Connections vs Complexity

In my first “basics” post on connections, I explained why a network that requires connections (or as North Americans call them, “transfers”) can actually get people where they’re going faster than a network that tries to avoid them.

But there’s another important reason to plan for connections rather than direct service, one that should be important to anyone who wants transit to be broadly relevant to urban life: Unless you welcome and encourage connections, your network will become very, very complex.

If a transit system is too complex for people to learn, it can’t be used spontaneously.  If you can’t use transit spontaneously, then transit is reduced to a very limited role in urban life, serving only trips that people make regularly and can therefore learn how to make even in a complex system.

In presentations on this point, I often compare a network based on connections — such as that of Los Angeles — with a network based on fear of them — such as the inner network in Sydney. Here are more or less comparable slices of the two networks, each image about 8 km wide and 3 km high. Both lie just south of the CBD of each city, so you can see routes converging toward the CBD on the north edge of the map.

Los Angeles:

La for syd

Sydney:

Syd for la

Network maps drawn by transit agencies are often hard to compare, because each is full of details and conventions that make sense only locally.  To make sense of the Los Angeles map, you must know that line numbers in the 300s and 700s are limited-stop or “Rapid” services that almost always appear on top of a local-stop service on the same street.  So for example on Vernon Avenue, the big east-west corridor across the center of the Los Angeles image, 105 and 705 are local-stop and rapid versions of the same line.  These two services are not duplicating each other; in fact, they are designed to complement each other.  So you should think of 105 and 705 as a single composite line.

Once you account for that, the striking thing about the Los Angeles map is that most streets have only a single line on them, even though we are immediately south of downtown and you would expect many lines to be converging.

By contrast, in a similar sized slice of Sydney, most streets have several lines on them.  In most cases, these lines are not designed together. In Los Angeles, if you see two buses right next to each other for many miles along the street, they are probably on the same line and one of them is probably late (a phenomenon called bunching.)  In Sydney, if you see the same thing, the two buses may well be scheduled to do that, usually because they’re on different lines that run on top of each other on the same street and no attempt has been made to co-ordinate their schedules.  For example, Sydney’s Oxford Street, for the entire 3.4 km between Taylor Square and Bondi Junction, has four local bus lines and one limited-stop line, ranging in frequency from 10 minutes to 30 minutes, with no particular coordination of their schedules.  On this street, bunches of buses are a common sight because, in many cases, they’re scheduled that way.

Sydney-style networks happen where the prevailing assumption is that people will resist connections at all costs, which means that there is a strong motivation to run direct services between every possible origin and destination.  Even where planners would like to introduce connection-based design, they are frustrated by a fare structure, out of their control, which penalizes connections.  So Sydney ends up with lots of overlapping lines, a high degree of complexity.

This problem is also the cause of others, such as the excessive volume of buses, most of them not full, choking the narrow streets of the CBD.  Obviously, if you didn’t feel constrained to run direct service from everywhere to the CBD, you’d organize some buses to feed others (or to feed trains).  This would give you fewer buses in the CBD, with heavier loads on those buses, and thus a stronger moral claim to transit priorities, such as exclusive lanes.

Just for fun, here’s an 8x3km slice of an even simpler network, also just south of downtown, though admittedly in a smaller city.  Portland:

Pdx for syd

I often show this map in Australia because the pattern is even more clear than in Los Angeles.  The simplicity is extreme.  Every arterial has exactly one bus line on it, with a simple 1-2 digit number.  The only exceptions are very close to downtown, where radial lines inevitably have to converge, and a few ghostly line numbers that appear on a white background, which signifies a handful of trips that run only on the peak.  (Note, by the way, how this notation helps the viewer see the all-day network on which she might rely.  This clarity is quite unavailable on the Sydney map.  The Los Angeles map also doesn’t distinguish peak-only services, though across much of the city there are very few of them.)

The Portland map also makes clear that even in a city with an exceptionally strong downtown, many frequent lines don’t go there, including 70, 71, 72, and 75 on the map above.  The focus, as in Los Angeles, is on anywhere-to-anywhere travel using connections.  Because these “crosstown” lines rely on connections more than the radial lines do, they all run as frequently as possible and as late into the night as possible.  Sydney, too, has non-downtown lines; with one exception, the 400, they are too infrequent for fast connections to be possible, and often shut down earlier than the rest of the network.  Again, the frequency is hard to afford in Sydney because resources are divided over so many lines, even those these lines are often overlapping.  Portland and Los Angeles, by contrast, are running the minimum possible number and length of lines, so that they can run each one as frequently, and as late, as possible.

Sydney planners will sometimes argue that Los Angeles and Portland are simpler because they already have street grids, while Sydney is an unplanned accretion with no large-scale order to its street network.  This is partly true; grids are certainly easier to remember, in both cities and transit systems. But you can take a structure like Sydney’s and design a more gridlike system, featuring fewer yet more frequent lines, all designed to work together to serve any origin-destination pair, rather than picking a few major destinations as today’s network tends to do.  I did a sketch of one here.

UPDATE:  An interesting series of comments, starting here, is applying this thinking to equally gridless London, including a very, very frequent grid bus network sketch by commenter Zoltan, which is here.  Remember that a polar or “spiderweb” grid, consisting of radials and circles, can also achieve good connectivity while still focusing on a center.  Most workable grid networks, including my sketch for Sydney, are a hybrid of polar grids and standard rectangular grids.

47 Responses to Connections vs Complexity

  1. teme November 21, 2010 at 1:57 am #

    Point taken, but the real issue is how good the connections are. Synchronized same platform connections are no problem, but they are difficult and often impossible to organize. For instance, and I haven’t been to LA, the image I get looking at the map is that in order to connect I would possibly have to cross up to two main roads. And this is not even considering all the issues that come with grade separation, or street corners one would rather not get off at.
    In abstract, way I would draw the routes in a grid would be to have a turn on two in the route to allow for easier connections. Say one block north in an east-west line for example, or L-shaped routes.

  2. Phil Jones November 21, 2010 at 2:07 am #

    I’d be interested to read your analysis of London’s bus network. My inclination is that it is more like Sydneys than LA’s. The network is very difficult to visualise, certainly compared to the tube network. Nevertheless ridership levels are high and increasing. How much higher would they be if the system was comprehensible to the occasional user?

  3. Zoltán November 21, 2010 at 2:57 am #

    @Phil Jones
    Firstly, much of the high and increasing ridership levels is a result of two things:
    – The fact it costs £8 to drive into Central London, even before you think of parking.
    – The inadequate rail network and its very high fares.
    But it’s also true that London’s network manages to provide high frequencies on a complicated network. As such, the complication doesn’t quite prove as damaging as it would be elsewhere, where the same complication would be to the great detriment of frequency.
    It does this in part off very high transit ridership and high density. But only in part. To maintain these frequencies, a lot of buses run unimpressively loaded – I find myself often sharing a double-deck bus with <20 others in Central London - and as a result, it requires more subsidy for buses than the rest of Great Britain put together. In other words, it's only a sustainable model if you're willing to throw money at it.
    And some issues remain despite high frequency:
    - Most tourists are intimidated by the complex bus network, and stick exclusively to the underground instead. London tourists are famous for leaving with very little understanding of the city they've just visited.
    - Most out-of-town visitors for business will take taxis because they are intimidated by the bus network.
    - A lot of anecdotal evidence I've got from people living in London and using transit suggests that even locals stick largely to the few routes that they know.
    - Cyclists often complain about menacing buses, largely because buses make a whole lot of difficult turns on London's narrow streets.
    -This has meant that introduction of articulated buses, with high capacity and far lower stop dwell times than double deckers, has been resisted, and now completely halted by the new mayor.
    So even with the amount of money London has to lavish on its network, a complex network isn't necessarily the best kind. In fact, if you can subsidise such high frequencies on a simpler, any transfers will only involve a minute or two's wait at most, less than your wait for your direct bus, or the five minutes spent at every stop studying the maps and trying to work out where the hell these buses actually go to.

  4. Zoltán November 21, 2010 at 3:09 am #

    I did, incidentally, a thought experiment (not based on the sort of analysis you’d actually implement such a thing based on) on how a simplified network through Central London would look, and sketched this out:
    http://maps.google.co.uk/maps/ms?ie=UTF8&hl=en&msa=0&msid=107489113689713493592.000494a08af03b1ed2315&ll=51.511734,-0.114498&spn=0.060041,0.153637&z=13

  5. Jack Horner November 21, 2010 at 3:16 am #

    Sydney is now creating a new light rail line on an old freight railway in the inner western suburbs about 6km west of the area pictured in Jarrett’s sample map.
    The bus network in that area is even more of a spaghetti than the area pictured.
    I have argued, and would encourage others to argue, that the light rail should be accompanied by a general overhaul of bus routes in the area to create a simpler network of more frequent routes to promote ‘anywhere to anywhere’ travel.
    So far the authorities appear to have no interest in that type of broader network planning. It’s much easier to throw money at physical infrastructure than it is to apply intelligent thought to getting the best value for money.

  6. Andrew November 21, 2010 at 3:39 am #

    I think that the emergence of trip planners online and on smartphones ought to make this issue less important. For instance in London the bus network and the commuter rail network are both very complex, but TfL has a very good online trip planner. Since most lines are very frequent this makes network complexity much less important than it would seem from looking at a paper map of Central London bus routes.

  7. Alan Robinson November 21, 2010 at 6:56 am #

    @ Andrew
    There are some of us without smartphones for which a trip planner is of limited utility for spontaneous trips.

  8. Alexander November 21, 2010 at 7:43 am #

    @Zoltán—that’s a lovely and legible network—I’d love to see it superimposed on the Tube map. 🙂 I’ve only talked to a Londoner about buses once, in Chicago, when he mentioned that Chicago’s full of “bendy buses.” For some reason this term made me giggle, and he asked what we called them. I replied “articulated,” which sent him into a storm of laughter.
    @Jack Horner—I wonder how much of a constituency each of those snaking bus lines has. Several years ago, when Milwaukee was considering putting a light rail or rapid trolleybus line along Wisconsin Avenue west of downtown. Even though Milwaukee’s bus system a fairly regular grid, along Wisconsin some bus routes will turn ninety degrees to run along Wisconsin to downtown (for example, one bus route goes serves N. 16th, another S. 16th, but head east along Wisconsin). With the light rail/trolleybus plans it was anticipated that these buses would become crosstown routes, “forcing transfers” (harshest possible language!) on current one-seat rides. Although it wasn’t a big factor in the plan’s demise, I do remember this being used as an excuse.

  9. ant6n November 21, 2010 at 9:22 am #

    @Andrew
    I would argue that even with trip planners, you still want to be able to understand the system – it’s like arguing that since you have a GPS, you don’t need to know the street grid at all. I would even argue that any transit app should help you to understand the transit system better, rather than try to hide it with point-to-point-at-some-time information.
    @Jack
    Rail lines probably change bus lines in many places. I wonder, even, whether the construction of light rail lines has forced the surrounding bus networks to not be spaghetti systems. Before rail, many bus networks would simply provide a bunch of one-seat rides into downtown. With the rail line, the planners would want to increase its ridership as much as possible – so they would turn most of the surrounding lines into feeder line, probably perpendicular to the rail line. The bus lines would not have to converge at one point downtown, but they would have to converge onto the light rail _corridor_.
    Maybe some rail lines become successful mostly because they change the surrounding bus network (Calgary C-Train), and turn infrequent one seat rides into two-seat rides along a simple grid. Of course, it’s much easier to sell bus+train to people, rather than bus+bus.

  10. anonymouse November 21, 2010 at 8:43 pm #

    A trip planner is even less useful than a GPS. A GPS can tell you how to get from point A to point B. A trip planner can tell you how to get from point A to point B at time T. And I do mean exactly time T, not time T plus one minute because you stopped to have a quick word with a coworker on your way out the foor at work. And certainly not at some entirely different time of day or day of week, unless trip planners get the quite useful feature of being able to stick to the frequent network. Which, in many cities, isn’t all that helpful anyway.

  11. M1EK November 22, 2010 at 5:44 am #

    @teme:
    “Synchronized same platform connections are no problem”
    Not true. They are clearly a problem; they’re just less of a problem than non-synchronized connections. New transit starts in the US over the last couple of decades clearly show that requiring transfers, no matter how easy, turns off a large number of choice commuters – as much as people would like to wish otherwise.

  12. M1EK November 22, 2010 at 5:46 am #

    And, by the way, for those of you falling into the trap of groupthink; it’s not just me saying this; here’s Christof Spieler in Houston (now a board member at Metro by the way): http://www.ctchouston.org/intermodality/2007/03/28/where-are-more-people-going/
    “Successful rail transit systems serve dense employment centers. Our light rail line is successful because it serves two, and it goes right into the middle of both of them. People don’t like to transfer; they want to be able to get off the train and walk to work.”
    […]
    “Notice a pattern? Passengers don’t want to transfer to a circulator service to get to work, even a high-quality circulator like Denver’s.”
    […]
    “If we want to maximize the number of people who will take transit (which should be the goal) we need to find places where transit will serve as many people as possible as conveniently as possible. That means serving density, particularly employment density, directly.”

  13. Tom West November 22, 2010 at 6:21 am #

    Anotehr big difference betwene the two: LA’s bus map does not use colour to distinguish between routes – it usses colour to signify otehr things (frequency/type of service). Sydny has to vary colour by route, becase you have multiple routes on one street.
    As someone who know almost nothing about L.A.’s transiet network, I was able to look at the map and gyess the thick red lines indicated some sort of “better” bus service than the thin orange ones, and that the blue line with named stops indicates some osrt of rapid(er) transit.
    Looking at Syndey’s map, I get none of that.

  14. John November 22, 2010 at 8:26 am #

    This has been presented as an either-or choice, connections vs. complexity. Obviously complexity is bad, but so are connections from the customer point of view. Isn’t there a way to design a system that offers many one seat rides AND is easy to understand? I have no doubt that changing the system design could help in Sydney, but maybe a combination of scheduling, nomenclature, and cartographic changes could help make the existing system more comprehensible?

  15. Rico November 22, 2010 at 9:40 am #

    I grew up in Vancouver, it was common (usual) to take a bus to the Skytrain and then a bus from Skytrain to the final destination. I can’t speak for other cities but provided the frequency is sufficient transfering never detered me or my friends (note: I was from the suburbs and the outlying bus was usually not frequent (every 20-30min) while the downtown portions were usually less than 5min frequency. This meant I only needed to know an approximate schedule for 1 leg of the trip).

  16. M1EK November 22, 2010 at 10:23 am #

    @Rico, refer to ‘mature transit cities’ and ‘transit-dependent by choice’. If you’re trying to get somebody who currently drives to consider your service, a transfer is a killer – unless parking is super-expensive at the core, and in that case, they’re probably already using your service anyways.

  17. rico November 22, 2010 at 10:50 am #

    @M1EK
    Vancouver has expensive parking at the core, but in reality transit in the core was more convient than driving (not true in the suburbs where transit sucked). I took transit downtown and drove in the suburbs. Transfers downtown where there was high frequency was a non issue, same with transfers to Skytrain that also had high frequency.

  18. rico November 22, 2010 at 10:55 am #

    @M1EK
    Oops implied in my last point is I did not take transit in the suburbs unless there was point to point service because transfering on low frequency routes sucked but that me and my friends used transit where high frequency service existed.

  19. Jonathon November 22, 2010 at 11:07 am #

    Toronto is lucky to have a very well-laid out grid system of roads interrupted only by natural barriers such as river valleys. As such, the TTC network is built on the idea that a bus route serves to get you either a) to another point along that road, or b) to the subway if you are going farther. When I walk out to Bathurst Street by my house, I dont even have to look at a map to know that the Bathurst bus runs along Bathurst street, north to the edge of the city, and south to the subway.

  20. Tom West November 22, 2010 at 12:58 pm #

    @John: Short answer is “no”.
    Longer answer…. complexity requires greaer resources for a given level of service – see https://www.humantransit.org/2009/04/why-transferring-is-good-for-you-and-good-for-your-city.html .

  21. bzcat November 22, 2010 at 1:10 pm #

    @Tom West
    The color coding is one thing that LA did right. I have introduced the bus system here in LA to many visitors and it was very instructive to tell people to “wait for the red bus” and look for the red bus lines on the map to see where you are going. The grid layout really helps but the color coding makes it easy to absorb.

  22. M1EK November 22, 2010 at 1:23 pm #

    @rico, then you’re arguing about a case where transit already had the deck stacked in its favor. If we only want to talk about the pitiful number of ANZUS cities and relatively small populations in those cities where that’s true, I suppose we could do so, but that’s not been the general thrust here, and those cities didn’t get that way by making everybody transfer on day one.

  23. rico November 22, 2010 at 2:15 pm #

    @M1EK, actually what I would argue is that in Vancouver there are many routes that run at high frequencies because they feed major Skytrain/Bline bus routes and that because of that transfers (and transit in general) is convinient (it could also be we have ‘learned’ that transfers are usually not painful). In a world with limited $ lots of short but high frequency routes feeding other high frequency routes seems like a better use money than low frequency routes going exactly where I want to go….but not when I want…and maybe Vancouver transit users are different than other N. American users (I doubt it though)but those I know will transfer as long as the wait is reasonable (note: my aquaintances are choice riders and own cars). The same has been true when travelling overseas (Europe, Asia) I don’t expect to get to my destination without transfering and as long as the connection is quick I tend to be happy with the service.

  24. Jarrett November 22, 2010 at 2:41 pm #

    @M1EK. I’m aware of the situation you’ve described in Austin, where a rail line ends not quite downtown and requires transfers to shuttles to reach any major destination. I’m very familiar with this from my years riding the Bay Area’s Caltrain system, which also ends just outside most people’s walking distance to the San Francisco downtown core.
    I would completely agree with you that such an arragement is undesirable. If you have a lot of people who are all going the same direction, then connections should and often can be avoided. It sometimes happens as an interim situation when an authority decides to get a rail line started and build the expensive downtown segment later; such decisions do get made and sometimes they reflect the political reality.
    But commenters in Vancouver aren’t connecting with you’re view because they’re talking about a completely different situation. They are talking about the inevitable role of connections in a system that’s designed to distribute people to many destinations, not just downtown. They’re referring to mature systems where rail transit does penetrate the core, but where people still transfer in large numbers to reach either their dispersed home locations or major secondary activity centers outside of downtown, such as the universities.
    Apples and oranges, really!

  25. Alon Levy November 22, 2010 at 3:28 pm #

    M1EK, the “high-quality” connections you’re talking about are still pretty shitty by European standards. These aren’t cross-platform, barrier-free, same-ticket, zero-wait transfers; transfers satisfying those preconditions, when they exist, are widely accepted even among car-owning Americans. “Change at Jamaica” is common parlance among LIRR riders, precisely because the transfer is so simple. It’s much more common than changing at Secaucus on the other side of Manhattan, because there the transfer involves climbing and descending stairs, passing through a faregate, and hoping you won’t wait too much for the connecting train.

  26. M1EK November 22, 2010 at 8:00 pm #

    @Jarrett, apples and oranges was the point I was trying to make to Rico. And the problem with Austin’s system is that a heavy railway running DMU cars is never going to penetrate downtown any further than it does now (which is to say it’s on the edge); nor will it ever go to the University of Texas or the State Capitol like the light rail line it killed would have.
    The reason we went with this commuter rail line is that a bunch of people bought into the idea that timed transfers (buses waiting for the train, Alon, closer even than ‘same platform’ in many cases) wouldn’t have any negative effect on ridership compared to the direct route (to Capitol; UT; etc) the light rail line that narrowly lost in a rigged election in 2000 would have had.
    Oops.

  27. Alon Levy November 22, 2010 at 9:57 pm #

    First, cross-platform and same-platform aren’t the same.
    Second, there’s a difference, if only psychologically, between a line that serves some destinations directly and requires transfers for others, and a line that makes everyone transfers. It’s especially different if the connection is to an unreliable shuttle bus.
    And third, there’s something very wrong with quoting Spieler as someone who’s supposedly anti-transfer, when he specifically commends the Yamanote Line for the fact that “all but 2 stations have connections to other rail transit lines” and explains that “a connection to a reliable transit service that runs every 5 minutes is different than a connection to an occasional shuttle bus that gets stuck in traffic.”

  28. SamSnead November 23, 2010 at 12:54 am #

    Here is the real horror story: for all the praise given in this article for LA’s grid system, our all-knowing MTA has begun the public hearing process of regressing back to the “point to point” hell system that pre-dates the current grind that was imposed back around 1980 or there abouts. I remember the big announcement and the massive changes to taking the bus on a grid system. When I saw the plans back then I thought, “What took them so long to get it?” And when implemented GONE were the meandering lines that were often detoured for the sake of a Los Angels city councilmen who wanted as close to a “door to door” service to get re-elected. Heavens, getting around by bun in LA improved about 99%. Almost all my varied commutes now took NO MORE than 2 buses, and they got there is fair time, and no bunching of lines and those line even “re-bunched” further into the neighborhoods
    There was some pain with the grid: instead of walking to the end of my block for the bus, I had to walk 2 short blocks and down a set of stairs. But the bus avoided the congestion that our public school at the other end of the block created.
    So it seems like our beloved MTA has exhausted in ever changing moniker, paint scheme, Metro M design, method of indicating color at rail stations, sliced the Red Line into a Red and Purple, pasting “Local” “Express” etc. on buses intending to make it easier for people to indentify, but–as I expected–often the assing a “Local” livery bus on an “Express route and vice-verse, are now moving to scramble our grid system just because they are bored again (they claim the point to point would increase ridership). Sorry, but it seems like planners just can’t leave something working well alone and have to justify their jobs. There are plenty of other matters that the MTA needs to address. Scrambling the grid bus system is not one of them.

  29. M1EK November 23, 2010 at 6:09 am #

    @Alon, you may pick those nits, but others are using your words (and especially Jarrett’s) to defend the concept of the transfer barn on the edge of the core.
    And Christof has been very vocal about NOT building a starter line that requires transfers. If you think I misquoted him, go read that whole post and tell me where I took him out of context.

  30. M1EK November 23, 2010 at 6:11 am #

    @Alon, and as for the transfer itself, it meets every metric you and Jarrett have laid out before for a supposedly penalty-free connection: it’s a VERY short walk (shorter than many of your rail connections I bet); the vehicle is ALWAYS waiting for you (so the ‘occasional’ comment is not applicable); and it’s free).
    Just like Tri-Rail’s shuttles. And in both cases, choice commuters stay away in droves.

  31. Alon Levy November 23, 2010 at 10:04 am #

    I have never seen Jarrett say anything positive about the Austin commuter rail system, or about the steam-era edge-of-town barn. Calm down. What he says is that a decent bus system should be based on a simple, easy-to-understand frequent network, with easy transfers.
    Not everything is about Austin. Austin doesn’t mean transfers are unworkable any more than San Jose means light rail will never succeed.

  32. M1EK November 23, 2010 at 12:19 pm #

    @Alon, I have seen Jarrett make recommendations here which are used to justify the Austin system’s design. He (and you) say transfer penalties basically don’t exist in most cases; and transit agencies desperate to believe that latch on to it (more from Jarrett, obviously).
    Transfers should be viewed as a necessary evil, not a positive feature. Yet everybody and their uncle these days extolls the virtues of intermodal centres in which everybody is routed for mandatory transfers to get anywhere worth going.

  33. rico November 23, 2010 at 3:55 pm #

    Throwing my 2cents back in transfers are about convinience. If there is too much of a penalty for transfering (time/money/stress) it will reduce ridership. But if the total trip including transfers remains more (or sufficiently) convinient than the alternatives a transfer will not reduce ridership. How much more frequent service can you run if you use a grid system feeding major corridors, and is frequent service more convienient than the hassels of tranfering? Of course it always depends but my experience says used wisely transfers can increase convience by increasing frequency.

  34. Alon Levy November 23, 2010 at 4:56 pm #

    First, it’s not really me saying transfer penalties don’t exist. I’m not completely sure what Jarrett’s view is, but the explicit ridership model I learned from is the MTA’s, which models transfer penalties as wait penalties. The MTA computes every minute spent waiting or walking between platforms as equal to 1.75 minutes spent in motion, whence timed cross-platform transfers are zero-penalty.
    Clearly, there should be some refinements to this system, and the above should be taken as a first-order approximation. I asked Jarrett, and he admitted the ridership models he’s heard of don’t go much further because it’d be too complicated. But these complications exist. For example, in air travel, it seems that passengers prefer transfers located close to the beginning or end of the trip rather than in the middle. On commuter systems, connecting from a suburban bus to a train has worked in Canada and Australia while connecting from a train to an urban bus has been much less successful, going back to mid-19th century London, where the Underground was built as a better way to connect the train stations to the CBD.
    Second, my point about Austin not being a representative example could be carried to other underperforming cities, too. For example, the insistence on direct LRT and BART service in the Bay Area actually undermines ridership relative to improving the mainline commuter networks. Ironically, it’s this insistence on BART that has choked off funds to Caltrain, forcing it to terminate right outside downtown San Francisco until California High-Speed Rail pays for a tunnel to the CBD.

  35. M1EK November 23, 2010 at 5:22 pm #

    @rico, @Alon, the problem is that the world in which transfers increase frequency without drastically penalizing trip length is illusory outside a few cities. And the idea that choice commuters will give up the convenience of a one-seat ride (their car) in favor of a three-seat ride just because one of the two seats runs more frequently is ludicrous. It’s hard enough to get them to switch to the two-seater.

  36. Alon Levy November 23, 2010 at 7:17 pm #

    On the contrary, it’s only in a few cities that you can provide direct service from anywhere to everywhere without cutting frequencies unacceptably. For smaller cities, even ones with Sunbelt development patterns, buses to rail is the normal way it works. That Austin’s rail-bus connectivity sucks doesn’t mean that Calgary’s does.

  37. anonymouse November 23, 2010 at 8:14 pm #

    On the topic of grid versus point-to-point: the latter can do a very good job of serving fixed commutes on fixed schedules. The former will do a worse job of serving those, but a much better job of serving a much wider range of trips.
    And for M1EK, a thought comes to mind: it’s not a matter of the bus waiting for the train but the train waiting for the bus. I assume the train runs half-hourly or so? For the outbound trip, is there a specific bus trip that meets each train? Because I imagine that it would be very, very frustrating to have to wait an extra half hour because the stupid shuttle missed the light and you got to see the train leave without you.

  38. Go Metro November 23, 2010 at 8:18 pm #

    Dear M1EK,
    Here’s the basic idea that Jarrett is trying to convey: riding transit should be about as simple as driving. For example, if all I need to know is that I am going to walk nearby to an arterial like Vermont Ave, take any bus to Wilshire Blvd, then take any bus on Wilshire to get to where I need to go, without having to worry about the schedule or the route because (a) the route is simple – it’s the street itself, (b) it’s frequent enough that I don’t have to wait very long, and (c) you don’t have to worry about which bus to take, then anyone can use it, not just those who carry timetables and route maps wherever they go.
    Where this type of system works best is on a grid, where the street system is very regular and thus you can think of the bus line as being the same as the street itself. Another characteristic of a city where this works well is where you have decentralization. Direct point-to-point service can be a viable option when there is a heavily disproportionate flow of traffic to and from one center, but for most North American cities a majority of jobs are now dispersed throughout the suburbs, where that model doesn’t work – plus many trips are not work commutes anyway.
    Bus frequency is very crucial to creating a reliable experience for a rider. Anything greater than about 10 minute frequency leads to significant deviations from a published schedule if the bus travels in mixed flow lanes with a lot of traffic. When the buses are more frequent, the schedule becomes secondary to the headway between buses.
    Note: LACMTA (“Metro”) is moving away from the red (rapid) vs. orange (local) color scheme. Aside from the obvious cost of having additional equipment redundancy, operationally there are times when bus runs have been canceled because none of the red buses had returned to the yard yet. Although this is going to undermine the “Rapid” brand concept, the operational benefit of having trips not get canceled is probably the right decision. The electronic signs on new LACMTA buses allow displays in different colors, so that might mitigate it partially.
    Without a doubt, forcing a transfer like the DMU line in Austin, or even between buses with anything less than high frequency headways, rapidly turns off any potential rider who has access to a car. Even if one of those buses is a high-frequency line and has a dedicated lane that moves faster than autos, most folks won’t put up with a long wait for the other part of their ride – you’re right about that.
    Note to Jarrett: one fantastic benefit of encouraging passes vs free transfers here in LA is that it reduces boarding times substantially on routes with heavy ridership, like Wilshire. Sometimes it used to take 3, 4, even 5 signal cycles before everyone could board (!); now it is rarely more than 2 signal cycles. The main problem is the horrible TAP “smart” card system; when people could buy a day pass on the bus, it worked like a charm, but now you have to buy the card beforehand, then load the fare when you board.
    Another benefit to encouraging passes vs. transfers is that it actually makes spontaneous travel more easy: you don’t have to worry about whether your transfer will expire or whether you’ll be breaking the rules by stopping along the way to buy coffee. The challenges with balancing needs vs. simplicity in transfer rules was one of the major reasons Metro chose to move to a simple day pass system.
    Moving from transfers to day passes (while retaining the weekly and monthly passes) was one of the best decisions Metro has made – about as good as TAP has been bad.

  39. Alon Levy November 24, 2010 at 3:56 am #

    I just now bothered to check Capital Metro Rail’s schedule. As you can see, there’s no midday or evening service and barely any reverse-peak service; even peak service is infrequent and has an inscrutable schedule. In other words, there’s no contradiction between this line’s utter failure and anything Jarrett has said so far about the importance of frequency, or what TPM will say about the importance of a simple clockface schedule.
    Thanks to Go Metro for reminding me in his comment that I should actually go checking these things.

  40. M1EK November 24, 2010 at 6:15 am #

    Guys, another piece of helpful advice: don’t get smug too quickly.
    The shuttle buses referred to here are dedicated to the train – you do not have a long wait for either part of the ride.
    The schedule for this thing is irrelevant – people know how to read a clock; and they know how to get to a park-and-ride on time.
    And their direct competitor, existing express buses that go on a far more direct route to the major central employment centers, also run on a less-than-frequent schedule.
    CM is spending a lot more operating dollars starting January to implement mid-day service on roughly the same frequencies as the express buses already do – and they are, and this is important, cancelling a couple of the more compelling express bus competitors. Without the latter, I would not expect any major ridership bump from the former – this is, after all, a work-oriented service (which means it has to serve choice commuters even better, relatively, than a typical all-day service would).

  41. Alon Levy November 24, 2010 at 3:06 pm #

    M1EK, after complaining about how frequency doesn’t matter because it doesn’t, you probably shouldn’t call people smug.
    The schedule actually is relevant. In Germany, they’ve learned that if you want people to be able to remember when trains run, you need the schedule to be clockface – e.g. trains leave at 15 minutes after the hour, every half hour. They would not do anything as stupid as implementing a 35-minute frequency, on the assumption that “People know how to read a clock.” They would also minimize the use of park and rides, because in a city that’s not completely gridlocked, choice commuters who need to drive for part of the way would need to drive the whole way. Instead, they’d focus on placing stops in every walkable area, with some timed bus-rail connections in the suburbs.

  42. Chief Clerk November 24, 2010 at 7:32 pm #

    Another complications of Sydney routes is the invention of a new route number to show the slightest variation from an existing route. e.g. routes 444 and 445 which differ by a few blocks somewhere along the way. Then there’s the substitution of E or L for the first digit, L90 or E86 in express / limited stop services. (Thereby losing the information of the region which the fist digit gives.) The online timetables are restricted to that route only with no cross-reference to the variants – you may find 3 trips per day where you’ve looked, and be unaware that another route number has the frequent service.
    Of course the trip planner will find you these services, but at the cost of minute demands to specify bus stops, street corners etc and instructions to walk 15 metres which will take you 2 minutes …

  43. M1EK November 29, 2010 at 7:35 am #

    Alon, the key is that I’ve observed frequency not making any difference on bad service here in the US on a line exactly like the Red Line (Tri-Rail).
    People who go to work every day pretty much at the same time (the ‘commuters’) know when the train is supposed to arrive. And even on good urban light rail lines in the US, they form the majority of riders.
    As for the location of this line and its stops, yeah, but you see, the planners at Capital Metro believed what everybody was saying about how forcing transfers on the work end of the trip wouldn’t really be a disincentive to ridership – and so they didn’t need to worry about the other end either (it’s just another transfer, after all; from car to train instead of train to bus).

  44. Alon Levy November 29, 2010 at 10:32 am #

    You’ve “observed frequency not making any difference”? Can you name me actual examples of high-frequency lines underperforming, or low-frequency lines overperforming? Because I could name you a ton of examples in the other direction – i.e. how new rail lines operating on a simple all-day clockface schedule have overperformed expectations, while the same service closed in the 1970s or 80s when it was infrequent.

  45. M1EK November 29, 2010 at 2:15 pm #

    @Alon, sure. Real simple. Two from personal experience.
    Tri-Rail underperforms despite having high frequencies for commuter rail. (Increasing frequencies made no long-term difference in ridership; there was a spike but it turned out to be gas-price-related).
    Capital Metro’s express buses operate on low frequencies and outperform most local routes.

  46. Alon Levy November 30, 2010 at 3:11 am #

    Okay. So, the Tri-Rail frequency only looks half decent by American standards. In the US, some networks get away with hourly off-peak service because they have so much peak congestion that people will suck it up and ride the trains. Elsewhere, the only lines that maintain just hourly service are rural branch lines, and those live off of timed connections to everything and integrated schedules.
    The various express buses you see in Texas play the same role as this essentially peak-only commuter rail. They attract trivial numbers of riders – just those who live and work on one of their direct routes, who enjoy high speed. Even LA’s bus system, based on gridded rapid routes, does better.

  47. M1EK November 30, 2010 at 8:07 am #

    @Alon, anywhere with frequencies much higher than those already has a huge population of transit-dependent (voluntarily or otherwise); thus you’ve pretty much made the whole thing moot if we use your standards.
    Nevertheless, in the real world we have to deal with here in the US, the claim is being made that increasing the frequency on the Red Line would make ridership skyrocket. I have to use Tri-Rail as my counter-example because, again, we don’t live in Europe (but neither do we pay $8/gallon for gas).