forces from one country into their neighbor.

-------
[1] Although a quick review of the text reveals that these tweets are primarily critical of Russia's actions.
[2] Of course, this shouldn't be particularly surprising. But for some people, it is.

August 18, 2014

Mapping Ferguson Tweets, or more maps that won't change your mind about racism in America

This post is the culmination of the Inaugural #IronWilson Map-a-Thon, held on Saturday, August 16th, and is the result of a collaboration between Matthew Wilson, Eric Huntley, Ryan Cooper and Taylor Shelton. 

A little over a week ago, the streets of Ferguson, Missouri, a suburb of St. Louis, were disrupted by the shooting of 18-year-old Michael Brown by Ferguson Police Officer Darren Wilson. While the details have been slow to emerge, the reaction to the killing of yet another unarmed young black man has been anything but -- whether in the form of street protests in Ferguson, or the online reaction to the news as seen on Twitter. The following graphic, produced by Twitter and published online by the Washington Post, demonstrates a typical representation of what we might call ‘#hashtag frenzy’, as people around the country take to Twitter to react to and comment upon the news. 


While certainly flashy and eye-catching, these so-called ‘animated ectoplasm maps’ tend to be short on meaningful insights. These visualizations show little more than population density in the US, and are remarkably similar from one trending topic on Twitter to the next. There is no attempt to normalize the data by population or overall levels of tweeting in a given place, thus obscuring both more detailed spatial patterns and broader social meanings that might be drawn out of such data. Still, maps such as this are useful in demonstrating the waxing and waning attention span toward issues of social importance, including the registering of yet another gun-related and police-initiated violent event, something that this blog post itself contributes to; therefore, in full admission of ‘yet another Twitter map of racial violence’...

We collected all geotagged tweets referencing a series of keywords -- 'Ferguson’, ‘handsup’, ‘mikebrown’, ‘dontshoot’ and ‘handsupdontshoot’ -- from Saturday, August 9th when the shooting occurred through the morning of Friday, August 15th, in an effort to provide a bit more resolution and, hopefully, insight into the ways and places people were tweeting about the protests. Starting with the first geotagged tweet referencing the shooting, we collected a total of 38,450 tweets. 'This tweet came from user Johnny__Tapia at 3:11pm Central Time on Saturday, saying “Ferguson police just shot a kid in the head in the middle of the street. 17 yrs old. Ain’t nobody saying what he did” [1].

In the several days following the shooting, news spread quickly over Twitter, with social media providing a key source of updates and information in lieu of any official reports or communication from the Ferguson Police Department. The map below, made by Eric Huntley, aggregates all the tweets in this dataset to hexagonal cells across the continental United States, and normalizes it relative to the overall amount of tweeting in that location at the same time. In other words it shows the relative focus of tweeting related to Michael Brown’s shooting (and the subsequent protests and police crackdown) compared to overall tweeting activity by location. In this map lighter shades indicate relatively more tweets about Ferguson than the national average.


The national and international media coverage of the story in Ferguson points toward the notion that this event transcends the local; there is something about it that speaks to people from any number of places and walks of life. For example, the aforementioned WaPo article ends with the relatively meaningless maxim, “People are watching from as far away as Fiji and Ghana. That's the world we live in now.” While discussions of increased police militarization and the persistent legacy of racism have certainly resonated strongly with a national audience, it is evident from our more-than-just-dots-on-a-map approach that the tweeting around this event is actually most prevalent in the general vicinity of where the shooting occurred: the St. Louis metropolitan area. The proportion of tweets on the topic is higher in and around St. Louis than anywhere else in the country, while other cities around the country have largely continued about their business, with lower levels of Ferguson-related tweeting relative to overall levels of Twitter activity [2]. While a few scattered and isolated areas throughout the country demonstrate a relatively high amount of tweeting about Ferguson -- mostly as a result low overall levels of tweeting -- the St. Louis region is really the only place that demonstrates a particularly concentrated and significant interest in the matter. In other words, "the world we live in now" is one in which spatial proximity and social connectedness remains incredibly important, even if people in Fiji and Ghana can follow along, too.

This lies in contrast to the aftermath of George Zimmerman receiving a ‘not guilty’ verdict last summer in his trial for the shooting death of Trayvon Martin, which is arguably the best parallel in terms of the public outcry and attention to the current Ferguson situation. Following Zimmerman’s verdict, large portions of the American South demonstrated a greater likelihood to use the #JusticeForTrayvon hashtag than other parts of the country, which we interpreted as indicative of Twitter users making connections between the events in Sanford, Florida and the broader legacies of racialized violence throughout the American South. Whether the different geographies of Twitter's reactions to these events are the result of different temporal evolutions (the immediate aftermath of the shooting vs. the trial verdict a year and a half later) or in divergent experiences, or perceptions, of racism between the South and Midwest [3], or something else entirely, is left to some level of speculation.

Despite the overall concentration of tweets in the St. Louis region, it also important to recognize that spatial unevenness exists at multiple scales, with respect to practically any phenomena. Indeed, the ability to examine such phenomena at a variety of scales is one of the major advantages to aggregating these points -- or individual tweets in this case -- to a uniform grid of hexagonal cells, as opposed to the more conventional, and largely arbitrary, Census-defined areal units. In the GIF below, created by Matt Wilson, you can see the spatial distribution of raw (i.e., non-normalized) tweets -- using the same dataset -- in the St. Louis metro area over time, beginning on Saturday when the shooting occurred, through the end of Thursday, August 14th [4].  


Given our lack of first-hand knowledge of St. Louis and its environs, we’re hesitant to draw too many conclusions from this data, though we certainly welcome any potential explanations from our readers. Because each of these snapshots is classified in the same way, we can see the diffusion of the news and growth in interest over time, becoming much more pronounced beginning on Monday. Tuesday is interesting insofar as it seems to demonstrate a much stronger clustering around Ferguson itself (the cluster of three dark blue hexagons north of downtown St. Louis), with the rest of the city actually seeing a decrease in tweeting about the event. This interest, especially in downtown St. Louis, ramps back up on Wednesday and Thursday, around the time of growing protests and the increasingly violent response from the Ferguson Police.

Ultimately, despite the centrality of social media to the protests and our ability to come together and reflect on the social problems at the root of Michael Brown's shooting, these maps, and the kind of data used to create them, can’t tell us much about the deep-seated issues that have led to the killing of yet another unarmed young black man in our country [5]. And they almost certainly won't change anyone's mind about racism in America. They can, instead, help us to better understand how these events have been reflected on social media, and how even purportedly global news stories are always connected to particular places in specific ways.

----------
[1] It appears that this user has since deleted all of his tweets back to July 10.
[2] There is still a significant absolute amount of tweeting in these places, there just also happens to be a generally massive level of tweeting about other topics, as well.
[3] Of course, St. Louis, like pretty much everywhere in the United States, has it’s own important legacies of racism. For example, please see: Deep Tensions Rise to Surface After Ferguson ShootingThe Most Racist City In America: St. Louis?, and The Century-Old Urban Policy That Divides St. Louis.
[4] These maps also use a somewhat smaller sample of tweets that have only exact latitude and longitude coordinates, so as to avoid using those tweets tagged to place names, such as ‘St. Louis’, which might give the impression that there were large contingents of tweeters at the geographic center of the city.
[5] Though data about racial profiling, as Ryan Cooper analyzed for us here, can point towards some potential explanations.

August 17, 2014

Mapping the #LouisvillePurge

The only way to introduce this post is to say that yes, a bunch of really naive and/or, in the case of the local television news media, willfully idiotic, people thought that there was going to be a 'purge' -- a 12 hour period where all crime is made legal -- in Louisville, Kentucky on the night of Friday, August 15th, 2014. Starting with a single tweet from a local high school student, things quickly grew out of control, with #LouisvillePurge becoming a trending topic nationally by the time things were all said and done. While the best tweets referencing the purge made light of the phenomena, there were many, many more expressing confusion, fear, bewilderment and a desire to save the poor souls who might have been convinced to participate in such an event. But for all the attention given to the role of social media in spreading the hysteria [1], there's been no attempt to look at the where some of these tweets were coming from, and how the news spread over space and time.

While the tweet that kicked the whole ordeal off was created at 8:32pm on Sunday, August 10th, the first geotagged tweet with the #LouisvillePurge hashtag didn't show up for another couple of days, at 11:33pm on Wednesday, August 13th. Beginning with that tweet, we collected all geotagged tweets with the hashtag through noon on Saturday, August 16th, at which point things were dying down.

The map below shows the overall distribution of these 4,351 geotagged tweets, aggregated to hexagonal cells across the continental United States. While Louisville and the surrounding areas clearly have the highest concentrations, the discussion of the Louisville Purge was truly trans-local, with less than 25% of the total number of geotagged tweets coming from the Louisville Metro area. Of areas further away from Louisville in absolute distance, Houston, Dallas and Los Angeles represent some of the highest concentrations of tweeting about the (non-)event.

All #LouisvillePurge Tweets thru August 16th at 12pm EDT

But perhaps more interesting than just the overall spatial distribution is how this distribution evolved over time, from the first geotagged tweet all the way through the cycle of hype and hysteria that led the Louisville Purge to be featured on any number of national news websites. In the series of maps below, we have divided all of the tweets in our dataset into a series of (more-or-less arbitrary) time frames that give a good idea of when and where the news spread to other parts of the country [2].

The lead up to the purge demonstrates a relatively localized phenomenon within Louisville, though it's interesting that there is some extra-local tweeting from the very beginning, with a very small number of tweets coming from outside the state in West Virginia, Kansas, Texas and Florida. There were only a total of 182 geotagged tweets referencing #LouisvillePurge in this 44-hour aggregate time span, with tweets originating in Metro Louisville representing 55%, 66% and 60% of the total number of tweets with the hashtag during the three periods, respectively. In other words, talk of the purge spread quite slowly over the course of the week.

Time #1: 42 tweets
From August 13th at 11:30pm to August 15th at 6am

Time #2: 36 tweets
From August 15th at 6am to 4pm

Time #3: 104 tweets
From August 15th at 4pm to 8pm 

The number of tweets with the hashtag exploded right around 8pm on Friday night, the 'official' start time of the purge. This four hour time period represents the peak of tweeting activity around #LouisvillePurge, attributed largely to the fact that this is when the event started to diffuse outward beyond the city's boundaries to places both near and far. One can see both a significant increase in the amount of tweets across Kentucky, as well as to far-off cities like Los Angeles, Milwaukee, D.C., Philadelphia and New York City. From 8pm to 12am, the 757 tweets from Metro Louisville represent only 30% of the 2,533 tweets across the country, further highlighting the spatial diffusion of information about, and interest in, the purge. In fact, this measure of locally-concentrated tweeting drops even lower to less than 10% from the hours of midnight to 6am (when most Louisvillians would be asleep), though it again rebounds a bit higher to 23% during our final time span of 6am to noon on Saturday the 16th, after the purge has 'officially' ended.

Time #4: 2,533 tweets
August 15th at 8pm to August 16th at 12am

Time #5: 1,420 tweets
From August 16th at 12am to 6am

Time #6: 216 tweets
From August 16th at 6am to 12pm

Like our earlier research on #LexingtonPoliceScanner in the wake of the 2012 Kentucky Wildcats basketball championship, we can clearly see an ebb and flow in the way the event originates in a fairly localized area before gaining a larger following and eventually slowing down and becoming more localized again as many users reflect upon the aftermath. But unlike the attention paid to the #LexingtonPoliceScanner in large cities around the country, and especially the South, the interest in the #LouisvillePurge tended to be somewhat more diffuse, without any single location outside of the city or state paying a disproportionate amount of attention to the events.

In the end, we're happy to report that all of the Floatingsheep emerged from the purge unscathed and thoroughly amused, and we hope the same can be said for all of you and your loved ones. And do remember, don't trust everything you read on Twitter [3, 4]!

-------
[1] Again, it's probably worth noting -- somewhat ironically, I suppose -- that despite the rumor originating and being passed around via social media, it was the traditional local television news networks whose willingness to believe and highlight the rumor drove further attention to the situation, which was almost obviously a farce from the very beginning.
[2] You can also access an animated GIF version of this time series map here.
[3] Especially if you are supposed to be a "real journalist"!
[4] For that matter, don't trust everything you see on the television news, either!

July 22, 2014

How Many Hobbits Could Chuck Norris Take In a Fight?

Inspired by the (relatively) recent Buzzfeed quiz, "How Many Five Year Old Children Can You Take In a Fight?" [1], we have been wondering about other potential battle royale matchups: Juggalos vs. Bronies, Juggalos vs. polar bears, Justin Bieber vs. Miley Cyrus and even goats vs. llamas

Perhaps our favorite attempt at recreating this kind of scenario is asking: how many hobbits could Chuck Norris take in a fight? The analysis was quite complex as we had to first set rules on the engagement (e.g., what kind of weapons? is mithril armor allowed or not? etc.) and decide which version of Chuck Norris (Walter, Texas Ranger Chuck Norris? Actual current Chuck Norris? Perhaps Delta Force Chuck Norris?) and what kind of hobbits (after all are we talking Brandybucks or Tooks? are these typical Shire hobbits or have they been abroad? etc.) we are talking about here.  

As you can suspect, there was a lot to sort out. But after much discussion and analysis we have come up with a clear answer but sadly, as the actual question has nothing to do with this blog, we've been forced to bury it in the footnotes [2]. What we can do, however, for the purposes of this blog is compare the distribution of references to hobbits, as opposed to references to Chuck Norris, in geotagged tweets. Starting from a 10% sample of all global geotagged tweets from July 2012 through March 2014, we collected all references to "hobbit*" and "Chuck Norris" to enable our comparison.

Hobbits vs. Chuck Norris, July 2012-March 2014

At the global level, there are actually quite comparable numbers of references to hobbits and Chuck Norris, thus making the location and scale of our hypothetical battle all the more important. There are 27,527 references to the man on Superman's pajamas, and 24,145 references to those short little guys with hairy feet.

What is evident, however, is that Chuck Norris isn't particularly popular anywhere but in the United States, as nearly half of the global references to him come from the USA, giving him a nearly 9000 tweet advantage over hobbits. Perhaps not everyone else in the world finds quite as much humor in the many Chuck Norris Facts as Americans do? Or perhaps other countries have their own Chuck Norris-like cult heroes to look up to [3]? The next closest country in terms of Chuck Norris appreciation is France, with just 250 more Chuck Norris tweets than hobbit tweets, followed up by South Africa, Nigeria and Puerto Rico in the top 5 countries favoring the man who predicted 1000 years of darkness were Barack Obama to be re-elected President of the United States.

Meanwhile, the top 5 countries favoring hobbits are Indonesia - where they hold a 2,141 tweet advantage - Turkey, Mexico, Spain and Malaysia, each of which have a greater than 500 tweet advantage for hobbits over Chuck Norris. A total of eleven countries have more than 100 more references to hobbits than Chuck Norris, a considerable feat given that only the top 3 Chuck Norris countries have a more than 100 tweet advantage.

In many ways, the pattern in this map is a replication of that from our recent map comparing references to Bieber and Miley; just as the only places with a real preference for Miley Cyrus were the USA and a smattering of African countries, so too are these the only places with a significant preference for Chuck Norris. Does this mean there is some sort of Chuck-Miley conspiracy afoot? Or that Bieber has taken command of an army of hobbits in his quest for world domination? We'll leave it to you to find out...

------
[1] See also: How many Justin Biebers could you take in a fight? How many 90 year olds could you take in a fight? How many hipsters could you take in a fight?
[2] The answer is zero.  Because hobbits are actually just fictional characters and Chuck Norris is a real living person. See? Sometimes there are clear and easy answers to tough questions.
[3] Ironically, of course, Kenya seems to display a slight preference for Chuck Norris over hobbits, despite Makmende's imposing presence.