Your Bus, Much More Often. No More Money. Really.

What if almost every bus in Seattle came every 8, 10, or 15 minutes? And gave you a fast, reliable ride?

That may sound like a pipe dream. But it’s entirely possible. And the best part is: we don’t need more money to do it. We just need some inventiveness, a lot of political courage, and the occasional willingness to walk a couple extra blocks or to make a transfer.

This post, together with the linked documents, sets out a proposal called the Frequent Network Plan—a new idea for the core all-day bus network for the city of Seattle.  This initial presentation is general and covers the whole city; specific neighborhoods seeing big changes will be addressed in more detail in future posts.

A small piece of the route map

A bit of the map, in a neighborhood seeing a lot of change

I built two versions of the Frequent Network Plan map: one where each route has a separate color, and one where each frequency level has a separate color.  The first shows where routes would go, while the second shows just how much more frequently buses would be running along any given corridor.  I also wrote three reference documents, linked at the end of the post.  Further explanation after the jump.

The key goals of this proposal are frequency and speed.  Within the urban parts of Seattle, almost every route would run at least every 15 minutes all day, with key routes that collectively serve every dense area of the city arriving every 8 or 10 minutes. Even in peripheral areas, many buses would run every 15 minutes, with the rest running at least every 30 minutes. There are no hourly routes.  Most routes are designed to run faster than current service, without deviations or bottlenecks. Routes are on straightforward, easy-to-understand corridors wherever the frequently odd geography of our city makes it possible.  Not only do frequency and speed improve the rider experience, they also allow more trips per service hour, improving the efficiency of the system.

To a fairly precise approximation, this proposal would not require any additional money. It is based on a well-educated estimate of the service hours required for the current all-day network, plus a small number of hours taken from current peak-only service that would be entirely redundant (in terms of both routing and frequency) with the planned all-day network. It does not address, or use any hours from, the majority of the current peak-only network.  The proposal is a 2021 vision; it relies on the completion of North Link as far as Northgate, and on the completed Seattle Waterfront project.

So what’s the catch? There are two. First, more transfers will be required. Some very heavily used one-seat rides would turn into two-seat rides, always with one or both legs on Link or an 8- or 10-minute bus line. Second, riders might have to walk a few extra blocks. Corridors in today’s network that are close together and not separated by steep hills are mostly consolidated. Many deviations that slow down service are removed. Service to some very-low-ridership areas is cut entirely, particularly if it requires a high number of service hours. For most of us, those changes should be a price well worth paying to get frequent and fast service throughout Seattle and North King County with no more money.

I should emphasize that it would be impossible to implement anything like this plan under the worst-case scenario of no CRC replacement funding and a 15%-17% service cut. If you cut 17% of the hours in the plan, most 10-minute routes would become 15-minute routes, and most of the longer 15-minute routes would become 30-minute routes.

Please forgive all mapmaking sins (or, better yet, offer feedback in comments); I remain a GIS and cartographic novice. Also, there is not yet a schematic map. Designing a schematic map covering the entire city is an enormously complex undertaking which I don’t have the skills or time to do.

I’ve written three reference posts covering the plan in more detail, which will likely be useful as you examine the maps:

  • A short list including each all-day route, its base frequency, and its key corridor or destinations.
  • A detailed route-by-route list of the all-day routes, with basic descriptions of each route and a listing of technical hurdles each proposed route would face.
  • A cross-reference with current service, explaining how riders of current all-day routes would be served by this proposed network.

Answers to many more questions are in the Questions and Answers post.

I should particularly thank members of the STB community who have helped me develop and refine, and sometimes suggested, the ideas in this plan.  Bruce, Zach, Martin, Adam, Brent, Matt, Mike Orr, d.p., Anandakos, and Aleks, among others I’m sure I’ve left out, all deserve part of the credit, and I’m very grateful to be part of such a well-informed and collegial community.




Comments

  1. Matt the Engineer says

    To make it clear to others (covered in the FAQ) this covers is all-day service only. I found my peak service route missing and was concerned, but the FAQ cleared that up.

    Great work David. I can only imagine the amount of effort this took.

  2. lakecityrider says

    Looks great, where do I sign? Can we have it in place by September’s service change? :D (I kid, of course, but I’d give my right arm–I’m left-handed–for this to happen by February, which it can’t, so my right arm is safe.)

    My only disappointment, which you mostly dealt with in the FAQ, is around non-day service. I see a lot of “hand waving” (for lack of a better term, not meaning to be disparaging) around service outside the 7a-7p time frame. Tolerable but not great night service is one reason why I keep a cheap car since that’s when I’m most active. I realize that tinkering with and improving the “all day” network gets the most bang for the buck, but darn it, night folks ride and pay taxes for transit, too, and I’m biased… :)

    Sill, I think it’s a great proposal. A++, would campaign for again.

    • lakecityrider says

      To clarify, the “hand waving” comes from seemingly everyone who has a great (or even horrible) idea to improve service on Metro, not just your plan. Here’s how much I like your plan: If it even gets to the drawing board stage and just half is slated to be implemented, I’m immediately having any vehicle or vehicles I own melted down and turned into souvenir ORCA cards. (Hmm, metal NFC card; difficult, but I’ll come up with something.) Yes, I see some warts (that is one really long haul for FNP 40; Lake City Fred Meyer to Ballard Locks via Crown Hill, what a tour), but it looks much more workable (and simple) than what we have now.

      • Mike Orr says

        My biggest regret after last September’s restructure was losing the one-seat ride from Lake City to Aurora and Ballard. The 75 was abominably slow, but an uncoordinated transfer at Northgate — with no compensating improvements — is even worse. Extending the 40 to Lake City recognizes that crosstown service in north Seattle is one of the most critical unmet needs, even with the more-frequent 31/32, 44, 48, and 40. I also agree with the longer-term goal of splitting the 40 in Ballard rather than splitting in Northgate.

      • asdf says

        The breaking up of the 75 was simply a case of Metro calling a spade a spade. Metro finally realized that the Northgate TC deviation was so time-intensive that everyone who needed to make such trips remotely frequently who could possible squeeze the money to buy any sort of car was already driving. So, Metro figured that since there were few thru-riders left to lose, they may as well abandon all pretense of carrying about Lake City->Greenwood trips and split the route to improve reliability for everyone else.

        Unfortunately, I see only two real options to make transit a viable option for cross-town trips that go through, but not too, Northgate:
        1) Reroute the 40 to go straight through on Northgate Way, without stopping at the TC.
        2) Keep the 40 on its current routing, but create a whole separate route that would be identical to the 40, but without the deviation.

        Option 1) would impose a 1/3 mile walk for connections to and from Link. If the number of people making this connection is small, I would say “do it” in a heartbeat, but given that the network, as shown, uses 40 as a major feeder route into Link, this would be likely be a very rare case where significantly more people would be transferring than riding through. I can easily imagine virtually everyone getting off the 40 at Northgate TC, to be replaced with a whole new crop of riders getting on. The shear numbers suggest, regrettably, that the deviation should probably stay.

        Option 2) would violate the principle of simplicity and frequency. A “40X” shadow route scaled to meet demand would likely run once an hour at best, and its service hours would ultimately have to come from making other core routes less frequent.

        So, the conclusion is – with any reasonable service restructure, short of completely moving Northgate Station or a major re-arrangement of the street grid – if you work in Greenwood or Lake City and want to take transit to work, you’d better live on the same side of I-5 as work, and more less frequent trips across Northgate, it’s time to bike or drive. This is an excellent example of a type of trip where Car2Go can save a lot of time over transit for comparatively little money.

      • David Lawson says

        asdf, it’s worth noting that my proposed 40 deviation is quicker and simpler than the current 40 deviation (which was also the old 75 deviation). I think using 1st NE would save 2-3 minutes over the current 40 routing.

      • biliruben says

        I’ve spent over a year scouting bike routes that I could reasonably talk my willing but wary wife into using in her commute from Lake City to Holman. Such a beast does not exist for regular humans. Back when the 75 went through, I talked her into bike one way, ride the other a couple times. Can’t even do that now.

        If they chose a 130th North Link stop, Running a through-bus Lake City via 125th/130th Greenwood-Holman would be a no-brainer. Northgate is a giant transit bog and should be avoided at all costs.

      • Scott Stidell says

        This plan should be permanently pinned to the blog and should be a starting point for all further discussions about restructuring bus service in the city. Well done.

        Since I am most familiar with the far NE quadrant of the city, I’ll just comment on that specifically. The only real issue I see is the lack of any reasonable service to the UW from the Meadowbrook area between about NE 95th and NE 120th (north of there is an easy walk to Lake City and the various options there), and from LCW west to Sand Point Way. This is a fairly large geographic area in a part of the city where historically there has been a decent number of UW students and staff resident due to its proximity. Yes, it is predominantly single family and hence does not warrant exceedingly frequent scheduling (and the 15 min service along 35th NE is a dramatic improvement), but none of the proposed transfers would make sense in a real-world situation. Nobody is going to ride from, say, NE 110th to the UW or U Village by traveling crosstown on NE 75th to the Link station, nor is a transfer at 75th to the 65 much more likely—and once you get to the 71 at 65th—a great route by the way—you’d be looking at two transfers to travel the last couple of miles.

        (If you’re not familiar with the area and are just looking at the map, thinking “Hey—the 69 and 75 are both nearby!” the area I’m describing is in a bowl and except to the north there are quite steep hills to get from Meadowbrook to any of the surrounding bus lines. Very few people outside of some of the Hale students ever walk it.)

        I think a couple of tweaks might fix the situation: what if the 78 continued directly down 35th NE via Montlake to Husky Stadium station; the 65 would travel on the proposed NE 75th portion (currently slated in this plan for the 78) from the Roosevelt Link station to 40th NE, thence south to Sand Point/NE 45th, thence east on NE 45th to the UW campus? The 65 would already connect to Link at Roosevelt, so the 78 should access Link at HS Station. The only area that would lose coverage with that change is the weird vestige of the current 71 on 40th NE from NE 75th north to NE 85th. That area has nothing other than single family homes and a ballfield; it is as flat as any neighborhood in Seattle, it has sidewalks and it’s only 5 blocks west to NE 35th where you would have very good N-S frequency—in fact, the country club limits any potential walk distance to the west to 10 blocks, and there would still be service at NE 75th and 40th NE so very, very few homes would be affected negatively at all by this change. Additionally, you’d gain another N-S line between 35th and Sand Point which would also feed into the 71.

        Again, David—very well done.

      • David Lawson says

        Scott, thanks.

        I don’t think the 78 -> Link transfer is as unrealistic as you think. To the U-District or the southern campus, it’s faster than the current one-seat ride on the 65, which is what drove my decision to truncate the 65. (Link is fast, and 75th moves quickly as well.) For someone going to U-Village, I would suggest a transfer to the 65 at 85th or 75th — this is exactly the sort of transfer that making everything frequent will make much easier.

        It’s also worth noting that there are relatively few non-peak, non-Hale-student riders in Meadowbrook to begin with. The peak riders would be very well served by the 78, which would make a downtown trip dramatically faster than the slog on the current 64, particularly during PM peak when the 64 does the Express Lanes/42nd dance.

    • David Lawson says

      Thank you for the compliments!

      To respond to readers who may not see the FAQ, I started with the all-day daytime network because it is the base on which all other service is built. Once an efficient daytime network is in place, the night network will improve “automatically” to some degree even without any new hours. I’ve been meaning to do some quick math to figure out roughly what night service would be possible with this network, but haven’t had time.

      One of the things I would like to do with this network, if only there were more money, is split the 40 in half. The south half would go between North Beach and downtown, and could be through-routed with something downtown. The north half would go between downtown Ballard and Lake City.

      As is, because the 40 is not through-routed downtown, reliability on it should only be somewhat terrible. It’s no longer than a couple of the longer through-route combinations.

      • says

        I’d really prefer to see the “15″ routed all the way to Northgate and the 40 use the current “15″ Terminal with timed/well executed transfers to the “15″.

      • says

        I missed that you extended the 40 over to Cedar Park, which kind of nullifies my earlier comment. Sure I’d like to have the “15″ connect to NTC/Link (which would make getting North out of Ballard SOOO easy. But perhaps I’m stuck in the mindset of RapidRide being the “core” service as opposed to the entire map.

      • Aleks says

        Velo, I agree with you, and I don’t think it has anything to do with RapidRide being “core” or not. The 15 “wants” to continue straight to Northgate; the 40 has to take a more complicated route to get there. I think the two transfer points at 15th/85th and 15th/Leary would provide a great way to get to Northgate from central Ballard.

      • Mike Orr says

        RapidRide D would probably have gone to Northgate in the first place but Metro didn’t have enough money to extend RR-branded buses and build more stations, so it extended the 40 instead. This plan just realizes the original goal.

      • Aleks says

        RapidRide D would probably have gone to Northgate in the first place but Metro didn’t have enough money to extend RR-branded buses and build more stations, so it extended the 40 instead. This plan just realizes the original goal.

        David’s current plan still has the 40 going to Northgate. It sounds like you’re agreeing with me that it should be the 15/D instead?

      • David Lawson says

        Extending the 15/D to Northgate would make it long enough that you would have to break the downtown through-route. And the 40, even shortened to Holman Road, is long enough that a through-route would be difficult. Extending the D in a reliable way would require quite a few extra buses compared with my current proposal.

      • Scott Stidell says

        The 15 continued to Northgate follows the route of the NW line of the earlier proposed (1968) rapid transit system; Link to Roosevelt + the 522 would be what was the proposed NE line. With decent crosstown service (and David’s post does better at that than anything we’ve seen in the north end, ever) those are still decent spines that hit a large number of the existing neighborhoods/urban villages.

        Adding a 125th/130th crosstown line once Link arrives there would only add to the north end’s connectivity. There is enough of a grid north of the Canal to start looking at these things. Great post and the amount of work David put into this is appreciated. I can see this as a good way to start people really considering the possible and how they might benefit.

    • Adam Bejan Parast says

      I like the route 7 routing as well but I think one issues that needs to be addressed for this to work is speed and reliability on Boren AVe in the NW driection from about Seneca to Olive from about 4-6pm. It becomes an absolute parking lot in the afternoon just like Denny.

      • David Lawson says

        Yes, that’s definitely an issue. The nice part about my suggested routing is that the issue wouldn’t affect too many passengers. Most of the ridership on Boren at that time of day would be southbound, so you could more or less deal with the problem by finding an extra peak-hour bus to lengthen recovery time at the South Lake Union terminal.

        I didn’t bother to find that extra bus for this proposal because recovery time on my 7 is already pretty good, but it would be an easy improvement.

    • says

      My quibble is that I don’t like your 7 terminating in a random spot in SLU. Drag it out to the Center on Harrison or Mercer; I prefer Mercer just because it provides an easy way to get to the Center’s west side.

      • David Lawson says

        Mercer is and will remain unusable for transit because of traffic.

        My 7 terminal isn’t “some random spot in SLU,” but the activity center of the whole neighborhood.

      • says

        I admit I’m not familiar enough with SLU to know what’s where there, but I would probably still prefer going through there to terminate near the Center.

  3. adam says

    Well done, this is great work that could affect actual route improvements with the right eyes on it. I particularly like a lot of the route simplifications in downtown and in the Central District.

  4. GuyOnBeaconHill says

    This seems to be a compilation of what’s been talked about for the last couple of years on this site. There will be some strong political and neighborhood opposition that will arise. Expect to hear that Magnolia residents will object that their service is being “gutted” and Queen Anne riders will feel “shortchanged”. Rainier Valley/Beacon Hill riders who don’t live within walking distance of Link won’t appreciate having to transfer to get downtown despite the greater connectivity to Duwamish. Writers on this blog will love the plan, but the real world won’t be so appreciative.

    • Will Douglas says

      Especially since there is no good reason to pit riders against each other like this. There’s no need to improve transit for one set of riders by making it worse for others. Transit advocacy should have as its goal to give everybody a useful, realistic option for getting around that does not involve a car. Too often on this blog transit advocacy instead looks like giving a few privileged people great service at the expense of everyone else. Seattle is very supportive of transit. They’ll vote for more money to expand frequency and speed without sacrificing coverage.

    • says

      “There’s no need to improve transit for one set of riders by making it worse for others.” – It’s called money. Whine all you want about how “bold plans” will motivate voters to approve more money – The reality is we don’t have unlimited funds and never will.

      Even in a world with unlimited money, it would more efficient to integrate taxis to serve the last mile or two in low density areas. Imagine living along the current 61 and being able to tap your ORCA card in a Taxi to get to downtown Ballard. I guarantee you that would be less expensive to subsidize that Taxi ride than running empty Orion Hybrids every 30 minutes until O dark thirty.

      • GuyOnBeaconHill says

        Note that the FNP increases bus service along most of the 61 route to 15 minute headways. How cost effective will that be?

      • David Lawson says

        GoBH, missed this one.

        The combination of the frequency increase and the restored one-seat ride to downtown would probably increase ridership on Sunset Hill, but the real purpose of the 24 extension is the connection to Ballard. Once you’ve reached the locks, the Sunset Hill extension costs just one more bus — really a no-brainer even if ridership is no higher than it was on the old 17.

      • asdf says

        Unfortunately, the one-seat ride to downtown from Sunset Hill is so slow it’s next to useless. Simply getting off the at 15th and transferring to the 15 would be much faster.

        I do have one suggestion I have as an alternative way to get coverage to Sunset Hill. As an alternative to extending the 24 to Ballard, how about simply adding an extra mile and a half of trolley wire and extending the #44 instead. If extending every #44 trip is too expensive, do it only for every other #44 trip.

        Doing it this way would provide Sunset Hill a one-seat ride to a much larger array of useful destinations, than the 24, including the U-district. As these would be one-seat rides in a straight line, unlike one-seat-rides-slower-than-competing-two-seat-rides to downtown and Seattle Center, this option should be much better at attracting riders.

        The extension of the #44 would also be cheaper than the extension of the 24 – fewer miles, plus less layover padding needed to deal with Ballard Bridge traffic. I also don’t like the idea of making the reliability of Magnolia->downtown service contingent on traffic at the Ballard bridge. With your proposal, every opening of the Ballard bridge will amount to a 10-minute service delay for everyone who wants to go from Magnolia to downtown. With my modification, very few people would be affected by Ballard Bridge openings except for those actually traveling across it.

        You would lose the one-seat ride between Magnolia and Ballard, but I don’t think too many people would miss it.

    • Mike Orr says

      Remember that people’s attitudes evolve. Magnolians were shocked that they lost late-evening service when they rejected the 2012 restructure. People also become more accepting of density over time, and start trying to channel it rather than deny it altogether. In Magnolia’s case I don’t see them ready for density yet, but I do see them recognizing that their bus service needs to change in order to keep density away. There’s also the pro-walkability, good-transit-network voices in Magnolia that are starting to speak up. All these trends will inevitably increase, maybe not precipitously, but steadily enough to make a reorganization in a few years a possibility.

    • David Lawson says

      The difference between my proposal and a lot of earlier proposals is that most riders get more frequency in exchange for their need to transfer. To my knowledge, no one has proposed two 15-minute routes for Magnolia before. Queen Anne would get a new 15-minute corridor right down Queen Anne Ave. North Beacon Hill gets a bump from 10 minutes to 7.5. Rainier Valley north-south riders get a bump to 10 minutes on the 48. The 7 stays the same, but the unfortunate fact is that today it runs a bit more often than is justified by the ridership numbers, in comparison with other current routes.

    • Luke S. says

      I’m on Queen Anne. Transit to Queen Anne is already pretty bad. This makes it slightly better. Having the 4 go through to Nickerson would be a godsend. I’m not sure it’s possible with the grade of the hills on the North side though.

      Magnolia looks actually transitable with this map, which is more than can be said for the present.

      • Eric says

        The 13 already takes this path down 3rd Ave W, so don’t worry about the bus being able to do it. 3rd Ave W is less steep than other streets up the hill.

    • Alex Bailey says

      While your statement must be given some credence in so far as it is practically a truism I actually think this plan did a very good job of synthesizing some of the best ideas into a surprisingly passable plan. In particular, I thought David’s Magnolia component was a very clever way of greatly improving service in that area, while only cutting off a tiny bit of service area coverage. Notably it seems Magnolia will get MORE service per hour than it does now, although not by much.

  5. RapidRider says

    I advocate changing the RapidRide D back to the 15. It can have its name back when it decides to truly become “rapid”.

    But why does Ballard only get every 10 minutes service? Does this assume that our miracle LRT will be constructed via gracious donations of Jeff Bezos and Bill Gates, and more frequent, duplicate bus service isn’t necessary?

    • RapidRider says

      Another quick comment: why not turn the 71 down 24th Ave towards downtown Ballard? It could layover on Leary where the old 75 used to layover. It’s a small addition of route time with a huge ridership advantage.

      And I also just noticed removing the LQA deviation on the RR-D/15…me gusta!

    • says

      I have long referred to RapidRide as “RapiderRide” and for the most part don’t view it as a failure. Overselling/underdelivering is the failure here. Hopefully we’ll see less of that in the future.

      • RapidRide says

        Oh, I agree, the line’s not a failure, it’s just the 15 with some much needed upgrades and some unfortunately downgrades. But I won’t take it seriously as a “RapidRide” until it fixes its problems, like waiting at a stop when it’s ahead of schedule.

      • Lack Thereof says

        “RapiderRide” is pretty generous.

        My wife calls it the SlowRide, and she’s not even a transit nerd. Her first time riding the C/D, she came back proclaiming “They named it wrong!”

      • says

        Yes, they named it incorrectly. However, having driven the 15, 54, C, and D lines extensively, I can assure you it IS faster. Obviously, there is more to be done: Off bus ORCA readers in the Downtown core will be a huge improvement as will Metro’s cashless plan. Is it BRT? No, obviously.

    • Mike Orr says

      This isn’t an “ideal network”; it’s what we can achieve in a revenue-neutral manner. It also doesn’t speculate about Link lines that haven’t been approved for construction yet. In an ideal world, we can see Link replacing most trips on the (current) 40 or (proposed) 15. That would relegate it to a secondary route, although it could remain frequent in the same way that most secondary routes in San Francisco rarely drop below 20-minute frequency.

  6. Brett says

    I like it. David, what input (if any) have you gotten from Metro planners? If so, have they expressed any opinions on why these changes could or couldn’t happen? And how do we start the political process to effect these changes?

    • David Lawson says

      I have deliberately gotten no input from Metro other than what it’s put into public documents in the past. (A few of the ideas here are based on past Metro restructure proposals that didn’t get implemented.)

      The best way to translate what you see here into political action is to ask for consolidated, frequent corridors in your neighborhood. The reason we don’t already have a network like this isn’t because Metro doesn’t know how to design one. It’s because of heavy political pressure to keep old, underperforming routes. A big part of the reason I did this is to motivate people to apply pressure the other way.

      • Will Douglas says

        But what’s so bad about that pressure? The purpose of a transit system is not to have “performing” routes, it’s to bring transit to everyone. Some routes will have sky-high ridership. Some won’t. That’s fine. The goal here ought to be to provide something for everyone. The greater cost is not doing so, especially given the costs of global warming, oil, and car ownership.

        You are pitting riders against each other when there is no good reason to do so, and plenty of reasons to avoid it.

      • Kyle S. says

        The purpose of a transit system is not to have “performing” routes, it’s to bring transit to everyone.

        You need to disabuse yourself of this notion pronto.

        If you constantly chase ludicrously inefficient service plans, when the time comes to make tough decisions, you make it harder to get useful, efficient service.

        Nothing like being your own worst enemy.

      • adam says

        This is the pretty classic clash of transit philosophies, one we’ve heard many times over on this blog before: efficiency versus coverage. It’s a delicate balancing act, one that puts Metro in a thankless position. Remember that the best compromises are the ones in which no one comes out entirely satisfied.

      • Aleks says

        This is the pretty classic clash of transit philosophies, one we’ve heard many times over on this blog before: efficiency versus coverage. It’s a delicate balancing act, one that puts Metro in a thankless position. Remember that the best compromises are the ones in which no one comes out entirely satisfied.

        Actually, it’s worse than that. You can consciously choose to prefer coverage over efficiency, but if you don’t understand that the real world involves trade-offs, then you can’t consciously build anything. You’ll just end up with whatever you manage to create before you run out of money.

        It reminds me of how people want free parking, and then they complain that they can’t ever find a space.

      • Mike Orr says

        “The purpose of a transit system is not to have “performing” routes, it’s to bring transit to everyone… The greater cost is not doing so, especially given the costs of global warming, oil, and car ownership.”

        Many people are being left out of the current bus network. People with half-hour service on parallel routes, and too many time-consuming turns, are cursing Metro for providing skeletal service in a major city, and are driving on weekends and demanding parking because of it. To cite one example, the 4S is redundant with the 3S and 48, and few people in Judkins Park would choose to go downtown via the ultra-slow Harborview routing. The 4S is what’s preventing a frequency increase on the 3S and 48. Those would benefit many more riders, who are going to a variety of destinations, not just to downtown.

      • says

        It’s also “worse than that” because a restructure like this doesn’t even cut coverage, except under absurd, circular definitions of coverage.

      • Josh F. says

        I actually agree with Will Douglas that transit should serve every part of the city. Even in very low-ridership places, there are still people who are unable to drive: for example, those who are too young or old to drive, those who cannot afford a car, those who are disabled, and those who simply prefer not to drive. There is definitely a very good reason for providing mobility to all these people, even if said service is not that productive. Although I am sure that many will disagree with me, I believe that forcing anyone to walk long distances (over 800m/0.5 mile, for instance) to access a bus is very bad and cannot be compensated by other ridership gains. In my view, making 1 person walk for a mile to reach a bus is worse than downgrading 10 peoples’ service from every 10 to every 15 minutes, because the amount of hardship is far worse in the former.

        The good thing about the plan (which is definitely a really good plan, by the way), however, is that very few people will have to walk more than 10 minutes to reach a bus.* To address the exceptions to this, the 24 could be extended into North Beach, the 85 into Broadview, the 31 into Laurelhurst, etc. These extensions would provide a crucial service to some people and could probably be paid for by dropping some of the 10-minute routes to every 12 or 15.

        *Yes, I understand that some people find walking for 5-10 minutes an extreme hardship. In that case, however, I believe that expanding paratransit or subsidizing taxis for these people would be more cost-effective, due to the very small number of people in this situation. I am not saying that it is ok to abandon these people, but operating fixed-route service to serve the small number of people who can walk for 1 minute but not 5 minutes is most likely a waste of money.

      • Aleks says

        Josh, what you’re describing is the classic “ridership versus coverage” trade-off. It’s totally legitimate to want to provide some amount of geographic/temporal coverage, even if the service hours could attract more riders if they were spent differently.

        In fact, your half-mile standard is actually pretty generous — I believe that Metro’s planning assumes a walkshed of closer to 1/4 mile, maybe up to 1/3 for really high-quality service.

      • asdf says

        For the past several years, I have routinely walked a mile to catch a bus because it goes where I want to go, moves quickly, and runs frequently. By contrast, the neighborhood that goes within a couple hundred feet of where, I use about once a year. Even in cases where I actually need to go somewhere it goes, it’s always less hassle to simply walk or bike than to mess with schedules, OneBusAway, etc. – especially when I can jog at the bus’s average speed anyway.

      • Nathanael says

        Coverage is bullshit.

        Either an area is populated enough to support frequent, good bus/rail service, or it isn’t. (Whether it is or not depends on the level of funding of the bus/rail system, obviously!)

        If it isn’t, it should have NO transit service. As for the transit-dependent… offer them subsidized housing closer to the mass transit lines. This would make more sense than running infrequent buses, slowly, nearly empty.

      • Nathanael says

        “There is definitely a very good reason for providing mobility to all these people, even if said service is not that productive.”

        If they’re not rich enough to afford their own chauffeurs, then *subsidize better-located housing for them*. Honestly, this is what we do with most elderly disabled people — they move into group homes.

        Transportation can’t be planned in isolation, and the missing factor here is housing policy. A transportation subsidy for one disabled person living out in the boondocks is, generally speaking, crazy; you would get better results with a housing subsidy so that said disabled person did not have to live out in the boondocks.

      • asdf says

        Even if you live in the city, coverage routes still greatly increase the range of destinations you can get you without needing to spend lots of money on a rental car. If there’s one particular place out in the burbs you need to get to once a week, it is not at all inconceivable that the presence of an hourly coverage route out there could be the determining factor in whether it is worth the money to buy a car or not. And, once you buy the car, you have little reason to ride transit anywhere except downtown, as most intra-city trips cost more in bus fare than they do in gas, once all the car costs are sunk.

    • David Lawson says

      It would be reorganized. Broadly speaking:

      1) the section between Broadway and Jackson Street is taken over by either the 7, the First Hill Streetcar, or the 35, depending on where you’re going on the north end;
      2) the section between Jackson Street and Georgetown is taken over by the 34 (with an assist from the 35); and
      3) the section between Georgetown and Westwood is taken over by the 59.

  7. chas redmond says

    I see the combination which replaces the 60 but which new route gets one from WS to CapHill?

    • David Lawson says

      You’d need a transfer, although the trip would be far, far faster than a ride on the current 60.

      For example, to ride from Westwood to Broadway, take the 120 downtown; transfer to Link at University Street or Westlake; and ride to Capitol Hill Station.

  8. aw says

    David, some of the links at the top of your FAQ page point to your personal website which I can’t access. I can of course access them from the links here or from the sidebar at your blog.

    I’m curious why you didn’t include cross-lake ST service in your maps. They aren’t involved in the restructure of course, but e.g., the 550/554 will still be a frequent corridor between 2nd/4th Ave. and Rainier Freeway Station in 2021. Assuming, that is, that the busway and Rainier Freeway Station are not closed down by East Link construction.

    I also notice that you didn’t include a crosstown route that would serve a future NE 130th Link station as you have discussed in recent posts.

    • David Lawson says

      Thank you for pointing out the bad links – they are fixed.

      I decided not to include ST Express service on the maps because I would not change it and the maps were already becoming very crowded, but you have a good point about the 550/554, especially since their existence helps allow for the new 7 routing. I’ll think about putting them on the next version of the map.

      Since the Link station locations north of Northgate aren’t settled yet, I didn’t include them. If there is a 130th station, I’d change the 75 to become a crosstown route; eliminate the 85; improve frequency on the 86; and extend the 67 north to 130th.

      • says

        Wouldn’t the 550 and 554 be cut and heavily modified, respectively, when East Link is built? Your plan assumes North Link, why doesn’t it assume East Link?

      • David Lawson says

        The proposal is at a particular moment in time, based on the known Link stations. When ST makes a firm decision on Lynnwood Link station locations, I can change that moment in time from 2021 (when there is no East Link) to 2023 (when there is East Link). Having East Link but not Lynnwood Link would be inconsistent.

  9. RossB says

    Brilliant. Absolutely brilliant. Excellent set of ideas and extremely well presented. I like how the frequency map not only had different colors for each frequency, but different widths. The nice thing about the widths is that they are intuitive (thicker line, more frequent service). The colors aren’t, but there isn’t much you can do about that, although you could maybe follow the traffic light colors for the first three (green is fastest, yellow is a bit slower, red a bit slower than that) but you would run out of traffic light colors anyway. But that is quibbling over what is, essentially, a top notch presentation. To put things in perspective, even though rail is listed in grey (the hardest color to see) this is the best map for Link Light Rail I am aware of. That sounds silly, but if you look at Sound Transit’s site, you will know what I mean. They give you a decent Bing map to work with, but only in a small window inside their other page. Not only that, but there is no way to add in the service that is planned. In other words, I will keep a link to this map just so I can see easily see the Link stations in the city.

    As to the essence of the idea (as opposed to the presentation) I completely agree. I’ve been riding buses in this city since I was really little. So, that is about fifty years, now. The system was easy to understand. If you want to go somewhere, first go downtown. It takes so long to go downtown and buses are so infrequent that rides would often take about an hour. As a bus rider, you got used to it. Once I was able to afford to drive, though, I was shocked at the difference in time. Of course, everyone else who had a car already knew this: unless you are going downtown, Metro is slow. This was always (and probably remains) the biggest complaint about our buses. Some people will never take a bus (inconvenient, smelly or rude passengers, etc.) but for those who might, speed is the key issue. If a bus can compete, people will take it. Unfortunately, we got ourselves in a terrible cycle. Bus routes are too long and too slow, which make them unreliable and make it difficult to add frequency (e. g. back to back buses followed by a long wait). The end result is that people hate transferring. So, Metro responded by adding more “single ride” buses. This is great, but it stretches the system. To make matters worse, we aren’t nearly as downtown focused as we used to be. Forcing everyone downtown isn’t such a bad thing if most people work there — but they don’t anymore. We work all over the greater Seattle area, which complicates things immensely. If you are trying to get from Ballard to Redmond, for example, you don’t want to go downtown. Even today, for such a ride, you don’t have to go downtown (you can do any number of north end transfers) but the system is flawed because it lacks good hubs outside the downtown area. Taking advantage of Link to redesign our bus system makes sense. It would probably make sense before then, but it will be a much easier sell once north Link gets built. I’ll get to the particulars on a different thread.

    • asdf says

      Even getting downtown along a major corridor takes far longer than it should. From where I live, door-to-door travel times to downtown on a weekend are something like this:

      Bus (71/72/73) – 50 minutes (includes 10 minute walk to bus stop and 5 minute wait for bus, assumes local routing with typical crowding)

      Bus (510/511/512) – 35 minutes (includes 20-25 walk to bus stop and 5 minute wait for the bus. Even though the total walk is 1.1 miles, the extra walk is well worth it).

      Bike – 25 minutes (30 minutes on the way back, due to hills)
      Car2Go – 15-20 minutes (includes a 5 minute walk to the car, plus another 5 minutes of walking downtown, from the nearest parking space)
      Taxi – 10-15 minutes (includes drive time only, not wait time)

      I find it funny how the routes between the U-district and downtown that are supposed to be the main transit corridor, comprise the worst option.

      (For peak-direction trips during rush hour, things reverse and 71/72/73 actually becomes the fastest option, thanks to the I-5 express lanes and the direct tunnel connection. Complicated!)

  10. Will Douglas says

    I’d need to see a before-and-after comparison to properly judge this. On balance it seems like a very risky and utterly unnecessary move that could backfire badly.

    Your starting point is a political problem: the state legislature is controlled by anti-transit forces, so more money for transit is a challenge. But what you potentially do here is substitute it with another political problem: you could alienate the riders and voters of Seattle. And then you’re worse off than before, because not only do you still have the Olympia problem, you might also erode voter support for transit within Seattle.

    Here’s why. This plan has two substantial problems:

    First, more transfers will be required. Some very heavily used one-seat rides would turn into two-seat rides, always with one or both legs on Link or an 8- or 10-minute bus line. Second, riders might have to walk a few extra blocks.

    You cannot dismiss these problems with a wave of the hand. Both are serious from the perspective of ridership, social justice, and public support.

    Transfers in Seattle are highly likely to reduce ridership. With each transfer the risk increases that a trip by bus could be delayed, and that would drive away riders. Right now people only have to wait once for a bus. A delay is frustrating, but once you’re on board, you’re assured of getting to your destination at a reasonable time barring an unusual meltdown. But with another transfer, the number of things that could go wrong is increased, the delays increase, and more people get frustrated with the system.

    Reducing the coverage of the bus system will also negatively affect public support and social justice. Many transit-dependent riders are not able bodied young men, as are the authors of this proposal. Seniors often cannot walk the distances you are asking them to walk. There may be safety hazards in between one’s home and a more distant bus stop. And those who lose close access to a bus may be of a lower income status. Of course, the people in the further reaches of Seattle that would be most affected by this are more frequent voters.

    I expect these objections to be dismissed either by reference to an ideological theory or with some sort of personal attack. But they are real and serious and just a taste of what would happen if anything like this were ever seriously considered by Metro. Perhaps the level of disruption is not so bad, which is why I’d like a before-and-after comparison. But we saw in Tacoma that shrinking the service area does not actually produce more public support for transit, it produces less.

    If one wanted to create an ideal transit system, one would build rail to serve the key corridors, reaching most areas of the city. Buses, cycle tracks, and sidewalks would provide the last mile connection from the nearby rail station. Rigorous studies would be conducted to understand current travel patterns and desired travel patterns. Ultimately the system would be designed to give riders what they want – even if it’s a one-seat ride – rather than trying to force riders to conform to an ideological theory.

    • says

      You’re right about about the political resistance this would face, but I think you’re much too pessimistic about how it would operate if implemented. Gridded networks that need transfers to get to more places but compensate for that with more frequent, faster and reliable service aren’t just an “ideological theory” – they’re a well tested reality. We don’t even need to go very far to see them: Vancouver and Portland both have reasonably good examples.

      I distinctly remember the first time I visited Vancouver without a car. I was staying in the airport hotel, for rather unfortunate reasons, and on my first morning I wanted to go somewhere that wasn’t in the core downtown peninsula or served by Skytrain. Google Maps told me I’d need to take 3 busses, with the first change very much in the ‘burbs, and I swore at it. But without any alternative I went ahead and tried the route, and when I got where I was going I realised that my *total* wait for all three busses had been less than I was used to for a single bus from the suburbs. Over the next few days I started to really see the benefits the city–especially the middle suburbs–was getting from concentrating busses on fewer routes so they could have much higher frequencies than their counterparts in Seattle.

      …all of which gives me a practical suggestion for Metro should they ever try implementing something along these lines. Obviously they can’t ship every complainer over to another city to see a gridded network in operation, but they could simulate it with a sort of “parallel” trip planner. Where the real trip planner tells people how to get around, the parallel one could show them what the same journey would be like under the proposed new system. Not everyone will be happy–I don’t think it’s ever possible to do a revenue-neutral change without making some peoples’ journeys longer and more hassle–but if the proposal is likely to work well in practice, enough people ought to see that their travel will be made easier to support it.

      • GuyOnBeaconHill says

        Transferring from a 10 minute route to a 10 minute route shouldn’t be hard to sell–if the transfer is easy. Service in Ballard looks looks like it will work beautifully. Every route connects to downtown with 10 or 15 minute headways and there are 2 crosstown routes with 10 minute headways. Northgate might deserve better than 15 minute headways, but the Ballard plan should be getting rave reviews.

        But compare Ballard to Rainier Valley where many of the transfers to downtown are concentrated at Mt. Baker Station. Yes, Link to downtown is faster than any existing bus route, but the time it takes to make the transfer to Link eradicates any potential time savings and leaves Valley riders with an extra burden and likely no faster trip times.

      • Will Douglas says

        I will say that implementing this in 2021 will make much more sense than trying to do it now. The north Seattle rail stations make all the difference. There will be a dedicated right of way from downtown to those stations and frequent service, both of which will significantly reduce the wait time downtown and make for easier transfers at the stations themselves.

        Some sort of service restructuring is going to happen anyway in 2021. That’s a good hook for a change. I don’t see any reason to junk the current system we have, at least not until rail to Northgate is open.

      • says

        Will,

        That’s a fair point for North Seattle, but then why not do the southern half now, and central – Fremont/UW/downtown Ballard when the Cap Hill & UW Link stations open? A restructure like this can’t be done one route at a time, but I can see some big benefits to doing a sector at a time.

      • Aleks says

        Eldan,

        Politically speaking, I think it’s important to present this as a unified restructure, or at the very least, to restructure the “rich” areas (e.g. Capitol Hill, Queen Anne, north of the Ship Canal) before restructuring the “poor” areas (e.g. the CD, SE Seattle). Right now, we’re in this weird holding pattern, where folks in the north are jealous that folks in the south have Link, and folks in the south are jealous that folks in the north have one-seat rides.

        So the only way this proposal flies is if we can convince people that a connection-based network is a service improvement rather than a reduction, and the only way we can do that is if the north “takes its medicine” at the same time (or first).

        Also, honestly, the benefits of a grid are magnified by the consistent frequency. It’s no good to have a bunch of 10-minute routes in one part of the city, if there’s nothing else to connect to.

      • says

        I love your “parallel trip planner” idea. Since true long-term planning is anathema to both Metro and Sound Transit, could someone set one up for David’s idea?

      • asdf says

        A big key to making transfers between two 10-minute routes work is reliability. If buses become bunched so that 10-minute routes turn into effective 20-minutes routes (2 buses back to back every 20 minutes), transfers stop working so well.

        In order to get the reliability we need, we really need to go cashless. The other day, I watched a 512 spend nearly 10 minutes at 6th and Olive to load about 15 people or so. By contrast, when I board the bus home from work every day, about the same number of people typically get on at my stop. But, because nearly everyone has an Orca card, the dwell time is as short as 1-2 minutes. Of course, with off-board fare payment and all-doors opening at once, we could be having 15 people board the bus in under 1 minute.

        (The RFA was never really a solution to this; it simply moved the change fumbling from downtown to Lynnwood. The only way to really get rid of the change fumbling is to go cashless.)

    • says

      …and all of that said, I think your complaint about elderly and mobility-impaired passengers, and areas with bad enough pedestrian infrastructure that walking further to a bus is a problem, is a real concern worth worrying about. Do any other STB readers know how Vancouver or Portland handle this? (I suppose Portland’s the more relevant comparison, since its covered by the exact same disability access laws as here)

      • Nathanael says

        “areas with bad enough pedestrian infrastructure that walking further to a bus is a problem,”

        (1) Change the sidewalk funding system. A lot of places have archaic sidewalk funding which inappropriately calls for property owners to pay for the sidewalks in front of their buildings, which makes no sense whatsoever. My small town is just now changing the funding system to a city-wide levy.
        (2) Make sidewalks a matter-of-course part of *all* street work, both maintenance and repairs.

    • Rico says

      The key is the quality of the transfers, if they are between two high frequency lines with a decent transfer point (comfortable, safe) transfers work well (as mentioned above see Vancouver or Portland or San Fransico etc.). It sucks deep in the burbs if your bus only comes every 30minutes, but downtown when the if you just miss a bus the next one comes in 5minutes it is painless and the ridership gains from having buses every 10minutes instead of every 30 will be more than the loss from people not wanting to transfer or from wider spacing on coverage routes.

    • David Lawson says

      Will, as a general comment, this network is not based on an “ideological theory” — it’s based on a careful effort to make the trips that real transit riders actually take faster and more reliable. While we should try to address the concerns of riders that this proposal would affect negatively, I feel 100% confident in saying that many, many more riders would be affected positively than negatively. Now, on to some of the specific concerns you raise…

      Transfers in Seattle are highly likely to reduce ridership. With each transfer the risk increases that a trip by bus could be delayed, and that would drive away riders. Right now people only have to wait once for a bus. A delay is frustrating, but once you’re on board, you’re assured of getting to your destination at a reasonable time barring an unusual meltdown. But with another transfer, the number of things that could go wrong is increased, the delays increase, and more people get frustrated with the system.

      The biggest reason to try to improve frequency — the core goal of this whole exercise — is to make transfers easier. Far, far less can go wrong when you are transferring to a bus that runs every 10 minutes than when you are transferring, as far too many people have to now, to a bus that runs every half hour. Even if a bus goes entirely missing, the next one will still be along in a reasonable amount of time.

      Reducing the coverage of the bus system will also negatively affect public support and social justice.

      With very narrow exceptions that affect very few riders, this proposal does not reduce the coverage of the bus system. Almost all of the neighborhoods and corridors that were served before would still have service, and the corridors that would lose it would have more frequent, alternate service very close by. If you compare my frequency map to Metro’s current system map, you will see that most corridors don’t just keep service — they get a frequency boost.

      The most difficult concern in implementing a restructuring like this is its effect on mobility-impaired riders; that’s the criticism in your posts that has the most resonance. But keeping the system frozen in amber, with truly spectacular inefficiencies and dysfunction, for such a small number of riders badly hurts the city as a whole. And, with a restructured system, the effect on mobility-impaired riders would diminish over time, as new mobility-impaired riders chose to locate along the new frequent corridors. In the meantime, I would explore ways to ease the transition for current riders who are mobility-impaired and would have to walk further than they can easily do. A literal “grandfather program” providing door-to-door alternative service, such as taxis, for some of those riders who are longtime residents along affected corridors might well be worth the temporary outlay, and I would support such a proposal if it could be implemented at reasonable cost.

      Perhaps the level of disruption is not so bad, which is why I’d like a before-and-after comparison. But we saw in Tacoma that shrinking the service area does not actually produce more public support for transit, it produces less.

      I think you will discover if you compare my maps with the all-day service indicated on Metro’s map, and if you review my cross-reference with current service, that the level of disruption is not bad.

      Also, my proposal barely “shrink[s] the service area” at all. If you assume a walk that is perfectly doable for almost all riders, all-day service is eliminated only from North Beach, a small and very wealthy part of north Montlake, a small and very wealthy part of western Magnolia, and the southeast edge of Laurelhurst. In most places, it would get more frequent.

      • Will Douglas says

        This isn’t an answer:

        But keeping the system frozen in amber, with truly spectacular inefficiencies and dysfunction, for such a small number of riders badly hurts the city as a whole.

        How does it hurt the city as a whole? Seattle has a good bus system that generally works well. What is needed is more – more frequency, more speed, more routes, more coverage, and more money to fund it all. As I said to Eldan above this map may well be workable in 2021 as rail to Northgate opens new opportunities, and that is your envisioned starting point. But there’s no reason to frame this as something better than what we have now or to denigrate the current system that works very well for a lot of people.

      • David Lawson says

        Seattle has a good bus system that generally works well.

        This is where we disagree. I think the current system is not effective for most trips that require a transfer, because transfers between the current infrequent routes are so time-inefficient. This proposal is largely an effort to fix that problem. It’s absolutely impossible to provide one-seat service to everywhere, and we should try to make two-seat rides work as well as they do in many other places around the world.

      • Will Douglas says

        I would add that I do not support reducing any transit service to anyone, anywhere, for any reason. Even if they are rich. (Which I most certainly am not, nor do I live in any of the neighborhoods mentioned. This restructure would neither benefit for hurt me personally, from what I can tell.)

        At a time when we face a serious climate crisis, an energy crisis, and an economic crisis brought on in large part by dependence on oil and cars, we should resist firmly any suggestion that involves reducing the availability or frequency of transit service to anyone.

      • Aleks says

        I would add that I do not support reducing any transit service to anyone, anywhere, for any reason. Even if they are rich. (Which I most certainly am not, nor do I live in any of the neighborhoods mentioned. This restructure would neither benefit for hurt me personally, from what I can tell.)

        I want a pony, too!

      • djw says

        I would add that I do not support reducing any transit service to anyone, anywhere, for any reason.

        This is simply an astonishing admission. I don’t understand how anyone can possibly take you seriously after you’ve made it.

      • says

        “At a time when we face a serious climate crisis, an energy crisis, and an economic crisis brought on in large part by dependence on oil and cars, we should resist firmly any suggestion that involves reducing the availability or frequency of transit service to anyone.”

        Even if they would never ever take transit in a million years?

      • Nathanael says

        And when you figure out how to get the legislature to authorize a cornucopia of money, then you can retain all the useless “coverage” services as well as offering good service.

        Given that the legislature won’t authorize money, you have to make choices.

    • Stuart says

      The reason people hate transfers is because they are inconvenient and unpredictable. If transfer stops had good facilities, and if the route one wanted to transfer to had good frequency, people might see transfers in a better light. Transfers aren’t ipso facto a bad thing.

      This plan isn’t about giving everyone what they want – it’s explicitly a zero-sum plan, making most efficient use of the scarce resources Metro has. And the best way to do that is with good, frequent, reliable routes that go from destination to destination. Not maintaining a bunch of higgeldy-piggeldy routes that only come every half an hour for the benefit of the few people whose front doors are passed by said routes.

      • RossB says

        Exactly. People hate transfers because they have been so bad in the past. For people who have a choice, it is often the difference between taking a bus and driving. In other words, there are lots of people who say “I would like to take the bus, but it takes too long. I have to transfer…”. You never say this when riding a subway system. It is just part of the process. You walk from one colored line to another, make sure you take the subway heading the right direction, wait a minute and get on. It’s when you have to wait 20 minutes, or find out that you just missed your transfer that you decide that driving might be a better option.

        I’m sure there will be people who would hate this. But there will also be a bunch of people who would love it. I think it would be great to see some modelling on this to see what percentage of rides go up and what percentage go down. My guess is that this would actually increase ridership, because more folks could get to where they want to go faster.

      • asdf says

        In subway-oriented cities like New York and Boston, people all time choose trips that involve transferring between subway routes over a one-seat bus-route. The reason for this is not just rail bias. It’s a preference for a transfer between two routes which are very fast, frequent, and reliable, over a one-seat ride on a route that is neither fast, nor frequent, nor reliable.

    • Aleks says

      Will,

      Transfers in Seattle are highly likely to reduce ridership.

      This is completely false.

      Metro’s recent “big bang” change reorganized West Seattle, to create a single transit spine (RapidRide C) and a number of connecting routes. It turned out that ridership on RapidRide was so high that Metro had to deploy virtually all of its spare service hours on increasing peak frequency, just to ensure that all the riders could fit.

      If all riders hated connections, this wouldn’t have happened.

      With each transfer the risk increases that a trip by bus could be delayed, and that would drive away riders.

      This is completely false.

      Reliability is something that can be quantified. Metro has lots of great data on the reliability of various bus trips and corridors at different times of day. For example, certain corridors are really bad for reliability, because they’re filled with cars trying to get on/off the highway, or because of bridge openings, or just because there’s too much traffic.

      Every inch you travel on a bus is an opportunity for delay. There’s nothing special about connections with respect to the time you spend on the vehicle. In fact, a well-designed grid network should *reduce* the amount of time you spend on a vehicle, by shifting buses away from congested downtown corridors, and towards more reliable crosstown corridors.

      It’s true that you have to wait for two buses, instead of just one. But depending on the frequency of the buses, that doesn’t have any impact on reliability. If you have to wait for two buses that come every 10 minutes, then the absolute worst delay you could have is a 20-minute wait. If you have to wait for one bus that comes every 30 minutes, and you just barely miss the bus, then you could have a 30-minute wait.

      Also, low frequency magnifies the effect of route unreliability. If a bus comes every 30 minutes, and one trip is running 10 minutes late, then you could end up waiting 40 minutes. But if a bus comes every 10 minutes, and one trip is running 10 minutes late, then you can just take the next one.

      Reducing the coverage of the bus system will also negatively affect public support and social justice. Many transit-dependent riders are not able bodied young men, as are the authors of this proposal. Seniors often cannot walk the distances you are asking them to walk.

      This comment makes me think that you didn’t actually read David’s proposal. There are very few corridors that have completely lost all service. The few that have — such as the 19th Ave segment of the 12N — have virtually zero ridership at all times of day. In other words, very, very few people are being asked to walk any further than they have to walk today.

      It’s also hugely misleading to pretend that frequency has no impact on “social justice”. Let’s say that a senior citizen needs to get to a doctor’s appointment at 10 AM, and the trip involves about 20 minutes of in-vehicle time. It’s very important that the person not be late. If the bus came instantly, and was perfectly reliable, they could leave at 9:40. But let’s say that the bus comes every 30 minutes, at 9:11 and then at 9:41. This person needs to take the 9:11 trip (since the 9:41 trip would be too late), *and* they need to leave the house about 5-10 minutes early in case the bus is ahead of schedule. The same thing happens on the way home. So the 30-minute frequency has effectively turned a 40-minute round trip into a 2-hour one. That’s enough to convince people to stay home for non-essential trips.

      I expect these objections to be dismissed either by reference to an ideological theory or with some sort of personal attack.

      David has probably spent on the order of 100+ hours putting this proposal together. You have made several objections to the proposal which suggest that you haven’t even read it. I think “reference to an ideological theory” and “personal attack” is a perfect description of what you’re doing.

      Frankly, I’m not even writing this comment to reply to you; I’m writing it for other people who may be reading this thread.

      Ultimately the system would be designed to give riders what they want – even if it’s a one-seat ride – rather than trying to force riders to conform to an ideological theory.

      One-seat rides are not what riders want.

      This is ridiculously easy to demonstrate. Go out and ask bus riders which they would prefer: a bus that comes once a day and takes 2 hours to get to their destination, or a two-bus trip where each bus comes every 5 minutes and the total journey takes 10 minutes. If you find a single person who chooses A, I’ll be shocked.

      What riders want is to get to and from the places they want to go, when they want to travel, with the minimum amount of stress. There are multiple ways you can build such a system, and different trade-offs you can make.

      If one wanted to create an ideal transit system, one would build rail to serve the key corridors, reaching most areas of the city.

      Okay, this really takes the cake. If you build a rail spine, with bus connections on each end, then you’ve turned a one-seat ride into a three-seat ride. If riders truly hate connections, as you have ideologically asserted, then building rail would never be a good idea — it would always be better to run point-to-point buses instead.

      And yet, you seem to understand that riders are happy to make connections between high-quality services. David’s proposal — which I urge you to actually read — simply proposes that we make our bus network as reliable and frequent as our rail network. By your own logic, riders will like this just fine.

    • Mike Orr says

      “Many transit-dependent riders are not able bodied young men, as are the authors of this proposal. Seniors often cannot walk the distances you are asking them to walk.”

      This ignores the transit-dependent, non-able-bodied people who are not served well by the current network. It’s not true that all transit-dependent, non-able-bodied people currently have convenient one-seat rides to most places they want to go. Rather, they are forced to live in a few places that are convenient to some of their destinations, and they spend an hour transferring to others, or avoid going to those places at all. And people with low incomes have little choice where they live: they have to live where an affordable vacancy exists, which may be in Rainier Valley or it may be in Lake City or Kenmore.

      • Charles says

        I also observe that this proposal would make east/west connections in the south Seattle corridor worse and does not factor having to traverse steep terrain.

      • David Lawson says

        Charles, I’m not sure how my proposal would make E/W connections in the south end worse, with the exception of deleting the Othello tail of the 50.

        The 50, which is already the main E/W route, gets more frequent.

        The 60 is the other current E/W connection, and it’s incomplete — unhelpful to anyone east of western Beacon Hill or west of Westwood. My 59 is slightly less frequent off-peak, but provides a complete E/W corridor from Rainier Beach to southern West Seattle.

        And, believe me, I spent a lot of time thinking about terrain. Terrain is why the 1 exists, why the 5 and 40 remain on separate corridors, why the Magnolia network looks the way it does, why there aren’t better crosstown connections in the north end, why there is a 52 loop, and why I kept the Beacon Hill deviation on the 34 and 35. If you can give me specific examples of places where terrain makes my network difficult to use, please offer them.

      • Nathanael says

        It’s also very common to have *inconvenient* one-seat rides. The disabled do not want to have to schedule their lives around the unreliable less-than-once-an-hour bus any more than anyone else does.

  11. says

    David,

    As you will have gathered from my other comments, I am a fan of this idea, having seen systems more like this work better than Seattle’s status quo does. I’ve often wondered how to fit something like a grid to Seattle’s complicated topography, and I’m glad someone’s done the homework – I keep looking at your maps wanting to make them more of a pure grid, then zooming in and understanding why that route turns there, etc. I do have a few concerns, though:

    1) The disability access point that Will Douglas raised.

    2) On the “common corridor” segments, how would the interleaving be maintained? I’ve seen this work really well with rail or fully traffic-separate BRT, but I don’t think I’ve seen it done well with in-traffic busses except where the overall frequency is *so high* that a little bunching doesn’t matter. Obviously that can’t be done with a revenue-neutral restructure, but getting this right seems crucial to making the whole plan work.

    3) There are some stretches of N and SE Seattle with a rather long distance between E-W routes. I know those routes will never be high ridership, but their absence seems to increase the number of not-that-long trips that will need two transfers – I wonder if it’s worth adding something there to pre-empt likely objections to the whole plan.

    • RossB says

      1) I think he already addressed the disability issue. Basically, this provides just about as much service as before. In other words, if you were two blocks to the bus before, then you will be two blocks from a bus now.

      • says

        That’s not true. In fact, I don’t think it *can* be true of a revenue-neutral plan. The route straightenings mostly take out deviation that existed to bring bus stops closer to particular groups of people, and the frequency improvements come at the cost of pruning less-used lines. On balance, these changes are good, but for both moral and political reasons we need a way to make them not strand people who are both transit dependent and mobility impaired.

        Personally I think paratransit should pick up the slack here rather than weakening the whole scheduled transit system to cater for a comparatively small number of people but (a) disability advocates tend to hate this idea and (b) it can’t provide a decent level of service without more money. I haven’t done the homework to figure out how much more money, or whether there are better alternatives.

      • David Lawson says

        As I mentioned in a response to Will, the issue of riders with disabilities is very real, and it’s something we need to think about. But it can’t be the only thing that animates the planning. Riders with mobility impairments should get some extra consideration, but not so much as to make the network materially harder to use for the far more numerous riders with full mobility.

        The first solution I would look to is to eliminate physical barriers to mobility. Improve sidewalks and ramps, look for flat and safe paths, make sure crosswalks are in good condition and signal phases allow for safe crossings, and so on. This would help many people who have trouble with hills, rough terrain, or dangerous crossings.

        But that won’t work for some riders, who have real difficulties moving even over short, safe distances. I would advocate alternative service delivery for many of these riders. It would not necessarily have to be paratransit in every case. Much cheaper taxi scrip would work well in the case where people can enter a normal car, but cannot walk a long distance, and would allow for greater schedule flexibility.

        And, as I mentioned, this is a temporary problem to some extent. If you restructure the system, new riders with mobility impairments will locate near the new routes, and some existing riders with mobility impairments will move to be near the new routes. A grandfather program to provide alternative transport to specific riders could serve existing riders well, and its cost would slowly decrease over time.

      • Aleks says

        Eldan,

        Suppose that you have three residential neighborhoods (A, B, and C) and three commercial centers (X, Y, and Z) and you want to provide transit from each residential neighborhood to each commercial center. (I know this is way oversimplified, but the basic point remains.)

        One approach is to run 9 bus lines with 30 minute frequency: A->X, A->Y, A->Z, B->X… C->Z.

        The other approach is to run 3 bus lines with 10 minute frequency and a common transfer point: A->Z, B->Y, C->X.

        Both systems provide the same geographical coverage. The only difference is that instead of waiting up to 30 minutes for a single bus, you wait up to 20 minutes in total for two different buses.

        Most of the changes in David’s plan have this flavor. For example, consider Fremont. Currently, Fremont is served by a hodgepodge of buses: the 26/28, the 5, the 40, and the 31/32. In the proposal, the number of bus routes serving Fremont is reduced by 50%. And yet, the number of effective destinations you can reach from Fremont increases dramatically; you can go from Fremont to Upper QA, and you can also go to Phinney/Greenwood without having to walk up the hill.

        As another example, consider what happens to the 43 — it’s deleted, and its service hours are used to boost frequency on the 8 and the 48. Not a single stop loses service, and the network gets simpler and better.

        Now, it’s true that *some* (but certainly not all) of David’s proposed changes do discontinue service at certain stops. In fact, David made a list of all such routes (look for “including the elimination of some all-day service”). But I’m not personally bothered by most of these changes, with the possible exception of the ones that take away service from urban hospitals (i.e. not the VA).

        Also, keep in mind that most of the services David proposes to cut are extremely low-ridership. That means, by definition, they aren’t benefiting very many people.

        I don’t think paratransit is the answer (mostly because it’s ludicrously expensive and user-unfriendly), but I do think there’s room for some “super-coverage” services — maybe 1 bus an hour along some of the corridors that would otherwise lose service altogether. And for some of the really sparse/low-ridership areas, I agree with Charles that it might make sense to subsidize ridesharing.

  12. community council member says

    Thanks for the idea and all of the work that went into this post. Recent history shows that some of the proposed changes would meet resistance. To improve chances for progress, I would hope that those who write for and comment on this blog would take part in their local community organizations. There’s no reason that community groups have to be the bastion of resistance if they have the advantage of participation by those who see the advantage of change and can make a reasonable case for it.

  13. Al S. says

    It’s always exciting to lay out a new routing map. While the map looks great, I tend to look at things from a community topological and accessibility perspective — particularly for those who are more transit dependent. The reason that I point this out is that the Central District residents would have to transfer more by splitting the 8 into the 8 and 6. The 8 is an important crosstown route and forcing transfers in Madison Valley is the wrong place to do it. I would move it to either the Capitol Hill or Mt. Baker stations. I also think there are some productivity issues, such as the pairing of high demand Route 2 in the Central District with the low demand Leschi district. In sum, I would suggest being more innovative at Central District routing because this area of high ridership needs fast, less crowded buses. Maybe split the 2 with a branch going down MLK or 23rd, replicating the old 4 for example, or a Route 2 that doubles back on Route 3?

    • David Lawson says

      I agree the 8/6 division is not perfect. If I could find two more buses to add to the 6, I would extend the 6 to Capitol Hill Station. As is, I tried to give it connectivity to as many super-frequent routes as possible. Almost any popular destination should be a two-seat ride with a time-efficient transfer from stops along the 6.

      As for the 2, the Leschi tail is “free,” in the sense that I wouldn’t be able to delete a bus if I chopped it.

      • Al S. says

        The ridership above 34th/Union on the 2 is very light. Most riders board between 14th and 23rd Avenues, so having Route 2 serve as a Route 4S replacement would seem easy. As far as the Route 8 goes, I would point out that there will be major ridership shifts with the Capitol Hill station opens, and ridership within walking distance of the station will fall significantly. Except for Group Health and trendy restaurants on 15th or in Madison Valley, I’m frankly not sure why Route 8 riders in South Lake Union would want to travel east of Broadway. It seems rather unfair to penalize thousands of Central District residents with a transfer so that South Lake Union residents can have a direct bus to Luc. Just have the meeting point at Capitol Hill Station rather than Madison Valley and that should solve the issue.

      • David Lawson says

        Al, current ridership statistics don’t support the idea (implicit in your post) that the 8 has more Central District riders than SLU riders. It’s the opposite, by a significant margin.

        My revised 8 would be the only reasonable way from the Madison Valley or central Capitol Hill to SLU. Mostly, riders wouldn’t be people going to Group Health or “trendy restaurants”; they would be people going to or from home.

        And I think you’re overstating the problems with my proposed 6. It has easy and quick transfers to the following very frequent service, going to a million places (and I’m not even counting transfers to 15-minute bus routes):

        - Link
        - Route 7
        - Route 8
        - Route 12
        - Route 48

        In an ideal world, I’d extend it to Capitol Hill Station, but I think it’s very useful as is.

        I don’t want to replace the 4S. It’s an inefficient route with low ridership (south of Jefferson), and its riders could have faster trips downtown by transferring to Link at Mt. Baker or to the 550/554 (later, East Link) at I-90/Rainier.

  14. Al S. says

    The rerouting of Route 7 is a curious idea. Still, the route is heavily used by many Asian transit-dependent people — many of whom are seniors and have mobility problems. The Streetcar is a good reason to move it, but I have to wonder whether or not it does damage to those riders that depend on it.

    • says

      “Still, the route is heavily used by many Asian transit-dependent people — many of whom are seniors and have mobility problems.”

      Ah, yes, the old 42 line of reasoning. They may lose a one seat ride to downtown in many cases, but they would gain frequent connections to many other areas of the city, plus a one seat ride to the hospitals on First Hill. Easier access to more employment centers for workers and hospitals/medical practices for seniors.

      Don’t forget, the last of the high floor coaches will be LONG gone by this date which will make boarding coaches easier. Add in implementing RapidRide’s proof of payment boarding across the entire system and cashless/paperless payment across the system and you speed the buses further.

      Seriously, how does all of this “damage” them?

      • Al S. says

        I sometimes wish that this board had more participation from people of color, particularly representing seniors and low-wage workers who are African-American, Latino or Asian, to keep us young and middle-aged non-Latino white guys in check. The proposed route map right now looks painfully geared to us.

      • GuyOnBeaconHill says

        Al S. — I wish that “low-wage workers” couldn’t be used as a synonym for “African-American, Latino or Asian”, but I get your point. I don’t think this plan will be well accepted on Beacon Hill or Rainier Valley. Transfers are fine if they reliably speed up the overall trip, but the proposed changes will make longer trips for most SE riders who want to go downtown. Bus service on Beacon Hill will connect to Broadway or the waterfront (which won’t be a hub of convenient connections). The 7 will be on Boren, which isn’t exactly free flowing during peak hours. The 106 will end at Rainier Beach and the 50 won’t run to Othello anymore. The 59 is a junk route that goes everywhere (and nowhere) and will take forever to get there. If this plan gets circulated in SE Seattle, the responses will be “Are you kidding?” in 50+ languages.

      • David Lawson says

        the proposed changes will make longer trips for most SE riders who want to go downtown.

        This is only true if you add the qualifier “one-seat.” Don’t fall into the fallacy that one-seat rides are always the best rides.

        For a quite considerable majority of SE riders, the quickest trips to downtown are already on Link; one-seat bus trips are extremely slow. This plan would improve Link-bus trips in the following ways, and would not slow any of them down:

        - For 6, 7, and 48 riders, the transfer at Mt. Baker would get more convenient. It’s impossible to see at the scale of my map, but I’ve revised routing on all three of those routes to drop passengers off and pick them up closer to the station. Only 14 riders would have to walk all the way to MBTC.

        - Riders at the south end of the 7 would have a vastly more convenient transfer to Link at RBS.

        - For 34/35 riders (current 36 and 60 riders) north of BHS, buses would have more reliable frequency, being scheduled in a 7.5 minute common corridor.

        - The 106 routing to Rainier Beach Station is 2-3 minutes faster than current routing.

        - The 50 won’t run to Othello (where essentially no one — less than 1 rider per trip — rides it), but it will run twice as often to Columbia City (where a lot of riders use it).

        - The 59 isn’t great, but riders along the eastern half of it will have the same speed and frequency of trips to downtown that they have now. It will actually improve matters in upper Rainier Beach, because (unlike the current 7) it will make a direct connection to Link without a walk. And it will vastly improve service for riders in several West Seattle neighborhoods that have been hurt by the restructure. Also, with more resources it could warrant 15-minute service — which would be a big upgrade for everyone along it.

      • GuyOnBeaconHill says

        VBD, Beacon Hill riders currently have a one seat trip to most of the hospitals via the 60. With the FNP, a transfer will be needed. Rainier Valley riders will get better service to the hospitals, but that’s not where most of them are headed.

      • GuyOnBeaconHill says

        I’d love to see details of the plan that makes transferring at Mt. Baker Station easier. I don’t see how it’s possible and without an easy Link to bus transfer at MBTC, there just isn’t a benefit to the transfer.

        I’m also very skeptical about the 34/35 providing better service. Because the 2 routes will be coming from different places and then heading in different directions, it seems that it will be hard to maintain consistent 7.5 minute headways on their common corridor.

        The 50 seems to be building ridership between Othello and Seward Park. It won’t ever be tremendous, but there aren’t any layover facilities on Orcas, so why not invest the 6-7 minute it takes to get to Othello?

        For 106 riders, I’m seeing schedules that say it takes Link 24 minutes to get to 3/University from RBS and the 106 is usually scheduled for about 29 minutes midday. Allow 2 minutes to walk from the bus and tap in, then an average wait of 5 minutes for the train and the forced transfer could be slower. If the 106 were moved to Cloverdale the bus trip would be several minutes faster than Link.

      • Al. S. says

        Guy I think we’re on the same page. I said both persons of color AND low-wage or senior; usually think of that as a double criteria rather than a synonym. I do think that there is a prevailing attitude that because MLK has Link, that SE Seattle can now be neglected, and the Central District is somehow served by Madison and First Hill Streetcar, which of course only skirt the communities (consider the TMP which prioritizes putting more frequent and faster service out to more empty Madison Street buses than to crowded Union or Jackson Street buses even though the Phase 1 analysis showed great needs on Union and Jackson and the effects of the Capitol Hill Light Rail Station and streetcar weren’t factored into the Madison load analysis). It amazes me how the push to take Route 2 service off of Seneca Street in First Hill is something that continues to be heavily advocated, even though that street segment has high rise residential buildings including seniors with mobility issues and the residents had an uproar just a year ago. Finally, I think that there is a tendency to think about dwelling unit density rather than population density (as if dwelling units ride buses – LOL), which tends to suggest more focused planning favoritism for single-person households and less for larger, extended family households more common in communities of color. While no single item is intentionally focused to hurt communities of color, I think there is an unintended effect of “privilege” to encourage routes and frequency to benefit “riders who look like me” among some of the advocates in the blog. That’s why I feel obligated to speak up for those not represented here as best as I can even though I don’t fit the underrepresented demographic on the blog.

      • Aleks says

        Al,

        I agree with the thrust of what you say, but I have to quibble with a few (okay, most :D) of your points:

        I do think that there is a prevailing attitude that because MLK has Link, that SE Seattle can now be neglected

        Can you cite an example of this attitude? I don’t know anyone who wants to neglect Link — just lots of people who want to restructure service in a way that allows people to take advantage of Link. That’s exactly what I would want if I lived near a Link station.

        and the Central District is somehow served by Madison and First Hill Streetcar, which of course only skirt the communities

        Can you cite an example of where someone claimed that the CD was served by either service? In David’s proposal, the CD is served by the 2, the 3, the 6, the 14, and the 48, each of which are heavily-used buses today that receive significant service upgrades.

        (consider the TMP which prioritizes putting more frequent and faster service out to more empty Madison Street buses than to crowded Union or Jackson Street buses even though the Phase 1 analysis showed great needs on Union and Jackson and the effects of the Capitol Hill Light Rail Station and streetcar weren’t factored into the Madison load analysis)

        Have you ridden the 12 through First Hill? It’s one of the most consistently busy corridors in the whole network. Madison and Boren is a blockbuster of a stop. And given that most of the people using that stop are heading to the nearby medical centers — and that it’s on the top of a hill — I think it’s highly unlikely that either Link or the streetcar will do much to affect that demand.

        But none of this has anything to do with the Central District. This is a First Hill service, and it isn’t coming at the expense of service to the CD.

        Finally, I think that there is a tendency to think about dwelling unit density rather than population density (as if dwelling units ride buses – LOL), which tends to suggest more focused planning favoritism for single-person households and less for larger, extended family households more common in communities of color.

        Can you cite an example of this? In all the conversations I’ve had about service restructures, the overriding goal is to improve ridership — whether it comes from single-person apartment residents, or extended families, or offices, or hospitals, or anything else.

      • David Lawson says

        GoBH, MBTC improvements would consist of the following:

        1. Repositioning of stops on Rainier to be as close as possible to the MBS entrance. This has already been done southbound. Northbound, I’d place a stop right by Starbucks, on Rainier farside MLK. Buses blocking the Starbucks driveway? Too bad. Deal with it. This is a major transfer point.

        2. Building of a new, well-marked, signalized crosswalk across Rainier directly in front of the MBS entrance. This would slightly slow down northbound car and bus traffic which would be held at the MLK intersection. Again, too bad; the transfer is more important.

        3. Revisions to routes 6 (currently 8S) and 48 to serve these stops on Rainier in both directions. For the 6, the revision would be to travel through the MBTC in both directions between Rainier and MLK, using Rainier between the MBTC and the intersection of Rainier/MLK. For the 48, the revision would be to extend the route slightly from the MBTC to the former route 8 terminal on S Walden farside Rainier, stopping at the new bus stops in both directions. The 7 wouldn’t require route revisions, just stop revisions.

      • Al S. says

        I think you might have misread my comments on neglect. I was referring to the neglect of any effective service improvements in SE Seattle, not neglect about Link. Route 50 is shortened, Route 8 is severed in Madison Valley and Route 7 is cut off from direct service into Downtown. SE Seattle is taking a big hit on accessibility here. Meanwhile, the often discussed Graham Link station isn’t shown in this proposal.

        David’s proposal of course includes routes in the Central District, but the hit on Central District service is profound and probably unfair given that service reductions further north are offset by new North Link service. Route 2 in the CD is severed and the frequency is reduced. Route 6 is severed. Routes 4S and 27 are removed (also noting the low frequency on Route 14 proposed here). Contrast that with more frequent service all the way to Madison Park on the proposed 12 for example. As far as Madison ridership goes, isn’t the Streetcar project somewhat justified because of providing new access to the medical center area, taking riders away from Madison? If it works (yeah I have my doubts) Madison ridership will fall.

        I find only five references to “population density” in the past four years on this blog. However, I find numerous references to dwelling unit discussions — from apartments to parking requirements per unit to zoning density per unit to building height to apodments. Consider the 2011 STB posting “Density, An Ill Defined Term” which only defines density in terms of units, not people. Meanwhile, my foreign-born friends (the fastest growing US population segment) expect to be able to have other family members stay with them for extended periods of time and often end up moving to the suburbs for that reason.

        Don’t get me wrong. I am glad that thinkers are willing to think of new alternatives! David’s work on laying this out is commendable. It’s just that transit riders are not all able-bodied males traveling alone and living in small one and two person households. If we design a transit system thinking that transfers are effortless, that riders all have OneBusAway on their phones and that no one has strollers, three grocery bags or a cane — we carry an implicit bias on our view of transit. Lets try to widen the “input” on these kinds of things before casting them in stone and defending them as a done deal.

      • David Lawson says

        Al, I think you must be misreading the map with respect to frequency. Routes 50 and 14 are both doubled in frequency (15 minutes rather than 30). Route 2 frequency remains the same; it isn’t reduced.

        Route 7 is a bad way to get downtown even today; it’s slower than Link even including transfer time.

      • Aleks says

        Many of the things that you describe as “neglect”, I would describe as “improvements”. Splitting the 8 would majorly improve reliability on the 6. Rerouting the 7 would enhance connections to First Hill and Capitol Hill, while Link continues to provide access to downtown. Splitting the 2 would increase reliability as well.

        In addition, frequency on the 14 is doubled. Frequency on the 3 is doubled. Freuqency on the 48 is improved by 50%.

        All of those represent real service improvements, and they’re not free.

        As a contrast, Capitol Hill is arguably losing service. The 35 replaces the 49, at the same frequency. The new 8 (6 buses per hour) replaces the old 8+43 (8 buses per hour, albeit poorly spaced). The 12 stops serving Capitol Hill. The 9 and 11 and 60 all go away. The only real addition of service hours comes from rerouting the 2 along Pine.

        As another example, consider Upper Queen Anne. The 1 has the honor of being one of the least frequent routes in the whole system. The 13 replaces the 2N/13 at the same frequency. The 4 replaces the 3N/4N at the same frequency. There are no new service hours in sight (with the exception of the cost to extend the 13 to Fremont).

        Or consider Wallingford. The 31 replaces today’s 31/32 at the same frequency. The 16 (4 buses per hour) is a service reduction compared to the combined 16+26 (5 buses per hour). The 44 gets a boost in frequency, but if you’ve ever ridden the 44, you know that it’s desperately needed.

        In fact, in terms of absolute number of service hours added, I can’t find another neighborhood that’s getting as much of a service infusion as the CD.

      • asdf says

        We already have a pedestrian bridge over Ranier. Time to add another 10 feet of elevated sidewalk to connect the bridge with the station, avoiding the need to go down and back up again. This is something that should have been done years ago, when Mt. Baker Station was still being built. The excuse that if wheelchair user can’t use the bridge to access the station then no one else can either is extremely flimsy. Wheelchair users would still have signalized crosswalks, so they would not be neglected.

      • Scott Stidell says

        Also not noted–because it won’t be open yet (although if that section of East Link could be prioritized for an earlier opening date it would help dramatically) is the Rainier Station. This will give a direct transfer from the platform to 23rd and a reasonably short, flat walk to the proposed 6 on MLK. For people traveling from that general area to the north end of downtown or SLU it may prove a popular alternative.

  15. Al S. says

    You have a creative mind, so can I suggest another strategy to a conceptual layout: a overlay “local” and and “limited” network? One of the biggest headaches in riding Metro is that the system seems to try to be everything to everyone so travel speeds for riders are painfully slow for many longer trips. I agree with other posters that the “RapidRide” designation is someone deceiving, but I do think there is a vital need to have a comprehensive limited-stop system in the city to compliment Link. It could really improve accessibility for many who don’t live near Link or a RapidRide line.

    • David Lawson says

      In general, I don’t think (and neither do a lot of people with much more planning expertise than I have ) that local overlays work well until frequencies on the express lines are very, very short. I would rather focus on improving travel time along the 10-minute lines, which can include wider stop spacing, BAT lanes, comprehensive transit signal priority efforts, and possibly headway management in the long run. We actually have a pretty good street network here in Seattle compared with most cities, and buses can run fast in most corridors (with obvious exceptions, cough, 8) if they have the facilities and stop spacing to do so.

      • RossB says

        The 44 is also a problem. With this layout, I think it would become even more important. Since rail and a few BRT lines provide the major north south service, the regular buses need to provide excellent east-west service. If they bog down, then the whole system becomes slow. Maybe it makes sense to move the 44 off of 45th, and onto 50th.

      • Aleks says

        I would prefer to see the 44 extended east to Children’s/Sand Point. It would avoid a major source of congestion (15th/Pacific), and it would provide desperately-needed frequent service to that corridor.

        Moving the 44 off 45th means that you’re adding an extra 3-block walk for pretty much 100% of current riders. For better or worse, 45th is where people want to be. I would prefer to invest in speed and reliability improvements for 45th, including measures to encourage cars to take alternate routes (like 50th and Pacific).

      • d.p. says

        The 44 needs to be a subway. Underground. The whole way. It’s pretty much the only corridor in this entire metropolitan region with such a cocktail of incurable bottlenecks, topographic impediments, and untapped network benefits as to justify an uncompromised, fully-buried rail segment. [Ot] But that’s a different pot of money waiting to be spent well or spent poorly. And I don’t mean to digress too far from David’s admirable proposal, which smartly addresses only the Metro pot.

      • David Lawson says

        What d.p. said about the 44 being a subway.

        For now, on balance, I think it works better to have the 31 be the route that goes all the way crosstown. You don’t need to build a bunch of trolley infrastructure (when I’m already suggesting way more trolley projects than will happen soon), the frequency matches up better with demand, reliability between U-District Station and U-Village is better, and you preserve the heavily used connection between Ballard and UW Hospital from the current 44.

      • Aleks says

        I completely agree about the 44 being a subway. That may be sufficient reason not to build lots more short-lived trolley wire. However, if a trolley expansion is something that the City of Seattle could fund in the next couple of years, then I think it could still be worth building, especially if the soonest we’d get a 44 subway is 2025.

        Regarding frequency and demand, you could make the same claim about the 71. According to the TMP, there’s almost twice as much demand on the segment west of Roosevelt (today’s 71) as the segment east (today’s 48).

        In the case of the 44, I would argue that there’s a huge amount of latent demand. It’s a royal pain to get to U-Village by bus. I think you’d see tons of new riders on frequent service to 25th, at the very least. This is doubly true when you consider that this route will also be the best connection to downtown (via Link) for the whole area.

        I also think you’d meaningfully improve the reliability and speed of the 44 by moving it away from heavily-congested 15th and Pacific, and avoiding the turn at 15th/45th. For people actually heading to the hospital, Link will provide a fast, frequent, and reliable connection. Pacific can be so congested during rush hour that the connection would probably be a net time savings, even if the 44 continued south.

      • d.p. says

        For the record, I tend to agree with Aleks that, with or without a subway, the 45th Viaduct east from the U District is longing for direct access to the frequent grid.

        It’s highly possible that it will not make sense to extend the desperately-needed Ballard-UW subway past Brooklyn station — it would add 50% to the length and cost, which is a heavy price to reach a sprawl-addled mall (with lots of parking) and an outpost hospital (an order of magnitude smaller than First Hill’s hospitals, which were denied a subway).

        But the demand absolutely does exceed the quality of today’s service, with its presumption that a 5mph tour of UW campus is a fun activity to be shared by all. The campus detours are ridiculous. UW is ringed by stops to the south and west, and a 45th route could stop directly at the primary northern entrance.

        Through-riders on core routes should never be forced into the campus gauntlet just to access the rest of the network; retaining the 31 for this purpose is a gross violation of the transfer-improvement principles under which David’s proposal was conceived.

      • David Lawson says

        d.p., just to be clear, the 31 in my proposal uses the viaduct, not the campus. The campus tour is reserved for N/S routes that switch direction at the U-District and have ridership that overwhelmingly wants to go to the campus.

      • d.p. says

        Ah… okay, then.

        As I mentioned just below, I haven’t been on a normal computer up until now, so I’ve only been able to view the maps in slivers. Will view them holistically, and will probably post my remaining general concern(s), as yet unaddressed in the discussions, way down at the bottom.

      • David Lawson says

        Aleks, your observation that serving U-Village might be as quick as slogging down 15th proved right. Thank you! Extending the 44 to Children’s cost… zero buses, a result which surprised me but made me happy.

        And I was able to modify the 31 to serve 55th with just one extra bus, which I was able to get from trippers on the 74 Express, which would be redundant with the latest version of the 31.

        Great comments and ideas. :)

  16. d.p. says

    One thing’s for certain: I need to look at this on a real computer, because it’s really hard to parse on a mobile!

  17. Charles says

    Laudable effort. But topology and the need for frequent east/west connections in south Seattle also need to be addressed.

  18. the358 says

    Love what you’ve done with the 345. Still wishing there was something that connected Bitter Lake with Lake City that didn’t require going to Northgate, but until the Link station opens at 130th and makes that a priority, I suppose this will do. It’s certainly WAY better than what we have now, which is basically a nursing home shuttle for Four Freedoms to get to medical appointments and community college classes. Which, I’m all for people in nursing homes having shuttles–just not ones that the rest of us have to ride because it’s the only bus service in the area.

    • RossB says

      I agree. All the more reason that the 130th station is a priority. When you look at a map, that section (that serves the Northgate station) seems out of place. A lot of curves and twisty lines for what? Oh yeah, there is a station there. It would be interesting to see how this map would change if the 130th street station is added. I think it becomes just a bit more elegant and thus much faster (fewer turns means faster buses).

    • Mike Orr says

      I was surprised that you kept the 5th Avenue deviation on the 348. Earlier you talked about consolidating the 347/348 onto 15th and deleting service on 5th. Did you find yourself unable to put the hammer down at the last minute? Or are you just postponing it until the Shoreline restructure? Do you see any potential ways to eliminate the U-shape and still provide service to 5th while also making 15th more frequent?

      • David Lawson says

        I would postpone it until Lynnwood Link. At that point, I would do a major restructure, which would involve splitting what is now the 347 and 348 into three pieces. I’ll cross that bridge when we have a firm decision on Link stops.

  19. says

    Despite above quibbles, I LOVE this.

    You have created something that, if implemented, will result in full employment for transit bloggers dedicated to improving transfer points. Start with Husky Stadium station and Montlake which, I’m sad to say, are likely to be disasters. I’m sure there are many others that we needn’t go into here.

    Bravo!

    • RossB says

      I was going to comment about Husky Stadium and Montlake as well. I don’t know what the eventual plans are. Are people excepted to walk from the train station to the bus station? Exactly how far is that? What about the changes to 520? How does that fit into all of this. For example, will buses use the express lanes which will dump them out close to the U-District station (at 45th)? If so, then maybe we don’t need to do much. Folks might transfer at the U-District station instead of the Stadium station. This would also work well for folks coming from Ballard (and lots of other places) riding the 44. They would avoid the last bit from the U-District to Husky Stadium (or jumping on and off Link) and would directly transfer to the bus headed towards 520. Personally, I think that makes the most sense. It’s a little less convenient for folks heading from Redmond to downtown Seattle, but that is a small price to pay. They will still get there pretty quickly by transferring to Link in the U-District.

  20. Ben Schiendelman says

    I love this too, and would love to help. I think that by building rail services in all the highest frequency corridors (the red ones on your frequency map), we’ll make a lot of the rest of it politically possible.

  21. Aleks says

    David,

    First off, this is fantastic work. I can’t imagine how many hours you put into this proposal. Thank you!

    I have a few minor questions about it, but I want to emphasize that these are all minor. The network that you’ve proposed, completely unmodified, would be a huge leap forward for Seattle transit, and I will enthusiastically give my support towards doing anything that would help make it happen.

    With that, here are my minor questions/comments:

    - It seems like your proposal assumes that there is full service on U-Link and North Link. Do you have any thoughts on how to phase in these changes, in advance of those services coming online?

    - As with prior proposals, it appears that Virginia Mason is losing direct service from downtown. Did you have any thoughts about how to mitigate this issue? I still wonder if it wouldn’t make sense to run the 12 (in your proposal) via Madison and Spring, crossing over at Boren, to at least provide front-door VM service in one direction.

    - Metro has already committed to decoupling the C and the D. How important do you think it is to maintain this coupling? How much more would it cost if these services were not through-routed?

    - Looking at the 15 and 40, I can’t help but notice two things. First, around 85th, it looks like the routes play musical chairs; second, the 40 takes a nasty-looking detour right around Northgate. I’m curious about a slight variation: [a] the 40 uses the same terminus as the 24/71; [b] the 15 is extended to Northgate; and [c] the 16, rather than the 40 (or 15), continues to Lake City. Did you consider an approach like this?

    - How come the 34/35 continue to use 12th/14th Ave S, rather than staying on Golf Dr/15th Ave S?

    - I noticed that several of your proposals involve stringing new trolley wire, and you’ve also proposed to create some new cross-town routes to replace radial ones (e.g. the 71 and the 35). Is there a reason that you haven’t proposed the same for the 44? The Seattle Transit Master Plan tentatively proposes extending the 44 to Magnuson Park along Sand Point Way; I think that would fit in very well with the theme of your plan, especially because it would remove buses from a congested street (15th Ave NE).

    - It’s been pointed out to me that the 50′s many deviations represent a significant time penalty for its riders compared to a straighter route. For example, imagine a version of the 50 that stayed straight on Admiral Way to the bridge, and that skipped the SODO deviation entirely. Did you consider a straighter version of this route? Thanks to the excellent grid that you’re proposing to create, it seems like a straightened-out 50 would work great. It would still be easy to get to from downtown (it intersects with plenty of downtown bus routes), and from West Seattle (it crosses the 54 and the 55), but it would provide a much straighter trip for people accessing it directly.

    • Mike Orr says

      Could alternative service delivery (e.g, an inexpensive First Hill taxi fleet) fill in the gap for Virginia Mason patients?

    • says

      I suppose one reason to leave the 34/35 on 12th/14th is the trolley wire that is there.

      It’s also got a school and a large public housing building. 15th doesn’t have much.

    • David Lawson says

      It seems like your proposal assumes that there is full service on U-Link and North Link. Do you have any thoughts on how to phase in these changes, in advance of those services coming online?

      U-Link is essential. Without it, we can’t restructure Capitol Hill in any way that makes sense, because without it we can’t delete the 43.

      As for North Link, you could still make some of the changes, but you’d have fewer hours available to boost frequency because you’d spend a bunch of hours running the crosstown buses in my proposal down to the UW.

      As with prior proposals, it appears that Virginia Mason is losing direct service from downtown. Did you have any thoughts about how to mitigate this issue?

      The new 7 routing is offered in partial mitigation, but I don’t see any way to serve VM’s front door without creating slow deviations (as with your 12 suggestion) or splitting consolidated corridors back into infrequent spaghetti. The walk from the 12 on either 9th or Boren is relatively flat, and should be accessible to most customers.

      Metro has already committed to decoupling the C and the D. How important do you think it is to maintain this coupling? How much more would it cost if these services were not through-routed?

      To my knowledge, Metro hasn’t committed to that. It was just a legislative proposal.

      As I mention in the Q&A, I’d rather have 10-minute service on both routes, rather than breaking the through-route — and the 10-minute service is cheaper than breaking the through-route and retaining 15-minute service.

      In the long run, I might suggest extending the 15/D to Lake City via Northgate, terminating the 40 at Whittier Heights, expanding the 40 to 10-minute frequency, through-routing the 40 with the 54/C, and leaving the 15/D on its own — but that would require quite a few more buses.

      and [c] the 16, rather than the 40 (or 15), continues to Lake City. Did you consider an approach like this?

      One thing I really wanted for this restructure was a true crosstown route from Northgate Station. The 16 couldn’t fulfill that function.

      How come the 34/35 continue to use 12th/14th Ave S, rather than staying on Golf Dr/15th Ave S?

      Large numbers of riders (many seniors) in the area of 12th/14th Aves S. This area is very dense and drives most of the extremely high ridership between Beacon Hill Station and the ID.

      Is there a reason that you haven’t proposed the same for the 44?

      For right now, I’m having the 31 do this, for several reasons. The biggest is cost; I’d need more buses to provide 10-minute frequency down the 45th viaduct, and I don’t think ridership would warrant it. The others are reliability (the 31 is far more reliable than the 44) and preserving the heavily used Wallingford/UW Hospital ride.

      [50 is too indirect.]

      The trouble is that the two deviations you mention each serve an essential function. The Alaska Junction deviation has worked very well in practice; Alki ridership, anecdotally, seems as high as it’s ever been, and many of the riders are either going to the Junction or using it to transfer. Using Admiral would take away a lot of connections. The Sodo deviation is necessary because, with the deletion of the old 22, there was no longer any other Junction/Sodo connection, and no other way to get there without going all the way downtown. Ridership using these deviations is far higher than ridership actually traveling from Columbia City to West Seattle.

      I do eliminate one deviation on the 50: the Genesee deviation. It’s served instead by the new 52, which fits it better.

      • Aleks says

        That all makes sense. Thanks!

        The 44 conversation is happening on a different thread, so I won’t recreate it here. :)

        The new 7 routing is offered in partial mitigation, but I don’t see any way to serve VM’s front door without creating slow deviations (as with your 12 suggestion) or splitting consolidated corridors back into infrequent spaghetti. The walk from the 12 on either 9th or Boren is relatively flat, and should be accessible to most customers.

        I’m probably missing something, but I don’t see why a Madison/Spring couplet instead of a Madison/Marion couplet would need to be either slow or a deviation. You get rid of three left turns (the live-loop), and replace them with three right ones. In exchange, you add one protected left turn, from Boren onto Madison.

        The bigger issue for me is that I think a hospital is one of the few places where “two flat blocks away” doesn’t cut it. There are plenty of people going to VM for whom that really will be too much of a walk. And it’s not like we’re talking about the VA hospital; VM probably has the most pedestrian-friendly hospital campus in Washington state. (Though it’s a shame that the Jones Pavilion’s facade against Boren is so hostile to the street — even an entrance there could have made a huge difference.)

        I don’t think the answer here needs to be a radical redesign. I just think that VM needs some sort of front-door coverage service to one of its many doors.

      • says

        Spring doesn’t go through at Boren.

        If Virginia Mason wanted front door service in the long term it shouldn’t have built a blank wall facing Boren with its expansion.

  22. shotsix says

    I’m curious why things are routed down E. Marginal Way instead of Airport Way? The latter is almost an expressway from Georgetown to the Boeing Access Road, and E. Marginal Way bypasses the a lot of the airport’s commercial activities.

    • Brett says

      There isn’t much to serve along Airport way. Yes, UPS has a facility and there is the King County Airport terminal building…meh.

      Along E. Marginal there are multiple Boeing facilities, the museum of flight, Aviation High School and a few industrial sites. The number of people who work on E. Marginal is at least an order of magnitude (maybe 2 orders) larger than the number along Airport Way.

  23. shotsix says

    Ha! I didn’t realize that’s the way things are organized now. For some reason I thought there was a route along Airport Way. Oh well, I guess I should have looked at Google Maps before commenting.

    • David Lawson says

      There used to be peak-only service along Airport Way. Ridership was low and it was restructured out of existence.

      The East Marginal routing is a historical artifact of Boeing’s former footprint at Boeing Field. Right now it’s mostly a way to get 124 riders quickly from Georgetown to north Tukwila. It’s marginally more direct than Airport Way.

  24. Gordon Werner says

    I have no problem with the 3 running up Yesler to get to Harborview … the only issue I have with this is the disappearance of the RT 4S … Getting from Downtown to Harborview already means standing room only a lot of the time when there are two VERY frequent bus lines passing it (3 & 4) … reducing that will only cause problems. And other buses are only a few blocks away doesn’t cut it when talking about access to the Public “County” Hospital. the 3/4 are the only bus lines where I see drivers refuse wheelchairs due to over-demand already …

    Same goes for the removal of the 60. One of it’s busiest stops is at 9th/Jefferson as it is the only bus route from the south that goes to Harborview (I am purposely ignoring the 9, as it only operates on non-holiday weekdays.)

    The only thing I don’t understand … is why you kept some routes with their current rt number and others are completely changed. Is your methodology that if a route remains mostly the same it keeps its current rt #?

    (note … this is all personal observation … not analysis of published boardings/alightings)

    • Aleks says

      David’s proposal doesn’t remove any service from the Harborview segment. It’s just that the buses turn back at 21st, rather than continuing south along the 4S’s tail.

      That said, I agree with you that I’m nervous about the amount of hospital service we’re potentially losing. The 7 is adding service to Boren, but I’m not sure that’s good enough. I think it’s essential to provide some sort of front-door service to Virginia Mason, and to provide as much service to Harborview as possible; a frequent 9th Ave route would help with both.

    • David Lawson says

      the only issue I have with this is the disappearance of the RT 4S

      As mentioned, frequency between downtown and Harborview would be the same as it is now. 7.5 minutes off-peak, roughly 6 minutes peak (when route 1 trips would be through-routed with additional short-turn 3 trips). It would also be possible to run artics on some of the short-turn trips, assuming that the new artic trolleys are cleared to go up and down the counterbalance.

      Same goes for the removal of the 60.

      The 7 would pass by only two blocks away.

      The only thing I don’t understand … is why you kept some routes with their current rt number and others are completely changed.

      You pretty much got it… if the route is recognizable as a successor to or variant of a current route, I kept the number, but if it was radically different, I changed it.

  25. Charles says

    I think one possible adjunct to a situation where significant chunks of the city no longer have direct bus service instead requiring people to hike over significant grades to reach bus service, is for the city to allow and encourage rideshare car services to operate in the void and perhaps consider city subsidized fares to reach those lines.

  26. Al S. says

    North Link to Northgate is there but Eastlink is missing including the station at 23rd Avenue. You might want to correct the map.

      • David Lawson says

        Al, the only part that would change after 2023 would be the far north end. East Link would improve the utility of the new 7 but wouldn’t trigger any changes to my proposed south-end network. It’s just a question of what to put on the map.

      • aw says

        This proposal looks to me like a good starting point for evolutionary changes that could adapt to changes in the rail system. Also, as some other commenters have pinted out, the evolution toward something like this could happen sooner in some areas where Link is already operating.

      • Aleks says

        East Link would have a tiny impact on this network. The only new Seattle service is from the Rainier Ave freeway stop to downtown. That would make the proposed 7 more useful, but otherwise, I can’t see anything that would make sense to change.

      • d.p. says

        The best outcome of the advent of East Link would be that the ten-minute 7 and 48 would provide excellent connecting service for cross-lake commutes to and from points throughout the Rainier Valley, the Central District, and First Hill.

        To a much lesser extent, East Link will improve the downtown transfer for those between Judkins Park and about Walker Street.

        But as I’ve long said, and as Aleks says above, East Link is really quite irrelevant to 99% of intra-Seattle journeys.

  27. Mike Orr says

    David Lawson, thank you from the bottom of my heart. I wish STB had a Pulitzer prize so I could nominate you for it. This is so many changes at once I can’t comment on specific routes yet, but it shows the right direction. It’s something we can go to Metro with and say, “We want something like this, or something at least as good as it.” It’s also something people can rally around like the Seattle Subway vision. I was particularly intrigued by your innovations in Magnolia and Leschi, ideas which I’ve never seen before. It’s sad that Metro is currently headed toward a major cut that would gut the frequent network and make increased transfers unviable, but hopefully that can be forestalled and reversed over this decade.

    • David Lawson says

      Thank you for the kind words. Creating a pro-frequent-transit coalition and voice would be the best outcome of this exercise.

      Magnolia will be one of the first neighborhoods I spotlight in a separate post. It took me a lot of puzzling, but I think I’ve got a solution there (with some peak changes as well) that would give everyone what they want.

      • TransitVs.Climate says

        FYI, re Magnolia: W Emerson and 36th is a serious grade and so is W McGraw and 29th. Not insurmountable, but probably not ideal.

  28. Jeremy says

    Great ideas, awesome map. One suggestion I would have is to connect the 48 to the 71 so that the 48 as it is now would be fragmented into 2, not 3 transfers.

    • David Lawson says

      Thank you!

      There are few through riders headed from west of Roosevelt to south of UW Hospital; most 48 through riders, who are headed to and from either UW Hospital, UW, or Roosevelt HS, would probably be better served by transferring to Link than to an extended 48 stuck in traffic.

  29. mic says

    Mr. Lawson: Rarely have I seen such attention to detail you have laid out in an easy to understand concept – even to the point of calculating layovers, cycle times and number of buses needed to keep the system filled. We all too often skip the grunt work and go right for the map and colored pencils to make a case. Your work is exceptionally ‘Credible’.
    With that said, it would be easy to slip into my favorite pet peeve or tweak on a particular route, but that’s not where the genius of your plan and writing shines. It’s in the very fact that Metro and King County apparently have it in their power to make significant improvements to the quality of our transit system while dealing with an impossible funding gap.
    YOU GIVE US ALL HOPE THAT TRANSIT IS FIXABLE.
    That’s the message this post should leave us with. Let’s fix this damn thing, and here’s a pretty good idea to start the conversation with.

  30. Bruce Nourish says

    Of course, I love the ideas behind this.

    A few specific things:

    * Your extension of the 55 would require a new or upgraded off-street turnaround and draw possibly-insuperable NIMBY objections from the people in North Admiral. There are also weird county politics regarding the Water Taxi Shuttles, which are paid for by KCFD. The additional connectivity probably isn’t worth it.

    * The extension of the 2 misses the only real pocket of density down by the lake, which is near the tail of the current 27, while mostly extending coverage along a tiny, upscale, not-dense strip of Madrona. It’s not worth it, and it’s especially not worth fighting about trolley wire down there.

    * The extension of the 10 similarly is likely to draw more neighborhood opposition than it could possibly by worth.

    * The new loop on the top end of the 59 also isn’t worth it. We had a bus like that, namely the 51: it was deleted because it had Will-Douglassian levels of ridership; i.e. nil.

    * I’d go back to the drawing board with Magnolia. West Viewmont isn’t worth the sacrifice of the directness of the 24, and you’re still not providing good coverage of the much denser but steeply sloping east side of Magnolia. I don’t think there’s a good way to serve Magnolia with less than two radial routes, unless and until there’s real rapid transit to someplace nearby.

    * Lots of commenters on this blog have a thing for the “all Madison” 12, but it’s a terrible idea. It doesn’t connect well to a light rail station or any other major transit hub other than Colman Dock, which is useless to most city residents, and it’s not worth the effort of extending trolley wire out there. The smart way to serve the east part of Madison is with the 8; turn the 12 back at 23rd.

    * I’d put the 16 through Fremont and the 28 on Aurora. Aurora doesn’t need that much service, and the improved connectivity for Wallingford is a net benefit.

    * Don’t bother sending the 13 up to the zoo, just loop it around on 36th St and Fremont Pl, counter-clockwise.

    * Reconnect the south end of the 50 to Othello. That connection multiplies the utility of the route — Othello is a real place, and the Link connection effectively almost doubles the frequency to downtown for a person out by Seward park.

    Great work, though. I look forward to seeing our system evolve along these lines.

    • Bruce Nourish says

      One more thing: your 16 in northeast Wallingford probably won’t work. You’re missing some of the density by Green Lake, and still not providing a realistic level of coverage to the southeast side. A better solution there is something like what was proposed by Metro: connect today’s 30 to today’s 26 and give people a good connection to Link at Brooklyn; keep the 16 on its current alignment. This would also bring back coverage of NE 55th St, which, again is probably required for this to be realistic. Of course, we’re not quite in budget-neutral territory any more.

      • GuyOnBeaconHill says

        Bruce, In the 1980s part of my job entailed planning transportation routes for school kids using Metro’s route structure. You don’t have to tell me how inefficient bus service was in the 1980s.

    • GuyOnBeaconHill says

      The 13 would be very useful if it headed north from Fremont and then east to the U District on the 44 wire. The 13 and 44 could combine to provide 7.5 minute service in Wallingford to the UD.

      • Bruce Nourish says

        I don’t mean to be mean, GoBH, but you are a fountain of terrible ideas. Did you design Metro routes in the 80′s? This is exactly the kind of “hey let’s just turn that bus around and send it out there” network design we need to get rid of.

        We already have a frequent bus that goes from Greenwood to the U-District and the center of Wallingford — the 44. We have a frequent bus from Fremont to lower Wallingford and the U-District — the 31. We would have a frequent bus from Fremont to Greenwood. What you are proposing would degrade the network with duplicative and confusing routes.

    • Aleks says

      I agree with Bruce about the 12, and about sending the 8 up to Madison Park.

      If we’re spending any money to extend the 10, I’d be curious about extending it north along Boston St to meet the 35. Probably not even close to worth it (in terms of money or neighborhood opposition), but I think the network effects are enough of a plus to make it worth considering.

      If the 16 goes through Fremont, I would vote to put it on Dexter, and move the 5 to Westlake. That lets you treat the 5/40 as a true common corridor. Otherwise, I bet you’ll have a heck of a time actually getting the 5/40 schedules to line up in a meaningful way.

      Also, I admit that I’m never a fan of intersections where every bus route turns. I love that the 5 finally goes straight, and I worry that the proposed 16/31 just creates another version of the same problem. I’d rather keep the 31 on its current routing, and have the 16 follow 35th between Fremont Ave and Wallingford Ave.

      Regarding Bruce’s suggestion, I understand the point about coverage for SE Wallingford, but I also admit that I’m not crazy about how we currently run a bus on what’s effectively a pair of 1-lane streets. I don’t think the loss of coverage here is any worse than the loss of coverage from deleting the 2N; in both cases, alternate service is about 1/4 mile away (at worst).

      As an aside, I wonder if it would be possible to put a 358 stop at 41st and Aurora, right where the pedestrian bridge is. Seems like that wouldn’t be much more complicated than building a shelter…

      • Bruce Nourish says

        I’m not sure the 2N deletion is realistic, giving my last experience on that rodeo. One idea shopped around internally at Metro was the following relative to today’s real-world network:

        * Extend the 3/4 to SPU.
        * Make the 13 every 15 minutes.
        * Make the 1, 2N every 30 minutes during the weekday, stagger them with the 13 for 7.5 minute service in Uptown.
        * Evenings, weekends: 1, 2N combined operate as a loop shuttle from Uptown. Riders transfer from the 13; timed transfer when the 13 is infreqent.

        This was approximately budget neutral versus today’s service.

      • Aleks says

        You may be right. Even so, the gap between the 44 and 31 is no wider and no steeper than the gap between the 10 and the 48, or between the 3 and the 14, or between the 4 and the 13, or between most of the N-S routes in North Seattle. It seems like the extra service hours needed for a Latona/Thackeray route would be better spent making the 31 more frequent.

      • J. Reddoch says

        Bruce,
        I was wondering about the possibility of extending the 1 along the 2 routing, then left on Queen Anne Ave to the current 3N terminal.

    • David Lawson says

      Bruce, thanks for the comments, which are very perceptive indeed. You are the first to pick up on some of the smaller details.

      Your extension of the 55 would require a new or upgraded off-street turnaround and draw possibly-insuperable NIMBY objections from the people in North Admiral. There are also weird county politics regarding the Water Taxi Shuttles, which are paid for by KCFD. The additional connectivity probably isn’t worth it.

      Physically, the upgraded terminal would require only a bit of extra pavement in the current parking lot. I’ve scheduled the new 35′ coaches for the route. Of course, shoreline development and county politics could make that a lot harder than it sounds. Having talked to some residents in the Admiral and Junction areas, I think they would love a new direct frequent connection to the water, and there would be a substantial constituency fighting for it if it were seriously proposed.

      The extension of the 2 misses the only real pocket of density down by the lake, which is near the tail of the current 27, while mostly extending coverage along a tiny, upscale, not-dense strip of Madrona. It’s not worth it, and it’s especially not worth fighting about trolley wire down there.

      Have another look; I think you may have the wrong idea of where the terminal of my extended 2 is. It’s smack at the south end of that little pocket of multifamily density. South of where I put the terminal, density drops off.

      I wouldn’t have bothered anyway, except that I was able to extend the route without adding another bus. It’s 4 minutes in each direction (which is generous) from the current terminal. I agree with you that the trolley wire would be a fight, but this is about proposing new ideas.

      * The extension of the 10 similarly is likely to draw more neighborhood opposition than it could possibly by worth.

      It would also have a lot of neighborhood support. The neighbors on the existing loop (of whom there are more than there are neighbors on Olin Pl) *hate* the buses, and for good reason — Grandview Pl just isn’t wide enough for bus service.

      The new loop on the top end of the 59 also isn’t worth it. We had a bus like that, namely the 51: it was deleted because it had Will-Douglassian levels of ridership; i.e. nil.

      The 51 had low ridership, but not zero (it was higher than ridership on the 50 south of Seward Park). In the long run, I’d hope the extension of the 59 would do better thanks to being two-way (it’s not a loop), and would allow restructuring of the 57, which is not really optimal.

      I’d go back to the drawing board with Magnolia. West Viewmont isn’t worth the sacrifice of the directness of the 24, and you’re still not providing good coverage of the much denser but steeply sloping east side of Magnolia. I don’t think there’s a good way to serve Magnolia with less than two radial routes, unless and until there’s real rapid transit to someplace nearby.

      I think perhaps I haven’t labeled the map correctly enough in western and central Magnola. Only the 31 goes to West Viewmont, and the 24 is rather direct (Magnolia Village -> 28th -> Gilman -> Emerson -> Ballard Bridge).

      On the east side, the 33 would still be there at peak, and Manor-area residents riding off-peak could either walk up to the 24 on 28th, or down to the 24 at Emerson and Gilman. Either walk is short, and there is the ability to walk downhill both ways. I think the most challenging part of this is asking southeast Magnolia residents to be satisfied with the 31, or a very steep walk up to 28th, as their only off-peak service. There are few enough off-peak riders there that I’m still comfortable doing it.

      Lots of commenters on this blog have a thing for the “all Madison” 12, but it’s a terrible idea. It doesn’t connect well to a light rail station or any other major transit hub other than Colman Dock, which is useless to most city residents, and it’s not worth the effort of extending trolley wire out there. The smart way to serve the east part of Madison is with the 8; turn the 12 back at 23rd

      This would be easy enough to implement (it would have no extra cost compared to what I’ve proposed), but I still haven’t gotten comfortable with it. I’d like to see more debate on this point, and I think I’ll make a separate post about it at some point.

      I’d put the 16 through Fremont and the 28 on Aurora. Aurora doesn’t need that much service, and the improved connectivity for Wallingford is a net benefit.

      Putting the 28 on Aurora without a big infusion of hours would be hard when I’ve turned it into a shuttle. ;)

      Don’t bother sending the 13 up to the zoo, just loop it around on 36th St and Fremont Pl, counter-clockwise

      A very good idea. If I can save a bus by doing it, I’ll do it. I didn’t think of 36th as a potential turnback.

      Reconnect the south end of the 50 to Othello. That connection multiplies the utility of the route — Othello is a real place, and the Link connection effectively almost doubles the frequency to downtown for a person out by Seward park.

      In theory, it’s a great connection. The problem is that for some reason — which I honestly don’t understand — no one rides it. Taking it away saved me 2 buses, which were deployed to the 52/55.

      One more thing: your 16 in northeast Wallingford probably won’t work. You’re missing some of the density by Green Lake, and still not providing a realistic level of coverage to the southeast side.

      I’m not sure I agree. The Green Lake density pocket is no more than 3 blocks’ walk from my proposed routing, and the benefits to getting buses out of the Kirkwood/Woodlawn rabbit warren are substantial. The southeast edge is more of a concern, but I think the good connections to Brooklyn on the 31 and 44 will help with it. I’d like to avoid bringing back the 26 (except in peak express form) because it’s so close to other service.

      55th (actually, really, Ravenna and University Heights) is a big hole and one I’d like to address with a 31 extension. If only Santa Claus would give me 4 buses…

      • Bruce Nourish says

        Regarding the 2 and 10 extensions: I’m much more interested in restructures that enable new connectivity between high-performing routes, or better access to urban centers, than I am in extending the outer tendrils of the Seattle network a little further out, and this is especially true when expensive fixed infrastructure is involved.

        Nobody’s going to sell their car because the 10 runs a few blocks further north, but a few blocks of wire to connect Union to Pine around 12th Ave could facilitate 8 minutes service all day. A few blocks of wire to Fremont would radically improve the way Upper and North Queen Anne connects to the rest of the city in a way that only a billion-dollar subway or new Ship Canal crossing could, in turn, radically improve upon. Those are the changes that might allow some people to sell their cars, and those are my first, second and third priority.

        It’s also extremely difficult to get new layover in residential areas. Metro policy more-or-less requires the consent of the adjacent property owner, and no single-family homeowner wants a bus parked on the other side of their fence.

        (I’m sure you know all this and we actually agree on the underlying issues).

        I mistook your 28 idea, so yeah, that makes sense.

        I plan to discuss Madison and the 8-11 in some of my own posts, so I look forward to that debate.

        I expressed myself poorly on Magnolia. Here’s what I should have said:

        I don’t think the contortions you’re going through to try and serve both a little bit of West Magnolia and a little bit of east Magnolia are worthwhile. I think we should just chuck West Magnolia under a bus (off-peak) and focus on the rest of it. There are two pockets of density in Magnolia: along Government Way to the north, and the Village at the south. The wiggling around on the 24 does not serve either of them well.

        I think we have to eat the reduced frequency and do the restructure Metro proposed last fall, before West Magnolia people got upset about their maids’ commute: keep the 24 on 34th from the Village, terminate the 31 at the Village (as today), and have the 33 terminate in a loop that serves east Magnolia.

      • Kevin R says

        First – great job. It’s clear you’ve looked at this comprehensively.

        My only concern about the 16 would be the loss of service on Stone between 40th and 45th. That’s a major density target for the neighborhood and there are tons of new housing units as well as uses that have a lot of all day transit demand (mental health facility, retirement home, etc.). I recognize the challenge with that is then you’re making the service too far away from folks in the Southeast quadrant of the neighborhood but that’s why some version of the 26 (paired with the 30 as Metro proposed in the past) might make sense once the U District Link Station opens.

        I would definitely favor keeping it on the construction reroute (no Seattle Center) and staying on Aurora.

      • David Lawson says

        Bruce, I’m not sure keeping the 24 on 34th would be significantly more direct than my route on 28th. The jaggies up at the north end are not as bad as they look in terms of travel time. And if we reduced frequency to half-hourly in order to reconstitute the all-day 33, the 24 and 33 would redundantly cover Gilman — running in opposite directions and on a schedule with only rough coordination, a situation that was horrible in South Park, is horrible in North City (even though I didn’t change it for now), and would be horrible here.

        I think Magnolia Village has real potential as an urban center and I think getting it a 15-minute route warrants a slightly (1-2 blocks) longer walk for a few off-peak riders in the east, especially since that walk would be to a frequent service corridor. I’m not sure why you think my proposed route doesn’t serve the Village well — it pretty much treats the Village as first priority, serving it straight off the Magnolia Bridge without any deviation. My route to the Village is more direct than the one Metro proposed.

      • Bruce Nourish says

        Both routes would briefly be together on Gilman, and that looks bad on the map, but providing useful service to all the dense areas in Magnolia is more important than how things looks on the map. This is the arrangement I’m talking about:

        http://seattletransitblog.com/wp-content/uploads/2011/10/suggest_network.png

        The 33 would be loop that went clockwise in the AM and counter-clockwise in the PM; there would be no schedule coordination. It provides good downtown service for all the apartments on Thorndyke, coverage for 28th, and allows the 24 to go straight up 34th. It has more immediate practical advantages, too, like only using streets that are already transit-classified.

        I think this proposal has practical and political viability that your current one probably doesn’t have.

      • David Lawson says

        Point taken about the non-transit-classified street (Clise Pl), although I think in the long run making that street transit-classified would be well worth the effort to get more direct service to Magnolia Village and to avoid the Condon Way neighbors (who imposed a 15 mph slow order because buses were vibrating objects in their houses).

        If there is a 33 loop, it should run clockwise at all times, to ensure the riders are walking downhill at all times.

        Personally, if I lived in any dense area of Magnolia except along Thorndyke, I’d prefer my proposal to Metro’s because of the frequency and consistency. But others may disagree.

      • Nathanael says

        Next question: what infrastructure work would it require to make that street transit-classified?

        (Same question applies to all other routes on David’s map which are not currently transit-classified.)

        It’s not clear to me what the actual standards are of Seattle DOT for transit classification, and I’m not sure it’s clear to anyone else either.

  31. alexjonlin says

    One major issue for interlining routes to get that 7-8 minute frequency is that, if they have very different cycle times, you may have to have some routes laying over at the terminus for a long time, creating a lot of waste. From your website, it looks like you may have accounted for this but I can’t totally tell – could you address that? Thanks!

    • David Lawson says

      Yes, I’ve thought this through for all of the common corridors I’ve planned. The trickiest one was the 2/10 corridor, because the routes are both live-looped at one end.

  32. JohnS says

    A very minor point (but hopefully informative): Current 23rd Avenue rebuild thinking has the parallel greenway on 22nd. Assuming this holds true, the greenway route would run along 22nd right in front of the Safeway (to take advantage of the light at 22nd/Madison). Not sure a layover and the greenway and the primary entry to the Safeway parking lot would do well sharing that street space – but would be happy to be proven wrong.

    • David Lawson says

      Thank you; that’s a very good point. There are other potential options for the 6 north-end layover and I will explore some of them.

  33. Ambarish says

    Wonderful work, David; I can’t imagine the number of hours that went into this! Readers much more knowledgeable than I have said lots. I’ll just pick a minor nit; your “how your current service change” page has an entry missing for 32. Towards UDist, I suspect it would involve transferring from the 15 to the 31, which is easy and something I do if I just missed a 32 on 3rd Ave W & Mercer. Towards LQA, though, the transfer from the 31 to the 15 would be terrible, unless some stop relocations and pedestrian improvements are made at Nickerson/15th Ave W.

    • David Lawson says

      Oops! Thank you. I will add an entry.

      Route 32 riders headed all the way between Uptown and UW would do better by heading downtown and getting on Link than by trying to use the former 32 path.
      Riders between Uptown and SPU or Fremont would use the extended 13.
      Riders between Uptown and Interbay could either walk to the 15 or use the 24 and transfer to the 15.

  34. Matt L (aka Angry Transit Nerd) says

    You got picked up by Publicola:

    “The system he proposes would also eliminate some very low-ridership routes, which is pretty much a nonstarter if you believe mobility is a human right.”

    Le sigh.

    • Nathanael says

      Yeah. I find that attitude idiotic in the extreme.

      If mobility is a human right, why don’t I have a government-paid chauffeur picking me up at my front door?

      Uh, no.

      Disabled people, first and foremost, deserve better housing locations. That has nothing to do with “mobility is a human right”.

  35. fare strike forever for youth liberation says

    I drew up an optimal bus network for West Seattle a few weeks ago; it’s certainly one of the most geographically-challenging regions of the city. I would like to compare our maps and present my loose judgment of areas in which either network is preferable. However, I did not operate under the assumptions of revenue-neutrality; my network would almost service require increased service hours. This caveat holds for all of my comments:

    -I like your idea of stitching together the 22 and the Genesee segment of the defunct 51. Expanding coverage is a good thing for a comprehensive network.
    -I like your idea of improving service frequencies to South Park on the 132. Ideally, I would increase frequency on the 131 as well to create a 7-8 minute corridor on 4th Ave S.
    -I agree that the C-Line/54 and the 120 should be priorities for service improvements to 10-minute frequencies, 7 days of the week, ideally, well into the evening.
    -I agree that the North Admiral service should be extended down to Seacrest Park (Water Taxi terminal)
    -I, too, considered speeding up the 50 by eliminating the Genesee deviation, but this makes Alki a 3-seat ride from the areas of Delridge lacking easy access to E-W services.
    -The live-loop of the 52 around the Junction is hideous. A better option for serving this area would be revising the 37 to head up Jacobsen from Beach Drive, then heading north on 49th and east on Alaska to the Junction. Introduce an off-peak 37 loop that heads back to the Junction via Harbor, Avalon, 35th, and Alaska. Time it with water taxi departures and you only need the 773 and 775 shuttles during peak. I realize this is similar to the barely-used 53 of yore, but with better frequencies, this route could attract some riders in very populous areas.
    -I have to say that your 16th Ave SW is an unsightly patchwork of services. There’s a certain logic to making South Seattle CC a hub/terminal, and I see that the 128 is less frequent on the segment that has pedestrian connections to Delridge, but the route structure could be simplified. I prefer Metro’s idea for revising the 128: sending it east from the Junction to Delridge on the same route as the current 50, then assuming the role of 16th Ave service, with a deviation to the on-campus stop at SSCC. This route should have 15-minute service, meaning that the 7-8 minute service frequencies from the Admiral District to the Junction is extended east to Delridge.
    -I would make the Morgan St./Sylvan Way service more of an E-W one. My route starts at the Alaska Junction, travels down to the Morgan Junction, goes east to 16th Ave SW, but then continues on Holden and Highland Park to South Park and Georgetown, thereby doubling South Park-Georgetown service. I would consider a deviation to SSCC.
    -Why did you axe the Thistle segment of the current 22/proposed 59? Your routing of the 22/59 down 35th Ave SW offers no appreciable time savings, and a transfer to the 21 is already available. Direct service to Chief Sealth High School is necessary, considering Seattle Public Schools’ abrogation of the responsibility to provide transportation to high school students. Students are left to rely (and pay for) Metro, and eliminating direct service is adding insult to injury. The alternative service for riders on California north of the Morgan Junction is taking the 54/59 to the south side of Westwood Village and (a) navigating the pedestrian hell of mall parking lots and incomplete sidewalks or (b) taking an indirect path around the mall, then (a+b) climbing a staircase and walking along the side of a sports complex. There is no alternative service for other riders. Either walk is 10+ minutes. School trippers have limited utility because the bell schedule varies by day, and students have to occasionally arrive or leave early or late. I realize that STB is relentlessly complicit in marginalizing transit-dependent youth riders (claiming that all riders can choose to live near better services when routes are eliminated, holding events at 21+ venues, faulting riders for avoiding ORCA when youth ORCAs are nigh inaccessible), but c’mon, try to pretend you care. Then again, I’m not sure I care so much about on-time school attendance.

    • fare strike forever for youth liberation says

      sorry for the block o’ text, adding spaces would have only made it longer tho :(

    • David Lawson says

      I, too, considered speeding up the 50 by eliminating the Genesee deviation, but this makes Alki a 3-seat ride from the areas of Delridge lacking easy access to E-W services.

      The areas of Delridge that are not accessible to either the 52, the 55, or the 59 in my proposal have no appreciable ridership. Any area that can access any of those three routes has a two-seat ride to anywhere west of the Duwamish.

      The live-loop of the 52 around the Junction is hideous. A better option for serving this area would be revising the 37 to head up Jacobsen from Beach Drive, then heading north on 49th and east on Alaska to the Junction. Introduce an off-peak 37 loop that heads back to the Junction via Harbor, Avalon, 35th, and Alaska.

      As you mentioned, the 53 was infrequent, and no one rode it. Making it frequent would mostly duplicate various other routes and cost a lot of money, and I’m still not sure you’d get any ridership. (Also, I’m not sure Jacobsen can handle a 35′ coach — I’d have to research that.) The 52 loop has just one goal: get people up the hill from 49 Ave SW. It would do that frequently and quickly for essentially zero cost compared with laying over the bus at the Junction.

      I have to say that your 16th Ave SW is an unsightly patchwork of services. There’s a certain logic to making South Seattle CC a hub/terminal, and I see that the 128 is less frequent on the segment that has pedestrian connections to Delridge, but the route structure could be simplified. I prefer Metro’s idea for revising the 128: sending it east from the Junction to Delridge on the same route as the current 50, then assuming the role of 16th Ave service, with a deviation to the on-campus stop at SSCC.

      Just to make sure you know (I’m not sure it’s clear on the map), the 52 and 55 are through-routed, so they are really just one route, with different route numbers to avoid total confusion near the Alaska Junction. They serve the E/W market between SSCC and the rest of West Seattle as well as frequent connections to downtown, while the 128 handles the market going south (which is almost totally a different set of riders). I don’t think the 128 south of SSCC justifies 15-minute service yet off-peak. But Metro’s (and your) idea is not bad at all if you can find the funding.

      I would make the Morgan St./Sylvan Way service more of an E-W one. My route starts at the Alaska Junction, travels down to the Morgan Junction, goes east to 16th Ave SW, but then continues on Holden and Highland Park to South Park and Georgetown, thereby doubling South Park-Georgetown service.

      It would be a great connection, but I don’t think it would have much ridership now, given relatively low ridership on the existing east/west part of the 60. For a revenue-neutral proposal I think there are more important needs.

      Why did you axe the Thistle segment of the current 22/proposed 59? Your routing of the 22/59 down 35th Ave SW offers no appreciable time savings, and a transfer to the 21 is already available.

      Because if you retain the Thistle routing it becomes very clumsy to use the 59 to serve Arbor Heights, which really is a long walk from alternative service. Thistle has service within easy walking distance from any point along it. Chief Sealth is only one to three blocks from Delridge, where the 120 runs every 10 minutes. Students headed to the Junction can transfer from the 120 to the 55 or 52; students headed to 35th can transfer to the 21. I don’t feel bad about making students, who are overwhelmingly able to walk, walk three blocks.

      • fare strike forever for youth liberation says

        Thanks for the reply. Fair enough. I think I’m out for the day.

  36. nope says

    There’s nothing in here about special needs riders. You’re proposal involves making people walk or travel further to get to stops, but you haven’t included how you would handle that for the many special needs riders that currently use public transit as their primary means of getting around. I’d like to see quicker commutes as well, and this post presents interesting ideas towards that end, but special needs riders (there are many in Seattle) are going to get hit very hard if you start forcing people to travel further to get to stops or transfer more often.

  37. Peter N. says

    David, this is a terrific post. Thank you.

    Does each route’s cycle time take into account for increased load/unload time? In a network with higher transfer rates, we would expect the time spent at each stop to increase (on average). It wouldn’t be 2x, but it certainly would be higher than 1x.

    • David Lawson says

      I did not specifically account for increased dwell times, but running times (and thus cycle times) are fairly generous. My hope would be that if we can implement this program we can also implement a cash penalty that will speed boarding…

  38. Lack Thereof says

    I feel like the 8/6 split is in the wrong place; you should try and preserve the eastern CD’s 1-seat-ride to the Capitol Hill Link station.

  39. Bruce Nourish says

    Two more things occur to me:

    * I’m surprised you took the 48X path on Banner Way. You’re giving up access to a lot of people and facilities near North Green Lake, and creating a couple of odd situations: people a long way from a station will have better frequency than people who live 3/4 of a mile away (and thus, at 15 minute headways, will probably walk), while the decent number of riders on the north side of the lake, who are on a pretty decent pathway, will see their travel times to Roosevelt get worse than today.

    * I’m also surprised you sent the 31 up to Children’s rather than just straightening the 44. I bet the 44->Link trip to UWMC will be time competitive with the slog on the 44, especially as UWMC has an employees-only tunnel under Montlake Boulevard. That would free up some buses to send the 31 up to 55th.

    • Bruce Nourish says

      Hmm, I guess it’s not quite the 48X pathway, but I still think the time savings aren’t worth the lost coverage of Green Lake. Traffic is terrible on 85th by that freeway exit.

      • Josh F. says

        I would agree–I think that it’s important to have a one-seat ride from NE/E Green Lake to Roosevelt Station. If the 71 is to be a proper grid route, it should probably link all of the developed areas on the way, instead of randomly skipping over some. To compensate, the 78 could terminate at Roosevelt Station.

    • David Lawson says

      I see your point on the 71 path. Neither option is appealing, but I prioritized speed. The current Green Lake 48 routing (or the 16 routing, the other choice) is just so ridiculously slow, particularly trying to fight its way through the 5-way intersection and then up and down Wallingford Ave, that it would have a significant negative impact on Link accessibility for everyone north and west of Green Lake.

      Some access for east Green Lake would be preserved along 5th Ave NE, but it’s not the same as the slow-but-useful 48 routing.

      • d.p. says

        The 16 routing is at least slightly better than the Green Lake Drive N routing, as it avoids the epicenter of the 5-way gridlock-fest, has year-round activity on both sides rather than one, and doesn’t have to wait for the northbound light to cross 80th on Wallingford Ave (that light takes forever).

        The passing-bus standoff on Wallingford only seems to happen between 80th and 82nd. The simplest solution: extend the curbside parking ban (currently on the bottom 1/8 of the block) up to about the halfway point of the block. That should take care of it.

      • Bruce Nourish says

        The 16 routing has serious problems, too. That part of 80th is basically a freeway ramp, and the outbound bus has to turn left against traffic, which can cause multi-minute delays in the peak. SDOT doesn’t want to add a stop sign on 80th, because that would piss off drivers, and that would be terrible. A signal might be possible, but would be expensive.

        I’ve unofficially batted around several ideas for fixing the 48/16 in this area with SDOT staff and Metro planners. There are no solutions that are both cheap and viable; if there were, I would already be raging at Metro and SDOT to implement them.

      • asdf says

        A stop sign on 80th may not even be a net win for transit. It would benefit the northbound 16 at the direct expense of the southbound 16. The best improvement I see would be a stoplight that would ordinarily prioritize thru-traffic on 80th, but have working TSP that would make it change quickly when a northbound 16 bus approaches.

  40. Bruce Nourish says

    One more thing, really the last. Overall, I think the showstopper in your plan (in terms of network design, not logistics or politics) is the 28 shuttle idea. Here’s the daytime ridership on this route:

    http://seattletransitblog.com/wp-content/uploads/2011/10/king_county_metro_28.png

    You’ve got 20 riders/trip in the midday before that bus reaches Fremont, and neither your 40 nor 5 will have enough room for them without a major boost in frequency, above and beyond what those people will have to say about being forced to transfer to a bus that’s no faster than the one they have today.

    People tend to assume the 28 is a coverage route like the 61, and that was true of the former 3rd Ave NW section north of Crown Hill, but the Ballard segment is solid.

    • David Lawson says

      Bruce, I’m genuinely surprised by those numbers; they don’t match my experience driving midday 28s, which was reasonably full buses through extended Fremont (40 territory) and then dropping off a cliff north of that point, with maybe 5 riders headed to 65th, a couple more to 85th, and then a straggler or two the rest of the way.

      Definitely worth a second thought. It would take quite a few buses to run the 28 downtown, but maybe I should look for them. My initial thought would be to run it along Westlake, in a common corridor with the 40. I prefer that to the 16-to-Fremont 28-to-Aurora idea for two reasons: 1) because the 16 already suffers so much slowness trying to snake through Wallingford; and 2) because it would keep the Fremont connection, which in my experience is what a lot of those riders along 8th NW want.

      • David Lawson says

        On second thought, if I’m going to have a 15-minute 28 going all the way downtown, the right solution is to run the 28 along Dexter and through-route it with the 21, while interlining the *5* and 40 and running both of them along Westlake. Time to go look for some buses.

      • Aleks says

        David, I’ve been thinking about the same question. The more I think about it, the more I see the appeal of putting the 28 on Westlake, and keeping the 5 on Dexter. This lets you schedule the 28/40 as a common corridor, with 7.5-minute service all the way up to 8th. Which is pretty amazing, when you think about it. I ride the 40 daily, and I can tell you that there’s tons of demand on that segment of Leary, which the current uncoordinated schedules of the 28 and 40 do a poor job at meeting.

        Making this work would probably require truncating the 40 at Whittier Heights, so that it’s about the same cycle length as the 28 would be. In turn, that requires extending the D/15 to Northgate, and possibly reassigning the Lake City segment to some other route. I don’t think that extension would be prohibitive — if I’m reading your cycle times correctly, your proposed 40 is nearly 45 minutes longer than the *combined* time of the C/D.

        Doing this also allows you to “earn back” some service hours. Right now, the span of the 40 between 85th and the 15′s terminus is “wasted”, in the sense that it adds buses on that stretch of road without adding useful frequency. Those are some extra service hours that can defray the cost of running additional service between downtown and Fremont via Westlake.

      • Bruce Nourish says

        The 16 is indeed painfully windy and slow north of 45th, but keep in mind that its importance as a radial route will dwindle after North Link: Riders between 85th and East Green Lake will go to Roosevelt; riders north of there will go to Northgate. Many riders to downtown on the east side of Wallingford near 45th St will be attracted to Brooklyn station at those times of the day when 45th isn’t completely screwed.

        For that reason, I think it makes sense to focus the 16 more on making connections to the west side of town, rather than working to get it downtown faster. By contrast, North Link does nothing for 28 riders, and putting the 28 on Aurora gives faster trips downtown, admittedly at the expense of only hitting Fremont at the periphery.

      • David Lawson says

        Answer: It would take 5 buses to extend the 28 to downtown, assuming that either it or the 5 terminates downtown without a through-route.

        I can’t save even 1 bus by shortening the 13 to lower Fremont. I could only improve recovery time a bit.

        I could save 1 bus on the 59 by chopping both the Genesee Hill and Upper Rainier Beach tails. That’s a painful coverage-for-ridership trade, particularly in Upper Rainier Beach, but I expect it would be justified by the numbers.

        I could save 1 bus on the 6 if I truncated the Boeing Access Road tail, which would be a lifesaver for a few riders but would in all likelihood have pretty low ridership.

        I could save 1 bus on the 69 if I were to truncate it at or near 145th. That would be painful in the interim between Northgate Link and Lynnwood Link but would be the correct move once Lynnwood Link was open. The problem is that I expected to use that bus in my 347/348 restructure.

        I’d have to look into this a bit more, but I might be able to save 1 bus by interlining the 131 and some 120 coaches downtown. The problem there is an equipment mismatch; the 120 absolutely needs artics, which I don’t really want to send pinballing around the long-cycle 131/132/24 world.

        Beyond that, saving buses gets even more painful. Ideas are welcome.

    • d.p. says

      neither your 40 nor 5 will have enough room for them without a major boost in frequency

      .

      And there’s the elephant in the room.

      The 40 is more than just another median core route. It has, in its brief life, already become a heavy lifter. It should be part of the ten-minute map. It needs to be part of the ten-minute map. And it needs to avoid the massive evening ramp-down in service that is the Achilles’ heal of every transfer-based restructure proposal so far……………

      • David Lawson says

        If the 15/D actually worked, which it would be likely to with my changes, I think it would siphon off some riders from the 40. Also, my plan gives the 40 artics on every single trip.

        I’d love to increase the frequency to 10 minutes but that would require 5 buses…

      • d.p. says

        Well, I’m certainly psyched to see all-day daytime ten minutes on the D plus the death of the Uptown detour. Ballardites would be dancing on that detour’s grave.

        Meanwhile, I’ve just posted a long reply about my most pressing worries — evening network viability; and the implications of “saving” today’s network from the 17% cuts — down at the very bottom of the thread.

        It was the 200th comment. Which is an achievement in and of itself.

    • Eric says

      I am a regular 28 rider. I frequently ride it to Fremont, and ride it downtown a bit less often.

      I’m intrigued by the shuttle idea. It bumps up the off-peak frequency from 30 minutes to 15 minutes. If the 5 and 40 have well-interleaved schedules between downtown and Fremont, and are running on time at a 7.5 minute combined frequency, that means an average transfer wait of only 3.75 minutes. Not too bad.

      That means if you show up for the 28 at a random time, your expected total wait time for a downtown trip will decrease from 15 minutes to 11.25 minutes, and your expected wait time for a Fremont trip will decrease from 15 minutes to 7.5 minutes. Pretty nice!

      I do share Bruce’s concerns about downtown-to-Fremont capacity issues, though. Currently that stretch is served by the 40 on Westlake (15 minute mid-day frequency) and the 26/28 pairing (15 minute mid-day combined frequency). In this proposal the 26/28 would be replaced by the 5, giving the same capacity as the existing system but adding all of the 5′s riders from Phinney/Greenwood into the mix.

      Another potential problem is that the Fremont/34th St bus stop is already quite well-used in both directions. If you make downtown-bound 28 riders transfer in Fremont, and also add direct connections to upper Fremont/Phinney/Greenwood and SPU/Queen Anne, there will be even more people using this stop. Will all these people physically fit on the sidewalk at peak hours with room for pedestrians to pass? It’s worth a look.

      From a travel-time perspective, the proposed 28 shuttle sounds like an improvement despite the forced transfer for downtown-bound passengers. If that’s the price I have to pay for higher-frequency service, I’ll gladly take it. I would want Metro to make sure that the downtown-to-Fremont corridor has sufficient capacity and that the Fremont stop is large enough for this new service pattern before the changes are implemented, but overall it sounds like a promising plan. Nice work!

  41. M says

    Has/does Metro have a route pair that connects on both ends like your proposed 73/75 do? Would it save any buses to have it run in both directions as a loop as opposed to linking them in the U District and then sending them to the same destination?

    • David Lawson says

      I looked at that and there is really no difference whether you run buses in one direction or two. The important part is that you send all or at least most of the buses alternately on both sets of routes (73/75 and 65/67).

  42. Josh F. says

    I have to say that I’m very impressed by this proposal–not only all the work you’ve put into it, but also the fact that it is in fact possible to cover almost the entire city with frequent service without increasing revenue. I think that one way to graphically show how much an improvement this would be is to create before-and-after maps depicting how much of the city is covered by service every 15 minutes or better, kind of like what has been done with Auckland’s network redesign (http://www.humantransit.org/2012/10/auckland-how-network-redesign-can-transform-a-citys-possibilities.html).

    Here are a few random minor suggestions (some of these might not be good, but I’m interested in hearing why):
    -The 7 should be extended down Mercer St to connect with Uptown/Seattle Center. It might be worthwhile to cut a route from 10-minute to 12 or 15 minute frequency to accomplish this.

    -The post-viaduct waterfront might need more service than every 15 minutes (which is inadequate for short distances). Perhaps the 14 could also be rerouted to the waterfront in order to extend the 7.5 minute service corridor? Connecting to a very-frequent Link or 3rd Ave buses into central downtown shouldn’t be too hard.

    -A crosstown bus down NW 65th St (with a detour around the south end of Green Lake) might be a good idea to address the lack of east-west connections in that area. Past Roosevelt Station, this bus could continue onto Ravenna/NE 55th to address the grid gap there. Of course, the problem there is money, but this seems like an important connection once funding is found.

    -I know that this proposal is for North Link only, but when Lynnwood Link opens, would it be a good idea to make the 75 go to 130th St Station and then across to Broadview (instead of the 85)?

    • J. Reddoch says

      Not sure how a 65th Street bus would make it over Phinney Ridge to get from NW 65th Street to Green Lake.

      • Bruce Nourish says

        +1. A potential crosstown on 65th is a subject that’s close to my heart, but the infrastructural obstacles to it are daunting. I could see some ways of making it work, at least between Greenwood and Sunset Hill, but I don’t see it happening in the near future.

      • Jon says

        I asked Metro about this exact issue earlier this year, hoping that they were considering this option. I received the following reply:

        “Thank you for the suggestion and there was a mini-bus route that used that street back in the early 1990s. However, there was very little use of the route. Also, our safety department has become much more conservative with routing buses and is unlikely to approve operation on N 65th Street at the intersection of Phinney Av N. Further complicating matters is our funding situation in which we have no additional resources and may have to cut service up to 17% in 2014.”

        Very disappointing.

  43. Drew Dresman says

    This is absolutely fantastic and wonderful to see a plan where you have considered the full extent of the City. The explanations really show how well thought out this network is. The route changes like keeping the 7 on Boren and the 35 routing are particularly good ides though there would be serious growing pains for people if you made sweeping changes like this in a short period of time. I think this represents the general direction that we are going though it is hard to say how long it will take.

    Many people continue to be confused by the concentration of service to create more frequent networks and believe such proposals are insensitive and ignore transit’s goal of serving the least fortunate, but I think many of the situations people fear will occur when we consolidate service or stops (such as someone who can’t make it an additional two blocks losing access to a bus stop) could be addressed in better ways- ride share, subsidized taxis, van pools, and a variety of other tools could do a better job at serving people with such needs than fixed route transit service. If we want transit service that the majority of the city will use we need to move to a more grid like, frequent system such as this.

    I will also add that investing in new express service in a grid like pattern would really make this a world class system and perhaps make it more palatable for people who generally oppose service changes.

  44. asdf says

    I really like the idea of the 69, something sorely missing from today’s network, excluding two peak-period trips a day on the 243. It would be even better if, unlike the current 372, we could scrounge the service hours to keep this route running on evenings and weekends. One obvious place to look for additional service hours is the fact that on Bothell Way, north of Lake City, the 69 and 81 duplicate the 522. Perhaps if the 522 gets truncated at a Link station and made more frequent, we can get away with having the 522 be the only service on Bothell Way.

    Other opportunities to squeeze some more service hours for all-day service include getting rid of the 243 (riders can use 69->271 instead).

    I would also support a restructuring of the 271 in which the U-district->downtown Bellevue segment is upgraded to every 10 minutes all day, 7-8 minutes peak. To pay for it, we could cut the Bellevue->Eastgate segment of the 271 to every 30 minutes all day, 15 minutes peak, and cut the Eastgate->Issaquah segment to hourly all day, half-hourly peak. The gains would be much appreciated, and the cuts would be barely noticed. Once all the 520 bridge construction is finished, this segment of the 271 should become more reliable, allowing for shorter layover periods that would save even more service hours.

    • says

      I thought about sending the 522 to Northgate on the 40′s routing, but then read upthread where Dave really wanted a straight-through crosstown route in the area.

    • Scott Stidell says

      The 522 should probably terminate at Roosevelt – this routing has been at the core of one of the proposed HCT lines for at least 50 years. If you want to go from Kenmore to Northgate a frequent transfer will be available at Lake City (even better after N 130th station is open). Many more riders will be going from Bothell/Kenmore downtown or UW Bothell–UW than would ever be going from that far NE to Northgate, and shoving that route into the traffic hell that is Northgate would eat up a bunch of service hours.

  45. d.p. says

    It should not come as a surprise to you, David L, or to anyone who has read my years of level-eleven outrage at the deficiencies of the present urban bus network, that I really like this proposal. In its entirety. A lot.

    There are only two outcomes that seem to need more thought: the evening effects of this restructure; and the political path to this outcome.

    1) Evenings and Sundays:

    I won’t lie: As happy as the daytime proposal makes me, it scares me to see the rows of parallel green that fill your final map, threading across many of this city’s most urbanized districts, and to know that any (or most, or all) of those “15 minute” markers will drop to half-hour waits the moment the clock strikes six, plus all day Sunday.

    When you promise a smoother, simpler, more freedom-enabling experience through a transfer-based network, the only way to assuage the Will Douglases of the world is by stalwartly upholding your promise. Those transfers must work. Always. That means no evening abandonments and no Sunday horrors.

    It isn’t enough to vaguely hope “the night network will improve ‘automatically’ to some degree” on account of the simplified route structure. Of course it should! But will it do so enough to prevent the “nightmare transfers” that you’ve worked so hard to rid us of in the day? If not, then it isn’t good enough.

    The threshold between a transfer-based network worth using, and one that fails and infuriates, is tangible and measurable. When Portland TriMet decided to spread service cuts evenly across every route in 2009, it effectively smothered the 27-year success of its gridded network. Transfers are now bad enough that a multi-leg journey across Portland is usually more trouble than it’s worth. At night, you shouldn’t even attempt it.

    The greatest botch of Metro’s last Ballard restructure wasn’t the wasteful, ill-coordinated 61. It was the attempt to convince the public that low-frequency last-mile transfers and barely-there night services were things to crow about. It was the treatment of the 40 — the agency’s newest multi-hub-linking ridership powerhouse — as the bastard child of the restructure, just because it wasn’t painted red. (Frankly, it pains me to see the 40 remain green on your map; you and I know the new 24 will never hold a candle to it.)

    So again, the numbers must be crunched, the evening/Sunday frequencies must be deduced and plotted. If most routes will be reduced to the point where the transfers fail, then your network plan isn’t yet finished.

    2) The Politics:

    You and I want the same thing: for a frequent, all-hours network like the one you’ve devised to become the new-normal expectation for intra-urban movement. We want the stubborn one-seaters to be proven wrong through experience. We want the remaining detour- and quirk-defenders at Metro to see the light. We want every future expansion and revision to treat frequency and legibility as baselines, not as ancillary benefits when they happen to get snuck past the special interests.

    So how do we get there?

    It seems you would fight for a full restoration of the threatened 17% funding — knowing full well that Metro has promised no changes to the network for the forseeable future. Then, sometime around 2021, you expect a collective epiphany to will your (objectively better, in every possible way) network into existence.

    I have far less faith in that outcome. Metro’s rush to put all of its funding-advocacy eggs in the “save the status quo!!!!” basket scares immensely.

    If the 17% cut goes through, Metro has promised to do the exact opposite of what TriMet did. Metro won’t spread the cuts evenly, but will embark on a massive and immediate restructure. And that restructure might look more like yours in 2014 than an emergency-funded network would on the cusp of 2021.

    As you well know (and as discussed above), an 83% version of your network is risky business. Make the wrong cuts, and your transfers fail across the board. But cut carefully — maybe the 28 shuttle disappears for a year; better yet, cannibalize some middling unidirectional peak runs — and the rest of the new network can tantalize Seattle with the possibilities of what could emerge post-crisis.

    How do you reconcile your admirable design efforts with your support for Metro’s advocacy of the precise opposite?

    • Mike Orr says

      Again, this is not purporting to be an ideal network. It’s the most that can be achieved in a revenue-neutral way. Only a very few corridors currently have frequent evening/Sunday service, so it’s not like you’re taking something away from somebody that they have now. Consolidating and straightening routes will only improve frequency. Not to the level that meets people’s needs and maximizes ridership, but better than the status quo.

    • Mike Orr says

      To be clear, I mean that these kind of consolidations can only improve evening frequency. Not 15-minutes until 10pm across the board, but maybe on one route till 8pm and another route till 9pm instead of both routes dropping to half-hourly at 7pm. That would make a material difference for people doing after-work errands and attending evening events.

      • d.p. says

        The difference, Mike, is that as infrequent (and for the majority, offputting) as our evening service is today, those who do use it take some solace having a one-seat ride all the way home from some well-trodden destination such as downtown. They might hate having to dilly-dally when they’d rather be on their way to bed, but once their bus finally comes, it actually takes them there.

        If their trip home starts to require transfers at random points in Little Saigon or the U District, and perhaps makes them wait for two should-be-frequent-but-now-it’s-9pm-so-oops-they’re-not buses, and no longer allows them to dawdle in a warm and safe space… that’s probably a dealbreaker.

        One of the concessions of a transfer-based network is that the minimum headway at which transfers work becomes a dealbreaker for riders if violated. That dealbraker minimum is significantly shorter than the dealbreaker one-seat headway. You can’t just ignore that fact after a certain hour.

      • Mike Orr says

        The only way to get from here to a robust network is to build the core of it. We can’t just do nothing until we can make it fully frequent everywhere 24 hours. Transfer locations are important, and the U-District is a good transfer location. Conversely, 23rd & John is a bad transfer location because there’s nothing there, and I’m concerned about how well the 8+48 will really substitute for the 43. But most transfer locations are in neighborhood centers, and there the center also has partial responsibility, because it should have a wide enough variety of destinations to make a transfer less unpleasant, and sometimes quite convenient because you can get something else done that you were going to do sometime anyway. The U-District, Fremont, and Ballard succeed in this. Some other places don’t currently, but that just means they’re another challenge to work on.

      • David Lawson says

        Mike, for what it’s worth, I’d expect many more riders to transfer from either the 8 or the 48 to Link than to transfer between the 8 and 48. That said, anywhere you have two 10-minute lines meeting there should be upgraded passenger facilities.

      • adam says

        The big ticket transfer here will be between the 8 and the 12, not the 8 and the 48. Riders coming up from Madison Valley and Madison Park will change here to go across town on Denny, and vice-versa.

      • adam says

        And yes, the best place to transfer between the two is down the hill on Madison at 28th, but the last place to transfer to the westbound 8 is at 23rd. I believe its completely possible that the first or last opportunity to connect can sometimes trump the best possible place to do so.

    • David Lawson says

      d.p., I agree with much of what you write. And I particularly agree that the evening situation is important. Eventually I will come back with an update about what would be possible in the evening. My back-of-the-envelope read is that my 10-minute daytime routes could all be 15-minute evening routes; the question is how many of the 15-minute daytime routes can remain 15 minutes in the evening, and how many have to go to 30 minutes. I’m hopeful that no 15-minute daytime routes would have to be hourly. (When you’re down to 30-minute headways, you can help things a bit by identifying the most used transfers involving 30-minute routes and timing them. The old, pre-2004 night network had some brilliant timed transfers on half-hourly routes.)

      I have more faith than you do in Metro’s ability — admittedly, over a glacial span of time — to move the network in the right direction without forced radical surgery. They do few restructures, but most of the ones they do are either good (120, central Bellevue, Kent) or put down network bones that could be good with more frequency (last fall’s restructure). Forced radical surgery that imposes a transfer-based network without the frequency for transfers will just make people all over the network feel like people in Ballard did immediately after RR D started… in other words, “Metro sucks.” They won’t be seeing future potential, they’ll just be angry, and less receptive to anything but the restoration of the old network.

  46. says

    When I read this, my first instinct was to write “You’re kind of late to the party. Martin and Zach already proposed similar revamps for the Rainier Valley and Capitol Hill respectively, and Metro caught a ton of flack just for moving the 2 to Madison (though admittedly a move to Pike/Pine would get it a lot closer to a Link station).”

    Specific points not covered upthread:

    I see a lot of seemingly random terminations; the 8/6 and 13 were covered upthread, but for example, why are you truncating the 65 to 85th-ish (deviating from 35th at 75th) and having the 78 (which I believe has the same frequency) inherit the remainder? I see that you’re having the 78 serve a similar purpose to today’s 330, but conversely, how much ridership will the segment on 75th see, really? (Perhaps that segment could inherit a version of the 71′s present-day View Ridge routing?) Someone mentioned making the 522 the only Bothell Way route and truncating the 69; how viable would having the 69 inherit the 330 routing be in that case?

    I see you’re keeping the joke stub of a route that is the 47 intact, only improving it by running up and down 1st.

    Why does the 71′s terminus loop go on the south side of 85th?

    • David Lawson says

      Addressing your points in order:

      Yes, this looks a lot like Zach’s Capitol Hill restructure in particular, although, oddly, I came up with most of the network before I ever saw his proposal, which came out before I became a regular STB reader. Gridded frequent service implies a certain type of result up there.

      The 65: The changeover between the 65 and 78 occurs roughly where it becomes more advantageous to transfer to Link than to ride directly onto campus. I originally had the 65 going all the way to Jackson Park, and the 78 being just a stub route to connect Roosevelt to Lake City, but then realized that actual north-end 65 riders would be better served by a fast and frequent Link connection than a straight N/S route.

      The 75th segment of the 78 will attract a good deal of ridership it doesn’t today, because it gets to Link so quickly. It will also pick up most of the existing riders from both the 68 and 72 in the north Ravenna area.

      The current 71′s View Ridge routing has almost no off-peak riders, particularly in the area that would be too far to walk from either my 65 or my 71. Serving it off-peak would be a waste of hours. A truncated 76 could presumably continue to serve it during peak.

      Having the 69 take over the current 78 routing would work OK, but I’m reluctant because I like having it on Lake City all the way to 145th, and the Lake City -> 145th left turn is a delay nightmare which I want to avoid.

      The 47 will no longer be a joke if you cancel the 43 and make the 47 frequent. Ridership between the Summit area and downtown is very heavy, but most of it is on the 43 now because the 47 is so infrequent. And the area is so dense — the single densest neighborhood in the city — that if you build a frequent bus riders will come.

      • adam says

        I like the idea to kill off the 43 at the expense of better 6/8 and 48 service and a high-quality connection point at the MadCap triangle (23rd/John/Madison) with the 6/8 and 12 (present day 11).

        I live near this intersection, and I have my own set of objections to the way the stops are currently organized between the 8, 11, 43, and 48 even without this idea. Alas.

      • says

        So, you want the 69 to stay on Lake City all the way to 145th but you don’t want a bus that stays on 35th all the way? I thought you were trying to create a gridded, transfer-based network.

        You’re misunderstanding why I call the 47 a joke; making it more frequent will make it MORE of a joke, not less. Okay, maybe a little less because it’s so infrequent today, but a frequent route that goes less than a mile away from downtown just seems laughable to me. I know there aren’t much in the way of options for extending it, and sending it down 1st gives it a bit more reason for existence, but it just looks funny to me. (It was more justifiable when it was the north end of the 14, so finding a through-route partner for it could help, but that has its own problems.)

      • Aleks says

        Honestly, I think the 47 is ridiculous for one reason: it runs down a street with traffic circles!

        Either the traffic circles need to go, or the route needs to change (to use Bellevue Ave in both directions).

      • Aleks says

        It was more justifiable when it was the north end of the 14, so finding a through-route partner for it could help, but that has its own problems.

        Here’s an interesting and probably terrible idea:

        - Restore the 7′s routing to downtown, along the corridor currently assigned to the 14. Through-route it with the 47.
        - Send the 14 up Boren, along the corridor currently assigned to the 7. Continue through-routing it with the 70.
        - In the south, cut off the 7 at Mount Baker, and reassign the southern portion to the 48.

        This should save some amount of service hours, since you’re deleting the layovers/turnarounds for the 7 and the 47, and deleting the downtown segment of the 70. It also has the side-effect of boosting frequency on the 47 — but since that segment is so short, it shouldn’t really cost very much, if anything.

        By swapping the tails of the 7 and the 48, you keep the 7/47 to a manageable size, and you avoid letting Boren/downtown traffic interfere with the southern segment of Rainier.

      • David Lawson says

        Aleks, that’s an interesting set of ideas, but probably impractical for a few reasons. (In case you missed my reply above, though, I did adopt your 44 to Children’s! Check out the latest map.)

        - The 7 needs artics. Full stop. But the 47 routing can’t handle them.
        - Until we get a streetcar extension, we need a bus along the full length of Jackson. That’s a major travel corridor, and will only get more major with the deletion of the 27. There’s a reason why I’ve assigned all artics to the 14/70 — requiring me to chop the Mt. Baker tail.
        - The highest-ridership portion of the 70 is between downtown and SLU. Something needs to connect Fairview with downtown and Link.

      • Aleks says

        I’m glad that my 44 argument stood up to scrutiny! :)

        You make very good points about capacity and artics.

        I still have to admit that the 7 and 70 (on your map) _feel_ like they should be connected. I admit that I don’t have data to back up this claim. But let’s say you’re standing at Fairview and Harrison. You can walk north to the streetcar, or west to the 5/40, or south to the 8 (which will take you to Link), or you can take the 7/70 north or south (either of which will take you to Link). So there are still plenty of ways to get to downtown and Link.

        In return, you get to cheaply boost frequency on the 70, because the unique segment (i.e. north from Mercer) has relatively less traffic than the section that the 70 already shares with the 7.

        At any rate, it doesn’t seem any worse than diverting the 7 from downtown.

      • Nathanael says

        “- The 7 needs artics. Full stop. But the 47 routing can’t handle them.”
        Double-talls? Or is this a case where the many-door-boarding of artics is needed?

  47. mic says

    Kudos David: Well over 200 responses and not one [ot] or [adHom] needed.
    Amazing, when the conversation is so well focused on making transit better for everyone

    • David Lawson says

      mic, I noticed that too, and I’m very grateful for it. The conversation my post sparked was incredible, and not a single post needed moderation. Thank you, readers.

  48. Martin H. Duke says

    As everyone else has said, great work. A few comments about my neck of the woods in the Southeast.

    - I love the new 6, although frankly I’d rather have the 48 back on MLK. There are much more interesting destinations on 23rd than MLK in the Central District. That said, splitting it and extending it to Boeing Access Road fix many, many problems vs. the status quo.

    - This is really worth a post of its own, and I’ve done one before, but it would be great to kill the 50′s tour de Sodo to make it a worthwhile East-West connector.

    - I’m somewhat flummoxed by the 59. It seems like David has used it to clean up a bunch of unserved areas (the Prentice Loop, the south end of Beacon, Arbor Heights, the south end of California, etc.), but the result is a horribly indirect milk run. In particular, Rainier Beach to South Park going the long way around Boeing field seems especially punishing. It’s really two routes split at Georgetown, and it’s frustrating that it’s not a true east/west connector. That said, I don’t see how you serve all those neighborhoods and have a direct route without spending more money.

    The real way to do it is to get from Rainier Beach to South Park via Boeing Access Rd and SR599, but then you have to do something else to connect South Park and Georgetown, as well as the South end of beacon. Perhaps this could at least cover MLK south of Henderson and re-truncate the 6 to somehow serve these corridors?

    -

    • mic says

      Or at least shorten the Sodo tour-de-farce to Lander only before heading south again. The E3 busway combined with Link at Sodo is a good frequent transfer to many places.

    • David Lawson says

      Martin, there are all kinds of interesting ideas in my head about how to improve the 59 situation, but they all cost money. Ultimately, the truth about what I’ve proposed is that it says “give frequent and direct service to all the neighborhoods with high ridership, and make the coverage route for low-ridership neighborhoods as good as reasonably possible.” The only area that has the numbers right now to justify better service than the 59 offers is South Beacon Hill. And with a few extra dollars, I could fix it. As you can see from the posts above, though, dollars are already stretched.

      The trouble with straightening the 50 is that, right now, it’s attracting a lot more Sodo commuters than through riders. Realistically, I think the Sodo market remains bigger, and a Junction-Sodo connection in particular is still needed.

      • Martin H. Duke says

        As my comment suggests, I’m not sure you could do better without spending money, although making the 50 direct would save money while still providing a decent transfer opportunity to Sodo via the 21.

        I would like to the see the stop-level 50 data, but it wouldn’t surprise me if it attracted more Sodo riders than crosstown riders, as that’s what it’s good at doing!

        There are three possible east-west corridors from West Seattle to Southeast Seattle: the West Seattle bridge, Georgetown/1st Avenue Bridge, and Boeing Access Road to South Park. It would be nice if at least one of those was direct. The 60 (using the South Park bridge) is direct but because it’s diagonal it takes a long time. In your plan there are two east-west paths but they’re both terrible. The cheap option that induces a non-horrible transfer is a direct 50. But this is a post I’ve been meaning to write when the 50 stop-level data arrives.

      • GuyOnBeaconHill says

        Another sip from the fountain of bad ideas….

        The Prentice loop had lots of riders back in the days when the only other service to the Rainier View/Skyway area was via hourly legs on the old 42. In those days, the 7 offered the most frequent and fastest trip to downtown. These days the 106 and 107 cover the legs formerly served by the 42 and continue to Renton and connect to Link. Predictably, ridership on the Prentice loop has fallen dramatically. I once suggested that the Prentice loop would be best served as an extension of the old 34/39 bus, but that ship has sailed. The loop around Rainier Beach still needs to be served, but the ridership that is generated by the 7 on the hill above Rainier Avenue is very, very small. A 106 running every 15 minutes would likely cannibalize the last of the Prentice loop’s ridership.

      • David Lawson says

        GoBH, that’s not a bad idea at all. I debated whether to keep the Prentice loop, and still might delete it from a revised proposal. An issue is that deleting the Prentice loop, alone, wouldn’t save me a bus on the 59 — I’d also have to delete the Genesee Hill loop in West Seattle to get enough time savings. Both of them have just enough ridership that it’s painful to do that. In the case of the Prentice loop, the part that matters is actually far away from Prentice — it’s the bit of 57 Ave S/Waters Ave S immediately up the hill from Rainier Beach. That’s where the remaining riders are, it’s far away from the 106, and it’s a bit of a steep hike from Rainier.

      • David Lawson says

        Martin, if you straightened the 50, how would you connect Alki and Admiral riders with Link? They already lost their one-seat service to downtown (and were the biggest group of users to do so), and one of the things they were supposed to get in return was a good connection to Link. If you kept the 50 on the freeway, they would have to ride all the way to Columbia City, or transfer to the 21 and have a three-seat ride — obviously, neither is ideal.

      • GuyOnBeaconHill says

        You might have another problem in Rainier Beach. The 106 appears to be the only bus to cover the Rainier Avenue loop, but how will the 106 cover both directions? Currently the 7 and 8 cover the RB loop in different directions–how will the 106 cover the loop in both directions?

      • says

        If you look at the post he linked, the 50 would get off just long enough for downtown-bound riders to transfer to a bus headed north. That should also take care of Sodo-bound riders. Beacon Hill and Mount Baker stations would be a single transfer away as well.

      • Martin H. Duke says

        Unfortunately, the redo of the Spokane St. viaduct may very well have foreclosed the possibility of easy interface with the busway, unless the 50 were to use the low bridge into West Seattle. The mapping sites are useless to verify this, so I’ll have to see for myself the next time I’m down there. Score another own-goal for SDOT.

        David, I’m really not sure what to make of your question about Link interface. From West Seattle, if I’m heading to the Rainier Valley or points south on Link, then clearly the connection at Columbia City is adequate or superior. If I’m headed downtown, that’s what RapidRide is for. And if I’m going to Capitol Hill and beyond, or to the Eastside, a RapidRide connection to the DSTT is perfectly adequate.

        More philosophically, the whole idea of the transfer-based system is direct, detour-free routes that rely on frequent transfers to vary off the path. It’s clear that our horrible road grid forces deviations from that plan, but there are ample alternatives to get to Sodo, particularly if we can figure out a way to use Spokane Street and the infinity buses on the busway.

  49. Brett says

    As a north Beacon Hill resident, I would feel utterly spoiled by this restructure. My commute to Renton becomes much easier with LINK -> 106 with the increased 106 frequency and reliability (thanks to RBS origin). I currently avoid the 106. My trip to Capitol Hill is also improved by the 35 versus the meandering 60 I have now.

    I understand this is meant to be implemented in 2021, but as a South Seattle resident, I’d like to see the changes South of the ship canal implemented sooner. Particularly the 106 :-)

  50. says

    Good work! Your plan is excellent… it make so much sense, it will never happen, sadly.

    King County Metro Transit still operates under the 1950s-60s premise that public transportation is for the underclass, so things like convenience, ease of use and service frequency simply don’t matter. There is simply no good reason to run buses to, say, North Bend, but they do it. Likewise, routes with one or two runs per day.

    This is in stark contrast to Portland’s TriMet, which easy to figure out, easy to use, and offers nice little touches like rear doors on buses that require no driver intervention — when the light’s green, push the handle, and the door opens! They had a synthetic voice calling out stops for YEARS before Metro ever even thought of it. The people who run TriMet know full well their primary responsibility is to get people from Point A to Point B as quickly and efficiently as possible. Metro still seems to be running mobile homeless shelters…

    • asdf says

      I have ridden the bus to North Bend multiple times. In spite of being hourly, it does carry a good 10-15 people or so per run. It’s great way to hike Mt. Si and access the Iron Horse Trail. Without the 209, the only options left to reach north bend are a 2-hour, very hilly, bike ride, or a car.

  51. Mike Orr says

    David, you may want to do the next logical step of prioritizing the unmet needs into groups, so that we can get an idea of “X more service hours costing $X1 would fill in gaps A, B, and C; and Y more service hours beyond that will fill in gaps D, E, and F; and ultimately a complete network would require Z hours costing $Z1.” That would help to ease some people’s concerns about the gaps, and also give a rallying point for groups like Seattle Subway that want to campaign for more service hours. (I’ve got a potential name for such a group too: “Better Metro”.)

  52. FDW says

    I know I’m a bit late to the party, but why didn’t you consider having the Streetcar on 1st ave operating? With that the segment of the 47 on 1st become redundant, and it could be shifted over to the spring/seneca couplet. (which would probably serve to bring a ton of people on board that might otherwise be reluctant)

    • David Lawson says

      The streetcar on 1st would be just fine with me. At that point I’d find something else for the 47 to do.

      I’d be reluctant to put it on Spring/Seneca because that would require turns on and off 3rd Avenue on both ends. Now adding it to Madison/Marion for even more frequent short-turn service on the 12… now we’re talking.

  53. mic says

    With such a ground swell of support for Metro to significantly change it’s core route structure, it would seem to follow that they will begin to embrace this concept through some serious community meetings with stakeholders – an all day brown bag, with breakout sessions by neighborhood (or something like that). They won’t do it by themselves, BUT
    Several Council members from both King Co. and City of Seattle could make this a reality.
    Leaders with a vision of saving Metro can sign up below (flak jackets will be issued):
    1.__________
    2. _________
    3. _________
    On behalf of transit riders, employees, and citizens, we thank you!

  54. threeandfour says

    Looks like effectively for me this service change would probably result in an additional 30-40 minutes added to my commute out of Wallingford to South SLU (all the additional time being on foot). The upside I guess is it that you make it quicker and more reliable to go to a bunch of places I never want to go to. Not much of a bargain for me. I’d have to have a lot of reasons to go to random neighborhoods on a daily basis to make up for the daily commute cost. Bummer.

  55. Limes says

    This post, for me, epitomizes the disconnect between transit wonks and the lived experience. (It’s part of the problem of a self-selecting audience of cheerleaders.) Everyday riders’ opinions are put down for the elite here know what’s better for them.

  56. Nathanael says

    Here’s a question.

    *Given this restructure*, would “closing the gap” on the wires on the #48 pay for itself (in better acceleration, electricity being cheaper than diesel, trolleybuses lasting longer than diesel buses, etc.)?

  57. sally says

    Montlake to downtown requires a transfer to an every 15 minute bus or a sojourn through the Montlake bridge. The irrational hatred of the 43 continues. But we will get great service on lots of routes with low demand (like the bus in front of my place on Summit).

    • William C says

      Fortunately, this plan is predicated on the construction of Link from Northgate to downtown through Montlake every seven minutes!

    • David Lawson says

      Montlake to downtown typically would be a short trip north on the 48 (every 10 minutes) followed by a transfer to Link, which takes 6 minutes between Husky Stadium and Westlake. The Montlake Bridge might occasionally create a few minutes’ delay when it’s open, but it’s really not bad when it’s closed.

      That trip will almost always be far, far faster than riding all the way from Montlake to downtown on the 43, usually just short of half an hour. The reason to hate the 43 is because it may be the single slowest route, on average, in the entire Metro system. (It shares that honor with the 2S, another route that gets radically changed in my proposal.)

      • sally says

        From Montlake (if things go well) it will be a 5+ minute wait for the 48, a 5+ minute ride to campus, a 5 minute walk to the station, a 3 minute walk to the platform, a 3 minute wait for the train plus a 6 minute ride to Westlake. Currently with no schedule it is a 7 minute wait for the bus and 20-25 downtown. So if all goes well then it may be about the same as the 43 you hate so much. But rarely does all go well when I am trying to get across the Montlake bridge.

      • David Lawson says

        A 5 minute ride across the bridge? More like 2-3, unless the bridge is up.

        And a 5 minute walk to get across the street from the stop to the station? Even given the less-than-ideal design of the station area, that is unrealistic. You could walk all the way from the Stevens Way stops to the station in 5 minutes.

        I think the worst case with the bridge down is that someone in north Montlake will save 5 minutes vs. the 43. The best case (later in the day, when the 43 takes half an hour) is that that person will save 15 minutes. With the bridge up and in the worst case, they’ll break even. (And don’t forget that if the bridge is up, it will also be delaying that southbound 43.)

      • sally says

        Apparently you have never crossed the Montlake bridge northbound in a.m. rush hour or walked from the stop to the IMA. The connection to the southbound 48 will be even worse. But keep repeating to yourself “it is just six minutes to downtown” and the crappy connections between Metro and Link will magically disappear.

  58. TransitVs.Climate says

    David,

    Why does the 31 not use the Emerson street entrance to Magnolia (like the 24), thus providing the majority of magnolia’s density (between Emerson and Dravus) with all day service and still giving a 20th and Dravus stop for Interbay?

Trackbacks

Sign in or create an account to save your credentials and make commenting faster.



You may want to read our comment policy.