Washington DC’s transit agency WMATA is preparing to publish a frequent network map, and wants your comments. To start, they’re taking the approach of showing frequency as a highlight on the main network map, as this makes it easier to show how multiple overlaid routes combine to form a frequent line.
For over a decade now as a consultant I’ve been emphasising the need for transit maps to provide clear visual signals about frequency. I laid out the case on the blog here.
I have one comment: WMATA is defining “frequent” as every 15 minutes all day, but on the draft map their legend refers to “buses/hour,” which is implicitly an average rather than worst case frequency. Nobody cares about average frequency! Frequency maps are about a guaranteed worst case, so they should show only corridors where the longest scheduled gap between trips is 15 minutes. Four buses an hour may not guarantee 15 minute service if they’re not scheduled with exactly even spacing. In fact, you can meet a standard of “four buses an hour” with four buses that all arrive in a bunch at the top of the hour, but in reality, that’s a 60 minute frequency, not 15!
I’ve working on one for DC, as it happens. I found that using the 15 minute standard didn’t work that well for DC. I analysed the daytime frequencies, and I found that they fall into two categories-quite-frequent-routes at least every 20 minutes, and really-frequent-routes at least every 10 minutes.
A lot of routes work on exactly those frequencies, with a lot of 5-10-min routes branching to 20-min routes, and a lot of others are very close to those (including some thoroughly annoying 21-minute frequencies, which I was torn as to whether to include).
So I’d have gone with a map that marks 10 minute routes and uses a noticeably thin line for 20 minute routes.
My first reaction to the WMATA’s map is that it is very, very difficult to trace a single route though that map! It’s lost in a mess of highlights and tiny, tiny colour lines, and makes for one of the least legible attempts at a diagram I’ve ever seen.
On the Washing DC FSN map – what a mess! I think they’ve missed the point in making such a map, namely to make it easy to see frequent routes. There is too much information that the FSN is completely lost on me.
I also agree with your point – the map should show worst case waiting, not averages.
I really hope they completely rethink that map as I don’t think it will do anything to boost ridership or make the network easier to navigate.
I had written an awesome comment, but twitter login ate it.
This is the best map we could hope for in DC. They previously told me that this kind of map is not useful because there’s already the system map, and customers already know the routes they use.
For the 15 minutes between buses, WMATA picked routes where the buses are coordinated to have even scheduling. Of course, WMATA considers buses on time if they’re anywhere from 2 minutes early to 7 minutes late, so the buses could be as much as 24 minutes apart and still be called “on-time”.
Having three different colors for 15, 12, and 10 minutes is overfussy. When I tried to draw a New York map with multiple standards, I used 15 and 7.5, and even that proved unworkable. If the agency doesn’t mind having multiple highlight colors, it should use much larger separation between standards – say, 20 and 10.
I would assume that 4 buses per hour are spaced at even 15-minute intervals, but it would be nice to see that explicitly stated, especially since so many of the corridors seem to be a agglomeration of multiple routes.
I think the map is readable at the regional level, but very difficult to understand for the downtown inset. I’ve always had trouble making or finding good maps of downtown transit systems. It almost always gets very complex and difficult to depict.
One other complaint, I don’t like the term “corridor-member route.” I understand what they mean, but I think many people will be confused.
If your 4 buses/hour don’t come every 15 minutes, you have to ask why… one possibility would be if they connect with a subway running every 6 minutes – then the buses might run at (say) 0, 18, 30 and 48 minutes past the hour.
The single best thing this map does is highlight the common trunk lines were various services share the same route. The branches of those lines might not add up to high-frequency service, but the trunk lines through the center of town do.
These lines are often labeled with a common convention. For example, the 30s lines all share a common route through downtown – the 32, 34, 36 and 39. They each branch out towards the end of the line, however.
One big thing this map does is show exactly what that common segment is. This allows a user to know they can catch any particular bus with a 30s number if they’re traveling along that common segment, and if they’re going beyond, then they need to research exactly what route to take.
I agree that the 3 tiers of frequency are too much, and that the frequency needs to be stated in absolute terms, rather than an average.
The graphic design is also a little busy and hard to read, but this is an excellent start for an agency who’s current bus map is indecipherable. See:
I’ve got a comment for WMATA: Stop being dumb about your GTFS data. When developers create applications using that data, it helps you increase your ridership…there is no need to scare developers away. A good application like Google Maps will do a hell of a lot more for your ridership than your print maps.
I lived in DC for 14 years.. all on the 14th st or 16th st lines. It was very, very common to wait for half an hour for a bus. Then, as you gave up, 3 buses would come along together. I somehow doubt, in the year I’ve been away, WMATA has gotten a handle on this problem.
This was common to see along CT Ave and WI Ave as well.