I live in an area of Brooklyn rich in transit options. If I want to get into Manhattan, I could take the 2 or 3 from Flatbush Ave. or the B and Q also from Flatbush Ave. I could walk to 4th Ave. and grab the R (and then switch to the D or N one stop later) or I could head south to catch the F at 7th Ave. If I’m going to the Village or Midtown, none of these four rides are that much quicker than any other. How then do I know which ride is the best?
The answer to that question is an easy one for people familiar with the subway system and a complicated one for underground novices. From experience, I know that F and R are the worst options, that the B and Q dump me into Midtown quickly and that the 2 and 3 are usually the best bet. Perhaps someone not used to the system could eliminate the F and R based on the map, but the current subway map doesn’t hold the key to my decisions. That key is frequency.
The 2 and 3 are generally my trains of choice into and out of Brooklyn because of their frequencies. During peak hours, the trains each run every four-to-six minutes during the peak hours, and no train is ever that far behind. The B and Q, on the other hand, each run only every eight-to-ten minutes during peak hours, and thus, the wait is longer. Because it generally takes the same amount of to get to, say, 34th St., it makes sense for me to take the train that arrives more frequently, but only those with a close familiarity with the subway system will know that.
Last month, Jarrett Walker at Human Transit considered the frequency map. He is critical of the way maps do not differentiate between routes based on frequency. “A transit map,” he writes, “that makes all lines look equally important is like a road map that doesn’t show the difference between a freeway and a gravel road.” While my example in this post marks the difference of only a minute or two, Walker’s point applies more to service that runs less frequently than every 15 minutes, bus routes and the way subways and buses can interact. He writes:
I contend that transit agencies have an obligation to push back against that complexity, to make their systems look as simple as possible, to give their citizens (not just their riders) the clearest possible image of how their system works. We may not need this understanding to follow your website’s directions, but it will help us if our trip is disrupted and we have to improvise, and it’s essential if we ever want to feel free to use the transit system spontaneously, for our own purposes.
OK, but how to we decide which routes are “major”?
This question may sound like a recipe for decades of focus groups, but in my experience, the answer may not be so complicated if we just consider the basics: frequency, span of service, speed, reliability and ease of access.
The rest of Walker’s post explores the theory in detail and applies it to the Minneapolis-St. Paul transit system. A few weeks ago, Cap’n Transit offered up his frequency map of Queens’ bus routes. It is a very simple idea that brings clarity to what can be a very confusing aspect of the transit system. Have you ever tried to make heads or tails of the Brooklyn bus map? It’s nearly incomprehensible.
For subway maps — and, in particular, for New York City subway maps, a frequency map could strike at the heart of the current version’s biggest shortfall: It provides a schematic of only peak-hour and midday service. It’s true that the map admits as much in the upper corner, but Transit doesn’t make available a map of weekend or overnight service. The changes — from shuttered lines and shortened routes to express trains making local stops — are significant enough to warrant one. A frequency map would be a great step in the right direction as the subway map would become even more user-friendly.
17 comments
Nice idea, but I think our system is far too complicated for something like that. Frequencies vary between the AM and PM rush hours; frequencies vary within a single rush hour; and of course frequencies vary between rush hour and midday.
On weekdays, everything runs every 10 minutes or better, except for the two branches of the A. That’s pretty much the same on weekends, with a few exceptions (I think just the 2, 3, and 5 – 12 minutes each). Only at night do headways grow beyond that.
Incidentally, the lines that go through DeKalb each run every 6 minutes during rush hours (in the peak direction).
Because there are few occasions when scheduled train frequencies are longer than 15 minutes, I don’t know if a frequency map is really worth the hassle. We already view the subway as NYC’s high frequency and inter-borough network, making a more accurate frequency map will probably not help a lot of people, especially because commuters likely start in a neighborhood with fewer transit options than Park Slope.
The idea of a night/overnight map is a good idea. Weekend service changes too frequently to warrant a permanent map, but the night map may be useful, the MTA does few service changes at night, and the service patterns change drastically. The frequency map idea is best for the bus system, which will highlight corridors in each borough that really have strong service, which is not as clear with buses as it is with subways.
Here’s the thing about the weekend map: If Subway Weekender can produce one every week, so too could the MTA. They could stick it on their website for those folks who hunt it out or they could eschew printing myriad confusing service change posters and condense the information into a map. Maybe it’s a printing cost issue, but considering the upcoming overhaul of their advisory posters, adding a map to it wouldn’t be that much of a challenge.
As a semi-frequent user of the Minneapolis-St. Paul bus system, I remember they added the “hi-frequency” designation not too long ago. Even though I already knew, as a semi-frequent user, that the lower-numbered routes were generally the most frequent routes, it was nice to see a visual confirmation of that.
Trains run at higher frequencies because they need to either because of their popularity with the riders or because they run through the most congested neighborhoods. Trains that run at lower frequencies tend to be routes that are used by fewer people. Adding a frequency designation would just further pull people away from the lesser used routes and add more to the already crowded ones. We already have a problem with riders waiting for packed express trains while empty locals pass by. This would be a terrible idea.
I’ll take the B division over the A division any time I have the option. Yes, the headways are usually longer, but the cars are wider and quite often less crowded.
The holy grail is a live-feed, digital map in every car and every station that gives directions, adapts to service changes, etc.
I was thinking, it’s really time to put such a map in key stations at least. It may be too much to expect in the entire system, until it can be built into future rolling stock anyway, but places that would make sense include Grand Central, Times Square, Fulton(Manhattan), Flushing, Atlantic Avenue, Penn Station, 125th Street, stadia stations, and maybe even Myrtle-Wyckoff.
Most people figure out which is the fastest route after several trials and errors if it is a regular trip. If its only a one time trip, what does a few minutes matter and why bother showing frequencies so much needs to be provided on the map anyway? It’s as cluttered as is. The new version which eliminates much information did not declutter the map because the print size was enlarged and the angles of the atation names were shifted so that now you have to turn your head to read them. They should have left those names as is.
Another problem with showing frequencies, is that it would tend to encourage riders to make crowded lines even more crowded, slowing them down and encouraging the MTA to reduce service even further on the lines passengers would be taken from. Frequency Map – Bad idea.
I agree I don’t see the need on the subways, but I think it’s valuable for bus maps.
Bus maps have schedules on the back. As for the maps in the buses, I don’t see how it would be possible to incorporate all that information on the map itself.
hmmm, again i second that it sounds like a cool idea, but not one with a constant enough time table to be represented on a map. if frequency can be every 4 minutes during peaks, and then every 20 during off peaks, and what if the common scenario occurs when somebody could take the 1,2, or 3, or the 4, or 5 at around 3 am, in manhattan just below 14th? when looked at on a train by train basis the trains are 20 minutes apart, but combined all trains could be every 7 minutes, and how could that be represented. I think this could only be done on a software program, an app of course for an iphone.
Los Angeles prints and makes a PDF map available of all bus lines in the city that run on 12-minute headways all day, giving riders the option to choose one of these routes over a slower one. The subways, however, run on a 6-12 minute headway most of the day, so they don’t bother to make a subway map available showing headways.
Interesting idea. Not sure I agree with your example at all though. I know it’s a little beside the point, but I’ve done my fair share of commuting to Penn Station from that same region of Brooklyn, and in my experience it’s MUCH faster to get there using any of the lines that go over the bridge (I use the R to the D usually), rather than plodding underground through downtown BK and Manhattan on the 2/3. You may gain 3-4 minutes by having the 2/3 come sooner, but you lose 8-10 on the trip. To each his own, though.
Jarrett would focus on making a frequent bus map, and publishing it alongside the subway map. A frequent subway map wouldn’t matter too much; as rapid transit, subways are already the top level of transit available in the city. His argument is that the main distinction for service is frequency within the class of non-rapid transit; to talk about rapid subway map is to turn him into a strawman.
As if the depiction of the system on the current map isn’t complicated enough already, you’d want to add even more clutter to it. Is it really that important to know whether you’d be waiting two minutes at a platform or four? And would that really help you to make a better decision? What if there’s more than one option to reach my destination from my point of departure? Based on your map, I’d take the most frequent line. However, what if that same line actually takes me longer to reach my destination than if I’d opt for the less frequent line? That wouldn’t make the frequency map more user-friendly, just more misleading.
What I would like to see in the (near) future is personalized travel information. This information system should be able to pick up my current location, ask me for my destination (which could be a subway station, but also an address) and then, based on live transit data, offers me the quickest and/or most convenient travel option to my destination. That could be line X or line Y, depending on my current location and the time of day I’m travelling.
The technology required for such a system exists as we speak: we’ve got GPS satellites, smart phones, huge databases with travel data we could link our application to. What’s really missing is an incentive to put those elements together into a comprehensive application (and some $$$, too).
I think the MTA should create a map that shows weekend service and print it on the back of the existing map. There are several lines that either have their routes truncated, or don’t run at all, and this information has been made virtually undetectable on the new map now that they’ve eliminated the table that used to show each line’s schedule for weekday, late-night, and weekends. Moreover, this map wouldn’t be all that different from late night service so with a few minor notations here and there, it could serve both purposes. I understand that every weekend brings all sorts of service changes for construction, but at least one could get an idea of just how much less service is available during off peak times–particulary for tourists.
While it’s nice to have a LIRR/Metro-North railroad on the back of the current map, shouldn’t those customers simply have their own map?