Metro Connects Concepts

141 Replies to “Metro Connects Concepts”

1. A Boren Ave bus is a no-brainer, and I quite like the idea of a bus using Aloha, Roy and Belmont down to SLU as an alternative to the 8.

   1. I studied all these routes the first time around but haven’t been able to get through them again yet. So we need to figure out what has changed since 2020. This reminds me that the original maps had an MLK route on Boren like this (1074). I’m thinking it was only in the final, so maybe it has been advanced. But it wasn’t in the RapidRide G restructure as I expected, and neither was the 8-Madison Park (1061), nor the Broadway north-south route (1064).

      Aloha Street has never had service in my memory. The previous map had two Lakeview-Aloha-Garfield HS routes, one like this one (3208 to West Queen Anne), the other I think to Magnolia. I thought they were two of the weakest routes because how many people want to go to Aloha Street, so they’d be the first to be deleted in a recession. But maybe this indicates Aloha is stronger than I thought. This mapping interface doesn’t show multiple routes on one one street as well as the previous one did, so it’s easier to miss some routes, but I don’t think there’s a second one on Aloha now.

      The previous Interim split the 8 and deleted service on northern MLK, then the final re-added MLK and attached it to another route going south, so that it terminated at Madison & MLK in the middle of nowhere. That looked like it would depress ridership and kill the route. This Interim (3997) keeps MLK and extends it to 15th (Kaiser Permanente), making it a stronger concept. I wonder if it should be extended further to Broadway and Capitol Hill Station, rather than gratuitously terminating six blocks short of them.

      Other changes on Capitol Hill:

      This has the 10-Pine (1213). The 10 was on Pine Street until U-Link, when it was moved to John. The RapidRide G restructure intends to move it back to Pine. So that’s reflected here.

      And the 2-Pine-12th-Union route is gone. It looks like the 2 will remain on Seneca-Union long-term (1068). That will bother some people who think it’s too close to the G on Madison (1059). But it will help disabled people going to Virginia Mason, for whom a 2-block longer walk to Madison would be difficult (like my relative).

      1. But it wasn’t in the RapidRide G restructure as I expected, and neither was the 8-Madison Park (1061), nor the Broadway north-south route (1064).

         True, because I think they largely ignore these proposals.

         To be fair, some of it was lack of money. Hard to propose a new route when the old routes are running infrequently. I get that. But one of the “no-brainers” was having the 2 dogleg to Pike/Pine. So from Madrona, head west on Union, as it does now. Then go north on 14th to Pine. Then follow Pine like the current 11 (to downtown). This increases coverage in the area, which means instead of competing with the G (and losing badly, since the G will be a lot more frequent) it would complement it.

         The changes for the RapidRide G were way too timid. It is much easier to keep the routes the same, and then just shrug and say “if only we had more money” when few take transit. Hopefully they fix the routes some day.

      2. But it will help disabled people going to Virginia Mason, for whom a 2-block longer walk to Madison would be difficult (like my relative).

         That is what access vans are for. Seriously. Don’t get me wrong, it would have been nice if RapidRide G had a stop on 9th (between Madison and Spring) but it really isn’t that bad of a walk (https://maps.app.goo.gl/35bY1oot8GXNrbaC8).

         It just doesn’t make sense to run buses so close to each other unless all the buses are running very frequently (as in every two minutes). Otherwise you end up with what we have: buses running infrequently. If people wonder why buses will run every half hour on Capitol Hill (Capitol Hill!) or we don’t have a “direct, all-day, bus between First Hill and South Lake Union” it is because we are afraid to move the buses a couple blocks, and create a better network.

   2. The lack of a direct, all-day, bus between First Hill and South Lake Union is ridiculous, and the Boren Ave. bus should have been created years ago. Maybe the lack of it made sense back in the 1990’s when South Lake Union was just a dump of abandoned warehouses, but when the area boomed since then, the route should have been created.

      Yes, it is not possible to have a direct route between every possible neighborhood pair across the entire city (it would require too many routes, which would spread service too thin). But, when you’re talking about two huge neighborhoods one mile apart from both each other and the city center, there is no excuse.

      That said, the bus network today is still a lot better than it used to be. Until 10 years ago, there was no all day bus between Fremont and Ballard. And I was told that Metro staff were deeply skeptical about both the 8 and the 48 at the time these routes were created – they thought that because the bus didn’t go downtown, nobody would ride it. Those people were proved wrong. The impact of a Boren bus would be similar.

      1. Yes, it is not possible to have a direct route between every possible neighborhood pair across the entire city (it would require too many routes, which would spread service too thin). But, when you’re talking about two huge neighborhoods one mile apart from both each other and the city center, there is no excuse.

         Furthermore, it is a grid! Boren is parallel 3rd, our main transit corridor. If it wasn’t for the stupid freeway, you could run buses on 3rd, 7th and Boren for an ideal grid. The area is essentially downtown now. There are skyscrapers on First Hill as well as South Lake Union. Without a doubt they should be connected.

         I have a much bigger problems with buses like the 348 (after the Lynnwood Link restructure). The bus will head down 15th which becomes Roosevelt, heading straight towards the Roosevelt Link Station and the U-District. Then it will abruptly turn, thrusting itself into traffic. Why? Because they want to connect that corridor directly with Northgate? OK, but why not connect that corridor directly with Maple Leaf, Roosevelt and the UW? That would actually save money. It would enhance the grid, instead of going against it. The 348 is basically an example of “connect all the neighborhoods” but in a very inefficient way.

2. Does anyone understand the reasoning behind giving local routes four-digit numbers? Are these just “planning” identifiers, or are people going to be saying, “Take the 4218 to Aurora and then get on the E northbound”.

   Why?

   1. They’re planning numbers. Four-digit numbers can’t be confused with existing routes or other proposals, and the first digit indicates the type. A couple routes have a different color than their digit, which may mean they were originally conceived as one type but are now being considered for another.

      1. Thanks, Mike. That’s better.

3. On the 2050 map, I noticed the Route 40 (identified as a RapidRide corridor) reroutes onto 14th Ave NW to Market St rather than continuing along Leary Ave to Market. I assume this is in anticipation of Ballard Link terminating near Market and 15th?

   I also wonder what improvements, if any, would be made along 14th Ave NW and Market, especially seeing as the city is moving forwards with the Route 40 TPMC project along Leary between Market and 15th Ave NW, which isn’t utilized in the 2050 map.

   1. > I also wonder what improvements, if any, would be made along 14th Ave NW and Market, especially seeing as the city is moving forwards with the Route 40 TPMC project along Leary between Market and 15th Ave NW, which isn’t utilized in the 2050 map.

      Well the city doesn’t know for sure where the link station will go and it’ll be 15 years away. But anyways:

      14th Ave NW: most likely some greenway/ or something similar to their plans for thomas street in slu.

      > This project will connect people using the Ballard Link light rail station to the Burke Gilman Trail, nearby neighborhood greenways and protected
      bike lanes, and schools. The project will make streets and public spaces around NW Market St safer and more enjoyable for people walking, rolling, biking, and visiting local businesses.

      NW Market: Probably adding or extending the bus lanes on Market St. Beyond that normal rapidride improvements of sidewalks etc…

      > Repaving and redesigning some areas of the street to better support transit and freight vehicles, including potential bus- and freight- only lanes
      > Repairing sidewalks and improving street
      crossings. Planting new trees

      From the draft transportation plan
      https://www.seattle.gov/documents/Departments/SDOT/STP/Aug_Draft/STP_ProjectsPacket_DRAFT.pdf

   2. The 40 was one of the half-dozen RapidRide corridors in Move Seattle. It was later downgraded to “some street improvements” due to budget limitations. Those are about to start construction on Westlake, Fremont, and Leary, which some neighbors are complaining about. It could still get upgraded to RapidRide again later, and it’s one of the top ten routes in the city so it certainly qualifies. So Metro is probably just leaving it “potential” as its fortunes go up and down, since it might be revived in the future.

   3. I noticed the Route 40 (identified as a RapidRide corridor) reroutes onto 14th Ave NW to Market St rather than continuing along Leary Ave to Market.

      Good catch. Yes, it definitely is designed to get close to a future station, that it assumes is at 14th. It is a great rendering that shows how bad the Ballard Station placement is. Notice it creates a transit hole in Old Ballard, one of the busiest place in the area. In contrast, if the station was at 20th NW, the bus would only have to change a little bit (if at all).

4. > For those who remember the 2020 version of these maps, we need to identify what has changed since then.

   They still have the pdf’s of the older plans (2025 map and 2040 map). I’ll call it 2040 for the original plans

   ## Express routes Changed

   2021 (2040 plan): Kent Station, Alaskan Junction, Burien TC
   => it’s been split in half at burien:
   2021 N (2050): Admiral District, Burien.
   2021 S (2050): Burien, SeaTac, Des Moines, Kent. (Note: Kent-KDM Link connector.)

   Honestly most of them all seem the same.

   ## no changes
   2003 (2050): Westwood Village, Fauntleroy, WSJ, 99 tunnel, SLU.
   2012 (Interim): Mercer Island, Issaquah Highlands, North Bend.
   2016 (2050): Seattle, Burien.
   2020 (2050): Auburn, Maple Valley, Snoqualmie.
   2022 (Interim): Renton, Issaquah.
   2028 (2050): Auburn, Enumclaw.
   2203 (2050): Redmond, Duvall.
   2204 (Interim): Bothell, Duvall.
   2205 (2050): Redmond, Fall City, North Bend.
   2206 (Interim): Mercer Island, Issaquah, Sammamish, Redmond.
   2207 (Interim): Seattle, Federal Way.
   2402 (Interim): Seattle, Kent, Auburn.
   2515 (Interim): Seattle, Roosevelt, Lake City, Bothell. (2050: Extend to Woodinville.)
   2516 (Interim): Smith Cove, SLU, 405, Totem Lake.

   https://kingcounty.gov/~/media/depts/metro/about/planning/metro-connects/appendices-011717.pdf

   I’ll write about differences in rapidrides/frequent ones in a different comment.

   1. It is hard to remember, but Mike and I both mentioned one already: The 2. From Madrona, head west on Union, as it does now. Then go north on 14th to Pine. Then follow Pine like the current 11 (to downtown).

      A lot of the changes are just variations. For example I think the Boren bus was different (but I’m not sure). I remember the 49 going from the UW to Beacon Hill (via Broadway) but I don’t remember it going to Othello. I don’t remember the 1061 (the replacement for the 8) going all the way to Nickerson (and then ending abruptly). That is weird, really (I’m not sure if someone forgot to finish their drawing).

      I’ll admit, it is tough to remember. It was a while ago, and I confuse it with variations that folks have come up with (including me).

5. Comparing rapidrides, For the 2025 versus new “Interim map” generally most of these routes are still suggested in the 2040 but slight rearrangement in priorities.

   ## interim map changes
   * 1049 Kent (route 150) proposed earlier
   * 1056 federal way to des moines via green river college: some slight deviations to the route, but overall the same
   * 1064 u district to capitol hill (via 10th) to beacon hill and finally othello. Originally a 2040 plan, but most prominently recently talked about in the seattle transportation plan
   * 1028 B line restructure to merge with 271. on the older plan only went from bellevue to crossroads mall.

   ## 2050 map
   I already mentioned some of it previously but I’ll repeat it here
   * Both 1047 (rapidride A to federal way) and 1049 (route 150 to kent) would end at rainier beach station
   * West Seattle it seems they’re having a hard time deciding, in the original 2040 plan the 1041 (H line) went to downtown Seattle, while the 1043 the C line went to alki beach.
   * In the new 2050 plan they’ve swapped it with the H line heading to alki beach. While the C line starts from alaskan junction and ends at burien on a slightly different route
   *1075 (105/106) route truncation from renton to rainier beach still in the plan, but no longer a rapidride. dropped in priority.
   * 1515 (route 183+901) from kent to twinlakes has been elevated in priority
   * 1999 ( redmond to overlake to eastgate)
   * kirkland they are having a really hard time choosing how to route the busses whether to go to UW, or bellevue, or slu. It seems to switch every iteration of the plan

6. Conceptually speaking, I believe that we will have autonomous bus capabilities by 2050 that will have a notable place in service delivery. They may not be flying down the freeway, but I could see them being used as slower but more frequent feeder routes.

   To me, this should be presented as a potential different network scenario.
   It’s not particularly forward looking when the only 2050 scenario demonstrates a bus layout is based on a continuation of a 20th century service model.

   1. I agree autonomous busses will exist by then. However, the problem is we are unsure of how capable they are and King County Metro should rightly just plan with what they know for now.

      > They may not be flying down the freeway, but I could see them being used as slower but more frequent feeder routes

      I mean it’s hard to tell. Perhaps it’s the other way around and only freeway hot lanes have autonomous busses since it’s mostly pretty easy to contain, while all the feeder ones are done by humans. After all, I was pretty surprised AI ended up being used for creative tools first before say automating construction etc.. It’s just not realistic to have king county metro a transit agency to guess the ai developments 30 years down the line.

      1. I agree, it is really hard to tell. At this point though, it appears that buses have the most trouble around pedestrians. So freeway running may be relatively easy. The tricky part is the transition.

         A bit off topic, but I was thinking about Brisbane the other day. They have a big busway. They will expand it. Originally they were going to convert to trains, but will stick with buses. This will save capital, but presumably cost more in operations. Until, of course they automate it. It may be that the buses struggle with the “open” part, and do just fine with the “BRT” part. This would mean something like this: A regular driver starts a couple miles away from the busway. They then pick up riders in a normal manner until they get to a busway station. Then they get off, press a few buttons, and the bus “drives itself” the rest of the way. Going the other way it is simple. A bus arrives and a driver is there, ready to get on. She taps her special card to access the driver seat and puts in her code. The seat and mirrors automatically adjusts to her height. She does one final check and away they go. The whole process takes less than thirty seconds. A longer than usual stop, but better than a lot of stops. This seems quite plausible to me. We have automated trains. The buses know how to steer. It seems like this wouldn’t be too difficult.

         How that would work for a system like ours would be a bit more of a challenge. It seems especially well suited for longer trips. I expect this to happen before anything else, starting with trucking. A truck is taking goods from Spokane to Seattle. It gets all the way to the truck stop in North Bend and stops there. Then a driver gets on and takes it the rest of the way. I could see buses between cities operating the same way.

   2. I don’t think self-driving busses are likely by 2050, at least not on a large scale. We can likely automate some aspects of bus driving (freeway driving, for example) and maybe some very basic routes but I doubt we will be able to move away from one driver per bus on a large scale. The technology is just really far away. Take a look at how far autonomous vehicles have progressed in the two decades (Waymo, Cruise, Tesla, etc), and keep in mind that autonomous driving will need to be very mature before it starts getting picked up by transit agencies

      I do think it’s very reasonable to automate the at-grade sections of Link in the coming decades

      1. “I doubt we will be able to move away from one driver per bus on a large scale. The technology is just really far away. Take a look at how far autonomous vehicles have progressed in the two decades (Waymo, Cruise, Tesla, etc)”

         Waymo has had cars self driving without anyone in the vehicle since 2017. They have been transporting passengers on a limited basis since 2019, done so without safety drivers since 2020, and began offering fully driverless rides in San Francisco almost two years ago. The technology is not really far away it is basically here. Programming a bus to run a set route over and over (with the possibility of a remote command center stepping in to intervene if needed) is less difficult that programming a car to run in SF traffic.

         It isn’t reasonable to upgrade Link to run autonomously at grade in the coming decades, it should already be doing so, programming a system to run on rails, with no need to turn, should be orders of magnitude simpler than what Waymo has been doing for half a decade.

         A transit agency that isn’t planning for a fully autonomous bus network (with smaller vehicles coming much more frequently, and perhaps some level of demand response, i.e. if you are going to the UW, the bus with a rider going to Wallingford passes you buy, and instead you are picked up by one with a passenger going to the U Dist) is an agency that is planning for its own obsolescence instead.

      2. While automating Link trains may be technologically possible in theory, there are numerous obstacles to doing it in practice. Public transit agencies have to be much more risk averse than private taxi companies because it’s a lot more people impacted if anything goes wrong. Opposition from driver unions is another problem. Then, there’s the 3rd problem that hooking up a computer to the train’s control system is probably not a configuration supported by the manufacturer, which means mucking with the trains at their own risk and voiding warranties.

         Plus, I don’t think reduced operating costs through labor savings is even something that Sound Transit really cares about. Whether the train runs every 3 minutes or every 10, the bragging rights from having built it is the same.

      3. As far as I can tell, Waymo is the closest, but they’re still (at minimum) years away from mass deployment. Driverless cars have had a pretty big setback recently; Cruise got suspended from operating in California and San Francisco opened a lawsuit against driverless cars.

         Automated buses aren’t going to be financially feasible until and unless driverless technology is very mature. Automating a bus is less difficult than automating a car – but who is going to do it, who is going to pay for the research, and how much will the final bus cost? It doesn’t matter if it’s technically possible if it’s prohibitively expensive to deploy.

         And yeah as @asdf2 brings up, it’s not going to be possible to automate existing trains. A manufacturer would need to step in and sell reasonably priced trains with the requisite technology and Sound Transit would need to procure a new set of trains. The technology seems to be coming in the near-ish future (automating an at-grade train is way, way easier than automating a bus), but it still needs to be proven/affordable and line up with the procurement process

      4. since the articles been out a couple days i’ll comment some more on automated busses hehe

         > Automated buses aren’t going to be financially feasible until and unless driverless technology is very mature. Automating a bus is less difficult than automating a car – but who is going to do it, who is going to pay for the research, and how much will the final bus cost?

         I think it’ll start out on two opposite sides of bus types then slowly meet in the middle. Aka there’s
         slow shuttles, collector/coverage busses, regular bus routes, express freeway busses, and express freeway with dedicated (hov) lanes busses.

         Slow shuttles are already starting to be done by Via and I could see waymo tech being used for shuttles as well. Aka like college campus shuttles 10~15 mph. Maybe even a downtown kirkland to 185th interchange shuttle.

         On the opposite side there’s freeways and we’ve seen how tesla already has automated it pretty well since it’s standardized on a freeway and basically zero pedestrians. I could see say Stride 2 in hov/toll lanes being automated.

         The middle scenarios of collector busses, regular bus routes, and regular express busses weaving in and out of traffic (aka 520 over to mercer street) are probably the hardest to automate

      5. While automating Link trains may be technologically possible in theory, there are numerous obstacles to doing it in practice. Plus, I don’t think reduced operating costs through labor savings is even something that Sound Transit really cares about. Whether the train runs every 3 minutes or every 10, the bragging rights from having built it is the same.

         I think that is the biggest issue, by far. Automated vehicles are generally safer. Not always, of course. Boeing managed to set back plane safety decades. But the last big train accident in the state occurred because a person — not a machine — was in charge. The Amtrak accident that killed three people (and injured many more) would have been prevented if a robot was driving. When it comes to Rainier Valley, I don’t think the driver of a train is really doing that much. There have been plenty of accidents, which means there probably have been plenty of “near misses” but I doubt the driver has made much of a difference. There is only so much a train can do — basically the key is to avoid doing something stupid (and the Amtrak driver failed).

         Even for Rainier Valley there is an alternative: Remote monitoring and operations (if necessary). Have a team of small team looking at cameras as the train goes through Rainier Valley. It is a relatively small stretch of the system, and will become an even smaller portion soon. It is a gradual process. You want to increase the electronics to the point where the driver is basically doing very little (just sitting there, looking at the various lights to see if anything looks wrong). At that point, the machines are doing most of the work, and you can consider taking that last step.

         The unions may not care as much as people think. At worst you phase people out, but more than likely you find other jobs for them. You adopt the model used by the longshoreman. In that case it was private industry, but it is the same idea. The fewer people there are, the better they get paid. Through in a healthy severance for those that are laid off and it should be a fairly easy negotiation.

      6. My guess is that the biggest factor in automating transportation on public rights-of-way is the inherent liability and associated insurance costs. If a operator-operated Metro bus crashes, that’s on Metro (or the operator if they broke protocol) because Metro trained the driver and provided the bus. If Metro purchases an automated bus system, would Metro still take on the responsibility if the bus crashes and Metro wasn’t the one who “trained” it? Would any tech company be willing to take legal liability for ensuring that their automated bus system is able to operate safely in variable weather, with variable passengers, and maintain headways without stopping in the middle of the street because someone strapped a cone to the front of it? Would any agency be willing to operate a bus full of the General Public without that warranty? It’s not like they’d be able to put a sign on the door that says “by boarding this driverless bus, you waive all rights to sue Metro or its [Driverless Bus Service Provider] for emotional or physical damages in the event of an incident”.

         Seems like a legal nightmare on top of the extreme technical complexity.

      7. Automated vehicles are coming. Heck they are already here!

         The issue isn’t whether or not they will ever happen. The issue is instead how fast can they safely travel in a particular urban environment, and what decade will they scale any remaining hurdles like ADA compliance to free up widespread application.

         2050 is 26 years from now. Look back 26 years to 1998! Vehicles did not have many of the responsive safety technologies that are found on many new cars.

         The whole notion of any microtransit was pretty far-fetched in 1998. Uber was founded just 15 years ago and Lyft 12 years ago. Now it is already in the consciousness of transit planning.

         Technology has transformed our culture and our travel behavior. Anyone who believes that technology that is already going through trials to hone its application won’t be ready well before 2050 is being a Luddite.

      8. John D, some trains are already automated. Skytrain is automated.
         BART has been automated since 1972! Yeah, they’ve got a baby-sitter riding in the cab to make sure nobody gets caught in a door and dragged to their death, but they only push buttons.

         You don’t think it would be worth ST’s money to put platform doors on every station, expensive though that would be, to save millions of dollars a week of operator wages? Hint: of course it would.

         Humans are needed to drive buses, at least, ones that go faster than about 25 miles an hour. Let staff drive local buses.

         As WL said, low speed shuttles are already being converted.

      9. @Tom Terrific

         Those systems are fully grade separated. As far as I can tell at-grade automated rail is very feasible, but it still needs to be proven and cost-efficient. I think the biggest limiting factor is that Sound Transit doesn’t have a good track record of adopting newer technologies – but that could of course change in the coming years

         @Al S.

         “Already here” is different from “ready to be deployed on a wide scale”. I agree that they are coming, but there are way too many unknowns with the timeline, cost, and functionality to do any realistic planning around them. I’m also not sure why planning for automated buses would change the network; as far as I can tell the biggest benefit is that it opens up more operator hours for more frequent routes or wider coverage

   3. Let’s leave an autonomous bus network for a separate article. I’ll make a stub article soon for it. Phasing in autonomous routes would require a lot of speculation of what their future capabilities will be, and we don’t have any recommendations from Metro to start from. I’m assuming the first autoroutes would have to be in short, uncongested, predictable corridors; e.g., not the B.

      1. At this stage, I’m merely saying that broad scenarios like this are the situations that deserve a few broad alternative outcomes. To imply that the only alternative is to merely hone the status quo is being lazy at best and naive at worst.

         The result may end up the status quo in 2050. But at least alternatives would have been developed. In particular, fixed route frequencies are not high enough in suburbia to do anything more than give much of suburbia only token transit service.

         I’ve noticed how willing locals are to accept a single alternative and accept very minor variations as the alternatives instead. At what point in the long-range planning process should a broader set of choices be discussed? To me, it’s this point in the process.

      2. I think it is extremely difficult to predict what a 2050 network will look like. Way too many things can change. Yes, some level of automation is definitely possible. It is already happening to some degree: https://humantransit.org/2023/09/a-next-step-for-autonomous-buses.html

      3. Rather than a fully autonomous system, I see a partially autonomous one. I could see major trunk routes still with drivers, and these may stop only at “hubs” every mile or so. At each hub may be a waiting autonomous vehicle (pre-ordered) to go the last mile. Those last mile vehicles only need to go 20-30 mph since the distance will be so short. Technology at that speed exists today. It may be that the short autonomous shuttle buses have a defined route or at least defined major stops — but the could also deviate as needed. It may not be practical in much of Seattle — but 2/3 of King County residents live beyond the Seattle city limits.

   4. Another reason why the 2050 plan won’t look like what is proposed: Bridges. Take the Magnolia Bridge, for example. It is very close to the end of its life, and building a new one is extremely expensive (and not worth it). Yet this shows buses going over the Magnolia Bridge. Why wouldn’t they? Seattle hasn’t made a decision yet on what to do. My guess is there will be a bridge close to Dravus, or maybe even next to Dravus. I would put it a bit south of Dravus, then add bus lanes to Dravus.

      That one is obvious, but so much of the network depends on the potential for change. If the buses can faster, they can go farther. They can run more often without bunching. This influences the network.

      1. We don’t know what will happen with the Ballard Bridge or Fremont Bridge either. Will new bridges go in the same place or next to them? If there’s a bridge at 15th and at 14th, which one is the car bridge and which one the non-car bridge? The same at Fremont Ave and 1st or 3rd Ave NW. A plan can’t depend on another plan that doesn’t exist yet or is undecided, so the only thing it can do is assume the existing bridge will remain for the foreseeable future. And there’s a likelyhood that will happen, because the Viaduct wasn’t closed until years after they planned to, and I-5 is only being refurbished now after needing it for a couple decades.

         Anyway, I don’t want the comments to get lost in whether the plan as a whole is good or whether Metro will look at it again. I want to talk about routes. These routes or alternatives. These routes are a starting point for suggesting alternatives.

   5. FWIW automated trains have existed since the 70s, and the region is still not using them for link light rail, or even for the monorail, which is just two stops completely grade separated.

      Vancouver will likely have automated busses by the 2050s, but I’m less optimistic about Seattle.

7. Metro Connects (MC) has limited value. It is a collection of concepts without much rigor. It is not a blueprint. It is over promised on many margins. There is the opportunity for much more rigor during restructure projects. In network design, agencies may attempt to maximize ridership (or overall benefits) subject to constraints (e.g., budget, coaches, rights of way) with some allocation to coverage. (See Jarrett Walker formulation). MC had much more budget than today’s network. So, the network is too rich and unaffordable. Network design is also about tradeoffs between competing uses of the scarce service subsidy: between fixed route and Metro Flex; within fixed route, between coverage, frequency, span, and speed. Some of those tradeoffs are expressed in the service guidelines.

   Just one Seattle example. Route 8 is successful, but has issues on Denny Way. Near the east-west leg, MC adds a route on Harrison Street; SDOT will study it; the corridors are quite close together; that shows the frequency v. coverage tradeoff. Can SDOT improve either pathway to move transit through SLU traffic? The original Route 8 was between Uptown and 16th Avenue East only. In 1997, it was extended south. That south leg on MLK Jr. Way, so another example of the frequency v. coverage tradeoff. In fall 2024, the G Line will be implemented on Madison Street; so, routes 8, 11, and the G Line may be atop one another. Is the south leg worthwhile given current budget constraints?

   Consider the Bellevue example. They are not a transit provider, but their transit plan had three levels of service subsidy to show the tradeoffs. Walker was their consultant.

   There was also an earlier version of MC before the adopted one that included ST3 and the SDOT seven RapidRide lines.

   1. Unlike others here, I look at these maps only very broadly and not literally. I could mention several corners where buses can’t easily turn for example. However, I’m avoiding talking about route level changes because I don’t think it’s wise to put this in a “recommended plan” perspective.

      Certainly we will see lots of changes by 2050 — lifestyle differences, shifts in traffic congestion, bus technologies, new developments creating more riders especially at new Link stations, pricing methods, supplemental funds targeted to small areas, commute travel demand variations by time of day, etc.

      I’m actually disappointed that Metro staff hasn’t laid out several different 2050 service scenarios based on different future outcomes and technologies. This kind of concept piece is exactly when an agency should explore how transit service could be approached in different ways. Instead, we get something closer to a mere boring continuation of the status quo.

   2. It’s the only thing we have; the only expert-level integrated recommendation. Otherwise we’re just amateurs drawing lines in the dark, not knowing what factors or tradeoffs we’re missing. Also, it’s important to have an ideal network even if it’s not fully affordable, to quantify what the transit needs are. Every route identifies a transit need and grades it at one of four levels. That’s important, however much the particular alignment and termini may be flawed. They illustrate the needs and tradeoffs. We have to look at it to see whether we agree the implied transfers and frequencies help more people than they hurt (relative to the current network). We can express this as deltas to this network.

      Every concrete restructure looks only at a certain subarea and certain routes, so while it may be integrated within its routes, it doesn’t address the routes outside it, which also have pending suggestions and tradeoffs, so we have to think about the restructure relative to the current outside routes, and relative to multiple potential outside alternatives. We need an integrated vision of one possible countywide network, and since we’re not smart enough to do that ourselves as well as Metro planners can, it’s better to use this as a starting point than to use nothing at all.

      It also helps when giving feedback to Metro and the politicians; we can say, “Take route X and two surrounding routes, and make this change.” Even if the Metro Connects comment period has ended, we can send feedback anyway, and bring it up during concrete restructure routes. The first 90% part of the suggestion has been validated by Metro and the politicians, so we just need to convince them of our 10% change.

      Some of these routes will come back in future restructures, so we can be ready for them. For instance, I asked Metro why the 8-Madison Park (1061) isn’t in the RapidRide G restructure, and whether that meant Metro had abandoned it, or was waiting for another specific restructure, or is keeping it to do or potentially do later. They said it may come back later but wouldn’t commit yes or no. So it’s something that might happen someday.

      1. Otherwise we’re just amateurs drawing lines in the dark, not knowing what factors or tradeoffs we’re missing.

         Right, except when we actually know people who work for Metro, and they can correct us. I realize that not everyone can (I’m lucky — I know a guy) and so you make a good point. All of the routes here are plausible. They all make turns that someone in Metro thinks is reasonable (although untested). They include layover spots that are already in use or are plausible.

         But that doesn’t mean that it is especially likely. Many of the proposals that I (and others) have made are just as likely. Metro just doesn’t work like this. This is not a “master diagram” that planners use when they start their next restructure. As we’ve seen, they basically throw this (and damn near everything) out when doing a restructure.

         Often that is a shame. There are various areas — like the greater Central Area — where the ideas for future bus routes were better in the previous Metro Connects than they will be after Madison G. This has little to do with money, either. For example, they proposed moving the 2 over to Pike/Pine. Now it appears the folks making this want to leave it alone. Maybe they figure that if you don’t move it when implementing the G, you’ll never get around to moving it.

         It is best not to take this too seriously, and treat it the way you would if I came with a similar proposal. I don’t mean to sound arrogant, but I’m saying my ideas are as likely to be implemented as the ideas represented here. That doesn’t mean it isn’t worth discussing. Far from it. It is important that we discuss these ideas (as well as others) and try to get the good ones implemented.

      2. The network has things Metro wants to do, that keep coming back in restructures. Metro wanted to split the 2, to reroute the 3N and 4N to Seattle Pacific, and to delete the 4S tail. These came back in multiple restructures: the RapidRide C/D restructure, the E restructure, the 2014 cuts, etc. Splitting the 2 always generated the most opposition, so in one restructure Metro proposed it but then backed down, and when it backed down it also canceled the 3, 4, 5, 26, and 28 proposals. But then some of them came back again in later restructures. The 3N and 4N finally went through. The others never did. The 2 split is now in the Interim map again.

         I think it’s not Metro ignoring the long-term plan, but that a few certain routes are right at the threshold of approval/non-approval, so they can easily go one way or the other depending on how bold Metro and the Council is at one particular time. So sometimes Metro backs down and withdraws it; and other times it goes through. We don’t know when Metro wants to keep trying to do something, vs when it abandons it — except when it disappears from the long-range plan. So everything in the long-range plan, I assume Metro still wants to try to do, and it may come back in a future restructure. For instance, the B split was in the East Link restructure, then it was out, but now Metro still wants to do it later. Sometimes they may not get into a restructure because of limited budget or Metro timidity, but those all go up and down over time, and we don’t know why something isn’t there. The only definitive indication we have that Metro has abandoned the idea is when it disappears from the Metro Connects map. Like the 2-Union has disappeared.

         It’s also useful to know what the planners’ preferences are, and what the initial political balance is. That’s all implied in these routes.

         The 177/577 is another example. We think there’s significant opposition to deleting them, because of Link’s particularly long travel time to Federal Way and Tacoma Done. Metro can’t control whether ST deletes the 577 or not, but it can either replace it or not replace it. Whether it does or not depends on whether economy outweighs complaints or vice-versa when a proposal is made or the final decision is made. It’s likely right at the threshold, and could go either way. We’ll know more when Federal Way Link opens and people either vote with their feet to it or don’t. That might give more pressure to delete or keep an express bus at that time. In the meantime, Metro seems determined to keep a 577 replacement, because it has remained in Metro Connects for eight years now. That might be something it actually tries to do, or it may be a nominal “we might do that”. But it’s presence in the map shows its likelyhood is above 50%. Maybe 51%, maybe 80%, maybe 100%, but still above 50%. And that’s good to know.

      3. The network has things Metro wants to do, that keep coming back in restructures.

         Metro doesn’t operate like that. A few people at Metro want to do certain things. Later on, a few other people want to do certain things. Sometimes they line up. As eddie wrote up above: It is not a blueprint. It is a collection of concepts without much rigor.

         If you look at the Lynnwood Link restructure, you will realize they were all over the map on that one. The Metro Connects Plan had the 75 going all the way across to Greenwood Avenue, and then north to Shoreline Community College (SCC). That is not what they proposed. They initially proposed sending the 65 across. Then they settled on the 77, a brand new route that takes over for the current 522. In none of those cases does the bus actually get to SCC. Even with this proposal it doesn’t get to the SCC. Or consider the 72. I distinctly remember it being extended to the 148th Station, then continuing, following the stair step pattern to SCC that the 330 takes (east of 5th & 155th). Metro never proposed that (much to my chagrin). Not only that, but the current proposal doesn’t match that or the proposed routing! They changed their proposals quite dramatically every time, and yet not once did it resemble the Metro Connects Plan. It is quite likely the planners never read the Metro Connects Plan. They never saw it, their bosses never told them to read it, and no one said “Hey, we need to adhere to the Metro Connects Plan more” as you are suggesting.

         For transit route nerds (like myself) the map is great. There are plenty of ideas that are worth discussing. But this isn’t the real plan. The real plan is the dry, boring document that has oodles of pages that set long-term, purposely-vague goals. This time around, it has a lot more about “equity” than last time. But any connection between this and a real proposal is either because it is obvious or coincidental.

      4. “The Metro Connects Plan had the 75 going all the way across to Greenwood Avenue, and then north to Shoreline Community College (SCC). That is not what they proposed. They initially proposed sending the 65 across. Then they settled on the 77, a brand new route that takes over for the current 522. In none of those cases does the bus actually get to SCC.”

         That’s one of the cases with a lot of changes. I don’t understand why the changes, but neither do you. Maybe it identified some need it’s trying to address. Maybe some political factor intervened. Maybe there weren’t enough service hours to do what it wants. I don’t think they’d do it at random for no reason. That’s what Metro doesn’t tell us, so we don’t know.

         “It is quite likely the planners never read the Metro Connects Plan.”

         They’re the ones who wrote it.

      5. “But this isn’t the real plan. The real plan is the dry, boring document that has oodles of pages that set long-term, purposely-vague goals.”

         That’s not a plan. Or at least it’s not the kind of route-specific plan we’re interested in. It’s more like a mission statement.

      6. “It is quite likely the planners never read the Metro Connects Plan.”

         They’re the ones who wrote it.

         I really doubt it. What makes you think this is the same group that planned the Lynnwood Link restructure, or the group that planned the RapidRide G restructure? Do you know someone at Metro that said that? I can find out, but does it matter? Seriously, just assume it is. Assume it is the exact same people (what are the odds?). So what? There are different goals. This is a throw away project. It is hypothetical and bureaucratic in nature. Notice that we can’t even find the old one. Why? Because they threw it away! Someone says “Hey, we need some pretty looking routes to go with the planning document”. Then it is assigned to some group, or someone volunteers and they hack something together. A proof of concept, if you will. They dust off ideas they’ve had in the back of their mind (split the 8, bus service on Boren). They cover various areas (why not?). They double check to make sure it doesn’t look too stupid, and is the opposite of what was just implemented, but even then, they fail. It doesn’t go through an approval process. No public feedback. No testing by the drivers to see if it will actually work. Likely very little in terms of peer review. It is just a sketch of ideas. That’s all.

      7. @Ross

         > No public feedback. No testing by the drivers to see if it will actually work. Likely very little in terms of peer review. It is just a sketch of ideas. That’s all.

         I understand there are flaws with long range planning but that is is for all organizations from local governments to small coffee shops to large tech companies and just part of the process from trying to look so far ahead.

         I’m not exactly sure what you want Metro to do any differently here? Like do you want them to just say “improve this corridor” but don’t actually specify what it is like the SDOT draft transportation plan? Or you talked about rebuilding the bridge — how in the world is King County Metro supposed to know where the bridge will go?

         They have to plan the best with what they do know.

      8. I’m not exactly sure what you want Metro to do any differently here?

         Nothing.

         I’m just saying that we shouldn’t take the map seriously. It is not a blueprint. It is merely a set of ideas (some good, some bad). It is great for discussion, but it should not be read as “Metro wants to do this”, because Metro doesn’t operate that way. It is not how they do restructures.

         It makes a difference. For example, I’ve been very disappointed with the more recent restructures. Not because they’ve largely ignored the Metro Connects ideas (although they have) but because they just aren’t very good. I’ve lost faith in the planners, and think that we need to do a major restructure, with the help of Jarrett Walker (or a similar firm). Note: I have no connection with Walker or anyone involved with his firm. I just see plans like this and get jealous. I want that.

         So imagine I try and start a movement for a major restructure, using an outside firm. Someone says “But Ross — there already is a long term plan. They’ve already done the work. It is just being implemented piece by piece. As long as they have adequate funding, we’ll get something like that”.

         Nonsense! This map is irrelevant. It doesn’t drive the planning. When they do restructures it may incorporate those ideas, but that is coincidental.

         Nor would I do anything different. I suppose it might make planning easier if they had some sort of “master plan” — a future blueprint. Then when there is a change (e. g. Lynnwood Link). They would dust it off, and apply the various changes. But it should not work that way. Things change. Areas where a deviation makes sense no longer do. The city fixes certain bottlenecks (or not). Folks on here have repeatedly mentioned the Boren bus (from First Hill to South Lake Union). For good reason — it is a very worthy route. Not not so long ago, it would have been pointless.

      9. I’m trying to stay out of these meta-issues but:

         I see Metro’s planning/restructure quality as going up and down. Sometimes it’s bold, sometimes it’s timid, sometimes the status-quo opposition gets its way, sometimes it doesn’t. It had an especially bold streak from 2012 to 2016 (RapidRide C/D to U-Link).

         “think that we need to do a major restructure, with the help of Jarrett Walker (or a similar firm)”

         Can we get some people together to hire Walker to draw his own countywide map? That would give us a second opinion. If Metro wants to bring him on for a revision, that would be great too, but I’m not holding my breath waiting for it to happen. I’d also ask Walker how he’d compromise if we just had the existing service hours, vs what he’d do if we had a successful Metro Connects-sized levy to do more.

   3. > Just one Seattle example. Route 8 is successful, but has issues on Denny Way. Near the east-west leg, MC adds a route on Harrison Street; SDOT will study it; the corridors are quite close together; that shows the frequency v. coverage tradeoff. Can SDOT improve either pathway to move transit through SLU traffic?

      SDOT has the harrison street transit way idea. https://www.seattle.gov/transportation/projects-and-programs/current-projects/harrison-mercer-transit-access

      > The original Route 8 was between Uptown and 16th Avenue East only. In 1997, it was extended south. That south leg on MLK Jr. Way, so another example of the frequency v. coverage tradeoff. In fall 2024, the G Line will be implemented on Madison Street; so, routes 8, 11, and the G Line may be atop one another. Is the south leg worthwhile given current budget constraints?.. Consider the Bellevue example.

      Route 8 and RapidRide B are actually kinda similar. They both currently form an L.

      KC Metro has proposed for both of them to be split into east-west and north south variants. The rapidride B split is quite natural place to end at as it’s crossroads mall a major destination and beyond that is just single family homes. The 8 split is a bit more complicated. I guess it could just end at miller park, or extended out to madison park

      1. Yeah, the idea of having Harrison be a transit street goes at least to when Bertha was digging the tunnel. It might have been one of the first things someone thought of when they realized that you could go across Auroraa for a few extra blocks. I think the problem is after that. 5th Avenue North is used by the 3/4, so there would definitely be value in adding bus lanes there. Mercer isn’t used by buses right now (at least not there). That means that in terms of trying to add bus right-of-way, we would spread ourselves thin.

         Likewise, with service on Denny and Harrison, we could be spreading ourselves too thin. That is what eddie is getting at. New bus routes are very expensive. The money has to come from some where. The question is, where? I can come up with some ideas, but they involve changes that a lot of people — including apparently those in charge of these restructures — aren’t interested in. I want this instead of this mainly because the buses could run reasonably frequent (every 15 minutes). If we get more money, then adding service on Boren should be a high priority. After that we can talk about Harrison.

         The 8 split is a bit more complicated. I guess it could just end at miller park, or extended out to madison park

         Yes, that is how I would do it. That is what this has, and it is what the last one had as well. Serving MLK is more challenging. What they propose (the 3997) is good. I would be OK with that. As I wrote a while ago (when the last one of these came out) I was quite happy with their set of ideas, and would only make some minor tweaks. From what I can tell, this takes a step backwards. Anyway, another option (which I like, because I thought of it :)) is to branch the 27. The tail of the 27 can’t justify high frequency, but the part west of 23rd definitely can. Thus it is a fairly ideal branch when it comes to relative demand. This would connect Yesler with 23rd (e. g. Garfield) a very natural connection. But like other things, I would have that later. You’ll notice my map doesn’t have service on that part of MLK. Frequency is more important than adding routes that are fairly close to other routes.

8. # Harrison Street – SLU

   I guess one thing to highlight that I didn’t quite understand from the pdf’s before but can now see on the map is how heavily they lean on the Harrison Street concept. I thought it was just to reroute the 8, but actually they want to reroute many routes both frequent and express to slu through the corridor.

   Going through the routes clockwise:

   Magnolia: The 31, 33, 34 (to Magnolia) is rerouted to end at SLU as route 3025.
   Interbay- Madison Park : Route 1061 (route 8) extended to interbay and then ends at madison Park
   Upper Queen Anne: Route 3028 is a new one from Upper Queen Anne to SLU to Capitol Hill, going up Eastlake down back down Belmont
   UW: existing rapidride J
   Kirkland: Route 2516, rerouting kirkland busses ((255??) to go to SLU.
   Rainier Beach: Route 1074 from Queen Anne down denny way then down Boren Avenue to Rainier Avenue then MLK extending route 106.

9. Wow Mike, thanks for saving/finding this again! Really interesting to see this again in light of Metro’s inability to deliver even 2020 service.

   Since we’re essentially at the interim network, it’s interesting to see what Metro Connects “predictions” have come/are coming true. Obviously nearly all of it isn’t, but it’s interesting to see what has. Lots of East Link stuff here.

   215 and 269 from East Link Connections (ELC) Metro nailed in Connects. You can quibble with frequency but they got the route changes exactly. 554, again nailed. All-day 111, they almost nailed but it will end at S. Bellevue and not go as far east. 931, basically correct except minor routing differences. 222, yes, *almost* the same but with slight differences. It’s almost eerie!

   One interesting find is today’s 225 is in there, as a frequent route. That’s one of the routes that Metro cut down to hourly all the time last year, which is a shame because that route needs good transfers to work well. Even Metro Connects’ accurate predictions make actual Metro look like Pierce Transit.

   1. One headscratcher that stands out is frequent route 1030, route 240 combined with 245 south part (plus adjustments here and there). They got the frequency of both segments right, but the 1030 would have no connection to Link until Redmond Tech. I don’t know how that makes sense. Even the slow Link connection to the 240 in the ELC is still a decent Link connection. Seems like a haphazard disassembling and reassembling of existing Metro routes, perhaps to try to make them straighter. Which is a good goal in the abstract for some of the reasons discussed in this thread, but not all routes can be straight when East Link is very much not straight.

      1. That looks very similar to the new 245 from the East Link restructure.

         I think that area is a challenge. There are several parallel streets with a mix of density. You have pretty good pathways going north-south, as well as the potential for decent spacing. But in between you are forced to zig-zag to cover the crossing streets (e. g. Main). There are a mix of destinations. The college is a big one, and the routes all basically converge there. That also connects riders to Seattle (albeit with a bus then the train). To the north is Redmond and to the west is Downtown Bellevue. They could have just treated Bellevue as the main hub, which would mean those buses would turn on NE 8th. Similarly, they could peel off to connect to Link. In both cases this makes frequency weaker. It also means extra transfers to those heading to Redmond.

         The East Link approach seems to be to build a spine of sorts along 156th, to go with the perpendicular (relatively) frequent service offered by the RapidRide B. The Metro Connects approach seems a bit muddled. I would not be the least bit surprise if they did exactly what you suggest. They looked at it a while, and then punted. There is no obvious solution (at least not that I can see) so there is no point in trying to pursue one. In contrast, I think when they made routes like the one on Boren they were pleased.

      2. > One headscratcher that stands out is frequent route 1030, route 240 combined with 245 south part (plus adjustments here and there)

         They’ve kinda already deprioritized the 240 section (eastgate to renton) from being a rapidride candidate to just “frequent” in the interim map when they redid the map.

         > They got the frequency of both segments right, but the 1030 would have no connection to Link until Redmond Tech. I don’t know how that makes sense. Even the slow Link connection to the 240 in the ELC is still a decent Link connection

         It’s just part of the tradeoff over here. It’s not only for busses but also for the future Link. Whether to connect factoria and eastgate heading west to south bellevue or to head north instead. And then for others coming from new castle now they’d need to back track from redmond tech center, or transfer at eastgate.

         If I remember correctly, the 240 in early phases of east link connections was going to use lake hills connector, but they wanted to provide an easier way for eastgate p&r users to reach south bellevue station so one could reach seattle. I’m a bit curious though how many will opt to park there versus just parking at south bellevue though

      3. “1030 would have no connection to Link until Redmond Tech. I don’t know how that makes sense.”

         The problem is the location of Overlake Village Station. Instead of being next to Safeway at 24th & 152nd, with all the retail around it and a bus-accessible intersection, it’s four blocks northwest next to 520. This affects the 226 and 245 too, which will affect me. The 226 gets the closest to my destination (164th & Main), but its closest stop to Overlake Village Station is east of 156th & 24th, a half mile away. The 245 serves Redmond Tech but not Overlake Village, and the 221 serves Redmond TC but neither of the others. Rote 3080 (the 226’s successor) would be even worse. (Nearest transfer: 156th & 20th. It also wouldn’t serve Interlake High School students.) I don’t have the East Link restructure handy, but I think the 226 would be straightened out on 164th but keep the two 24th stops mentioned (156th and Interlake HS). The 245’s and 221’s successor may detour to Overlake Village station; I don’t remember.

         I take it as a sign that there’s a lot of uncertainty and contradictory thoughts about this area and some others, so the proposals keep changing every six months as one factor/constituency gets the upper hand, then another, but they all cycle back and forth.

      4. Oh, Ross has a link to the East Link restructure. There the 226 does something different: it goes from Bel-Red to Northup and 164th. That’s the straightest way and I’ve often wondered why it didn’t. (Northup Way was also how we drove from our house to then-Safeway (now Trader Joe’s) and the rest of Bellevue and 520. The 226 wasn’t on Bel-Red then, but on NE 8th Street before the B started.) The current detour to 24th seems specifically to serve Trader Joe’s and Interlake HS. The 223 will detour to Overlake Village, while the 245 will remain on 156th bypassing it.

         That leaves me with an issue of whether to transfer from Link to the 223, 226, or 245, all at different stops and Link stations. Most likely I’ll continue what I’m doing: eastbound take the 226 (which will go practically to my destination, closer than it does now), but westbound walk downhill to 156th and take whichever of the 223 or 245 comes first, to whichever Link station it goes to. That would have the eastbound transfer gap between Overlake Village and the 226 stop, but at least it won’t be in the middle of nowhere and there may be a bus shelter with a bench to sit on. (There is for the current stop at 156th & 24th, but I don’t remember about Bel Red. I don’t use that stop now because I have to continue to downtown Bellevue: there’s no ST Express service at Overlake Village.)

      5. P.S. The reason I don’t go south to the 554 is that, when I tried it twice, the 554 is half-hourly even peak hours, and the 226/245 zig-zag a lot south of the college (and through it and north of it), and terminate short of the 554 stop. I’d like to go that way it weren’t so bad. And I don’t understand the zig-zagging in Eastgate. Is there some permanent reason or is it just for construction?

      6. “The East Link approach seems to be to build a spine of sorts along 156th”

         That is the most logical place. The retail and apartments are mostly north-south. The destinations are too: Overlake Village, Microsoft, Bellevue College, and Eastgate TC. West on 8th is a grid corridor. West on Bel-Red is a secondary grid corridor. East is all residential (single-family and some apartments) — that’s where I lived.

      7. The East Link Connections links in Ross’s 12:19pm post above are for *Phase 2* maps. Here is King County’s website with Phase 3 maps:

         https://kingcounty.gov/en/dept/metro/programs-and-projects/east-link-connections

         Specifically, the Phase 3 map for the 245, and the Phase 3 “East subarea” map:

         https://cdn.kingcounty.gov/-/media/king-county/depts/metro/documents/projects/east-link-connections/routes/245.pdf

         https://cdn.kingcounty.gov/-/media/king-county/depts/metro/documents/projects/east-link-connections/maps/p3-eastsubarearoutes.pdf

      8. Thanks John. So the 226 will have its 24th detour and the bus shelter at 24th & 156th, and it will remain straight on 164th between NE 24th and SE 24th (except a slight detour at the south end, which is probably required by the street). The 245 goes straight on 156th between NE 70th and Main, serving Redmond Tech but bypassing Overlake Village. The 223 zig-zags. I don’t know why it goes to 148th between NE 24th and NE 8th. That serves Fred Meyer, but at 8th it’s the middle of nowhere with only two one-story single-use buildings. Will those buildings ever grow up, and is the empty land a protected natural area?

         I think the straight alternatives are winning out, but not completely.

      9. Thanks John. Sorry for the outdated references. I need to update my old links.

      10. “The 223 zig-zags. I don’t know why it goes to 148th between NE 24th and NE 8th. That serves Fred Meyer, but at 8th it’s the middle of nowhere with only two one-story single-use buildings. Will those buildings ever grow up, and is the empty land a protected natural area?”

          This zoning map for Bellevue suggests that 148th between NE 8th and Bel-Red Rd isn’t likely to be a big traffic generator:

          https://apps.bellevuewa.gov/gisdownload/PDF/Planning/complan2str_b_11x17.pdf

          On the Phase 2 East Link Connections map, no route runs on 148th between Northup and NE 8th; for Phase 2, the 249 was routed to serve the Fred Meyer area. I understand the desire to maintain bus service to the Fred Meyer area — when I’ve ridden the 221, the area was among the route’s most-used stops.

          The Phase 3 ELC network is, IMHO, a rather timid plan, with a tendency to retract from Phase 2 back toward the existing network. IIRC, the Phase 3 223 has worse frequency, especially in the evenings, than the Phase 2 223. Although putting the 223 on 148th preserves more of the existing 221 (and even its predecessor, the 221 that ran straight up and down 148th prior to RapidRide B), riders are saddled with a couple awkward problems:

          (1) In the afternoons, 148th jams up southbound as it approaches NE 8th. This is not unique to 148th; the same happens afternoons approaching NE 8th southbound on 140th, 156th, and 164th (albeit not necessarily at the same timing and/or scale). What makes this egregious for the current 221 — and would be the same for the Phase 3 223 — is that the route must turn left onto NE 8th. That left turn can take a few *cycles* for a bus to get through, and the 148th/8th cycle length is understandably long. This was a weak point in the otherwise-clever RapidRide B restructure, but was easier to overlook because the 221 is a comparatively minor component of today’s network; by contrast, the 223 has the potential to carry more ridership in the East Link network.

          (2) The stretch of density from the Overlake Village light rail station to Crossroads Mall, taken together with with Crossroads Park and Interlake High School, has the potential to operate as a single transit-oriented neighborhood provided the component destinations are effectively tied together with bus routes. With the 223 shunted over to 148th, the Phase 3 map does a clumsy job of this. Examples come to mind readily:

          (i) Right now there’s a frequent one-seat ride between Crossroads Mall and Overlake Village: RapidRide B. On the Phase 3 network, if you want to go from the Overlake Village area (2400-2800 152nd Ave NE) to the Crossroads Mall area (1000-1300 156th Ave NE), your options are the 223 and some walking, or the B/245 and even more walking.

          (ii) As students leave Interlake High School in the afternoon, many head to Crossroads Mall. On the current network, this is easy: board the southbound 226, a one-seat ride. The Phase 3 map fails to preserve that link. (For comparison, the Phase 2 map did preserve this link, via the 223.)

          (iii) If you’re at Crossroads Park and want to go to Overlake Village, that’s essentially a two-seat ride on the Phase 3 network (or ~10 minutes of walking).

          For each of (i), (ii), and (iii), I’d say the Phase 2 ELC network would provide better connectivity. Not only this, taking a quick look at the Metro Connects Interim network, it also seems to satisfy (i)/(ii)/(iii) better than the Phase 3 network.

          Offhand I like the general goal of making the Eastside’s bus routes straighter; but in Bellevue, with East Link traversing from southwest to northeast, I would think that in practice this does mean having some ‘L-shaped’ routes. I’m also okay with some restrained zigzagging with the intent or benefit of avoiding chokepoints. For neighborhoods like Crossroads and Factoria that are a couple miles away from a Link station, I’m partial to the concept of having a couplet of routes run from the closest Link station into/through the neighborhood along a single shared routing (timed such that the couplet has frequent service), then having the routes split and fan out into the lower-density areas even further from rail.

      11. One-word revision above to avoid confusion:

          isn’t likely to be a big traffic generator -> isn’t likely to be a big *ridership* generator

      12. John, I’m putting together my own Bellevue network. Assuming RapidRide 156th (1999, Redmond-Eastgate), and the 226 on 164th, what would you do with the 223? And for north-south transit on 145th Place, 148th, and 140th? What do these areas most need access to?

          What about a route from Overlake Village to Fred Meyer, 140th/145th Pl (Safeway, Sammamish HS, student apartments) to the college? Or must it detour to Crossroads (156th) or Main Street? If that route stayed on 140th, could another route serve those other parts?

          If the 226 doesn’t connect Interlake HS to Crossroads, could students reasonably walk to 156th and take the north-south RapidRide intead?

      13. The easiest question first :-)

          “If the 226 doesn’t connect Interlake HS to Crossroads, could students reasonably walk to 156th and take the north-south RapidRide intead?”

          By my personal conception of “reasonably”: yes, albeit on the upper edge. It’s roughly 0.5 mile (~800m) from 164th to 156th, about a 10-minute walk (and the south side of 24th is fully sidewalked there).

          Alternatively, there’s a pedestrian corridor carved out from the south edge of Interlake High to the north edge of Crossroads Park, which does get some use; it’s about a 1-mile walk to Crossroads Mall going this way. (I recall reading some sort of news article about the corridor a few years ago, but I’m not finding it at the moment.) The corridor has only one major street crossing, a crosswalk on Northup Way at 160th Ave NE.

      14. Now for the tougher questions, which for the sake of readability I’ll split into one post per quoted question(s). (Doubt I’ve ended up truly answering any of them, but it made for some mental gymnastics at least.)

          “John, I’m putting together my own Bellevue network. Assuming RapidRide 156th (1999, Redmond-Eastgate), and the 226 on 164th, what would you do with the 223?”

          A glib answer is that 1999 essentially *is* the (Phase 3) 223. It connects Eastgate, Bellevue College, Crossroads, at least one Microsoft-area light rail station (Overlake Village on the Interim map only, and Redmond Tech on both Interim and 2050 maps), and a substantial portion of Redmond beyond the reach of Redmond Link, including the City Hall campus.

          The Phase 2 223, for comparison, combined some interesting traits: it performed a ‘neighborhood circulator’ traversal of various useful destinations in the Crossroads area; it hooked those destinations up to the closest light rail station, Overlake Village (even jutting up to Redmond Tech as well); and it stayed ‘frequent’ (by Eastside standards) throughout evening and late-evening service. It’s even (more or less) present on the Metro Connects maps, as #3096.

          ‘What to do with the 223’ or ‘what to do with a portion of service hours to get 223-like benefits’ seem hard to answer without examining other routes in the system — notably the 245. The 245 has strong ridership by Eastside standards — I’ve been surprised by how well it has rebounded ridership-wise — and whether to somehow refashion the route is also not an easy question. I’ll come back to this as I move on to the next question.

      15. “And for north-south transit on 145th Place, 148th, and 140th? What do these areas most need access to?”

          I tried to think if these street segments could be examined/discussed separately, but I think it ends up merging into an overall discussion of Bellevue’s transit destinations. What follows is rather plodding, but I think it’s worthwhile.

          IMO Bellevue has sufficiently few major ridership generators that they may as well be listed out (including a couple locations that are important to cBellevue yet outside city limits). I’ll also take a stab at ordering them by prospective boardings/alightings (upon full opening of East Link to Seattle). This ordering is super-rough, from raw intuition with minimal forethought.

          1) Bellevue Downtown Station
          2) Bellevue College (incl. high-density housing to the north, and perhaps Robinswood Park as well)
          3) South Bellevue Station
          4) Eastgate (transfer point, bus depot; not being on light rail hurts Eastgate as a transit focal point)
          5) Crossroads (Crossroads Mall, Crossroads Park)
          6) Redmond Tech Station
          7) Factoria (incl. Newport High School)
          8) Sammamish High School
          9) South Kirkland P&R (parking and transfer opportunities)
          10) Old Bellevue and Bellevue High School (far enough from Link to warrant separate consideration)
          11) Overlake Village Station
          12) Spring District/120th Station
          13) Interlake High School
          14) Wilburton Station
          15) Bel-Red/130th Station
          16) Overlake Hospital (far enough from Wilburton Station to warrant separate consideration)
          17) East Main Station
          18) 520 freeway stations at Yarrow Pt, Evergreen Pt (transfer opportunities)

          For deciding which one-seat rides to give locations on each of 145th Place, 148th, and 140th: I’d take the list above, and bias the ranking toward nearby destinations, but only mildly.

          For 145th Place, for instance, check out Metro Connects Interim #1999: it actually reaches over to South Bellevue Station. Now while I personally find the service to Robinswood Park an attractive feature, if the route ran from Lake Hills Library over to QFC on 145th Pl, then the high-density housing on 145th Pl gets a one-seat ride to South Bellevue Station, and the ample transfers available there.

          Going up to 140th, the Metro Connects maps have a route, #3116, that serves Totem Lake, 140th/24th, 140th/Northup, Bel-Red/130th Station, Sammamish High, and Bellevue College/Eastgate. I doubt the Eastside is ready for a route this grid-disciplined, but the route does address some legitimate issues. The gap between Bel-Red/130th Station and Overlake Village Station is so big that bus service (beyond what the 226 would provide) is needed within that gap. A specific example of planned densification in this area (with the last sentence calling out the light rail stations):

          https://www.exxelpacific.com/highland-park-to-break-ground-later-this-year/

          The densification further west on Northup Way (between 124th and 130th) presents similar issues.

          Returning to the topic of the 245, for while I’ve considered the possibility of dividing the 245 into two pieces, with overlap in the middle where ridership is strongest: a Kirkland-Microsoft-Crossroads piece, and a Microsoft-Crossroads-Eastgate piece. Each of the two routes would follow a different path Microsoft-Crossroads for better coverage. The Kirkland route could then be extended beyond Kirkland, and similarly the Eastgate route beyond Eastgate. This seems to have occurred to the Metro Connects designers as well, although their plans assume far more service hours than I see Metro coming up with anytime soon.

      16. “What about a route from Overlake Village to Fred Meyer, 140th/145th Pl (Safeway, Sammamish HS, student apartments) to the college? Or must it detour to Crossroads (156th) or Main Street? If that route stayed on 140th, could another route serve those other parts?”

          I think this sort of route does fill a gap, and does so without being too duplicative of Link. There’s an ‘induced demand’ possibility with a route like this, as there’s a substantial high-density area on the west side of 140th between 8th and Bel-Red Rd that has either the B line or the 226 available to them, but nothing else. If I had to guess why there’s been hestitancy to run a route down 140th, traffic backups in the afternoons may have been part of it. That still happens, but if memory serves, I don’t think it’s as bad now as it was 5 years ago.

          I don’t see connecting either Sammamish High or the 148th/Main area specifically to Crossroads as essential — IMO those two areas may as well be evaluated against the ranking rubric above, with only a slight proximity bias for Crossroads, to guess at ridership demand. In the case of the high schools, I would also bias a bit on the district boundaries:

          https://bsd405.org/wp-content/pdf/boundaries/high-school-attendance-area.pdf

          A minor side note: while Metro has seemed hesitant to put a regular bus route onto 140th Ave in the vicinity of NE 8th, this portion of 140th has nonetheless been designated (for several years — don’t recall how long exactly) as part of the 271’s snow route:

          https://kingcounty.gov/~/media/depts/metro/maps/esn/metro-system-map-esn.pdf

   2. “Wow Mike, thanks for saving/finding this again!”

      The credit goes to WL. I just made an article of it because the comments would be so long. And to see more evaluations of all the routes in it and other route ideas.

      1. Good call. I’m glad they preserved the plan in some publicly accessible form. I’m kind of surprised actually, since I’m sure Metro by now would like everyone to forget all about this. I hastily made a backup of my own in 2020 with the Chrome browser tools, but this one is much better.

         Really interesting to see what concepts they were thinking of again, and what they would do if they assumed big growth in service.

      2. I hastily made a backup of my own in 2020 with the Chrome browser tools, but this one is much better.

         Really? That is fantastic if we are talking about the same thing. The other map was interactive, like this one. It wasn’t a PDF, nor an image or static web page. As a result, when the Internet Archive (AKA the Wayback Machine) tried to archive it, they got basically nothing. If you have a working website, is there any way to share it?

         I know that some of it was essentially copied onto the PDFs, but there isn’t much detail. Not enough to really tell what they had in mind for Capitol Hill, for example. I distinctly remember the 2 doing a dogleg over to Pike to get downtown — that is missing from the current map(s).

      3. “I distinctly remember the 2 doing a dogleg over to Pike to get downtown — that is missing from the current map(s).”

         That’s what I said; that concept has disappeared. It’s not in the RapidRide G restructure either. So it looks like Metro has abandoned it. There has always been controversy between some people saying the 2 is too close to the G; others saying Pike/Pine would be faster and serve more people’s trips; others saying Virginia Mason must have a very close stop and only a route on Seneca can do that; and others not wanting any changes to the 2.

         In my experience at Metro hearings, any change to the 2 generates the largest opposition movement of any route. Opponents don’t want it moved off Seneca, and they especially don’t want it split downtown. One person living in the eastern part of the 2 said, “The 2 goes everywhere I want to go, and doesn’t go where I don’t want to go. I wouldn’t feel safe transferring downtown to get to Queen Anne.” Metro Connects was bold and suggested both changes. Now it seems the opposition is getting the upper hand about Seneca.

      4. Just a reminder about Pike and Pibe Streets: they now have multiple stop signs that weren’t there five years ago. It’s great for pedestrians but it’s terribly slow for buses and their riders.

   3. @Kevin Kvenvolden,

      “ Since we’re essentially at the interim network, it’s interesting to see what Metro Connects “predictions” have come/are coming true. Obviously nearly all of it isn’t, but it’s interesting to see what has”

      LOL, thanks for bringing this conversation back to reality. Because discussing in minute detail what Metro’s route structure will look like in 2050 is a bit “silly”, and doing so without discussing the elephant in the room is even more pointless.

      The elephant in the room is of course Sound Transit’s Link Light Rail. Whatever Metro’s route structure looks like in 2050 will largely be determined by what Sound Transit builds, or doesn’t build, for ST3 and ST4. Discussing bus routes in 2050, without knowing what LR will look like in 2050, is little more than an exercise in self-flagellation.

      “ Even Metro Connects’ accurate predictions make actual Metro look like Pierce Transit.”

      That statement seems a bit overly harsh to me. But hey, there you are, I said something positive about Metro.

      1. Sorry, that should be “Alex”

      2. We aren’t anywhere near the Interim network. The target date was 2025, but that didn’t take into account that Link is years late, RapidRide G and J are years late, RapidRide 40/44/48/62 got downgraded, the levy to fund the additional service hasn’t happened, and a pandemic intervened. Northgate Link was supposed to open in 2020 (2 years late), East Link in 2021 (5 years late), Lynnwood Link in 2022 (2 years late), KDM Link in 2023 (2+ years late and counting). The continuing relevance of the Interim network is some future time when Lynnwood, downtown Redmond, and Federal Way are all open, and adjusting for the nonexistent levy. When we reach that time (maybe in 2027), and if all the implied restructures happen, we’ll probably have something that bears a family resemblance to the Interim plan but with a lot of differences.

         The relevance of the 2050 plan is to have some reference of what we’re trying to get to. Even if it’s not exact, or even if there should be more than one alternative, they’d all generally say, from any one location, which Link stations and shopping centers you can get to, what your frequency level might be, and thus where you can move to and work and have reasonable transit mobility in all four directions.

         “Whatever Metro’s route structure looks like in 2050 will largely be determined by what Sound Transit builds, or doesn’t build, for ST3 and ST4.”

         That’s not the point. If Ballard and West Seattle Link and Stride 1/2/3 aren’t built, of course the “2050 plan” will have to be substantially revised. That doesn’t mean it’s useless now. Much of ST3 is in Snohomish and Pierce Counties, so it’s irrelevant to Metro’s network.

         The 2050 plan doesn’t assume anything about ST4: that would be a 2070 or 2080 plan. ST4 isn’t even outlined yet. We have no idea what would be in it, how large it would be, whether the subareas are unified enough for it, when the vote would be, whether the public is ready to vote for it (it looks like “No” now), or anything. We just have a few subarea requests that don’t add up to a full plan, and they may change by then.

      3. @Mike Orr,

         The 2050 plan is pretty much irrelevant. It’s a shot in the dark by an agency that doesn’t know what it should be shooting at, nor even which direction it should be shooting in.

         It’s good to have contingency plans, but this is little more than a fever dream.

      4. The 2050 plan is pretty much irrelevant.

         Agreed.

         But hey, there you are, I said something positive about Metro.

         That didn’t last long, did it.

         Look, every agency has to make long range plans. For a general purpose transit agency like Metro these include things like coverage, frequency, equity — all of that. Now look at the actual report. Notice that they spend quite a bit of time describing different levels of service (RapidRide, Express, Frequent, Local). They spend a fair amount of prose describing each type, but readers may still wonder what that actually means. They want examples.

         Thus the map. The map is an example of the type of service and the type of network they want to build. But it is not meant to be an actual proposal that anyone (within Metro) thinks will become reality. It just a set of examples of the various levels of service they expect to provide someday.

         I’m very suspect of long-range plans. I think they are largely BS. Things change too often for them to matter. I’ll admit, I am biased — I come from a software world. I used to be the king of long-range-plans. Before I was a programmer I was a tester, and I used to love spending all day producing or critiquing the documents. I came up with the quote “A good tester finds a bug before it is written”. This is known in the industry as the waterfall method.

         Hardly anyone uses it anymore. They use something called Agile Development. It is an iterative and incremental software development process. You make small changes, then do it again, and again. The short term goal is specific, but the long term goal is fairly vague. The idea (as the name implies) is that if you screw up, and for some reason your approach doesn’t work, you can adjust and build something better. You don’t try and anticipate every problem (from a user, system or coding standpoint) you solve them bit by bit (before the code gets released).

         It makes sense for programming, because programming is not a “measure twice, cut once” proposition. It is the opposite. It doesn’t make sense for something like Link. Failing to have a long range plan (for say, a branch at the UW for trains heading to Ballard) will end up costing us a fortune (if we even build it). Subways really are “measure twice, cut once”.

         Bus networks are more in the middle, but I would say they are closer to software. There are exceptions. Trolley wire is hard to move. Moving bus stops isn’t trivial. But overall, it is fairly cheap to restructure the routes. Furthermore, there is little to be gained by trying to anticipate the future network. Sound Transit didn’t anticipate Ballard to UW rail, and this is really bad. But if Metro didn’t anticipate, say, a frequent bus on Boren? Big deal.

      5. I would say that the long range plan is actually more useful in the short term, as it gives us some insight into how Metro is thinking right now. For example:

         Increased mobility services of all types to complement the high-capacity transit network. These will include flexible services that can provide fast and easy connections to the larger public transit system

         This suggests a continued investment in micromobility, at least for now. At some point I expect the micromobility fad to collapse.

      6. “It’s a shot in the dark”

         You don’t have a shot in the dark . What’s your 2030 bus network plan, or 2040, or 2050? If you had that we could discuss it, and compare it to other plans, and get more ready for whatever we want to convince Metro/public/council of when the opportunities to make decisions come.

      7. “We aren’t anywhere near the Interim network.”

         In some ways we sort of are. I make the comparison based on the implicit (here made explicit) assumptions that:
         2025 network = after East Link, Federal Way, Lynnwood, and Stride
         2050 network = after ST3 is done

         I didn’t even look at the 2050 network this time, and it’s way too early to say when ST3 minutes parking will be done (2050 seems optimistic TBH).

         But to respond that we’re nowhere near the interim network, yes and no, depends on the area. Of course we aren’t near the level of service, and won’t be for a long time, but in terms of what we’re going to get, we have essentially final restructure plans for East Link and Lynnwood. We should expect changes for Stride, but not too many. So I’d say for east and north King, we basically have enough to make a meaningful comparison. Next up should be Federal Way, which I’m eagerly awaiting to see what concepts Metro has come up with and how they compare to Connects. (FWIW I have my own FWLE fantasy map, but more realistic than the interim network: https://t.co/mtgfpcVVCu. Might write a page 2 about it since it’s still pretty unrealistic).

         I don’t accept that it’s meaningless to make the interim network comparison because Metro hasn’t passed all the funding needed to get to it. In some ways that’s not the point, of course they didn’t plan in Covid, etc. The broader point here is a critique of how they see transit needs now vs what they saw as the need in 2016. (If you were hoping I wouldn’t be so negative on Metro, sorry about this next part) It’s insane that Metro dropped the ball so badly during the 2021-2023 part of recovery that 2020 service levels are effectively a fantasy map still. Maybe it’s pointless because what can they do about it now? But at the same time, they seem to call progress promising people less transit so they don’t disappoint anyone. That’s not progress, and they shouldn’t be content with it.

      8. The broader point here is a critique of how they see transit needs now vs what they saw as the need in 2016.

         Yes, and the report (not the maps) explain that better, in my opinion. A larger focus on equity. More microtransit (which was barely mentioned back then, despite DART being quite successful).

         As for funding, I think several things happened. The loss of tax revenue from COVID, the smaller Seattle STBD, and the county failing to pass anything similar. Then you have the driver shortage on top of all that.

         But that doesn’t really explain the differences in the routing. It is pretty easy to “scale down” the previous map to fit less funding. In many cases what they suggested is more efficient — and thus better suited for less funding. Running a bus (like the 2) next to another, very frequent bus is fine if you are rolling in money. Not so if you are struggling.

         I think it is just a matter of different people. Various groups take over a project. Sometimes it is great, sometimes it isn’t. Personally, I think the various restructures of late are really bad. But it will likely ebb and flow.

         I’m a sports fan, and this is common. Pick a professional franchise and you usually see great ebbs and flows. The coaches are important, but they can only do so much. I’m a big NBA fan, and I think anyone who knows the game knows that Gregg Popovich is one of the greatest coaches ever. Yet for the last few years, they just haven’t been very good. I don’t think he has lost his touch, I just think the players aren’t as good. Given the young phenom from France they have now (Wembanyama) it would not surprise anyone if they get back on top fairly soon.

         The same thing might happen with Metro. Or Metro planning may continue to be like it has been recently — not very good (in my opinion).

      9. “I have my own FWLE fantasy map, but more realistic than the interim network: https://t.co/mtgfpcVVCu. Might write a page 2 about it since it’s still pretty unrealistic).”

         I’m starting to think we need more ideal-network maps if so many people think Metro Connects is too flawed for this. The first step to fixing something is identifying a goal. Not just an abstract “frequent straight service between activity centers or along grid lines”, but what routes where. For somebody at any particular location, what can they access? We can simplify it by focusing on the Rapid, Frequent, and Express routes. The Local routes can be more vague: what we need to know is, “Which hub can a neighborhood access?”, or “Which hubs” if there’s one at both ends. That’s more important than all the streets it goes through, which segments are strung together, or where the final termini are.

         So if you can do that for Federal Way, that would be a good contribution and might be suitable for Page One. The point is to identify what kind of network we need, that gives people adequate mobility. If it’s unrealistic, just say what’s unrealistic or what we can’t accomplish. The biggest thing probably is how many service hours or bus-shifts it would require relative to current service or Metro Connects. That would tell how much resources we’d have to advocate for compared to the politicians’ existing expectations. Even if it’s not realistic, it’s still worth documenting the total mobility need. That’s the politicians’ biggest blind spot: they don’t understand what all people need in order to get around reasonably without driving. That’s the missing key to reducing vehicle-miles-traveled and carbon emissions, and getting to a level of transit use comparable to other industrialized countries in similar land-use conditions.

      10. “Yes, and the report (not the maps) explain that better, in my opinion.”

          The report doesn’t explain anything. High-level goals is not a plan, it’s a mission statement. People can’t ride high-level goals from point A to B, they can only ride routes. People need to know what they’ll be able to access from their location, at what frequency, with what travel time. Sprinking fairy dust at unspecified places doesn’t help with that.

          “A larger focus on equity.”

          Yes, that’s been a government-wide emphasis since 2020.

          “More microtransit”

          It’s still only a small part of the total service.

          East Bellevue had a “Crossroads Connects” microtransit pilot, which started and ended before I needed it. In the East Link restructure round 1 and/or 2, Metro proposed deleting all fixed-route service east of 164th or 156th (I forget which) and replacing it with a Metro Flex area. I grew up in that area, so I’m glad I had a fixed route then and didn’t have to depend on an app-taxi every day. But in round 3, the 222 goes east of 164th, and I thought there was a route to Northup Way & 8th although I don’t see it now, so maybe Metro has withdrawn the idea of an app-taxi there.

      11. However, it is valuable to have the four service types in the report. That shows a commitment to get to 18-hour frequent service and all-day expresses, which is more than we have now. It counteracts the watering down of “frequent” to only until 7pm weekdays and Saturdays. It remains to be seen how much Metro lives up to it in the future, but at least it’s a goal.

      12. “So if you can do that for Federal Way, that would be a good contribution and might be suitable for Page One.”
          I very well might do that then. Most of my free time these days (which is less than it used to be) is taken up by a transit-related passion project, and I’ve accumulated some good article ideas once I get around to them, but I’ll probably expedite this one. FWLE is a special one for me because I grew up in FW and have thought a lot about how it might look after Link arrives. Living car-free through college biases me a lot towards coverage (because car-free) and span of service (because college), so as a concept it’s very much anecdotal around me, but I think an important perspective nonetheless. So if you look through the maps you’ll notice more adding routes and aggressively leveraging Link than straight up cranking frequency up, but I do sneak in a good deal more frequency around Link stations by overlapping routes.

          As for microtransit, I’m not sure how I feel about it. The common critique is that it’s much less efficient than regular transit, but in a lot of cases it’s probably the right tool for the job, and could still replace a lot of car trips. I viscerally feel negative toward it because it seems to be employed by an agency when it either gives up or fails abysmally at providing attractive fixed-route service. And paradoxically, it’s harder to market even if it’s often more convenient to individuals. Buses advertise themselves, but vans are less distinct. Though with Link openings, it may be easier to do, and once P&R lots start getting full again, that’s going to be a high potency place to advertise micro transit (hate circling around for parking? Why even drive when a van will come to your street and pick you up for no additional fare!!).

      13. An ideal network can also go beyond the agencies. What matters to us and to passengers is what service they can use, not which entity operates it or what tax or budget funds it.

      14. I’d like no to be contrarian, but it seems to me that “Microtransit” does not “reduce car trips”, it merely provides them under a rubric of “public”. Once in a while a DART will have two people in a neighborhood who want to go roughly the same place at the same time and turn into a tiny vanpool, but it’s pretty rare.

10. To quote Jarrett Walker:

    All other things being equal, long, straight routes perform better than short, squiggly and looping ones.

    He mentions this on his blog and explains why in his book. There is an obvious corollary to this:

    Try and avoid turns.

    Now consider the 1061 and 1074. These are both good routes. They increase coverage in an area of high transit use. It may be difficult to afford this, but the routes are a worthy goal.

    But look more closely at where they meet, at Fairview & Denny. Rather than cross each other (both going straight) they make 90 degree turns. This is very bad! It is clearly a mistake. A rookie mistake. All of the various reasons that Walker gives (the geometry if you will) apply here.

    There is another reason why turns are often bad: They take longer. If there is a stop sign or roundabout, a turn doesn’t matter. But with a traffic light, turns usually take significantly longer. In urban areas, right turns are problematic (unless they add a right turn signal) as pedestrians can prevent a bus from turning. Left turns usually require their own signal. The period for left turns is usually much smaller than for going straight, which means that over the course of a day, a bus will spend more time waiting for a left turn. It is common to have left turn signals that simply can’t handle demand. When push come shove, the city will (generally) favor vehicles going straight. Thus a bus may wait multiple light cycles to turn (as long as the backup doesn’t impede cars going straight).

    There are times when making turns in this manner could be justified. For example, there could be a service mismatch. In this case, let me start by defining the segments (broadly) by attaching neighborhoods to them:

    1) Denny west of this intersection (Belltown).
    2) Denny east of this intersection (Capitol Hill).
    3) Mercer and Harrison (South Lake Union).
    4) Boren (First Hill).

    This routing pairs 1 and 4 along with 2 and 3. If Belltown and First Hill had way more density than South Lake Union and Capitol Hill (or vice versa) then this turn could be justified. One bus would run often, while the other bus would be coverage in nature. But that clearly isn’t the case. All areas are high density and have high transit use.

    There are other cases where buses make similar mistakes because of the way they evolved. Maybe an area had only a handful of destinations back in the day (e. g. the 60 on First Hill). Maybe one route is RapidRide, making a restructure more difficult (the northern tail of the 40 and D). Usually — as with the 60 and 40/D — there is a reasonable trade-off. In this case though, they just made a mistake.

    1. > Now consider the 1061 and 1074. These are both good routes. They increase coverage in an area of high transit use. It may be difficult to afford this, but the routes are a worthy goal.
       > But look more closely at where they meet, at Fairview & Denny. Rather than cross each other (both going straight) they make 90 degree turns. This is very bad! It is clearly a mistake. A rookie mistake. All of the various reasons that Walker gives (the geometry if you will) apply here.

       1061 is trying to use Harrison Street to avoid the traffic west Fairview where on Denny Way it doesn’t have the bus lanes.
       1074 it can’t travel straight on Boren to Denny Way because that’s an unprotected left turn across two lanes of traffic. It’s honestly impossible to make that left turn for like most of the day even for regular cars let alone a bus.

       Or actually I think you’re saying to swap the two last legs of them to lower queen anne. Honestly it looks more like they would prefer to move both routes to Harrison, but if they had to keep one bus route on denny way they placed a ‘lower’ priority line there and moved the 8 to harrison to keep it more reliable. Whatever time savings from not making left turns is completely swamped by the i-5 traffic backups on denny way.

       1. Or actually I think you’re saying to swap the two last legs of them to lower queen anne.

          That is exactly what I’m saying. That is the whole point of that comment. Keep the buses going straight.

          I thought it was pretty clear. Here is what they do in this proposal:

          1061: https://maps.app.goo.gl/DqddDHXLXsJMVGKB6
          1074: https://maps.app.goo.gl/xk8JnsERVyKWR8Fu8

          Both turn at Fairview and Denny. Instead they should go straight.

          Bus from Boren: https://maps.app.goo.gl/cDayghtnqD5LKZKLA
          Bus from Capitol Hill: https://maps.app.goo.gl/jgktZmRUPfwcGE3N8

          if they had to keep one bus route on denny way they placed a ‘lower’ priority line there

          Except a bus that cuts across First Hill is not a “lower priority line”. There is no service mismatch. First Hill, Capitol Hill, Belltown, South Lake Union — all about the same. (I covered that in my comment.)

          As far as traffic goes, the other bus runs on Mercer, which is not exactly a traffic free zone. The answer is to fix Denny. Take a lane. If anything, that is easier. You basically leave Mercer for the cars, while you add a bus lane on Denny. If Mercer is stuck, so be it. This is much easier than adding bus lanes on both.

          This is another example of why these long range plans are just guesses. Imagine they add bus lanes on Denny. As nice as it is to run a bus on Harrison, do we really want to? Maybe we run the truncated 8 (between Madison Park and Uptown) like we do the 7 (every 7.5 minutes midday). Then the Boren bus gets truncated in South Lake Union. Or maybe the new 8 “only” runs every 10 minutes, but so does the bus from Uptown. That means that this section would be entirely in bus lanes, and have combined five minute headways throughout the day. Buses would basically split at Denny & Fairview, with one bus headed towards Boren, and the other headed straight. So much depends on what SDOT does.

11. Highway 99 is basically one long corridor from downtown Seattle to Marysville.

    When my friend lived in the Shoreline area, the hour long slog on the E to get to Seattle was annoying, when the distance could be driven in 20 minutes. Furthermore, it was annoying to have the break at the county line. There’s significant cross border traffic there.

    So it would be nice if there were an express from Northgate to Aurora, then north with stops only at the major cross corridors, and extending some significant distance north of the county line. I’m not sure where the best point would be.

    Maybe Northgate is too far south of a starting point, but it needs to be far enough south of the county line to tie together currently disconnected parts of the Aurora corridor.

    1. The best way would have been to have the link divert to the sr 99 in shoreline. But anyways that’s passed.

       Alternatively just use the east west bus lines to reach the link and then head down.

       Lastly there was an alternative proposal for a rapid brt on aurora to northgate station. I’ll see if I can find it. It would only have 3 stops between Lynnwood and northgate, however I think it was just going to use the existing bat lanes so the travel time wouldn’t be faster than the Rapidride e

       1. Using the east-west buses means a 3 seat ride just to travel anywhere point to point on the Aurora corridor. That makes traveling the Aurora corridor worse, not better.

       2. The comment section told me that my Kirkland to downtown trip moving from a two-seat ride to a three-seat ride is an improvement. A step up. A higher level of service. That I should even be thankful. So why am I now seeing the comment section complain about three-seat rides? I thought they were a higher level of service?

       3. I never really quite understood why they removed the montlake freeway station without adding a westbound hov direct access ramp.

       4. WL: re Montlake. The transit agencies use the highway provided by the state. The Legislature told WSDOT and its contractors to build a six-lane freeway. One step was to close the Montlake Freeway stops. Both SR-520 and Link were and are delayed. Three players; three schedules.

       5. > The transit agencies use the highway provided by the state. The Legislature told WSDOT and its contractors to build a six-lane freeway. One step was to close the Montlake Freeway stops. Both SR-520 and Link were and are delayed. Three players; three schedules.

          I mean they aren’t rebuilding the montlake freeway stops afterwards. The final lid will only have the eastbound direct access ramps, but no west bound ones.

          I did some digging:
          https://data.wsdot.wa.gov/publications/SR520/SDEISComments/item_report_15377.pdf
          > Removing the Montlake Freeway Transit Station would minimize the width of the freeway through the Montlake area, reducing the width by up to 40 feet compared to keeping the station
          ” See Attachment 8 to the SDEIS, Range of Alternatives and Options Evaluated, for further discussion of how and why removal of the stops was considered.”

          Currently trying to find that document I’ll update if I find it.

    2. One solution that is still constructable is to put three lines through North Seattle, and have a branch connect to Aurora (Tacoma – Aurora Village Line). The place where right of way could seemingly most easily accommodate a connection is SR 104 at the County line. A terminal station at Aurora Village would be the result of that.

       As the place where both Metro and CT could end major rapid bus lines, it would have strategic transit system value. It would also seem to be doable with aerial tracks keeping the costs down.

       While it’s a sea of parking and big box retail today, that actually offers an easy opportunity for upzoning since it’s one story concrete pads and asphalt surfaces that can be redeveloped without heavy neighborhood disruption.

       Of course, that’s an idea that is merely in my head. There is no funds available to build or operate this. But since the idea is to connect SR 99 to Link closer to Seattle, I thought I would put it out there as a long-term possibility. .

    3. The E is already an express. It gets a very high percentage of its riders from along Aurora. Without those riders, it wouldn’t run that often. There have been rush hour expresses in the past (the 301). But in the middle of the day it just doesn’t get enough riders to justify the trip. It is always annoying when your trip takes so long. It is easy to assume that lots of other people are taking that trip. But in the case of long distance trips (e. g. Shoreline to downtown) there really aren’t. Not enough to justify the cost (or benefit). For sake of argument though, I will call your bus an E-Express, or EE.

       The best way would have been to have the link divert to the sr 99 in shoreline.

       Yes, or really pretty much anywhere. Even a stop at Fremont (above the troll) could be a tipping point. That would make the trip from anywhere on Aurora to the UW much faster. With the increased ridership on Aurora to the UW, an express on top of the E (the EE) might be justified.

       Alternatively just use the east west bus lines to reach the link and then head down.

       >> Using the east-west buses means a 3 seat ride just to travel anywhere point to point on the Aurora corridor.

       Not really. Here is what I would do (from north to south):

       1) County Line to 185th: Swift
       2) 175th: Crossing route connecting Shoreline Community College (SCC) with 185th Station.
       2) 155th and 160th. Crossing route connecting SCC with 148th. Does a dogleg on Aurora like the 330.
       3) 145th and 130th. Crossing route to 130th Station that loops around via 145th and ends at Linden).
       4) 105th. Metro 40
       5) 85th: Metro 45

       You’ve got some gaps, but not a lot. Keep in mind, many of the buses I propose will be added. That means that many of the riders who might otherwise be interested in an express to downtown will take a crossing bus instead. Trips to Link destinations, like the UW, Capitol Hill or Rainier Valley. Or maybe transfers to other places like First Hill or the Central Area. Right now the EE would be great for a lot of riders. Once Link gets here, it will be less valuable.

       Also keep in mind that while this has gaps, so does the EE. That is the point. If you only serve “the major cross corridors”, then by definition those riders can take a direct bus to get to Link. You suggest skipping everything south of Northgate, which means that a for a lot of the riders who take the E downtown, the EE offers nothing.

       The biggest difference is really that section between Aurora Village and 185th. If Swift covered it, then riders would have a fast seat to Link where it saves them the most time (way up north). South of there the gaps are unfortunate, but not worth worrying about. There are a ton of gaps in our system, and Aurora really isn’t as special as folks thing. For example, consider Greenwood Avenue. South of 130th, they are in the same boat. The 5 is a great bus — it gets a lot of riders. But as you get farther and farther north, it takes a long time to get downtown. The 45 offers an alternative, but it slower to get over to Link than to just stay on the bus (if you are headed downtown). It just isn’t worth it to add an express on top of the express (outside of rush hour) and right now that bus is suspended (because of lack of ridership). Metro has trouble running decent service on Capitol Hill — hard to argue that an all-day express is worth it.

       Now Sound Transit is another matter. Their service is regional by nature. They spend oodles of money on poorly performing buses (while spending even more building trains that will perform poorly). So if they wanted to run an express like that — be my guest.

       1. “t is easy to assume that lots of other people are taking that trip. But in the case of long distance trips (e. g. Shoreline to downtown) there really aren’t.”

          There are quite a lot of people traveling the Aurora corridor north of 200th on Swift Blue, and south of 200th on the E. The density is pretty much the same on both sides of the border. There’s no reason to think a cross-border bus wouldn’t also have a fair number of riders connecting to places along the corridor. There’s no magic decrease in density at 200th. It’s all one corridor, just currently split into two routes.

          The 301 stopped a few times on Aurora and wandered into suburban nothingness west of Aurora before the county line. It didn’t do anything to link the two separate pieces of Aurora.

       2. I think we are talking about two different things Glenn. A very large percentage of the riders who take the E get off before downtown. The farther north you are, the less likely you are to head downtown. The trips are very spread out as well. Ridership is a bit higher at the major crossing routes, but still not that high.

          It surprised me when I saw the numbers. It is pretty much the opposite of the 512. The 512 has a good northern anchor in Downtown Everett. I figured there would be plenty of people in Mountlake Terrace and Lynnwood taking the bus north, to Everett. Not really. Only a handful. It has always been about Seattle, and mostly Downtown Seattle.

          Not true with the RapidRide E. Which is why I doubt that an express to downtown would do very well. Even after Swift connects to Link at 185th (and gives those riders an express to downtown) there will be plenty of people who transfer from Swift (or the 101) to the E. That’s because they aren’t going downtown. Which gets me to “link the two separate pieces of Aurora”.

          It is actually SR 99, but that is the idea. I think that has merit, but I think the Swift approach I mentioned would be adequate. Really the key is to have Swift stop on Aurora, at 200th (the E already stops there). That would mean a same-stop transfer. A little overlap (between 200th and 185th) would be even better. More overlap seems excessive. There is only so much the agencies can afford. I’m not sure what it would look like. Can you explain in more detail what you have in mind? For example, would you send the E up to Edmonds Community College?

       3. Ross,

          As a rider of the E, I can say that your description of the route is spot-on. We live close to the stop pair at 46th, and ride the bus north as often as we ride it south, and being able to transfer to other frequent buses like the 40 and 45 is a big plus.

          Even when we ride it south, we’ll sometimes get off at one of the stairway stops at Queen Anne and walk up the hill to go to stores at the top. Sometimes we’ll bus back downtown, but sometimes we’ll just head back to Aurora to save some time, especially since the 5 and 28 work as well as the E for us.

          An EE would maybe satisfy 1/3 of the trips that we take on the E, and certainly none if it skips 46th. If it comes at the expense of frequency on the mainline, it would be a loss.

       4. “The farther north you are, the less likely you are to head downtown. The trips are very spread out as well. Ridership is a bit higher at the major crossing routes, but still not that high.”

          That’s what I’m getting at. I don’t wish for the E itself to change. I just think it would be really helpful to have something better to connect the two sections of the corridor. It’s too long a route for one bus to continue all the way as E and Swift. Shoreline to Seattle trips get better with Lynnwood Link.

          Whenever I took the E or Swift, there were only 2-3 people that got on near Aurora Village, but further along either one got pretty busy.

          It suggests to me there’s a missing set of trips near the county line not being met. The two counties currently don’t cross the line, and SoundTransit is focused on much longer trips.

          It’s obviously not a vital thing, but it seems like there’s a group of riders on the Aurora corridor that currently don’t have great options due to the break in service.

       5. I’ve attended events along Snohomish 99 at Edmonds College, a church at 176th in north Lynnwood, a store at James Village at 196th, and gone to downtown Edmonds, all coming from Seattle. People along Snohomish 99 might go to Aurora for Sky Nursery or Fred Meyer or the ice rink at 185th, the shopping village at 155th, a BJJ school at 95th, a friend’s apartment at 90th, or an event in Greenwood at 85th.

       6. Here’s another thought:
          Rather than run the E through to Montlake Freeway station, have it continue somewhat further north on 99? Maybe to somewhere around 228th?

       7. Editor’s correction: Glenn said Montlake Freeway Station. He meant to say Mountlake Terrace Freeway Station.

    4. The built-up part of Aurora ends at 73rd, so that’s where the overlappingest trips end. The closest bus-sized arterial is 85th, so that would be a good endpoint if there were a Link station there, but Roosevelt Station is too far away. So Northgate Way may be the next-best. That would at least unify trips between 105th and Everett, even if it leaves out 73rd to 105th.

       The issue is not just that Swift+E requires a transfer, but that there’s no limited-stop route south of 200th. This would address both of those.

12. ## East west routes

    I’ve been rethinking direct east-west bus routes (converting our L bus routes) aka grid proponents. Looking at the 2050 plan, most of the east-west bus routes just look really hard to actually use without having to do an awkward 3 bus transfer.

    I’m aware about Jarret Walker being a proponent of grids https://humantransit.org/2010/02/the-power-and-pleasure-of-grids.html

    However, there’s also other transit expert’s who push back against grids being more a proponent of radial lines (or more accurately reach some downtown core) https://pedestrianobservations.com/2018/02/22/transit-and-scale-variance-part-3-grids/

    Alon Levy says: > The grid works only for surface transit and not for rapid transit, and only at a specific scale, so constrained as to never be maximally useful in an entire city, only in a section of a city. This contrasts with what Jarrett Walker claims about grids. Per Jarrett, grids are the perfect form of a transit network and are for the most part scale-invariant (except in very small networks). One of the impetuses for this post is to push back against this: grids are the most useful at the scale of part of a transit city.

    I won’t copy Alon’s entire article as you can read it for yourself but here’s snippets of what I find most important:
    * “The principle here is that a grid plan is useful only if the grid can be complete.” — Seattle and in general the metro area’s pretty hard to form a conducive bus grid with plenty of barriers. I don’t see why to force it.
    * “In the absence of a radial rail network to connect to, long grid routes are less useful. Cities have a center and a periphery, and the center will always get more ridership, especially transit ridership. The outermost grid routes are often so weak that they should be pruned, but then they weaken the lines they connect to, making it necessary to prune even more lines until the grid is broken.”

    Notably in North Seattle there’s many of “L” or C” pattern bus routes like the 40, 31/32, 62, 75 etc. Restructuring these routes as north south ones with separate east-west routes would end up cutting off their connection to downtown and with the loss in ridership and subsequent frequency, they’d be even harder to use.

    Near the UW/downtown Seattle there’s a couple bus routes going through dense cores 44, 8+11 idea, or 36+49 idea. But these are all basically travel through dense areas not really ‘circumferential’.

    West Seattle to South Seattle, the route 50 is practically useless. Honestly might as well just redirect and turn it into two lines(say columbian way and alki) that head to downtown. Whatever time savings from avoiding a transfer are swamped by waiting for the 30 minute frequency.

    Same for the F line, I say we should just convert it into another L route fashion. Cut the west half of F line over to the rapidride H. The east half is a bit harder to decide where to place. If I had my way I’d actually actually have stride 1 brt take that section. Or alternatively have rapidride I take it over.

    1. Grids are an ideal, and some cities are more grid-friendly than others. Most cities spread out in at least three directions uninterrupted, like Chicago, Toronto, and Los Angeles. There a grid works well, and they have it. Even Vancouver, Portland, and Tacoma are more grid-like than Seattle and have more crosstown routes to match it.

       Jarrett Walker once wrote that Seattle is “three peninsulas that meet”, and that a bus can’t go straight for more than a mile or two without falling off a cliff, plunging into a river, or being blocked by a freeway. He mentioned there are hardly even any roads between West Seattle and Southeast Seattle for a bus to take. Seattle is apparently the most fragmented city in North America. More than one person has told me Seattle is the most confusing city to drive in they’ve encountered.

       Another thing is Seattle villages aren’t in straight lines. A grid route on 15th Ave NE from the U-District to Mountlake Terrace bypasses Greenlake, Northgate, and the Crest Cinema area, serving mostly just houses and apartments instead. A grid route on 23rd from Mt Baker the U-District bypasses First Hill, Broadway, and 15th. People get frustrated that their home bus takes them to where they don’t want to go and where few people want to go, instead of where they do want to go and many people want to go.

       Other cities avoid this by having more destinations and villages along grid lines. Chicago’s Clark bus and Diversey bus have a lot of destination clusters along them. So does San Francisco’s Diversey bus and Fillmore bus and Mission bus. And Vancouver’s Broadway bus. We don’t have as much of that. Instead we have village islands scattered randomly in a sea of low-density residential-only.

       Jarrett also says bottlenecks like bridges tend to promote transit ridership, so that people don’t have to drive through the congested bottleneck. That’s one big reason Seattle/Pugetopolis has pretty high ridership per capita compared to most of the US.

       So the primary need in Seattle is to connect the villages. The 45 does that by connecting the U-District, Roosevelt, Greenlake, and Greenwood. The 62 connects downtown, SLU, Fremont, Greenlake, Roosevelt, and Sand Point. There are a lot of overlapping trips between those neighborhoods and the areas in between. Secondarily we can put strategic grid routes: the 8, 31/32, and 44 are very successful. Even the 48, while I dislike its residential-only middle, at peak hours every single stop has somebody getting on or off.

       L-shaped routes can be good. They’re either two grid routes interlined, or a strategic partial-grid-to-partial-grid connection. I used to look askance at routes like Muni’s 22 (Fillmore-16th) or 24 (Divisadero/Castro-Cortland/Palou) — why do they turn? — but when I spent time in the area I found they were useful. The 45 and 62 are like that. The 75 could be if Metro adopts the Sand Point-130th concept. The 45 can be interpreted as either an L-shaped route or a diagonal route. A route on the opposite diagonal — Lake City to Ballard — would also be useful and would complement it.

       Some routes are not so much about going from one end to the other, as going from the middle to either end. So it’s really two logical routes interlined. It can even be three routes: one third, one third, and one third. The routes can turn at those intermediate anchor points.

       Alon is more talking about something else, rapid transit lines on top of a bus grid. In this case the rail lines can connect the largest villages, and the buses can do their grid thing. Seattle doesn’t have enough rail lines to even have a rail grid. And many villages aren’t connected by either fast enough buses or rail — there’s a transit gap between them. If we can’t build enough rapid transit lines to connect them, then we must do it with buses, even if it offends grid purists.

       I’m not sure Jarrett and Alon are as far apart as they may appear. Jarrett comes from a route-geometry perspective; Alon comes from a numbers perspective. Jarrett has supported radial rail networks on top of a bus grid. Alon says Chicago’s buses aren’t a grid when they are. His argument “People don’t take a bus north and transfer to a bus west; they take a bus and transfer to the El” seems weak: even if people do that, the routes are grid lines.

       1. > Alon is more talking about something else, rapid transit lines on top of a bus grid. In this case the rail lines can connect the largest villages, and the buses can do their grid thing. Seattle doesn’t have enough rail lines to even have a rail grid. And many villages aren’t connected by either fast enough buses or rail — there’s a transit gap between them. If we can’t build enough rapid transit lines to connect them, then we must do it with buses, even if it offends grid purists.

          Not really, he is talking specifically bus grids for a large portion of the article. He specifically calls out Vancouver bus routes and Chicago’s bus routes.

          > I’m not sure Jarrett and Alon are as far apart as they may appear. Jarrett comes from a route-geometry perspective
          > His argument “People don’t take a bus north and transfer to a bus west; they take a bus and transfer to the El” seems weak: even if people do that, the routes are grid lines.

          I guess I failed to outline the bus grid proponents take. Jarrett’s grid idea is about how a bus route not entering the downtown core will still be useful as with a grid formation one can potentially only need to transfer once. This is specific for bus grids. For a hypothetical like if we had northgate way bus going east-west and the 45 on NW 85th continuing east to 65th/75th st to Sand point (route 62).

          Secondly in general the bus grid idea implies that east-west bus routes are ‘good’. But I think we should just embrace the L bus route idea.

       2. “he is talking specifically bus grids for a large portion of the article. He specifically calls out Vancouver bus routes and Chicago’s bus routes. ”

          And they call Chicago “not a grid” because he says people don’t typically do bus-to-bus grid transfers. Whereas to me the important issue is that the routes are in a grid structure even if the passengers aren’t.

          “Secondly in general the bus grid idea implies that east-west bus routes are ‘good’. But I think we should just embrace the L bus route idea.”

          Chicago’s problem is the buses crawl in congestion. The routes are good (grid). The frequency is good (5-10 min daytime, 20 min evening; 30 min night owls every mile apart). The problem is the speed.

          East-west bus routes are necessary. I’m mostly around Clark Street, at Fullerton, Belmont, Diversey, or Lawrence. To get to western destinations you have to take a bus. To get to O’Hare you have to take a bus+el. I’ve found that no matter which route you take from the Blue Line, the travel time is always 40 minutes, because the diagonal Blue alignment means that as the el segment gets longer, the bus segment gets shorter, and vice-versa.

       3. I think it’s unhealthy to present any bus route layout as inherently good or bad. It depends on the overall urban layout, density, difficulty in parking, general travel patterns, natural barriers, popular destinations, orientations of downtowns and commercial districts and on and on. A grid also works better if the major street network is laid out in a grid as opposed to the hub-spoke (or radial-crosstown) natural evolution of most of our US cities (Cincinnati has steep hills which resulted in how its core transit evolved, for example). The question is instead what works in a particular circumstance.

          If Seattle had uniform five-story buildings with ground-level retail and office, parking rates and supply uniform across the city, an equidistant length and width, and major destinations whoever the grid hit water a full grid system would look logical.

          The thing is: We don’t! Our city layout includes a mostly north-south downtown, mostly north-south oriented commercial districts and usually lower densities near water. Parking rates vary widely, with free parking in many areas and a high cost to park downtown. Also, Seattle generally has difficult east-west streets for buses to negotiate. The hills and water features also affect how far people will walk to a transit stop.

          I just whacked blackberry stalks. Like most of nature, the didn’t evolve fully perpendicular. The stalks crossed each other frequently, with branches of one stick intertwined with branches from another. It’s akin to what I think works for Seattle: Routes that weave radiating outward from Downtown, crossing often in places outside of Downtown that provide an overall stronger system.

          Along with that, through-routing through Downtown lets the buses serve all of Downtown with direct service (double duty through Downtown for corridors extending outward at each end) as opposed to forcing transfers in a true grid. The result of that is of course are L-shaped routes in most cases.

    2. The problem with the F line is all the turns between Southcenter and Renton. It’s fine between Burien and Southcenter, and between The Landing and Renton TC, but going from Renton to Southcenter or TIB takes an inordinately long time. The F tries to do too much: it detours to the Sounder station and Oakesdale Avenue. The Sounder detour is to avoid a second Sounder feeder route. But 95% of the F runs have no Sounder train to meet.

       Ideally it would just stay on 154th/Southcenter Blvd/Grady Way, which are all the same street. But the Southcenter pedestrian entrance is further south at Baker Blvd, and half the bus gets on/off there all day. So that’s a good detour. But the Sounder stop is only useful if there’s a Sounder train, and Oakesdale ridership is probably peak-only. An all-day RapidRide bus should focus on all-day destinations.

       1. It’s not just Sounder. The F goes all the way south to a couple of office buildings well south of the station.

          They could still serve the station if they could get permission to run the F for a short distance on the Boeing frontage road on the east side of the BNSF main line. That would shorten the route a little anyway.

    3. RapidRide I should continue north to Rainier Beach Station. Otherwise everyone in the major north-south corridor is a 2-seat ride away from Link. The fastest way would be to make like the 101 to Rainier Beach. Another way would be to make like the 106. But that would take rather a long time from southeast Renton or northeast Kent. And it would compete with the Rainier Beach-Renton Ave-Renton Highlands concept, which is also important. Eastern Renton needs better access to Link and the rest of the region, and this concept would do it.

    4. Walker is a transit expert. He simplifies things for the reader, so they can understand complex, scientific principles that have been well researched. Alon is a … um … I’m not sure what Alon is. But Alon reminds me of this comic strip: https://www.gocomics.com/forbetterorforworse/2024/02/04.

       Just consider the way the two address stop spacing. Here is Walker’s take: https://humantransit.org/2010/11/san-francisco-a-rational-stop-spacing-plan.html. Fairly simple. He addresses the subject, notes that it is full of exceptions (some will walk more than a quarter mile, some less) and then proceeds to address the various issues surrounding it (filling in the diamonds).

       In contrast, look how Alon deals with the subject: https://pedestrianobservations.com/2019/04/21/stop-spacing-and-route-spacing/. Immediately they launch into a complex set of mathematical equations that the average reader may be unable to understand or unwilling to deal with. This goes on for a very long time. And yet, when it is all done, the conclusions are similar (line spacing around 700 meters, stop spacing around 400 meters). Pretty much exactly what Europeans do, and what Walker said from the beginning! Jarrett has basically explained (in fairly easy to understand language despite there being some math) why Europeans have more or less standardized on a particular distance. Alon comes to much the same conclusion, but only after you dig through a ridiculous amount of math. (Oh, and I’m saying that despite being very good at math.)

       Or consider the critique of Vancouver. Walker claims it “an almost perfect grid” (https://humantransit.org/2010/02/vancouver-the-almost-perfect-grid.html). He then explains the importance of “anchors” and then deals with the Vancouver system.

       If you look closely at the system map, you’ll see that most of the north-south lines extend to the north edge of the grid and then bend east or west as needed to flow into downtown. There’s a bit of inefficiency in the resulting duplication, but it’s not too bad. It means downtown’s major attractions and connections anchor the north end of most of the north-south lines.

       Then he goes into how the lines converge to UBC on the west. The anchors are more spread out to the east (I’ll get to that). Then he sums it all up:

       1) It a grid.
       2) Lines are about 800 meters apart.
       3) It has strong anchors.

       In contrast, Alon claims that Vancouver doesn’t have a grid, simply because these routes converge at the ends. Bullshit! That is absurd. Take a piece of graph paper. It is a grid, right? Not attach triangles on each side. Draw lines so that they all converge in the corners. It is still a grid! Call it a “converging grid” if you want, but it is still a grid. Vancouver is so “grid like”, it even has “sub-grids”. New Westminster has a street grid that is basically at a 45 degree angle (where the rest of the city is North-South-East-West). Yet even for this part of Greater Vancouver, the buses form a grid. Not a radial pattern. Not a spine. A grid.

       The point being, if you are in Vancouver and want to get anywhere else in Vancouver, you have a straightforward path to get there. It is “almost perfect”. There is none of this bullshit: https://maps.app.goo.gl/oWLDPf6s9Myaiht16. Just consider: Noon. No traffic. A trip between the two largest urban centers north of Northgate (and top five north of the ship canal). Twenty minutes by bike, ten minutes by car. Over fifty minutes by bus (involving two transfers) even though you caught the first bus within a minute, and the last one is extremely fast and frequent. Why does that trip suck so much? Because we don’t have a grid.

       Back to Alon:

       In the absence of a radial rail network to connect to, long grid routes are less useful.

       Remove the double negatives and you get: A grid is more useful with a radial rail network. Yeah Alon. Duh. Walker mentioned that as well. Remember the “weak” anchors to the east? Walker again:

       The east end of these lines is a little more various, but many of them end at SkyTrain stations on the original Expo line, the blue-and-yellow line that runs southeast out of the city. This SkyTrain line extends well out into the suburbs of Burnaby, New Westminster, and Surrey. Most of its stations are forests of residential towers, or major commercial centers, or are trying to be. So there’s plenty of demand for access to SkyTrain; this keeps the east-west lines busy to the east end.

       What he didn’t mention — what is so obvious that it wasn’t worth mentioning — is that two of the three SkyTrain lines go downtown. Alon Levy again (same paragraph that contained the double-negative sentence):

       Cities have a center and a periphery, and the center will always get more ridership, especially transit ridership.

       I guess, sure. What are you getting at though Alon?

       The outermost grid routes are often so weak that they should be pruned, but then they weaken the lines they connect to, making it necessary to prune even more lines until the grid is broken.

       Right, except the bus lines you are so worried about — the ones you think will be pruned — are actually remarkably strong. Why? Because they connect to SkyTrain! Someone who is at the far edges of the grid (or part of some other sub-grid, like New Westminster, which has streets running at a 45 degree angle) does not have to worry about the fact that their bus doesn’t go out radially, connecting it to the one place Alon thinks everyone is trying to get to (downtown).

       Meanwhile, Alon completely glosses over one of the main benefits of grids: They save money! If your goal is “anywhere to anywhere” transit, then it is by far the best approach (https://humantransit.org/2010/02/the-power-and-pleasure-of-grids.html). Is Alon suggesting that there are some parts of the world where people aren’t interested in “anywhere to anywhere” transit, and will only take it on rare occasions, to the one major destination in the region? Again, duh. Walker mentions this from the very beginning. Grids don’t make sense everywhere. Transit doesn’t make sense everywhere.

       All of which is a very long way of saying: Who cares what Alon says about grids? Jarrett Walker has covered the subject eloquently and thoroughly. He would be the first to tell you that grids are not always appropriate — let alone possible.

       Alon has some interesting things to say. But those things are basically subjects Walker hasn’t written about. When they both write about the same subject, you will find that Alon offers little new information, and merely clutters up the argument. This is no exception.

       I’ll deal with specific suggestion later.

       1. > Walker is a transit expert. He simplifies things for the reader, so they can understand complex, scientific principles that have been well researched. Alon is a … um … I’m not sure what Alon is. But Alon reminds me of this comic strip:

          Alon Levy’s a “transit researcher” I guess would be the best description. They managed to get new york subway construction costs down by analyzing them using excessive station room. Alon was also on the study https://transitcosts.com/transit-costs-study-final-report/ They definitely could articulate their positions better for the general public it is a bit of a technical blog, but I haven’t seen anyone else even other transit experts analyze items this deeply. Or if others have then they aren’t sharing it with others lol.

          > Meanwhile, Alon completely glosses over one of the main benefits of grids: They save money! If your goal is “anywhere to anywhere” transit, then it is by far the best approach

          I’m talking specifically about Seattle in this case, we can discuss more broadly about transit grid’s in general if one wants. Anyways the main point

          > All of which is a very long way of saying: Who cares what Alon says about grids? Jarrett Walker has covered the subject eloquently and thoroughly. He would be the first to tell you that grids are not always appropriate — let alone possible.

          The main thing I am questioning the transit grid idea that Walker talks about ‘one transfer’ outlined here. That also implicitly suggests these east-west or weaker circumferential routes are useful.

          Walker: “In an ideal grid system, everyone is within walking distance of one north-south line and one east-west line. So you can get from anywhere to anywhere, with one connection, while following a reasonably direct L-shaped path.”

          That Alon questions:
          “However, I contend that these routes don’t really form a grid, at least not in the sense that passengers ride between two arbitrary points in Chicago by riding a north-south bus and connecting to an east-west bus. Instead, their outer ends form tails, which people ride to the L, while their inner ends are standard circumferentials, linking two L branches.”

          Are people really using it in such a fashion? Or more concretely is there really that many situations where that “one transfer” effect is useful for Seattle to propose creating more these east-west transit lines.

          For north Seattle if you are in ballard one will take D line, not take the 44 over to link. Same with fremont, just take the 40/62 down. One might argue that it takes longer between two arbitrary points in north seattle and it’d be easier with just ‘one’ transfer if it was more north/south and east/west routes — but if they end up with less ridership and less frequency it’d just end up worse than before.

          I mean even with the Link, it still doesn’t make much sense. To name a specific example, look at Renton to Seattle. None takes the F line to TIBS and then link up. They just take the 101. Or let’s say in south seattle (rainier valley) sure one could run more east-west busses, but honestly how much ridership would they have. They’d have so low ridership they’d run infrequently, or practically just run in an L fashion.

          tl;dr: embrace the L, don’t try to force east-west lines.

       2. Alon specifically critiques grids because they are inefficient in the peripheral:
          “The scale, then, is that rapid transit is so expensive that there’s no money for a complete grid, making a radial plan more appropriate. But surface transit, especially by bus, can be spread across a grid more readily.”

          Alon is making an important contribution here – scale matters. East/West travel within the Seattle isthmus is travel at bus scale, simillar to travel within a NYC borough, and therefore a proper grid is important. This is why the E line runs parallel to Link and does not bend towards it until downtown (where it is actually Link that bends west) – E and Link are separate lines on the grid. In contrast, Swift E bends towards Link at the end (185th Station) – trip pairs within the Swift line are “bus length” but traveling onwards to Seattle from a Swift stop is a “rapid transit length” trip.

          So yes, when designing a bus network within Seattle city limits, Alon isn’t really adding anything to Walker because Seattle exists at a bus scale. But designing a network across the county & and the ST district he makes an important point, which is 1. a radial network is a far more efficient use of scarce resources, and therefore 2. outside of the urban core, surface transit needs to feed rapid transit, otherwise the bus grid lacks sufficient ridership anchors to be cost effective.

          If your city’s arterial network is an unrelenting grid, it’s possible to build a good feeder network that also functions as a grid, but if the local network can’t do both, feed the rapid network first. This is why if a Link station is in an awkward location, it is still valuable to divert the bus to make the connection. This is why Pierce will dogleg route 1 to serve Tacoma Dome station, and why KCM has a cluster of route converging on 147th station rather than trying to build a set of routes that ‘pass by’ 147th.

       3. Alon is a mathematician who decided to apply their statistics skills to transit networks.

          (I believe Alon prefers “they”, and I corrected the pronoun in my previous comment.)

       4. “To name a specific example, look at Renton to Seattle. None takes the F line to TIBS and then link up.” That’s because Link through the RV isn’t “Rapid Transit.” Alon studiously uses the term “rapid transit” in lieu of rail. For a Renton/Seattle trip pair, Link is functionally a streetcar – the rail mode has greater capacity but no speed advantage over a bus.

       5. “Seattle exists at a bus scale.”

          Do you realize it takes 45 minutes to get from Westlake to Aurora Village on the E, the fastest bus route? Or 50 minutes to get from Westlake to Rainier Beach on the 7? That’s almost two hours from end to end. Or over one hour from Columbia City to 100th. Even just taking the 48+7 from the U-District to Columbia city takes forty-five minutes. What about going from 45th to Costco (S 44th) on the 28/131/132 interline? 40 minutes, and always 5-15 minutes late every day. Want to go to Bellevue? The 550 takes 30-45 minutes. 550+B to Crossroads? 55-70 minutes. People need a faster way to get around, at least for a portion of their trip.

       6. > Alon specifically critiques grids because they are inefficient in the peripheral: “The scale, then, is that rapid transit is so expensive that there’s no money for a complete grid, making a radial plan more appropriate. But surface transit, especially by bus, can be spread across a grid more readily.”

          I mean even for North Seattle, would it really be better if our current ‘L’ routes were turned into north/south and east/west routes? And one can still reach Link in most cases with the L routes east-west section.

          The 62 heading to Sand point. And the 40 heading to northgate. https://seattletransitmap.com/app/

          I’m not saying our transit routes should be circuitous for no reason but I’m kinda questioning the whole benefit of the ‘bus’ grid and idea the missing direct east-west bus lines are bad for seattle area compared to our current L routes

          Outside of seattle itself, there’s rapidride A and future rapidride I both somehow miss southcenter. and we have rapidride F awkwardly trying to fit in. Over on the eastside, the 245 and B line both form two L’s. Sure they are straightened out in the future proposals, but do they really bring that much more benefit than before?

          I mean even the Stride 3 is basically a straightening of bus route 522 to be more east-west rather than diagonal and that was decried by lake city.

       7. “Alon is making an important contribution here – scale matters.”

          It does matter, and I think Jarrett is either not taking into account a large city, or Alon is distorting Jarrett’s argument. Jarrett has spoken approvingly of a radial or radial+orbital subway on top of a gridded bus network — the same as Alon does.

          If we look at grid subway networks, they aren’t that common or extensive. Manhattan has east-west 14th and 42nd on top of the north-south Avenues. Brooklyn has one north-south G line, which is relatively underused. All the other subway lines in Brooklyn, Queens, and the Bronx are radial.

          Toronto, when construction is finished, will have two east-west lines, two north-south lines, two east-west feeder lines, an L-shaped line, and two streetcars. That’s one of the more extensive grids. It works well for Toronto, but is not necessarily replicable in a smaller, less dense city.

          London has a smorgasbord of lines, kind of like a grid on LSD. It works, but London is a very large city, like New York or Toronto. It doesn’t necessarily apply to cities like Seattle. The only crosstown routes that have even been proposed in Pugetopolis are a 45th line, a Renton-Burien line, and a vague 405 concept. The latter two are unlikely. Even with all three, it only marginally enhances train-to-train transfers, and leaves vast areas without any train-to-train crossing or even any train at all.

       8. “The 62 heading to Sand point. And the 40 heading to northgate…. I’m kinda questioning the whole benefit of the ‘bus’ grid and idea the missing direct east-west bus lines are bad for seattle area compared to our current L routes”

          Where are you getting the idea that we have to break routes like the 62 and 45? The 62 looks arbitrary but it’s very useful for getting from Roosevelt to Fremont, Greenlake to SLU, etc. If you split it, then given Seattle’s 15-minute frequency and congestion and unreliability, you might wait 15 minutes for a transter, That turns a 15-minute trip between two nearby villages into a 30-minute trip. Ballard to Greenlake or Lake City is practically hopeless now, since Ballard isn’t on the 62, the 40 doesn’t go to Lake City, and even when the 75 did go to Ballard it was slow. There could theoretically be a route from Ballard west to Greenwood Avenue, transferring to a route to Ballard, but a diagonal route really would be better to connect the southwest and northeast parts of North Seattle.

          “Outside of seattle itself, there’s rapidride A and future rapidride I both somehow miss southcenter.”

          Southcenter is on the other side of a ridge from the A. The I is too far east, and should be extended north to Rainier Beach Station instead.

          “Over on the eastside, the 245 and B line both form two L’s. Sure they are straightened out in the future proposals, but do they really bring that much more benefit than before?”

          There are a lot of overlapping trips in the B corridor. The 245 connects the second and third largest cities (Kirkland and Redmond), and connect them to Microsoft, Crossroads, and Eastgate. Both the B and the 245 are worthwhile, and their overlap is not problematic. The overlap is where a lot of trips on both routes start and end, and is an equity-emphasis area. I’m tentatively supportive of the proposed split (Redmond-Eastgate, Crossroads-Bellevue-UDistrict, Redmond-Kirkland), but I have some reservations about whether splitting the B and the 245 is really a good idea. The strongest trip pattern is the current B (8th+156th). There’s not much south of 8th except the college. And the 245’s L may be doing something useful (Kirkland to Microsoft and Crossroads). So I’m supportive of the split but somewhat ambivalent.

          “even the Stride 3 is basically a straightening of bus route 522 to be more east-west rather than diagonal and that was decried by lake city.”

          That’s not really a grid issue. The 522/Stride is mainly for Northshore cities and Shoreline. Lake City was only added because 145th station didn’t exist yet, I-5 is so congested, and its predecessor went through Lake City (307). Northshore strongly insisted it wanted Stride to go to 145th station, thinking that would be the fastest way to downtown Seattle, and it didn’t think Lake City was important. It also didn’t think an extension to Bitter Lake (Aurora) and Shoreline CC was important. And since they’re the ones paying for it, that’s what they got.

       9. Or more concretely is there really that many situations where that “one transfer” effect is useful for Seattle to propose creating more these east-west transit lines.

          Yes! Look at the engagement survey for the Lynnwood Link restructure project.

          Top Phase 1 Needs Assessment Themes

          1. Create new and improved east-west transit connections.

          These aren’t folks who have read Jarrett Walker’s books, or looked at the various transit studies in great detail. I would guess that most of them have no idea what a “transit grid” means. Yet the very first priority is east-west transit connections.

          That is because without, transit sucks. Not everyone is going downtown. True story: I was hanging out at the Beer Authority, a bar in Lake City. I was chatting with a woman who said she was a bartender. I asked her where and she mentioned some place up on Aurora (basically similar to Bitter Lake). I told her the commute doesn’t seem too bad and she started to tell me the various buses she needed to take. I felt terrible, and really stupid. I just assumed she would drive. But she doesn’t, and her commute sucks. Her commute, mind you. Never mind that she might want to just go somewhere else in the city that happens to be on the other side of I-5.

          For north Seattle if you are in ballard one will take D line, not take the 44 over to link.

          Not if you are going to the UW. Or Northgate, or Roosevelt.

          Same with fremont

          Yep, same with Fremont.

          tl;dr: embrace the L, don’t try to force east-west lines.

          So you mean get rid of the 40 and go back to running the 18 all day long?

       10. @Ross

           And to form those east west lines you’d end up breaking a lot the routes connection to Seattle. Wouldn’t someone else then lament about how hard it is to reach Seattle

           > So you mean get rid of the 40 and go back to running the 18 all day long?

           Confused here I said I’m a proponent of L routes. Unless you’re talking about a different route the 18 looks like just a north south route?

       11. So yes, when designing a bus network within Seattle city limits, Alon isn’t really adding anything to Walker

           Yes, exactly.

       12. > So you mean get rid of the 40 and go back to running the 18 all day long?

           Confused here I said I’m a proponent of L routes. Unless you’re talking about a different route the 18 looks like just a north south route?

           It is a north-south radial route. The combination of the 15, 17 and 18 are very radial like. They all follow the same pathway towards downtown. They offer some connectivity besides going downtown, but not a lot.

           In contrast, the 40 does something different. It goes east-west, connecting riders from Ballard to Fremont. Likewise, up north it goes east west, connecting Northgate riders to Ballard.

           The trip from Ballard to downtown is significantly slower. Yet the 40 works so well because so many riders are *not* going downtown. Consider the northern anchor — Northgate. Prior to the pandemic more people boarded there than any other bus stop. Do you think the are going downtown? Of course not. There was the 41. They must be heading west. Nothing else really makes any sense. The parts of the trip that are most “grid like” are the most successful.

       13. And to form those east west lines you’d end up breaking a lot the routes connection to Seattle. Wouldn’t someone else then lament about how hard it is to reach Seattle.

           What are you talking about? Everything I mentioned was in Seattle. West Seattle, Magnolia, Ballard, Lake City, Bitter Lake. They are all in Seattle.

           If you want to talk about the suburbs, leave me out of it. I’m not going to defend some sort of mythical greater-regional-grid connecting Covington to Issaquah. On that we agree — it just won’t work.

           What I’m saying — what a lot of people have been saying for quite some time — is that Seattle itself should have more of a grid, so that we can increase frequency. A lot of it has actually happened. Since David wrote this fantastic essay ten years ago a lot of the changes he proposed have been implemented. Or at least something similar. Notice that his proposal is not for the region — only for Seattle.

           Of course it might be a stretch for some areas to try and adopt more of a grid, but not much of one. West Seattle suffers from really low frequency — in part because they have too many buses going downtown. Rainier Valley could send more buses to Link if it was wasn’t for the fact that Link has terrible stop spacing (which then require a “shadow”). Even then you can make the case that the 106 should go up to First Hill (and then follow Boren to South Lake Union). If they could figure out how to run buses on 65th then Ballard could become much more of a grid. It needs work in the north (with the D/40). That shouldn’t be too hard to fix, although you end up with some redundancy simply because of where they put the Link stations and the street grid. Northeast Seattle will become more grid like with the addition of the 130th Station. I feel like the planners could do better, but that has little to do with whether they will send buses to downtown (they won’t).

           That leaves the greater Central Area. Should it become more of a grid? Hell Yes. By making a handful of modifications you could make it more grid like (https://seattletransitblog.com/2023/08/30/high-frequency-network-surrounding-rapidride-g/). With that proposal you have dramatically improved frequency. Now add service on Boren and you’ve got a great grid. It is challenging because you have two street grids (the one following the cardinal directions, and the one matching the waterfront). But it still follows the same basic principle: Whenever possible, keep the bus going straight. Only a handful of buses deviate to go downtown. Most of the buses that go downtown are heading that way (e. g. the 2, 3, G). The 49 keeps going south along Broadway to First Hill. The 8 keeps going east to Madison Park. The 48 keeps doing what it has been doing (quite successfully) for a long time. It is not a complete grid, but it is a lot more grid like than what we have now. Frequency would be a lot better. A lot of trips that are a pain in the ass right now become a lot simpler.

       14. I mean even for North Seattle, would it really be better if our current ‘L’ routes were turned into north/south and east/west routes?

           I’m not sure what you even mean by that. You give examples of the 62 and 40. I already covered the 62, but I’ll briefly summarize it. Yes, it would be better (albeit marginally) if the 62 was split. For example, Sand Point to North Beach, Lake City to Fremont/downtown. The strength of the 62 is not that it goes downtown, it is the fact that it goes east-west and connects to Fremont. The 40 is another hybrid. It goes east-west for much of its route. Even a lot of the north-south section is not especially efficient (especially if your goal is downtown). As good as it is, it would be difficult to turn it into more of a traditional grid like route. The geography is challenging.

           For an example of an “L” shaped route that I would fix tomorrow, consider the 49. I would send it to Beacon Hill. That would replace that part of the 60. So you’ve saved a considerable amount of money, which means that buses run more frequently. This bus would run opposite the streetcar. So we are talking maybe 12 minutes for each route, which means six minute combined headways from CHS to Jackson. On Broadway! All those people who want a one-seat ride from somewhere to First Hill? Sorry, but at least you have six minute all-day headways (in the middle of a driver shortage).

           What a more “grid like” trip? OK. How about this. You are up at Aloha & 10th, and you want to get to Swedish Cherry Hill. Here is what Google says: https://maps.app.goo.gl/Ny3Zfujzh1yC6xE39. I can walk twenty minutes and be there in a half hour. If I want to walk “only” 11 minutes, it will take me almost forty minutes. Wow, Uber is looking mighty tempting.

           Here is what I would say if someone asked me about the trip (and I didn’t have my phone): Let’s see from here you can take the 49, the bus stop is right there. Oh wait, that doesn’t do you much good. OK, take the 9 … oh wait never mind, it isn’t running now. The streetcar is a long walk … so forget that. How about this: Walk down this street until you can catch the 60. Then ride it 60 southbound. It will turn, as if it is heading downtown, but then it doesn’t really head downtown. It turns again. As it gets to James you can then catch the 3/4. Yeah, that will work. That will get you there. Eventually.

           Instead it would be: Catch the 49, transfer to the 3/4. Have a nice day.

       15. @Ross

           > What a more “grid like” trip? OK. How about this. You are up at Aloha & 10th, and you want to get to Swedish Cherry Hill. Here is what Google says:

           Idk what settings you have setup, but when I opened google, it said one can just use the 60 and the 4 and reach there in 32/33 minutes including a 5 minute walking time. I changed it to 12:15 and it suggested using the 60 and 3 again in 32 minutes. Later at 5/6 pm it slips to 37 minutes due to the transfer slipping.

           https://www.google.com/maps/dir/47.6269628,-122.3204899/47.6068595,-122.3107012/@47.6132527,-122.3241715,14.4z/data=!4m6!4m5!2m3!6e0!7e2!8j1707134400!3e3?entry=ttu

           Anyways not to say there aren’t definitely tough transfers with the L routes.

           > The strength of the 62 is not that it goes downtown, it is the fact that it goes east-west and connects to Fremont. The 40 is another hybrid. It goes east-west for much of its route. Even a lot of the north-south section is not especially efficient (especially if your goal is downtown)

           Yeah that’s the root of what I’m questioning. I’d wager that the strength of the 62 does come from it going downtown.

           > As good as it is, it would be difficult to turn it into more of a traditional grid like route. The geography is challenging.

           But that is the city and geography that we have. Providing east-west routes as tails of the L routes to downtown is a fine and good approach for the city we have.

    5. Notably in North Seattle there’s many of “L” or C” pattern bus routes like the 40, 31/32, 62, 75 etc. Restructuring these routes as north south ones with separate east-west routes would end up cutting off their connection to downtown and with the loss in ridership and subsequent frequency, they’d be even harder to use.

       There are two big reasons why it is difficult to make a grid like they have in Vancouver: Geography and Link. First off, there are places where water breaks up any chance of a grid (Green Lake, Lake Union, ship canal) or steep hillsides make east-west travel impossible (Queen Anne). Then there is Link. Link helps the grid immensely, but not as much as SkyTrain. But even Vancouver’s grid isn’t perfect — nobody has a perfect grid. The point is, generally speaking the more “grid like” your system, the better it is at delivering anywhere-to-anywhere transit.

       As to the routes you mentioned:

       62 — The 62 goes downtown, but few riders on 65th use it for that. It makes way more sense to take Link. It is basically two segments combined into one. It could easily be chopped up and put back together in a more grid-like way. For example: Take the 522, but start the bus at Lake City Way instead. Now combine that with the southern end of the 62. Combine the eastern end of the 62 with the 45. There would be winners and losers, but you’ve eliminated some three-seat rides. Is it worth it? Maybe. I would say the biggest flaw in the 62 is the Woodlawn/Kirkwood section. Metro tried to fix that a while back, but the city rejected it. The streets needed to be hardened to allow a more straightforward path. I’m not sure anything has come of that idea. Anyway, the fact that it the 62 goes downtown is not the key here — it is the places it serves “on the way”. It connects to places that Link doesn’t: Fremont, Dexter and South Lake Union. This would be true if the bus came from the north or northeast as well.

       31/32 — The 31/32 is fairly grid like. It is an east-west bus that stays pretty east-west. It deviates to go north simply because there is no Link stop between UW Station and U-District Station. If there was a Link station at Campus Parkway, it would end there. If there was a station inside the campus, it would end there (or continue and be part of something else). But overall it is pretty straight east-west, especially since it goes east-west across a segment that buses didn’t use to go on (NE 45th Street between The Ave and Children’s Hospital).

       The 31/32 is of course, two different routes that branch in Magnolia. How they branch is another subject, but options would neither detract or enhance their “gridness”.

       75 — The 75 basically just hugs the shore most of the way. It is mostly north-south but heads east-west because it has to. It is only in Lake City where it deviates. This section could definitely be handled by a different bus (and be done differently). The 75 could keep going north. It could go all the way to Kenmore, although then you have overlap. The overlap is caused by the geography (buses converge on Lake City in part because the street grid converges on Lake City). Meanwhile, the east-west route on 125th/130th does that abrupt dogleg because of Link. When the 130th Station is added, the bus that goes across 125th should keep going to connect to Bitter Lake, connecting to Link in the process. There is a case for that right now, but not as strong of one. There is north-south grid route. Only parts of it.

       40 — The 40 serves greater Ballard or Northwest Seattle (whatever you want to call it). South of Market it is relatively grid-like. It is a key east-west line between Ballard and Fremont. Without it, folks are forced to transfer, sometimes more than once. For example, imagine there was no 40 — just the 15, 17 and 18 in Ballard. Now try to get from 65th & 24th to Fremont. You have to take the 40 to Market, then the 44 east to 8th Avenue, then take the 28 south. Yuck.

       I suppose it could continue east (and be part of the 31/32) but that becomes a bit redundant. Unless the 31/32 took over the southern tail of the 40, but then what? You have one bus that goes along the north side of the ship canal, and another bus going along the other side, and they never quite meet. Sometimes you have to weave slightly to provide more connections — connections that are essential for a good grid.

       Which brings me to the northern section of the 40. Even though there is nice set of north-south buses in Northwest Seattle (greater Ballard, if you will) there are no crossing buses between Market and 85th. This is again due to the geography (running a bus on 65th is hard). You still want to make connections while serving the north-south corridors. To do that, the buses weave. You’ll notice that the 40, D and 28 all connect to each other. It isn’t perfect, but it at least allows riders to switch from one major corridor to another (https://maps.app.goo.gl/9gD5dqkckRyFZWXSA). I would say the only major weakness is the northern tail of the D. It connects to the 28, but that really doesn’t give you much. In contrast, if it could just go a bit farther and connect to the 5, that would allow riders to access 15th NW like they access 24th NW (https://maps.app.goo.gl/TCcpoUNBLiGuJCvd7). It basically ends before making the various connections (Greenwood, Aurora, Northgate, Lake City).

       So let’s take this concept and apply it without too much redundancy. Ideally it would work like this: The D takes over the northern route of the 40. The 40 straight across, then north to Northgate and on to Lake City Way. Now all the connections are made, even though you still don’t have service on 65th.

       Restructuring these routes as north south ones with separate east-west routes would end up cutting off their connection to downtown and with the loss in ridership and subsequent frequency, they’d be even harder to use.

       I don’t quite follow you. In northeast Seattle we have “cut off” the connection to downtown, and it is fine. People transfer to Link. In Northwest Seattle, it comes down to the physical geography. Would an east-west bus like this get a lot of riders? Yes! Definitely. That would change the nature of transit in Northwest Seattle. But how are you supposed to run a bus like the D without going downtown? If anything, the problem is that the D is trying to do too much. It should just go straight downtown instead of deviating to Uptown. The 40, meanwhile, is also taking a very roundabout route to downtown — acting more like an east-west bus for much of its journey. Meanwhile, the 28 converges unto Fremont, and ending there is actually quite reasonable given its low frequency. It has a fast connection to downtown that apparently a significant portion of the population doesn’t care about.

       Which leaves what, Queen Anne and Magnolia? With Queen Anne you can’t go very far east-west. The only logical place to go is south (towards downtown). You could send the 1, 2 and 13 over to Boren. That could work, except you would need to fix Denny first.

       That leaves Magnolia. I could definitely see the 31/32 being the only buses to Magnolia. You would need to bump up the frequency of the D (and in my opinion, straighten it out). Then increase frequency so that riders in Magnolia get 15 minute headways on each leg. Then riders take the bus and transfer to the D to get downtown.

       So yeah, even in Magnolia and Queen Anne you can make the case that the buses are too downtown centric. By no means is this the biggest weakness in our grid, but I would definitely be fine with those changes (as long as the buses aren’t stuck in traffic).

       West Seattle to South Seattle, the route 50 is practically useless. Honestly might as well just redirect and turn it into two lines(say columbian way and alki) that head to downtown. Whatever time savings from avoiding a transfer are swamped by waiting for the 30 minute frequency.

       Except if you send the bus downtown you are running each bus less often. That means what, hourly buses from Alki, but at least they run directly downtown? Sorry, no. That is backwards. We need to pick a small subset of routes from West Seattle to downtown (I suggest the 21, C and H). Then run other buses (like the 125, 50 and 128) more often. It is really the principle behind this: https://seattletransitblog.com/2013/08/19/your-bus-much-more-often-no-more-money-really/. People have to transfer — even to get downtown. But in return their bus comes a lot more often. For that matter it was the thinking behind the U-Link restructure. Major bus routes — the most productive in our system — were suddenly chopped. It was no longer possible to get from various parts of the north end to downtown without making a transfer. Even from the U-District it was a two-seat ride. And the transfer was not great, either — it is one of the worst in our system (that tunnel is deep). But guess what? The buses came a lot more often.

       [Ed: Fixed boldface that was bleeding into the following comments.]

       1. Anecdotal for sure: but the times I’ve taken the 40 from Northgate to Ballard, there were a large number of passengers that got off and transferred to the D as soon as that was an option. It seems like there is definitely unmet demand for RapidRide further north.

13. “1028 RR + 3101 RRIC” has a notable feature: it’s a route that has a terminus at Crossroads. These Metro Connects maps show the line running into the Crossroads Mall parking area, which I presume is just a marker-on-paper aspiration. In the current built environment of the Crossroads neighborhood, buses do not lay over; I suspect that this is in large part because there isn’t a convenient location for a bus to do so (at least with all paved areas as currently owned/marked). For comparison, the East Link Phase 2 and Phase 3 proposals both totally avoid terminating a bus route at Crossroads.

    If one were already committed to somehow carving out a place for an articulated bus or two to lay over at Crossroads, how much further of a step would it be to commit to a small transit station? I’m thinking specifically of a small facility mostly located on the eastern edge of 156th (but with buses still stopping along either side of 156th unless they intend to lay over), generally non-enclosed (open-air in the vein of, say, Kirkland Transit Center), positioned to provide a ped/bike bridge overcrossing 156th.

    1. “These Metro Connects maps show the line running into the Crossroads Mall
       parking area, which I presume is just a marker-on-paper aspiration”

       I don’t understand the fork-shaped tail. Continuing east on 8th would serve more of the Crossroads village. Going into the parking lot doesn’t seem feasible, but maybe there’s a Crossroads/Bellevue/Metro redevelopment agreement we don’t know about. The ends of the fork aren’t connected, so the bus can’t turn in a loop in that arrangement. But maybe there’s a missing line connecting the ends. Can a bus turn left from 8th into the parking lot across traffic? But the terminus is an insignificant issue: what matters is being able to travel from Crossroads to Bellevue TC, north Bellevue Way, and the U-District.

       1. “But the terminus is an insignificant issue: what matters is being able to travel from Crossroads to Bellevue TC, north Bellevue Way, and the U-District.”

          Agreed. (The Metro Connects maps show the RapidRide candidate running through Medina instead of on north Bellevue Way; even so, given Metro’s choice for the upcoming 270 routing, I would think that the north Bellevue Way routing will be favored going forward.)

          A bit of nostalgic trivia: it just occurred to me that a UW-downtown Bellevue-Crossroads RapidRide route looks rather similar to the UW-Crossroads portion of the old peak-only route 272. To my surprise, King County Metro still has a map online from the first proposal of the RapidRide B restructure (i.e. not the final layout that was implemented), showing the 272’s deletion:

          https://metro.kingcounty.gov/up/projects/pdf/2011EastN_peak272.pdf

          Trying out numbers I could recall from the pre-2011 network turned up some other old maps from the same restructure proposal. (Doubtful this is all of them; for instance, I think the old 225, 229, and 261 would have been mapped as well.)

          https://metro.kingcounty.gov/up/projects/pdf/2011EastN_221.pdf
          https://metro.kingcounty.gov/up/projects/pdf/2011EastN_222.pdf
          https://metro.kingcounty.gov/up/projects/pdf/2011EastN_233.pdf
          https://metro.kingcounty.gov/up/projects/pdf/2011EastN_240.pdf
          https://metro.kingcounty.gov/up/projects/pdf/2011EastN_249.pdf
          https://metro.kingcounty.gov/up/projects/pdf/2011EastN_peak256.pdf
          https://metro.kingcounty.gov/up/projects/pdf/2011EastN_peak265.pdf

          The proposed 233 routing in particular is an amusing look at what could have been.

          A little more digging shows that the website for the first iteration of the RapidRide B restructure is still up, albeit with many of the links dead or malfunctioning:

          https://metro.kingcounty.gov/up/projects/Bell-RedRoutes.html

       2. When I was in school in the 70s and 80s the 252 went from Eastgate to Lake Hills, 164th, NE 8th Street, Bellevue TC, Medina, and the U-District. The 252 express got on 405 and bypassed downtown Bellevue and Medina. I lived east of where Northup crosses 8th, so I took the 226 to downtown Bellevue or downtown Seattle, or walked up to 164th and took the 252 to the U-District or the Lake Hills library. After that I lived on Somerset Hill. Westbound the 252 started halfway up the hill on Newport Way so it was a 10-minute walk away, but eastbound it terminated at the bottom of the hill so it was a 30-minute walk. So instead I came back on the 210, 240, or 340, and walked variously from Newport Way, Factoria, or the Coal Creek Parkway exit. Those were longer walks but avoided the steep uphill, and the Newport Library was on the way. (That library had nice window seats.)

          Sometime later the 252 local was renumbered to 271 and rerouted to Issaquah, and the express was renumbered to 272. That may have been when the 550 started in 1999.

          Metro has gone back and forth about rerouting the 271 to north Bellevue Way. The 270 proposal to move it to Bellevue Way is, I understand, a precursor of the the RapidRide routing would be, like the 160 restructure prefigures RapidRide I. If some maps have the RapidRide in Medina, that may mean some paperwork hasn’t caught up with others.

14. Along the grid theme, I think there really should be a bus on 65th between Ballard and Phinney Ridge, even though Metro’s 2050 network omits it. They do have partial coverage of 65th to 3rd, but the bus really needs to continue to at least 15th or 24th and serve Ballard before ending.

    That begs the question of what the bus should do east of Phinney Ridge, as 2 miles for a bus route is normally too short. My idea is to implement the 65th St. route as a westward extension of the 61, rather than an entirely new route, thus resulting in a NE-SW route that zigzags from Ballard all the way to Lake City, connecting to lots of neighborhoods and N-S bus routes, plus Link (at Northgate station) along the way.

    I think such a route would have a lot of potential. Currently, places like north Ballard and Phinney Ridge, east/west travel on a bus is a real pain, this route would fix it – on top of all the benefits RossB cited that the shorter route 61 – from Greenwood to Lake City only – would provide. It doesn’t perfectly follow the grid lines – the geography won’t let it – but still a very grid-like route that complements all those north/south routes very nicely.

15. I just took a look at your website but I couldnt find a chat box anywhere. Are you still in business? If so, you’re probably losing out on customers that are trying to reach you. Go to google leads pro .com thats http://www.googleleadspro.com. It puts all your customer conversations on your phone. Hope that helps!

Comments are closed.