"Most photographed places in the world" promises the headline on a 10-image slideshow at CNN.com, a piece that starts out with the surprising news that
The Eiffel Tower isn't the most popular place for pictures in Paris, nor is the Empire State Building the top photo destination in New York.
Well, not exactly.
Once you've clicked through, the piece admits that opening statement is likely inaccurate:
At least, that's the case among users of Google's geolocation-oriented photo-sharing website, Panoramio, according to Sightsmap.com, which has ranked 15,000 cities and their attractions according to photos tagged.
Raise your hand if you send all your photos to Panoramio.
Yes, problem number one is that Panoramio isn't nearly as widely used for uploading and sharing images as, say, Instagram, Facebook or Flickr. But problem number two is that photos on Panoramio are very unlikely to represent the broader universe of travel snapshots, even with a large number of users.
Why? First off, you can only upload to Panoramio from a desktop system, Picasa Web album or Google+ account. All those people using iPhones to take their tourist shots who aren't on Picasa or Google+ ? Not counted. What about serious photographers who don't want their work available for others to co-opt? Not there. Those who don't geotag pictures? Missing.
Also troubling if you want to claim that this somehow reflects all photographs: Panoramio is a popularity-based social network. Photos are most likely to "succeed" and show up high on Google Earth if they're good. So that means many serious Panoramio users are unlikely to upload all their photos to the site, but instead only send what they consider their best. That's hardly a great way to count total numbers of photos taken even just considering people who use Panoramio.
The CNN slideshow admits that the Empire State Building may not show up as often as expected "perhaps because the Empire State Building is a tougher picture to take."
Alas, CNN wasn't alone in blowing this data set out of proportion. The Telegraph website in the UK followed a similar track with its headline The world’s most photographed locations - mapped. You don't see "according to a photo-sharing website" until you click through. And even that disclaimer seems to be ignored once acknowledged, as the Telegraph article earnestly reports all available data as if it somehow matters. And, not surprisingly, the Huffington Post wrote about this as well.
Interestingly, the actual source of the data, Sightsmap.com, didn't try to present the information as something that it's not. Their headline on the interactive map showing hotspots for most Panoramio photos? "Discover beautiful places." Much more reasonable.
I understand that slideshows like CNN's and stories like the Telegraph's aren't necessarily meant to be statistically valid scientific surveys. But it's one thing to stretch the idea of some data set allegedly representing the population as a whole. It's another thing to alter it so heavily that it becomes meaningless.