The ‘where people run‘ visualizations have been making the rounds on the Internet over the past week. Created by Nathan Yau of FlowingData, the data is gorgeous, mapping popular running routes around major cities like NYC, Boston, and Chicago. Yau used public data from RunKeeper to create the maps of where people run in these cities, and it’s a great example of how anyone can use open, public data.
But the data is incredibly flawed. Publications like Atlantic Cities and Core77 have heralded that the maps as urban planning tools to make cities more runner/biking friendly- a great goal, but it’s not possible with the data in its current state. Let’s focus on one of the maps: NYC. The map of NYC looks suspiciously similar to the NYC Marathon route. Compare them side by side, and the correlation is obvious:
I chose to analyze the NYC map because I live here and I run here. On the left is Nathan Yau’s visualization, on the right is the NYC Marathon map. The loop at the top of the map traces miles 16-26.2 of the marathon: up 1st avenue until the turnaround in the Bronx, then down back into Manhattan on 5th avenue before turning onto Central Park South and finally finishing the race in Central Park. On Marathon Sunday, over 45,000 runners fill the streets of NYC. While not every marathoner uses RunKeeper, that’s a high enough sample to skew the data to highlight certain routes that wouldn’t normally ‘pop’ as runner favorites. Which is pretty obvious: right next to the marathon route is Central Park, which has designated running routes and is one of the top places to run in NYC, but doesn’t look like a ‘favorite’ area to run according to this map. And from personal experience I can tell you this about the map: Runners don’t flock to 1st avenue… unless the street is blocked to traffic and surrounded by screaming marathon spectators.
So the data as it appears now is flawed, and not entirely helpful to urban planners. What would be immediately more helpful would be excluding Marathon Sunday from the data, and then redrawing the map. What would be even more helpful would be analyzing the data in correlation to other public data streams. NYC has made over 1100 data sets available for public use on NYC Open Data– let’s have a little fun, shall we?
After excluding Marathon Sunday data, let’s make the ‘Where People Run’ maps more useful by:
1. Overlaying the map with public restroom availability. This map would provide a helpful guide for runners who need to find a water fountain, or need to take a break. Even more helpful would be if the public restroom data set and the RunKeeper data set also included time: what time bathrooms opened and when runner traffic was highest. This information could create a map to help better inform the city when they should open/close restrooms in the morning and evening (because as an early morning runner, I’ve anxiously passed many closed ones…). The public restroom data set also shows that many public restrooms can be found in parks and playgrounds. These playgrounds are great for children, but can also be a source of free workout ‘equipment.’ An open jungle gym becomes a pull-up station, and steps become a spot for fast-feet and drills. Which leads me to my next point…
2. Overlaying the map with public health data and greenmarket availability. The public health data contains information about obesity, asthma, and diabetes, and the greenmarket availability data contains location and time/day information, as well as if the greenmarket accepts health bucks and/or EBT. This map paired with RunKeeper data could provide a general overview of neighborhood health- if people run in the area, if they have access to fresh fruits and vegetables, what health conditions are more prevalent in the area. Adding park data provides an even more complete story, and one that could be used as an action plan: in areas with high obesity, low run traffic, low presence of green markets, but available playgrounds, the city could create neighborhood wellness programs. In areas of low obesity, high run traffic, high greenmarket availability, health and fitness brands could position themselves to better serve this community, setting up classes or even a storefront.
These are just two ideas of how the ‘Where People Run’ maps could be improved to be more actionable. BRB, going to learn to code so I can put my money where my mouth is…