Posted by: jdantos | January 31, 2012

Where are Capital Bikeshare Riders Going? (Part 4)

Part 4 in a series mining a recently crowd-sourced pile of trip-level data on Capital Bikeshare. This post focuses on flows around the city, the top travel markets, and how certain areas are more in balance than others. If you’re bored, need sleep, or both, read Part 1, 2, or 3.

So, where are the Bikeshare bikes going to, and from?  While we can’t tell people’s final destinations, we can tell what docks they use.  So, I did a quick look at usage by dock, but unfortunately, with 142 stations, it’s impossible to visualize cleanly at the station level:

Can you make any sense of that? I can't really either. This is total trips by station. Time to aggregate.

But, I will say it’s interesting how the most popular station at Dupont Circle is the winner by a long shot.  As has been discussed elsewhere, this is likely because there are few other stations around, it’s a big station, and it was one of the first installed.

Dividing Into Zones: I divided up the Bikeshare stations into 12 big groups, kind of like “super neighborhoods,” to get a sense of where the riders are going to and from.  This is more art than science, but I basically drew lines based on my knowledge of biking around the city, and a healthy dose of arbitrariness.

I divided up the Bikeshare stations into these zones to make it easier to see trip patterns.

You can argue with how I drew the lines, and how I had to wing it when a station was right on the line, etc.  But I just needed some rough groupings of stations to make sense of travel patterns.  Wards seemed a little too blunt, e.g. Ward 6 encompassing everything from Shaw to the Ballpark. So, I drew Downtown as roughly the CBD from Constitution Ave. to Mass. Ave., Mid-City as everything in the middle south of Woodley Park and U Street over to Union Station, and then the others kind of fell in mostly logical groups.  I put the station at L’Enfant under Mall, and Woodley Park under Upper Northwest – eh, it was a judgment call. I also totally made up the zone names. The main thing is, we can now look at a manageable amount of numbers.

The numbers shook out so that every zone had between 4 and 30 stations, with most about 10-20.  More on that later.

Origin-Destination Matrix. Interestingly, this produces an 12×12 origin-destination matrix, or 144 different possible combinations of origins and destinations.  Here’s the full table, just showing 2011 trips, if you’re interested:

Capital Bikeshare origin-destination matrix by zone. Unpopular trip choices colored blue; medium and high colored white and red. (Click for larger)

Now, a few observations with this stuff.  A caveat – I haven’t accounted for the fact that stations were added over time, so neighborhoods that first got stations installed are going to look more popular than more recent arrivals, such as Rosslyn-Ballston.  And, different zones have different number of stations – although more station could be a symptom of demand just as much as cause. Nevertheless:

  • Trips to and from downtown dominate bikeshare patterns, at one third of all trips in 2011.
  • Trips tend to be short, as we saw earlier, making inter-zone trips popular (that’s the diagonal line running from top left to bottom right.  Especially in Capitol Hill, Mid City, and others.
Top Travel Markets.  This big table is hard to read, so I combined trips to and from into one table, and rank ordered. This kind of folds the matrix in half, so a trip from A to B and from B to A get counted in the same category.  The top three origin-destination pairs are downtown, and the neighborhoods just to the north (Dupont Circle, Logan Circle, etc.) which I’ve lumped into “Mid-City.”  In fact, trips within and between these two zones were 36% of all trips in the Bikeshare system in 2011.  The system is really anchored by this productive core.
After those two core markets, the differences between zones becomes less stark.  Trips local to Capitol Hill, and between Capitol Hill and downtown, are a somewhat distant second place, at 12% of all trips.   Then, trips between Mid-City North (north of U Street, roughly), and the Mall.  The table below shows the top Bikeshare travel markets:

Capital Bikeshare Top Travel Markets, 2011

I halfheartedly attempted to draw this table on a map, with admittedly third-rate arrows in PowerPoint.  It’s a little clunky, but it does (sort of) give a visual representation of the table above.  For this illustration, I suppressed all flows less than 30,000 trips or so.  I am clearly not a graphic designer – anyone want to help make this look better?

Capital Bikeshre trip patterns by zone, flows of 30,000 trips/yr and above. Thicker, darker arrows mean more trips.

In Balance, or Net Sender/Receiver? From the zone totals, I think you can also tell whether a zone is more or less in balance over the year (same number of trips beginning and ending), or whether it is a net “sender” or “receiver” of Bikeshare trips. If more trips ended at a zone than began at it, I believe the difference is the re-balancing crews.

Some zones are more or less in balance on net, while others are net senders or receivers of bikes.

So, over the course of 2011, 5% more trips ended downtown than began downtown, meaning that the rebalancing vans, on net, took bikes away from downtown.  Contrast that with neighborhoods to the north and west (places like Cleveland Park, Petworth, Mt. Pleasant, and others in Upper Northwest, Mid-City North) where it’s mostly downhill to downtown, and more trips began than ended. Here is the final column, graphed in terms of absolute trips:

Net trips supplied/(received) by Zone in 2011, i.e. trips originating minus trips ending. Some station groups are net trip "sources," others are net "sinks," and some are more or less in balance.

Okay, that’s all for now. Until next time, keep the comments and reactions coming! It’s nice to hear your thoughts.

P.S. – one further thought on the “balanced-ness” (is that a word?) of individual stations.

Advertisements

Responses

  1. This is amazing! Thank you so much for all of this work!

    • Thanks Lauren! Know a graphics person who could make a better visualization of flows? 🙂

  2. […] A short Part 3 in a series of posts analyzing Capital Bikeshare usage data. This post quickly looks at how much each bike gets used and churned through the system. See parts 1, 2, and 4. […]

  3. […] Part 1 in a (perhaps?) series of posts analyzing Capital Bikeshare usage data. This post focuses on system-level usage by a few dimensions. Check out parts 2, 3, and 4. […]

  4. […] Part 2 in a (now?) series of posts analyzing Capital Bikeshare usage data. This post focuses on system-level usage by trip duration. See parts 1, 3, and 4. […]

  5. Justin, this is your most interesting analysis yet. The Mid-City / Downtown fluxes are huge.

    Can you post the (O,D) table as data, not just as an image?

    Also, we’ve just started a forum for Bikeshare developers and data hackers here: http://bikearlingtonforum.com/forumdisplay.php?33-Capital-Bikeshare-Software-Developers

    Hope you’ll be able to join us there and post some of your results!

  6. […] call this part 5, even though it’s kind of a postscript to part 4. See parts 1, 2, and […]

  7. […] Capital Bikeshare data. This one focuses on net flows of bikes across the system. Parts 1, 2, 3, 4, and […]

  8. […] Capital Bikeshare data. This one maps a bunch of data by station. See also parts 1, 2, 3, 4, 5, and […]

  9. […] between temperature and usage. If you’re really bored, see also parts 1, 2, 3, 4, 5, 6,7, and […]

  10. サボイ バッグ モンサンミッシェルの画像 http://www.bagsoinspection.info/


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Categories

%d bloggers like this: