New Yorkers love the bike-sharing program Citi Bike, which allows riders to cruise around NYC and Jersey City without owning a bike. We know this because Citi Bike records just about every data metric you could think of and makes all 22 million rides since July 2013 publicly available.
But who wants to stare at a bunch of spreadsheets, right? Todd Schneider has crunched the data into a series of fascinating graphs and animations to help us visualize the massive numbers.
If you sit staring at Schneider’s animated map, showing a time lapse of one day’s Citi Bike traffic (Wednesday, September 16, 2015), you start to see some patterns. You can see many well-defined routes emerging from the speeding dots, and Schneider points out that the bridges between Brooklyn and Lower Manhattan, which have a traffic spike at rush hour, once in the early morning and once again at around 5:30 PM. Many Citi Bike rides seem to be commutes, with users opting to go to and from work by bike instead of by car or public transit.
Schneider also mapped the most popular roads, and–maybe surprisingly–they match the taxi pickup and drop-off maps he made last year fairly closely. One assumption made by Schneider is that most people follow the Google Maps directions between locations. But I decided to see how Schneider’s maps looked next to a map of Manhattan’s bike lanes. As you can see, they match to a pretty high degree. Bike sharing then, seems to be a utilitarian replacement for other forms of transport: more trips are made on weekdays than weekends, and while Citi Bike’s 10 million trips per year isn’t yet anywhere close to 175 million taxi trips, or 35 million Uber rides, it could be making a small dent.
It would be interesting to see a demographic breakdown of the users, too. Are the users rich folks ditching the car out of environmental conscience, or to get fit? Are they low-wage workers, for whom the $150-per-year fee appeals over expensive subway trips? We do, however, get data for rider age and gender, plus journey distance and time (and therefore speed).
Schneider’s graphs show that the average speed for weekday riders (he excluded weekenders because they are more likely to be taking a leisurely cruise) is 8.3 mph. The young ride faster than the old, and males go faster than females. The time and distance data also show that trips are, on average, a couple of minutes slower than Google Maps’ time predictions, although that might be due to pickup and drop-off times at either end.
If you’re due a coffee break, take it now and read the rest of Schneider’s fascinating post. It looks at how weather impacts usage, how rides vary with season, and even touches on privacy issues: you can identify individuals with a surprisingly small amount of data.
It seems that the availability of bikes, plus the infrastructure to use them safely (bike lanes), really does provide a popular alternative to transit and private cars. And while the health and environmental benefits are clear and easy to sell, the fact is, a city full of bikes is just a lot more pleasant than a city full of cars and buses.