No riders at all?

February 9th, 2014

I was taking another look at the old ridership dataset SORTA shared with me last January, when I realized: there are a good many stops that have an average daily ridership of exactly zero1.

No Riders teaser map

There are really a lot of them, and they’re pretty evenly distributed. About 1,000 of them by my count2, compared to ~2,650 with at least some daily riders(above in black). I seem to have missed this before by immediately visualizing all the stops with circles sized according to their total ridership…naturally, these stops simply failed to render.

Click the image above (or here) for a PDF that will let you look up close at the locations of ghost stops throughout the whole system. Red dots are ghost stops, black circles are stops with riders on an average day; their area is proportional to the number of riders. The average day, including weekends, has ~46,100 passenger trips, not counting TANK.

It’s important to note that the presence of these low-to-no rider stops may not be hurting anything if we’re OK with the lines serving them being there in the first place. If no one is getting on or off, the bus probably isn’t slowing down by stopping there.

Show 2 footnotes

  1. well, not quite exactly. Total passenger counts for a month were divided by the number of days and rounded to the nearest integer. But basically, if any of these stops had even one person using them more than half the days, they would have been rounded up to > 0
  2. …which may vary from yours. I’ve aggregated stops that share the same name and or exact location. That means that stops that are paired on opposite sides of the streets were usually lumped together.

4 responses to “No riders at all?”

  1. MW says:

    It’d be interesting to see count of ghost stops that are at most 400 meters from a stop with a larger boarding value (and of the same line). And on the flip side, how many are legitimate access points to the system.

    • Nate Wessel says:

      Wish and you shall receive…eventually ;-)

      I’d be interested in seeing the unrounded numbers on a log scale. Most of these stops have undoubtedly had riders in the past, just less than 1 a day on average. Logging would pull out occasional riders (0.01 riders per day = -2 logged, etc) and real dead zones in a way that would probably be more revealing.

  2. Tim Bender says:

    The 1900 block of Greenup Street in Covington has two bus stops on the same block. I used to love pointing this out in staff meetings at TANK. I felt if there was ever a compelling reason to re-visit the bus stop spacing policy it was this egregious example.

    If I recall however, the consensus was always that if one of the stops never got used, it was no detriment to the system, but if that stop allowed one more person to use the bus, it was probably worth it.

    This is a classic value statement, and I learned to accept it. One rider using one stop on one day didn’t negatively impact the service (because the incremental delay was so small). And it was important for TANK, an agency charged with meeting the transportation needs of the entire Northern Kentucky community, to do their best to meet the needs of EVERYONE.

    I’ve refined my view on this policy in time. If an agency is willing and able to monitor these types of situations to ensure that system-wide impact is minimal, that’s fine with me. I’d probably guess that 5% redundant stops system-wide is tolerable. But if you don’t monitor it, or let the redundancy grow much larger this kind of attrition takes over your service. Trips become longer for passengers and more expensive for an agency to operate.

    Unused stops create an opportunity for new riders to use a service. It’s not inherently wrong for a stop to sit unused, but it is wrong if that stop penalizes other riders. Honoring service standards like bus stop spacing is a good way to ensure that unused stops don’t impact your service.

    Final note: Metro’s data show that something like 30% of stops are unused on an average weekday. What if you did a 100% count for a month or year? I’m curious to know how many stops TRULY go unused.

    • Nate Wessel says:

      I wish I could answer your last question…if only I had access to the raw data! All I have are rounded daily averages and no one at SORTA seems to have the time…

      But yeah. I think for most of these ’empty’ stops, since they’re all on lower-frequency lines, it’s probably more useful to ignore most of the stops and think of the line itself as practically a hail-a-ride type service. I when I was first using transit(SARTA in Stark County), there wern’t any stops near me, but I could just wait by the road and hail the bus when it went by. For the amount of riders they had, it probably just wasn’t worth the bother or expense to put up signs except at major destinations. And of course the drivers were happy to stop anywhere they could safely do so.