The New Year's Eve Attacks in Cologne

Over 500 complaints of sexual assault have now been reported to the police in Cologne, all occurring during the city's New Year's Eve celebrations. This shocking number of attacks, the vast majority carried out by immigrants of Algerian, Tunisian and Moroccan origin, has caused a backlash against Angela Merkel's 'Open Door' policy on refugees (although only three of the 58 men arrested so far were actually refugees).

Die Welt has now released a CartoDB map visualizing the sexual assaults and other crimes which took place in Cologne on New Year's Eve. Übergriffen von Köln uses CartoDB's Torque library to animate through the course of the evening, showing where the sexual assaults and other crimes took place in the city.

The animated map provides a basic temporal view of the data. However the dots (representing the attacks) flash by really quickly on the map. This speed, coupled with the fact that the map offers no way to view all the attacks at once on the map means that the map certainly doesn't provide a clear spatial overview of where the attacks in Cologne took place. You might think that a spatial visualization was the point of using a map to visualize the data in the first place.

The result is that Die Welt's map of flashing dots seems to both sensationalize and trivialize the sexual attacks at the same time. CartoDB's Torque library has come in for a fair amount of criticism for what Andrew Hill dubbed burger cartography. This is a shame because the Torque library is such a powerful tool for adding a temporal analysis to time stamped spatial data. Perhaps CartoDB should create a few more Torque powered maps themselves to really show how a temporal analysis of spatial data can be done well with the library.

In defense of Die Welt the article which accompanies the map does provide a much better analysis of both the times and the locations of where the majority of the attacks took place in the city. It's just a shame that a little more analysis of the data wasn't present in their mapped visualization.