Here’s another variation of my Link stem-and-leaf schedule. If you looked at that schedule closely, you will see a consistent pattern to the departure times. Trains are always spaced at scheduled headways (7.5, 10, 15 minutes), even at transition points between peak and off-peak. You will see that trains depart Westlake for the airport “on the sevens”, at 7, 17, 27, 37, 47, 57 minutes past the hour, during off-peak hours and almost all day on weekends. Since the schedules rarely change and are simple enough, they are easy to memorize and communicate to others.

Realizing that, elements of this design could be simplified further while providing more details than the current Sound Transit presentation. Though the template for this was the schedule book, I think this would be more suitable for station schedules. The schedule book needs to have the timepoint-based schedules in addition to the basic summary.

The schedule data for my mockups came from One Bus Away. That may soon no longer be possible, as commenter Tim, suggests:

In the future, expect One Bus Away to show headways instead of exact times. (If available) real time information will help decide what time to add the headway information to. Example: 08:00 and the line runs every 10 minutes. With no real time available, schedule will show 08:00, 08:10, etc. With realtime available, and assuming the last vehicle passed at 08:02, you’ll instead see 08:02, 08:12, 08:22, etc. That is assuming that the operating agency plans on keeping headways and not schedules.

Eventually Sound Transit and Metro will no longer supply schedule data for both RapidRide and Link.

Let’s hope that both of them get their real-time arrival information systems in working order before deciding on pulling the plug.

85 Replies to “Imagine a More Detailed Link Station Schedule, Part 2”

  1. Am I to assume this schedule is based off of the pick cards Tim posted? (covering October 2010 – February 2011) Whether or not ST moves away from publicizing the times the operators are scheduled to start their runs, those start times are still public information, and will continue to be so.

    The question arises as to whether those are the times for leaving the station or for when the operator reports to the front seat. That is hopefully a simple question for which I hope ST or one of the Link operators can give a simple answer.

    I LOVE the format! If similar renditions for each station, in each direction, can be done (based on the real schedules), that would be awesome!

    And then, of course, a standard disclaimer should protect the publisher, ala “Disclaimer: This is not the official schedule published by Sound Transit, but is a precise extrapolation based on publicly-available materials, including operator pick information and published travel times. This schedule is not in any way the work of Sound Transit management or staff.”

    And then a permalink with an easy-to-remember address would enable us to spread the word of Oran!

    1. The question arises as to whether those are the times for leaving the station or for when the operator reports to the front seat.

      The time that the operator is scheduled to arrive at the stop (station).

      1. This brings up the question of when trains are supposed to leave Airport Station and Westlake Station.

        But the data seems like it should work for doing arrival-time schedules for every station in between.

      1. Yeah, after the Bellevue College update, I was about ready to fire everyone working at either agency and give all their jobs to Martin and Oran.

  2. Awesome Oran! Metro and ST need to have a more consistent policy about schedules. Routes like the 41 and 550 have complete schedules even when headways are around 5 minutes, yet Link and RapidRide don’t have schedules when services is every 15 minutes or more. That simply isn’t consistent and isn’t user friendly.

    1. I would add that this is a stem and leaf/clockface hybrid design and it works best when it is customized for a single station. This type of design doesn’t work as well for schedule books.

      1. But the Sounder schedule shows the complete schedule in a bus-route format, and then redundantly shows the arrival times at each station, taking up a whopping twelve pages! (not that I disagree with ST doing that)

        The Link schedule, er, headway listing, takes up one page. Oran’s schedule can fit in one page. Why not? … and then provide a link to the website wherein riders can find the customized schedules for each station, available in a downloadable PDF. Yes we can!

    2. I might add that Link sticks much closer to the schedule than any bus I’ve ever experienced, making the schedule extremely useful. It is odd that ST would be so disinterested in publishing a schedule for the service which most closely follows it.

      I’ve begged for it here in the blog, written to ST on several occasions, socialized it with friends and fellow transit users, and even offered to coordinate a student project to build it: PLEASE GET REAL-TIME TRAIN ARRIVAL INFORMATION!!! At the station entrance, on the platform, and of course in OneBusAway!!!

  3. On the topic of frequency, I would offer a small heresy here: The 7.5-minute headway is needed in the peak direction only. If ST were to be trying to shave headways to match ridership, they could shorten the 7.5-minute headway period in the off-peak direction up to 50 minutes at the front and back of the peak period.

    Likewise, the transition from 10 to 15 minute headway could start up to 50 minutes earlier in the off-peak direction.

    This may not work so well on M’s and Sounders nights, and would need to be re-thought once the next group of stations open, but in the dead of a winter like this one, it would have been totally justified.

    Now that ST knows seasonal trends, I think it would be good for them to adjust the winter pick to the seasonal ridership differences.

    1. The 7.5-minute headway is needed in the peak direction only. If ST were to be trying to shave headways to match ridership, they could shorten the 7.5-minute headway period in the off-peak direction up to 50 minutes at the front and back of the peak period.

      I’m confused here. The way this reads, you would have 6-7 trains piling up at the airport during PM peak if SB headways were 7.5 minutes and NB headways were 50 minutes. I’m pretty sure this isn’t what you meant–explain it again so my sleep-deprived brain understands it?

      1. I think he meant that the *duration* of the 7.5 minute headway period could be reduced by 50 minutes in the off-peak direction, presumably to 10 minute headways.

      2. I’m not suggesting piling up trains (although that could be done once tracks are laid toward Capitol Hill Station).

        I’m suggesting that the addition of trains into service and extraction of trains to out of service be done in a way that has peak-hour headway in the peak-hour direction at 7.5 minutes for the full length of the line by the time peak hour starts (which it already does), and not worry if the addition/extraction algorithm leaves the frequency for the reverse-peak direction at 10 minutes for a portion of the line during peak-hour periods.

        In other words, there would still be a couple trains dead-heading from the OMF down to Airport Station even after 6:30 am, with the in-service trains still transitioning from 10-minute headway to 7.5-minute headway southbound. Then, trains would start to be pulled out of service again at the stub tunnel around 45 minutes before the northbound trains start transiting back from 7.5-minute to 10-minute headway.

        What I’m trying to get at is that ST shouldn’t worry about advertising a *frequency*, with more service in the counter-peak direction than necessary just to back up the frequency claim, but should instead give riders some credit for ability to read schedules, and set the capacity in the two directions differently and according to ridership/need.

        The extra train service expense and operator wages to pump up reverse-peak service far outweighs any benefit of dumbing down the schedule.

      3. Frequency is what attracts some riders. The fact that you don’t have to bother with a schedule because it comes so often is a pretty good incentive to ride.

        I doubt TransLink made a schedule when they were running 108 second headways on SkyTrain.

    2. A corollary heresy: When the stub tunnel can handle many more traincars, ST should abolish the 7.5-minute afternoon peak headway, and switch to straight 10-minute headway, with the trains right-sized for ridership expectations.

      Having longer and fewer trains during the evening rush could be ST’s contribution to loosening up the tunnel congestion.

      As I’ve said before, ST is brutal about exacting savings on commuter bus routes. Train riders should expect no less.

  4. The 85 also makes a detour to Stadium Station to allow transfers to that first Link trip. It is even signed as “85 Stadium Stn”

    1. 81, 83, and 84 are signed that way as well. 82 heads south on 4th, takes a left on Royal Brougham, and then a right on the Busway (and then loops around to Holgate to Central Base), so I’m not sure why it doesn’t get signed up this way either, as I’m sure it would serve the zone across the street on request.

  5. Printed schedules should be customized for each station. At Westlake, between 9am and 2pm the trains leave southbound :07, :17, :27 etc. At Mt. Baker, between 9:12am and 2:12pm, the southbound trains depart :12, :22, :32 etc.

    The problem with not providing a full, printed Link schedule is that when someone is planning a trip that involves transferring to a bus, it is difficult to plan when to get on the Link train for a quick transfer to the bus. If a rider is in downtown and wants to catch a 107 at RB, how does the rider know which train gets to RBS in time to catch a bus that leaves at :22 and :52? Having to wait 29 minutes for the next bus because it was too difficult to figure out the Link schedule is a what some call a “transporttion barrier”.

    1. The problem with providing a full printed schedule is that riders expect the vehicles to keep that schedule. Having to wait 29 minutes for the next bus because it wasn’t possible for Link to maintain its schedule is what some call a “transporttation failure”.

      1. In the case of the 107, why can’t it wait an extra minute or two at RBS for the southbound Link connection, if it is timed for that purpose?

    2. I know I get grief for expecting people to do their own math but a schedule for each station is not necessary nor cost effective. Simply posting a timetable for the endpoints and using the graphic ST already has about travel times between stations would be enough. If I know it’s 18 minutes from Westlake to Columbia City, I can add that myself. The expense to make a timetable for each station would be, in my opinion, not worth it.

  6. Why does Link not have real time “Next train arriving in __ minutes” information at the stations? In many other cities even the bus stops have real time arrival information!

  7. Here’s a nice layout for a specific stop on a bus route:

    (Warning, this will cause a PDF to download)

    [See comment below]

    1. Please learn some basic HTML, namely the anchor. You make a link by typing <a href=”URL LOCATION”>this</a>, which renders as this. Using this allows you to include links seamlessly as inline text (great for comment threads), but more importantly it allows you provide the full URL without breaking layouts. That way I can see where I’m clicking to, whereas a goo.gl link gives me no idea what website I’ll be going to (you could be sending me to moviatrafik.dk or goatse.cx, I have no idea). And it also allows me to, say, right-click the link and Save As, which can be useful for large PDFs or image files. Also, if goo.gl’s services were ever to go under (not likely with Google, but possible with some of the other shorteners), your short link becomes useless; a full URL will persist. Heck, even if moviatrafik.dk itself dies or removes that file, with the full link I have clues for how I can search for the document elsewhere.

      1. Can you get the text below the comment box fixed so that it has the correct HTML grammer? Right now it is entirely different than the example given, and non-obvious.

      2. I am certain that if I ever sent you to goatse, or even tubgirl for that matter, Martin and Co. would send me packing, never to be allowed back.

        But I thank you for the suggestion and will do my best to follow your suggestions in the future.

  8. I’d rather get rid of the circle and put a square around all the Mt Baker trains. It’s confusing to look down and see that a circle in the middle of a line means the last through train. Even when I read it, I have a hard time believing it, because a circle connotes “this run is special”, not “this run is the last of its series”.

  9. What I find amusing about this thread is that ST did have a schedule very similar to what Oran has put here, in the Sept 2009-Feb 2010 booklet, in the Feb-June 2010 booklet, and in the June-Oct 2010 booklet. In each succeeding booklet, the schedule got briefer, until the current booklet has almost nothing, as noted.

    http://www.flickr.com/photos/51332149@N02/5407956049/

    It is obvious that ST can put out a schedule similar to the one Oran has devised, because they used to do that. The obvious conclusion is that ST just doesn’t want to do that any more.

    My guess is that Link trains were missing the scheduled arrival/departure times so often that it was an embarassment to Sound Transit, and that is why they stopped putting out a schedule. They were able to fit a fairly complete schedule onto one page in their booklets. At some point, they just made the decision to stop doing so.

    1. The first Link schedule booklet was a copy off the website. Then they put in the timetable for a while (I have the last edition). Then they removed it. They really should be embarrassed. This is no way to run a railroad.

      1. I just don’t understand Metro and ST. With all the good ideas from other operators around this country and the world AND with all the talent here locally…why can they NOT create useable maps and timetables? Blows me away. I also thought members of the city and county councils, as well as employees of Metro and ST were on this blog? Does anyone listen to public opinion and good ideas out there in the Seattle area? Aren’t they (the leaders in this community as well as at Metro and ST), interested in building the best transit systems in the country? Doesn’t seem like it.

  10. I don’t want to see detailed Link schedules. In the first place, with buses still operating in the tunnel, there is no way to bring Link to the level of reliability that such schedules would imply.

    And second, ST should adopt a more nuanced approach to train scheduling. On days when there are major events at the stadiums, for example, they should be able to add an extra train to the schedule and tighten headways as much as the extra unit would allow — still keeping even headways between all trains.

    They should use the flexibility that the current system allows them to, but instead, they schedule trains same as they do buses, which I guess isn’t too surprising since they hired bus people from Metro to run the line for them.

    1. Rag on the Metro planners if you will, but leave the operators out of it. The operators don’t set the schedules. They just do the work.

    2. Once the stub tunnel is lengthened, the headway debates may become moot. It will be cheaper to just add a couple cars onto each train during crush loads.

    3. I do want to see detailed Link schedules – at least when the headway is 10 minutes or longer. I want to know when the train will run, and when I need to be there. You’re welcome to come when you want, but I want to minimize my wait time.

      Even with the mixed tunnel operation, Link is far more predictable than any bus route running on freeways or traffic signals.

  11. Funny thing, Transit Voter: I sometimes hear Tunnel bus drivers complain about being forced to work with signals, procedures, and discipline designed for rail operations. The Downtown Seattle Transit Tunnel was designed for joint rail and bus operations. Whatever schedule problems originate there, the transit agencies have the means to remedy. Would be good if the same held true of schedule problems caused by conditions outside the Tunnel.

    Mark Dublin

    1. Mark, the problems are caused by bus breakdowns. I was watching one a few days ago, stalled at ID/Chinatown Station. It just wouldn’t start up. The driver cycled through various things, lights off, lights on, and so forth, and after 4 or 5 minutes, finally got it going. I happened to be standing by the driver of my bus and casually asked him about the stalled buses. It’s an all-too-common situation, from his perspective; he blames the computers. I noted that these buses have been in service for a few years and the bugs should be all worked out by now. He said yes, they should be, but they aren’t. Something with switching over from hush mode to street mode or vice versa.

      So, yes, the tunnel was designed for joint operations, but that was predecated on vehicles with a high level of reliability.

  12. “In the future, expect One Bus Away to show headways instead of exact times…. With realtime available, and assuming the last vehicle passed at 08:02, you’ll instead see 08:02, 08:12, 08:22, etc.”

    Um…………………….. why?

    As long as the following vehicle has left its terminal, and as long as OneBusAway has real-time info, there’s no reason it couldn’t or shouldn’t estimate each vehicle’s arrival separately.

    Especially considering all of the dilutions of RapidRide’s rapidity. Not to mention the hours per day of 15-minute headways. “Official headways” are going to be a joke, and there’s no reason for OneBusAway to incorporate them if it has better information available.

    1. The only thing different would be the absence of the words “on time” or “14 minutes delay” on OneBusAway’s screen.

    2. Way to not read the post (again)

      Eventually Sound Transit and Metro will no longer supply schedule data for both RapidRide and Link.

    3. Way to misinterpret plain text (as usual).

      “With realtime available…”

      Metro “supplying schedule data” only matters if the following vehicle has yet to leave the terminal. Once the next bus is in transit, the real-time info would supplant any schedule data. At this point, the only OneBusAway feature relying on the vehicle’s relationship to a fixed schedule is the “early/on time/min late” indicator, which would simply disappear for RapidRide or Link.

      1. Care to show me the place where realtime is available for Link? Care to count how many trips for RapidRide don’t have realtime info?

      2. Are you kidding me?

        “With realtime available, and assuming the last vehicle passed at 08:02, you’ll instead see 08:02, 08:12, 08:22…”

        …is a quote (not to mention a totally ridiculous conjecture) from your own post!

        No, Link does not currently have real-time info available. But RapidRide — which, in case you forgot, is essentially identical to every other Metro bus — does: http://www.onebusaway.org/where/standard/stop.action?id=1_60830

        OneBusAway real-time info for RapidRide fails at about the same rate as for any other route, because it uses the exact same system.

      3. …is a quote (not to mention a totally ridiculous conjecture)

        Here’s the source, since you refuse to believe anything I say.

        OneBusAway real-time info for RapidRide fails at about the same rate as for any other route, because it uses the exact same system.

        False, RapidRide coaches use GPS, whereas the rest of the fleet uses a signpost-based AVL system. The new GPS tracking system is completely separate from the old system. It was not “patched” in to the new system until last Thursday, which is why you saw “Scheduled arrival” for all RapidRide trips up until then.

        Even better, let’s look at link to call total b.s. on your “OneBusAway headway estimations should be adequate” hypothesis.

        Headways aren’t implemented yet. And where did I hypothesize they were accurate?

        Seriously, this has got to be one of your worst researched posts ever.

      4. Seriously, this has got to be one of your worst researched posts ever.

        Scratch that too–I forgot d.p. argues with opinions rather than facts.

      5. Your link makes clear that Brian @ OneBusAway fully intends to adapt his algorithm to continue providing real-time data — when available from Metro, which wasn’t the case when he wrote that post, but now is.

        “OneBusAway combines the timepoint prediction data from MyBus with the transit database from MetroKC to interpolate prediction data to each physical stop along a route.”
        – from http://code.google.com/p/onebusaway/wiki/FAQ

        OneBusAway therefore only relies on schedule information to provide arrival times when the bus is not yet in service. As long as the bus is moving, the arrival times are ostensibly derived from a more complicated algorithm that takes into account the average trip-tip time between the vehicle’s present location and your current one, as well as the current speed of the vehicle as a deviation from the average speed.

        So, with GPS data now available from each separate RapidRide vehicle, why in the hell would Brian not use it to make individual estimations?

        You’re always-in-error attempts to “QED” me are hilarious.

      6. “I forgot d.p. argues with opinions rather than facts.”

        Even if that were true, it would be better than drawing deeply stupid conclusions from poorly understanding the facts.

      7. So, with GPS data now available from each separate RapidRide vehicle, why in the hell would Brian not use it to make individual estimations?

        To answer that, I’ll quote you:

        real-time info for RapidRide fails at about the same rate as for any other route

        (I left out the part where you were QED’ed).
        I don’t think you understand how headways work in GTFS. All it does is say “from 9am to 3pm the bus comes every 15 minutes”. That’s it. That’s all OBA has to work with. What you see now has a unique ID for each trip. When we get realtime info comes in, it comes in with a trip ID. That trip ID gets matched up to what’s in the schedule and we know the deviation is for that trip. In the new system, there will be a trip ID from the tracking side but nothing to match it up to.
        My example included a situation where there was missing data. If we can calculate the time the next vehicle is arriving, of course we’re going to include it. But as you admitted yourself (indirectly) things break and we don’t always have info available for every coach out there.

        I’m not really sure why you’re arguing with me about this. It’s not my invention and I have no control over it. I’m just reporting the facts as the are and attempting to explain them where I can. You’re never satisfied with my explanations (or you get strangely silent when you are).

        So yeah, we didn’t initially have realtime for RapidRide. Now we do. But we still don’t have any for Link. So we’re still going to have to continue adding 10 minutes to the last scheduled arrival until (or if) Sound Transit ever gives us realtime info for Link.

        While I can’t back this up directly*, and I’m sure you’re going to argue with it, but Metro’s intention is to run RapidRide on a true headway system. Someone in the control center will monitor the locations of all vehicles, and if one starts getting too close to another, the control center will instruct the operator of one or more coaches to speed up or slow down as appropriate in order to ensure consistent headways. When this was/will be implemented I’m not sure. And I’m sure this is a “complete failure” on Metro’s part since you can’t plan a trip if you don’t know the schedule. Get over it; tight headways attract customers.

        *If I didn’t delete the video, I have a recording of it directly from Kevin Desmond. I can send it to you if you’d like.

      8. Even if that were true, it would be better than drawing deeply stupid conclusions from poorly understanding the facts.

        Wait, you think it’s better to argue opinion vs. opinion rather than fact vs. “misinterpreted” fact?

        Alright then, here’s my opinion: Your opinion is wrong.

      9. As is often the case, it’s not your facts that are wrong, or even your methodical analysis of systemic function. It’s the strange conclusions that you simply leap to, as if they inevitably followed, after you’ve laid out all of the above.

        The wheel-rotation/transponder system seems to fail for about 1/8 to 1/6 buses. The GPS that you say is on RapidRide A seems to fail at about the same rate, but hopefully the GPS system will improve.

        That means that you can accurately pinpoint 7/8 of the buses. If there are 5 vehicles simultaneously en route, why would you just read the location of the first and provide headways for the next 4, when you can likely estimate them individually?

        “There will be a trip ID from the tracking side but nothing to match it up to…”

        Who cares? As long as you know it’s a RapidRide vehicle, you can still estimate its arrival. Nothing changes, even in the vehicles bunch or pass one another — they’re still a certain distance away and traveling at a calculable pace! For the umpteenth time, the matching data is only necessary if you wish to declare it “early” or “late.”

        “Someone in the control center will monitor the locations of all vehicles…”

        So your solution is to run OneBusAway as if this were the case, even though it isn’t yet?

        “Get over it; tight headways attract customers.”

        Oy gevalt, Tim! Do you really think you have to tell me that? You’re the one eternally defending fractured Metro routings and 30-minute baseline headways. You also habitually claim 15-minute headways as “frequent.”

        Allow me to blow your mind: 5 minutes is a frequent headway. When you have 5-minute headways, even buses running off-headway rarely leave more than 10-minute spaces between them. No one in their right mind would radio a driver and tell her to drive slower, because even the “off” headway offers a very high level of service, and because artificially slowing down a vehicle is stupid!

        If your solution to one bus running slowly is to slow down all of the other ones, you’re doing it wrong!

      10. When you construct a mathematical proof, you can’t lay out steps 1-5 in painstaking detail, write “blah, blah, blah” in steps 6-9, then throw a total conjecture into step 10, and then claim it as “proven fact.”

        That is essentially what you do.

        So, yes, arguing an educated opinion is infinitely better then arguing facts-distorted-by-horrendous-logic.

      11. “There will be a trip ID from the tracking side but nothing to match it up to…”
        Who cares? As long as you know it’s a RapidRide vehicle, you can still estimate its arrival. Nothing changes, even in the vehicles bunch or pass one another — they’re still a certain distance away and traveling at a calculable pace! For the umpteenth time, the matching data is only necessary if you wish to declare it “early” or “late.”

        The system cares! Just because it’s a RapidRide vehicle does not mean it’s operating a RapidRide route. Vice versa is also true–we could have a non-RR vehicle operating on a RR route. And just knowing the position, speed, and (maybe) heading doesn’t tell us anything either–other than where it is. Just because the coach is on route does not mean it’s in service, and just because a coach is off route does not mean it isn’t in service.

        You’re the one eternally defending fractured Metro routings and 30-minute baseline headways. You also habitually claim 15-minute headways as “frequent.”

        You think I like 30 minute headways? Not at all! But when the agency is in debt, I’ll settle for 30 over 60 or 0.
        Please provide a reference where I claim 15 is frequent. I may have used it as an example in multiple places as a guideline for a headway, but I have used 10 and 7.5 in the same respect.
        Also, please provide a reference where I “defend” a particular routing. Then, please imagine that we take one or more routes that is within walking distance from your house and then move it. Then you’ll know what people bitch about to Metro, and then Metro caves and moves it back or doesn’t complete the proposed move because they think it’s possible to please everyone.

        5 minutes is a frequent headway.

        Allow me to blow your mind: 108 seconds is a frequent headway. TransLink did it. It’s possible. I know it. I’ve seen it, I’ve been on it. Their system is also grade separated, automated, and funded differently. If you can get Metro some more money, and the density to justify it, I guarantee you they’ll up the headways. Look at route 41 if you don’t believe me.

        If your solution to one bus running slowly is to slow down all of the other ones, you’re doing it wrong!

        It’s not MY soulution! I don’t work for Metro! I’m just telling you what THEY plan on doing with THEIR routes and THEIR equipment.

        Allow me to speculate here: if you keep consistent headways, customers will have consistent experiences. Consistent experiences theoretically make happy customers, and happy customers deposit their coins in the farebox with a smile. If you have inconsistent headways, you get bus bunching, longer than expected waits, and standing room only coaches. This is what you get when the 71/72/73/74 combined section isn’t scheduled properly, as Oran points out was the case last shakeup. People wait too long and then the bus fills up over capacity creating unhappy customers. So if we want to do the opposite of that we look at what causes it, and somehow Metro ended up with this “remote controlled headway” thing. Yeah, we’re basically turning them into RC cars with a Wienermobile paint scheme.

        Tim’s ideal solution would be to run the line as fast as density, demand, connections, and ridership projections allow. I don’t have data for any of those facets, so I cannot give a recommendation for any specific number.

        The number is not instantly five or ten. If we did that right now, these things would run too empty. I don’t believe the connections to RapidRide are good enough (save for the terminals). If your destination is on or very close to Pac Hwy, it’s really good, otherwise, it could use some improvement. I know you’re always pushing for East-West connections, and there are some of those along the route, but they serve relatively low density neighborhoods, so it’s hard to justify running them at higher frequencies.
        I think, that on the A Line, 15 minutes is enough for off-peak for now. Ridership doesn’t grow overnight, but once ridership does grow that we should definitely tighten the headways as the budget allows.

      12. “Just because it’s a RapidRide vehicle does not mean it’s operating a RapidRide route… Just because the coach is on route does not mean it’s in service, and just because a coach is off route does not mean it isn’t in service.”

        See this is the perfect example. All you know for sure is that there will be no individual Trip ID in the current sense. Everything else your own conjecture, and much of it is illogical. Why wouldn’t route information exist, even in the absence of individual trip IDs? Why would an out-of-service bus be reported identically to an in-service one? What use would that information be to the “control center?” Data reported in a way that would be of any worth to Metro operations could also be interpreted in a way that would make OneBusAway work properly.

        “I’m just telling you what THEY plan on doing with THEIR routes and THEIR equipment.”

        Fine. At some point, Metro plans on running a control center that will smooth at headways by ordering buses to slow down to remain evenly behind other slow buses. That’s stupid, but Metro does stupid those all the time.

        But then you jump to the conclusion that, because Metro will do this someday, OneBusAway is going to junk all real-time data now. Again, that’s a completely unsupported leap. Why would it do something to compromise its own effectiveness?

        “If you keep consistent headways, customers will have consistent experiences.”

        I think we both agree that having perfectly consistent headways is less vital at 5-minute frequencies than at 15- or 30-minute frequencies. The problem with the “remote controlled headways” solution is that customer may “consistently experience” her already sub-BRT bus (which is already a bit of a let-down) driving inexplicably slowly or dawdling at a stop for no apparent reason.

        The the customer will “consistently experience” the urge to say “this sucks” and get back in her car.

        “Allow me to blow your mind: 108 seconds is a frequent headway.”

        I’ve traveled much more than you, so sadly nothing has been blown. It would be fair to call <2 minutes "ultra-frequent" or "unusually high frequency," although it exists in many more places around the world than just Vancouver. 4-8 minutes is more of a "regular high frequency," even in New York or Paris or cities of the former Eastern Bloc.

        "The number is not instantly five or ten. If we did that right now, these things would run too empty. I don’t believe the connections to RapidRide are good enough (save for the terminals)."

        Of course, I know very little of RapidRide A. The problem is that when you apply the above logic to RapidRides C, D, and E, the result is a self-fulfilling prophecy and a catch-22:

        1. The lines are neither frequent nor fast enough to attract riders from further away — neither from neighboring routes nor new riders who would walk further for better service. The inadequacies of RapidRide fail to expand their walksheds in the slightest.

        2. Because you have failed to expand the walkshed or to make a change in ridership habits worthwhile, you have — and continue to have — insufficient demand for expanded service.

        If you don't start off with service good enough to be reward changing habits, you might never get there!

      13. [So many typos. If the “comment” box were somewhat bigger, it might be easier to edit for mistakes before posting…]

      14. Pretty much agree with everything you said, though I don’t know if you fully understand that Metro doesn’t have enough money to make headways as tight as you want them everywhere you want them. They’re doing the best that they can with what they have, and of course you disapprove of all of that.

        In the future, please STOP grouping me with Metro. I am not Metro. Metro is Metro.

        If you have Greasemonkey, or can get it, I could write you something that makes the comment box bigger. Not sure that would help, but possible.

      15. I know this is tangential, but I would highly recommend installing Google Chrome, which has the extremely convenient feature of allowing you to resize the comment box. I do this for pretty much every long comment that I write here.

      16. Aleks:

        Wow! I’ve been using Chrome, but I never had the slightest idea those two faint lines in the lower-right corner of the comment box meant anything!

        Look at that box go! Whee! Man, that’s going to make constructing long replies easier!

        Tim:

        I want to let you know that I appreciate your olive branch, and I’ll do my best to keep it next to my keyboard when I disagree with you in the future.

        “Metro doesn’t have enough money… They’re doing the best that they can with what they have.”

        This is, indeed, at the root of most of our disagreements. I so strongly feel that they’re not doing adequately with what they have (and I have such a litany of horrific transit experiences to back me up) that I can’t help but take offense to that assertion.

        I know the economy’s in tough shape. Yes, Tim Eyman has screwed transit over a few times. And yes, Olympia not only fails to recognize the need for and aid transit in its primary economic engine, but actively plays beat-up-on-King political games straight out of the 1970s.

        But I also know that Metro charges the second-highest fares in the country. It boasts the 6th-highest percentage of paying (peak commute) riders. It has (multiple and recently-supplemented) tax levies dedicated to it. It’s annual budget is over 1/2 billion dollars per year, and that’s a lot of money that can be spent wisely or poorly.

        To paraphrase Aleks, if 2/3 of Metro’s Seattle-proper routes disappeared tomorrow, but the remaining 1/3 continued to cover the same/nearby ground at 3x the frequency, the decreased average trip time and elimination of the transfer penalty would pay off for nearly 100% of riders despite having to walk further. And it would be revenue neutral — or even save money thanks to the corollary operating efficiencies of fewer, faster, more frequent routes. It’s that kind of bold and effective action that is missing from the Metro playbook, even when the success of RapidRide is at stake.

        For example, you wrote:
        “Then, please imagine that we take one or more routes that is within walking distance from your house and then move it. Then you’ll know what people bitch about to Metro…”

        That’s happening. RapidRide D (10 minutes walk from me) intends to replace the 18 (30 seconds from me). And you’re correct that I’m pissed… I’m pissed because, with frequencies ranging from 10 to 30 minutes — 30!! — and minimal speed improvements, that extra walk is never going to pay off! But if it were 5-10 minutes all the time? I’d be skipping-to-my-loo over to 15th with a grin on my face while I write the 18’s obituary!

        Metro favors tepid, half-assed improvements. That kind of improvement does cost more, as its effects are so negligible that you cannot justify siphoning the money from elsewhere. But when an improvement is substantial enough to ripple across the system, your funding priorities can ripple too. That’s how you do the best with what you have.

      17. Then, please imagine that we take one or more routes that is within walking distance from your house and then move it.

        It will probably come as no surprise that I agree with d.p.’s agreement with me on this topic. :)

        I currently live within a couple of blocks of the 43 and 49, so when I need to go downtown, I generally take one or the other of those buses. Sometimes I take the 14.

        But I would happily give up *all* of those buses (at least the portions going to downtown) in exchange for a Pine St bus that ran every 5 minutes.

        I don’t mean to suggest that Metro should make these changes immediately, or that doing so would be revenue-neutral. I’m just saying that, at least for me, the tradeoff you’re proposing (a longer walk for more frequent service) is exactly what I’ve been advocating, rather than an unfortunate consequence.

        Of course, I totally understand that Metro *does* get pushback from riders whenever they propose routing revisions. But, as I’ve said before, one of the things that bothers me most about Metro’s planning (and it’s not a problem unique to Metro by any means) is their bias towards existing users at the expense of potential users. That is, if a service change would cause 5% of existing riders to switch modes, but would also attract enough new customers to increase ridership by 10%, then Metro will sometimes choose not to make the change, even though (I believe) they should.

        Suppose that Metro announced that the June 2011 service change would be a complete restructuring of the system, with all-new grid-oriented routes that had 5-minute to 8-minute all day headways. Since the change would be revenue-neutral, a number of existing routes would be deleted. Undoubtedly, many existing riders would say ‘screw it’ and start driving. But I firmly believe that many more people would say, “You know, this whole bus thing isn’t as bad as I thought”, and the net effect would be a more efficient system with higher ridership than before. There are tons of potential riders out there — people who would use the bus (or use it more often) if not for low frequencies and transfer penalties — but their voice goes unheard, and so we stick with the current system.

  13. I love it Oran! Just one brief comment. The final hour of trains each day is a bit confusing, if you’re unaccustomed to where they terminate. You show a circle for the last train from Sea/Tac to Westlake, and a square for the last train to terminate at Mt Baker, but it doesn’t explain that the previous 3 trains also terminate at Mt Baker. If you’re new to the city or Link that isn’t intuitive. Also, being that there’d are a few peak-zone transition trains during the day that also terminate at Mt Baker, perhaps you repurpose the square to designate any train that ends at Mr Baker.

    My 2¢, but otherwise a beautiful design (as always)

    1. Good suggestion on the last trains. I didn’t look at which mid day trains terminated at Mt Baker. I think I did that for my previous design. So I regressed on that part.

    2. Do mid-day trains actually end at Mt Baker Station? After seeing some dead-heading from Airport Station, I kinda assumed they don’t. But hey, they may as well be in service, since they are headed north, and can’t go any faster just because they are empty.

      Bring those late-evening dead-heading trains back into service, especially for the Rainier Valley residents headed home from the airport.

      The same goes for the late-night trains deadheading from the stub tunnel. That 11% of Link trips that are intra-DSTT shows why they ought to stay in service. Indeed, they may as well make Link part of the RFA just to discourage slowing down the buses for intra-tunnel trips.

      1. Not everything goes back to the OMF in service. Not sure why.

        But many late night trips do run in service from the airport to Mount Baker, allowing you to catch the 7 to get all the way to downtown.
        …or if you miss it, and the next one isn’t for 45 minutes, you walk, like me, all the way to Westlake, because even though it’s 1:30 in the morning, it’s warm out and the walk is kind of enjoyable.

      2. I’ve done the Link-to-7-to-downtown thing for late-arriving flights. But if you miss the 12:50 bus and the weather’s bad (or you’d rather not risk walking that part of town with luggage), you are quite screwed.

        It’s crazy to me that they don’t hold the 12:37 and 12:52 route 7 buses (and the 12:50-ish route 8) for the trains that leave the airport at 12:20 and 12:35. That sort of thing is de rigueur in cities (e.g. Boston, Philly) where buses run later than trains and the same agency manages both. It’s also remarkable that the 12:50 train dumps its passengers at Mt. Baker with essentially no available connections in any direction.

        (This is yet another argument for restructuring our transit agencies by geography rather than by service type, with Link — at its core an urban form of transit — under the urban-transit agency’s purview. I know some would argue for combining all services under one umbrella; Metro’s predilection for running urban and suburban service on the same principles even when they serve completely different purposes demonstrates why that can backfire.)

      3. [Hmm… I even used Google’s comment-box-expansion feature to edit that one. I could swear I slashed out of the Latin itallics.]

      4. That’s the reason I don’t take the last Link train from the airport, because even though I’m not one of those people who thinks Rainier Valley is a really scary place, I still don’t want t0 miss the 7 and waiting an hour for the night owl. So I take the A + 124 instead.

        (Actually it was the 174 last time I did it, and that confirmed my dislike of long slow routes because it stopped at every single stop until it got to East Marginal Way. Put expresses or limiteds on long routes, I always say.)

  14. Last time I was in Switzerland I downloaded an app for my iPhone that lets me plan journeys on all public transit within the country; bus, train, tram, cable car, steamship, whatever, no matter what company operates it. I could also buy electronic tickets and day passes via the app that were displayed on my phone and then scanned by the conductor. The app also generates custom schedules for any stop or station within the country for any mode that serves it. It will also display a map of your entire journey via Google Maps. And oh yeah, it works in four languages. Meanwhile, here in the tech center of the universe, we can’t even get a schedule or real-time info for our shiny new light rail line, or a decent website for ORCA. Argh! I love Seattle, but I just don’t understand the way things are done here sometimes.

    BTW, nice work Oran, as usual!

  15. Man people really care about this…I just don’t put much thought into when trains arrive and depart. It’s frequent enough for me that I don’t need to and thus don’t try and time it just right. Not worth the effort. If Link were on hourly or half-hourly headways it’d be different.

Comments are closed.