Last year, at the beginning of 2016, I used the XML information provided by Swarm/Foursquare to generate some various maps on my whereabouts through the year. Using a fantastic blogpost on how to utilize Fusion Tables, I was able to visualize my average whereabouts in pins and heat maps.
To keep up my cartophilia. I’ve generated some new maps with my 883 check-ins of 2016. Outside of my usual spots in Austin, I was lucky enough visit several cities two years in a row. This was enough to generate some comparable location data of the app.
Austin
In general, my maps of downtown Austin and Riverside appear mostly the same, as I live and work in the same place I have since last year’s post. I have 2015 and 2016 side-by-side.
While I didn’t replicate the scale for my maps exactly, I’m still able to notice that I more or less spent time in the same areas around Austin. Just based on the heat maps I can see I spent more time downtown, East Austin and the South Congress area than the previous year.
Copenhagen Check-ins
Last year I wrote about how the geolocation data for my check-ins in Copenhagen, Denmark and Berlin, Germany were somewhat incorrect. After having the opportunity to go there again in 2016, I noticed that the check-in data has appeared to be more accurate:
On the 2015 map you can see there are more clusters of the check-ins, not because of their proximity, but more likely to rounded-off GPS coordinates for bars and restaurants. This year the check-ins are more distributed around the city than before. As my trips to Denmark were for the Mikkeller Beer Celebration Copenhagen (formerly Copenhagen Beer Celebration), I largely went to the same bars and restaurants two years in a row.
I’m not familiar with the tech behind how Foursquare works, but I wouldn’t be surprised if improvements to the app allowed for better data representation during check-ins. These could be improvements to the geolocations of places, or the end-user while checking-in. This could be aided by better data provided by location services on my iPhone as well. While I can speculate all I want, I can see that the data is improved.
Other Locations
I generated some pin maps of my other travels throughout the year, as well. I noticed from previous travels in places like Croatia and Turkey (where American social media penetration is light) that check-in geolocation can be off by several miles. However, when I was in Belize, I noticed that the locations were generally pretty accurate. The least accurate check-in I have is a totally incorrect check-in for the town of Ardmore in Ireland. On the map, you can see a cluster of check-ins on the central coast (Ardmore’s location). However, there is a totally inaccurate check-in near Waterford inland, nowhere near Ardmore. I tried updating the location in the app, but it looks like that it hasn’t been fixed.
