Route 8 Stop Level Data
Route 8 Stop Level Data

I’m sure most of you are familiar with these charts by now; if not, check out this post. After the jump, what I see in the data, along with a map of the route.

By popular demand -- A map of Route 8
By popular demand -- a map.
  • Ridership is solid at all times between Seattle Center and Group Health — the oldest part of the route — although a strong commute pattern from Capitol Hill to SLU and points west is evident. This is the only section with good night ridership.
  • Ridership is also strong between Mount Baker Station and Rainier Beach Station, except at night. Stops near stations — particularly Mount Baker — stand out as particularly busy, as do the stops at Graham St.
  • On-off activity through the Central District and Madison Valley is weak, except at the stops around the commercial district centered on 23rd & Yesler, where the 8 deviates from MLK…
  • …but, average passenger loads stay pretty solid throughout the same section, except at night, suggesting that many riders are using the 8 as a one-seat ride from the Rainier Valley to Capitol Hill and points west; basically, a very slow 9, but with a much better level of service.
  • Loads are greatest in the middle of the day throughout the route, except during the eastbound PM peak between SLU and Capitol Hill, a pattern we’ve not seen before in previous ridership pattern posts. I suspect this is because the 8 is the first route we’ve considered that doesn’t serve downtown — it primarily connects neighborhoods.
Share and enjoy.

46 Replies to “Ridership Patterns on Route 8”

  1. Another thing to consider: Denny is nearly impassible from Queen Anne Ave to Stewart each week day during evening commute (from 5pm). Are those morning west-bound boardings all elves who are setting sail from Elliott Bay? Are people going through downtown on the way home to avoid Denny? Are they staying near work for a drink?… or is it normal to have a strong morning commute not match up with a return home?

    Denny may be influencing the anomolous eastbound ridership pattern you mention.

    1. In my experience, Denny is nearly impassable in *both* directions during peak periods, and whenever there’s any kind of special event going on downtown.

      As much as I love the 8’s crosstown connectivity, Denny is just way over capacity. It’s a freeway onramp, and it connects fifty zillion different neighborhoods. Either we need transit lanes, or (more realistically) we need to pick another street (sounds like Harrison might be the lucky one), designate it as a transit corridor, and reroute the 8 to there.

      1. This has been kicked around a lot, and I think it’s a great idea. The 8 could go in a loop around Seattle Center, then take Harrison over to Fairview, then get on Denny to cross over the interstate. This would get it out of the line of cars getting on Aurora, and in the section of Denny from Fairview to I-5 the bus can stay in the left lane, out of the way of cars getting on the freeway.

      2. But Harrison doesn’t cross Broad or 99. Do you mean take the Mercer mess, then hop down to Harrison? Or Denny–>Broad–>Harrison (after the tunnel, so we can get past 99)? In that case, why not Thomas?

      3. Matt: I’m not making this plan up; it’s part of the official SLU Mobility Plan. There’s already a plan to reconnect the grid across Aurora (to be renamed 7th Ave N) up to Republican, and from the look of things, they’re planning to build a bridge over Broad as well.

  2. I think this data confirms that Metro could add in some additional runs of the 8 that just go between Queen Anne and Group Health, as they used to do several years ago. That’s the section with the most ridership and activity at all times, and it looks like the peak loads are really high and could use some relief. I’ll also throw in that this section of the 8 should be a priority for electrification, although they might want to wait to see if rerouting it onto Harrison is going to happen before stringing up the wire.

    1. That is one of the reasons I would like to see the 43 changed to go along Denny when Link opens to the UW. Coordinating this with the 8 would provide 7 to 8 minute service on the common portions of the route.

    2. Agree, and worth noting that Harrison will connect across 99 when the tunnel opens. Reconnecting Harrison, Thomas, and John should relieve some of the heavy volumes on Denny. Doing the same across I-5 would be even better.

  3. “many riders are using the 8 as a one-seat ride from the Rainier Valley to Capitol Hill and points west”

    I saw that one Saturday afternoon. I got on at Graham and off at Summit, and eight people remained onboard the entire time. Another 8-12 got on and/or off along the way. Several at MBTC, three on Jackson and Yesler, one on Madison, several along John/Olive. But only two along the entire stretch of MLK between MBTC and Madison.

    I have never seen the point of serving both 23rd and MLK, although I am glad to see more service in the CD. But I read that Metro likes the 8 because it avoids congestion the 48 suffers.

    1. Come 2016, I can’t possibly imagine the 8 being time-competitive with Link for Ranier Valley->Capitol hill trips, even if you have to walk a little further.

  4. This is one of the more severe cases of Seattle needing more EW capacity. In a dream transit world we could convert Denny to have transit lanes, but that’s unlikely at the moment. Are there any reasonable subway or elevated routes we could start dreaming up?

    You know what my suggestion is going to be. Queen Anne SLU Capital Hill gondola. Skip the crowded streets and highway/freeway roadblocks all together. We could keep the 8 as a local, but connecting these three neighborhoods and adding EW capacity would be really useful. It would take about 10 minutes for the entire route, versus (a scheduled) 40 minutes for this section of the 8.

    1. I actually think this would be fantastic, and probably the best case for a gondola anywhere in the city. No plausible alternative ROW is out there. Are there examples of gondola stations on the top of high-rises? The live load from a gondola is probably small compared to the dead load of a concrete high-rise, so it seems physically possible. You could have two conventional terminal stations and an intermediate station at a high-rise in SLU with a public staircase and elevator partitioned off from the rest of the building.

      Now if only we could figure out where to land it on Capitol Hill we’d be set.

      1. Yes. One of the oldest urban gondola systems has a station on the 15th floor of a building. It was built in 1974 in Singapore. They’ve recently rennovated the system, and there’s a VIP gondola car that costs $700 (though it comes with champagne and dinner for four). There are others, for instance the Medellin/Caracas line had civic buildings like libraries built into each station – however these buildings aren’t very tall.

      2. Here’s a starting plan for the gondola. We can call it the Harrison St. Gondola, and it’ll be an absolutely straight shot from the Key Arena, stopping at some building in SLU, right up to Broadway and Harrison two blocks from the light rail station. All over public right-of-way.

        As to where exactly we put the Cap Hill station, we can either demo & rebuild the building on the corner of Harrison & Broadway, the apartments next to it, or even build it over the street itself. There are wide enough sidewalks there to add an escalator and supports on each side of the street.

      3. Oh, and I was wrong about the time for that route. 6.2 minutes, probably a full 7 minutes after adding time for the stop at SLU.

      4. If we do a top-of-building station, then the stop would be wherever that building is. I believe Terry is the planned pedestrian N/S street for SLU and it has the streetcar, so if we’re adding non-building stations I’d probably put one there.

      5. If we are talking about a gondola to the top of Queen Anne, why not take it over the other side and take advantage of the height of the hill to get across the Ship Canal to Fremont or Ballard?

    2. Since we’re dreaming.

      For part of the route, I think a raised bikeway/pedestrian promenade from Dexter to Fairview connecting rooftop gardens on new highrises would be fancy.

      Gondola is more likely.

      1. Since we’re dreaming, the city should use eminent domain to seize the Space Needle and turn it into a multi-modal gondola/bus/monorail transfer hub. It already has the elevator! Another gondola could connect the Space Needle to the top of Queen Anne, and another to a downtown skyscraper. It could also be a zeppelin docking station, of course;)

      2. [zef] You’re a genius! You just made me realize how we get gondolas across to Bainbridge and Bremerton. The problem is anchoring towers in deep water. But who needs towers when you can keep the wire in the sky using zeppelins!!! Damn that would be beautiful.

      3. Like this. (I probably shouldn’t post things like this. Maintaining credibility while advocating for gondolas is hard enough without goofy almost certain to crash in the first windstorm, if I could even get zepplins large enough to carry heavy steel cable ideas like this.)

    3. I’d like to see signal priority on Denny, possibly queue jumps.

      It’s also worth noting that some of the stops in the graphs above have since been closed and consolidated with others, for example Olive & Boylston.

    4. Why build a gondola when we could spend 100 times as much and build a nice cross-town subway? Oh wait we are only a city of 600k and don’t have the money. :(

  5. (that was supposed to be E&tl;->W and Queen Anne &tl;-> SLU &tl;-> Capital Hill)

  6. I work near Denny right now, and it is annoying to try to get from the Valley to Denny without having to go through all of the Central District. The 8 is a painfully slow-moving route. It’s faster for me to take the train into downtown and walk the .5 mile to Denny.

    Another note: the on/offs on 23rd are likely not because of retail but because it is the main route students south of Mt. Baker use to get to Washington Middle School and Garfield High School. The only other option is to bus / light rail to the Mt. Baker station and catch a 48 from there. Easier and faster to just stay on the 8.

    1. I found I could take Link from Columbia City to Westlake and walk 20 minutes to Summit & Thomas faster than it took for the 8 to get there.

      1. Why would you take the 8 to get to Summit & Thomas from Link? Wouldn’t the 14 be a better option?

      2. It’s south of where the 14 terminates. Normally I’d take Link, but I wasn’t feeling well that day and wanted to see the 8’s route.

  7. I can always count on the route 8 to be 10 to 15 minutes late when it gets to 23rd Ave South and South Jackson heading northbound.

  8. What’s the reason for the 8 to take it’s detour at Yesler and Jackson? The walking distance between 23rd and MLK is only 1/4 mile and it’s relatively flat, especially along Jackson.

    So, if people getting on or off at 23rd save 4 minutes from the detour (5 minutes walking – 1 minute on the bus) and people going straight through lose 4 minutes from the detour (5 minutes detour route vs. 1 minute direct route), the detour only saves people time in the aggregate if more people are getting on or off at 23rd than going straight through.

    According to the ridership chart, the total number of people getting on or off at 23rd is just a little bit less than the number of through riders, if I read the chart correctly.

    However, when one considers that the added running times for route detours like this have to be paid for, the result is a route deviation that’s a wash for passengers on average (good for some, bad for others), and bad for the operation of the transit system as a whole. And that’s before one considers other factors, such as:
    – The “net wash” assumption is based on a 3 mph walking speed. If there were no detour, passengers who are in a hurry can still run and cover the distance faster. Similarly, passengers also have the option to avoid paying the 5 minute walk by riding a bike or kick scooter. With the detour, however, through riders have no option to avoid paying the cost of the detour because there is very unlikely to be a faster way to the destination than sucking it up and staying on the bus.
    – Frequent turns in the route make it difficult to read or do anything besides stare out the window without getting motion sickness. If the bus travels in a straight line, however, making productive use of the time on the bus is much easier.
    – The extra signals required by the detour, especially the left turns in the southbound direction introduce unreliability into a route that is already unreliable. Unreliability is always bad, in part because that means everyone who gets one the bus later, on average, has to wait longer at the bus stop.

    Now, if MLK had frequent traffic jams between Jackson and Yesler and the detour was doing a successful job at avoiding them, then I would concede that the detour is doing more good than harm. However, looking at the map, I see no reason why this would be the case.

    1. I suspect it was originally created for 4 reasons. Transfers to the 4, front-door service to a Seattle Library branch on Yesler, front-door service to an old-folks home on 23rd, and the 23rd & Jackson retail complex which includes the City of Seattle Neighborhood Service Center. That 23rd & Jackson stop is busy at all hours.

      These are all valid reasons for the detour, but I don’t think it’s enough to justify the deviation anymore. It only saves 4-5 walking blocks of relatively flat sidewalk, and it’s an incredibly expensive detour in terms of time.

      Getting on and off of MLK isn’t that bad, but the intersections at 23rd are both very congested. Southbound, Yesler has no protected left turn, and during the busy hours only 1 vehicle can make the turn per light cycle. At Jackson, it has a protected left, but usually has to wait through a light cycle for traffic to clear before it can get to the turn lane. In the northbound direction, it has to wait for traffic to clear to split the lanes and make the right onto 23rd. Heavy pedestrian traffic in this area makes the right turn very time consuming while the light is green, and heavy vehicle traffic makes a right-on-red impossible for a 60′ coach. It’s basically just as bad as an unprotected left.

      I live basically in the center of this detour loop, so it somewhat benefits me. I don’t get any of the drawbacks, because I can board/deboard on whichever side of the detour benefits me, and if I’m going to Mt. Baker station, I can go to the 23rd/Yesler stop and catch either the 8 or 48 with a combined 8 buses an hour.

      1. And the people who lived in the neighborhood and fought to get the original 8 created (trading off other service hours from other routes in the process) wanted it that way.

    2. System-wide, I really wish the Metro planners would go through every route one by one, look at every obvious route deviation from a straight line and ask the following questions:

      1) What purpose does this deviation serve? Which riders is it meant to benefit?
      2) What are the costs and benefits to the riders? How many riders actually benefit from the deviation and how many does it just waste their time? How many riders would the bus gain and lose if the deviation went away?
      3) What are the financial costs of the added running time as a result of the deviation? Does the added running time hit the threshold of requiring additional buses to be added to the route to maintain the desired headway? Does unreliability introduced by the deviation require longer layovers that drivers need to be paid for?
      4) If the route deviation were to be eliminated, are there any simple improvements that can to the street to make the additional walking that would be required easier, safer, or more pleasant? In particular, I’m thinking about things like sidewalks, lighting, signage, crosswalks, and landscaping. Could such improvements pay for themselves with the money saved in operating costs for the deviation that would no longer be necessary.

      Then, when the questions are all answered, Metro should ask the question “if we were creating this route today, would we create it with this deviation?”. If the answer is no, the deviation should be axed. Do this with every route and the number of less-than-useful deviations should be significant enough to add up to some serious cost savings. While not enough to solve the Metro budget crisis by itself, it should be at least enough to make the inevitable upcoming service cuts a little less painful than they would otherwise need to be.

      Yes, there will inevitably be some whiners who will put up a big hissy fit because there will no longer be a be a bus stopping right at their front door. But, that argument is easily countered by asking the question – is stopping right in front of your front door worth ending service for everyone along the route an hour earlier? Or having service run once an hour on Sundays vs. once a half-hour? Or not having a couple of extra peak-hour trips that are desperately needed to alleviate overcrowding? Or cutting service on some other route? As much as everyone likes to believe that their front-door is somehow sacred and deserves special treatment, 99% of the time, the answer is going to be “no”.

    3. It brings up the question of why is there a frequent route on MLK, which bypasses the Jackson commercial area. Why not saturate 23rd with a super-frequent route or two routes, which wouldn’t have to detour?

      1. I was wondering that myself, looking at the route and the ridership numbers.

        Why doesn’t this route run straight up 23rd north of Mt Baker station? Oh, the 48 does that… why isn’t there a nice frequent route up 23rd?

        It seems to be a bizarre hybrid of routes anyway. When U-Link opens, the “one-seat ride” function will vanish in favor of Link. The portion south of Mt. Baker is largely redundant with Link already. Apart from the “one-seat riders”, the portion west of the Hospital appears to have totally different ridership patterns from the portion east of it.

        So, 2016 plan: some sort of really high capacity east-west route on Denny or nearby. A different frequent route starting at Mt. Baker and running due north-south on 23rd. How’s that for an idea?

      2. Err, the 48 is, in fact, already more frequent (or to be more exact, there is a wider span of frequent service on the 48 than the 8).

        If anything, Link should cause us to rethink the existence of the 8 (east of Group Health) altogether, not add another route.

      3. You need a local route paralleling Link anyway. Perhaps the 8 doesn’t need to be the same route once the Cap Hill station opens, but you still need something there.

        The one-seat ride for RV folks going to Cap Hill goes away. And I would hope that taking Link and transferring to QA bus service would be faster than the 8. But all of us north of 90 and east of 23rd still need access. East/west right now on the 2 and 3 is…less than speedy. The 48 is known as the Forty-Late for a reason. Right now taking the 8 to Mount Baker is the fastest Link access we have.

  9. A general question for Bruce: should bus routes with two distinct peaks usually be separate routes so as to serve each peak appropriately?

    1. It depends entirely on the route and what the alternatives are. For instance, if the 48 didn’t exist, it would be essential to have the 8 provide north-south service in the CD, and splitting the current route into three segments would cost a bunch of money.

      On the other hand, if you believe that the 8 isn’t necessary for mobility in the CD, it being close enough to the 48 for most of its length, and you also believe that people going from the RV to Capitol Hill would be better served by making the 9 full time (it’s faster and much more reliable), then it makes sense to split the route and delete the CD segment, and you’d save money that way — particularly if you could find another bus to provide local service on the segment of MLK between RBS and MBS.

      You could, for example, have the 106 not go to downtown, and instead turn on MLK and run to MBS. This would force transfers for a lot of people from Renton, but their travel times to downtown would be only slightly longer, and it would save a boatload of money. You’d basically be quartering the run time of the 8, and that money could be used to bring the 106 up to the 8’s frequency, which would probably offset the slightly slower travel time. Some Metro people floated this idea for the 600k cut scenario. Unfortunately, the low extent of ORCA penetration and the general aversion of suburban riders to two-seat rides probably means that Metro wouldn’t try this now.

Comments are closed.