If a picture is worth a thousand words, how many is a selfie worth?
That’s what SelfieCity wants to find out. A new project by data visualization wunderkind in collaboration with Lev Manovich, Jay Chow, and Nadav Hochman, SelfieCity is an attempt to analyze the data of more than 3,000 self-portraits and, in doing so, extrapolate what a selfie is even meant to say in the first place. In the process, SelfieCity was able to figure out everything from the age of the average selfie takers to how much they opened their mouth and more.
“Selfies are an interesting thing to study right now,” Stefaner tells Co.Design. “Are they just a fad, or do they represent a substantial new trend of how we create and share photos? Are they a means of self-expression, a tool of self-promotion, or a cry for attention? And are there any cultural differences in the way people in different countries take selfies?”
Although people have indulged in self-portraiture pretty much as long as film has been a medium, the selfie–that smartphone self-portrait taken above head and at arm’s length–had a big year in 2013, when it was declared the Oxford English Dictionary’s word of the year.
To study the phenomenon in different cities around the world, the SelfieCity team needed to gather enough selfies to form what they called the “aggregate face of a city.” A super-selfie, in other words. But coming up with such an aggregate view of all of a city’s selfie data was still an enormous undertaking.
“Typically, data visualization is based on numbers. However, a single number can’t summarize a photo. It is not a ‘data point’ but a whole world, rich in meanings, emotions, and visual patterns,” Stefaner says.
In order to visualize what a city’s super-selfie would look like, Stefaner and his team needed to quantify their data. The team collected 656,000 Instagram photos from New York, Bangkok, Moscow, São Paolo, and Berlin taken between December 4 and December 12, 2013. Feeding these photos to workers hired through Amazon’s Mechanical Turk marketplace, they then identified which of these photos were selfies, and made best guesses as to the age and gender of the person who took them.
Armed with the top 640 photos from each city–those which at least two people identified as “selfies” during the discovery phase of the project–the SelfieCity team then fed these pictures into state-of-the-art facial analysis software, quantifying a number of measurements from the size of a subject’s face, the tilt of their head, whether they were smiling or not, and so on. From that data, SelfieCity was able to extrapolate a number of key findings:
• People take fewer selfies than you’d think. According to SelfieCity’s data, only 3% to 5% of the 300,000-plus images that they examined were actually selfies.
• Women take more selfies than men. “In every city we analyzed, there are significantly more women selfies than men taking, from 1.3 times as many in Bangkok to 1.9 times more in Berlin,” Stefaner says. In Moscow, the discrepancy is even more striking: 4.6 times more women take selfies in the Russian capitol then men. No matter where, if a man takes selfies, though, he’s likely to be older: the median age of men who post selfies on Instagram is more than 30 years old.
• Women strike more extreme poses in selfies (especially in São Paulo). According to SelfieCity’s research, women tend to take more expressive, sexy poses than men in their selfies. On average, the head tilt of a woman’s selfie is 150% higher than for men (12.3° vs. 8.2°). Translated, this means an awful lot of women take selfies holding their cameras way above their heads. But in São Paulo, it’s even crazier: there, the average head tilt for females is 16.9°! Guess they want to fit their bikinis in-frame.
• The younger a person is, the more likely he or she is to take selfies. No duh, but in SelfieCity’s estimation, most of the people who take selfies are pretty young, with an estimated median age of just 23.7 years old. Selfie takers skew youngest in Bangkok (21), and oldest in New York City (25.3).
• People are happiest in Bangkok and São Paulo, and more miserable in Moscow. “Our mood analysis revealed that you can find lots of smiling faces in Bangkok (0.68 average smile score) and São Paulo (0.64),” Stefaner says. “People taking selfies in Moscow smile the least (only 0.53 on the smile score scale).”
All of this data and more is freely available for perusal at SelfieCity. As an added perk, the system supports deep linking, allowing you to share, say, all the happy women from São Paulo who tilt their heads to the right in a single . Soon, Stefaner and his team hope that they will be able to expand SelfieCity to study more cities, as well as other selfie phenomena, such as multi-selfies (“me and my boyfriend,” “me and my best friend,” “me and my sister,” and so on).
But what is a selfie, and what’s it trying to say? At the end of the day, a selfie can be a means of self-expression, a tool of self-promotion, a cry for attention, or all of the above. No matter what it is, though, the typical selfie is more honest than you would think from Googling “selfie” and examining the images that come up, Stefaner says: “One thing we really found striking was how much more complex our authentic selfies are compared to the cliché-laden selfies you’ll see in Google results.”