J-Church, 14-Mission Reliability Improving But Riders Aren’t Seeing It

Flickr photo: ##http://www.flickr.com/photos/brandondoran/4667424156/sizes/z/in/photostream/##Brandon Doran##

Riders of the J-Church know all too well what it’s like to wait for a packed peak-hour train without any guarantee they’ll be able to squeeze on board. If you look at the data, though, the SFMTA says the picture isn’t as bad as it’s been made out to be. Still, two city supervisors aren’t buying it.

The J-line has recorded a 76.8 on-time performance rate since January, according to SFMTA Transit Director John Haley, but Supervisor Scott Wiener wonders whether that statistic reflects the reality of the daily riding experience.

“I think a lot of people who use it regularly would look at that number and laugh,” he said.

Supervisors Wiener and John Avalos held a hearing yesterday to address frequent rider complaints about poor reliability on the J-Church and switchbacks on the 14-Mission line at a City Operations and Neighborhood Services Committee meeting. Haley came to the table with statistical data highlighting Muni’s recent progress, but the numbers were cold comfort to supervisors and riders.

“In a system where [switchbacks are] a common occurrence, I have no real leverage in my district to encourage people to get out of their cars and use Muni. I want to do that desperately,” said Avalos, who admonished the disproportionate impacts of switchbacks felt by those living in the outer neighborhoods. “But…it’s impossible for me to speak about Muni being a reliable service that people should use rather than their cars. The reality does not meet their needs.”

Although Haley included a wish list of potential strategies to improve service, he defended the agency’s progress. Trends are reportedly looking positive since the last hearing, with switchbacks on the 14-mission declining from 180 in January to 116 in March up to the 25th, most of which are done just after peak hours, said Haley. On the J-Church, only 0.38 percent of trips were reportedly turned around early this year, and operator absenteeism has improved.

However, Wiener and riders who spoke at the hearing referred to commonly heard stories of regular extensive delays and wildly inaccurate NextBus predictions. Roughly a dozen students from Mission High School who came to the hearing said having to rely on the J-Church means high truancy rates.

Speakers also bemoaned switchbacks for the lack of communication to riders and long wait times, as well as their particularly harsh impacts on the disabled at non-accessible stops and confusion for residents who don’t speak English.

Haley said the discrepancy between anecdotes and data on the J could be partly due to confusion with an irregularly scheduled 7 to 13-minute headway, coupled with poor NextBus feedback for arrival times. As one solution, he recommended setting an easily expectable 10-minute headway.

“I think the data shows what it shows, but I think there’s also a big concern with the way people perceive how long they’re waiting, and I think we’ve contributed to that,” he said.

The most frequent cause for delays on both the J and 14 lines continues to be problems with poorly maintained Muni vehicles, while double-parked private automobiles remain at the top of non Muni-related causes. Increased parking enforcement, Muni signal priority, stop consolidation, and faster boarding, including ticket machines at J-Church stops, remain among Haley’s top priorities for strategies yet to be used.

The continued use of switchbacks as a technique to manage Muni service has yet to gain much tolerance from Avalos, but Wiener showed relative patience with the issue.

“I understand that Muni needs to use switchbacks periodically to re-balance the system,” said Wiener. “I also understand Muni would like to reduce the number, and that’ll be a good thing, but they’ll only be reduced as we improve the overall flow of the system, so that gives us extra incentive [to do that].”

  • Alex

    What a crock of shit. Haley is about as bad of a liar as Ford at this point. Perhaps Haley should set an easily OBTAINABLE headway instead of an unrealistic one like 7 minutes. Seven to thirteen minutes is not difficult to comprehend, but it basically implies that the next vehicle may or may not show up at all (that’s a margin of nearly 100% of the headway… for being late).

    Yesterday afternoon I watched so many L’s turn back at Sunset that they had to stop between stops to wait for all of the traffic to clear.

    But, again, this lack of looking at switchbacks as a systemwide problem points out how naive and unprepared the supervisors are to deal with the actual problem. For example, the stretch of track inbound from Forest Hill to Castro runs at about 5mph in auto mode no matter what kind of traffic there is in the tunnel. What kind of service can you realistically expect with a giant chokepoint like that? I haven’t had a chance to take the Metro outbound in a while, but if that same speed limit is set… bam… you’ve got extended backups in the subway and the J will feel the consequences. But… you know… let’s focus only on two lines. It’s easier.

    Yawn.

  • guest

    I challenge you both Mr. Wiener, and Mr. Haley to take this discussion to the next level and implement the changes you are proposing. I agree with Mr. Haley and I believe that expectations vs. reality are somewhat disillusioned, but seriously, I want to SEE the push to implement changes and get the community support required to make those changes. I challenge you to look at the ridership data and make the hard decisions that have been lacking all this time. It’s not sexy to consolidate bus stops and people wont realize that they made it through the traffic signal in less time than before or even in one cycle instead of two, but let’s start here and get it done!

    Let’s use the J-Church and the 14-Mission as examples of how we need to push for REAL MUNI reform and stop the incessant bickering. We’ll get to where we need to go, faster and we’ll experience a more reliable service. It’s a win-win for everyone involved, so make it happen in small steps. We dont expect that you remove and relocate stops overnight, but do it, one, two, three stops at a time. Keep us informed and make it happen.

  • Mario Tanev

    As far as I understand, the reason the headways are irregular on the J is to synchronize the schedule with the N, under the assumption that both lines operate on schedule. But that is a futile exercise given that no Muni lines really operate on schedule.

    The problems with irregular headways are many-fold:
    1. They cause delays due to uneven load. Let’s picture this.

    At the start of the line:

    First run in 0 – X people boarding
    Second run in 7 – X people boarding
    Third run in 12 – 1.7*X people boarding, boarding is slow, which means that this run will actually be slower, and arrive almost at the same time as the next run, causing bunching
    Fourth run in 7 – <X people boarding on all stops after the first stop, since the previous run was actually slower and some riders boarded early (if the previous run were on time, they would have boarded this run). Thus this run may be actually faster, and arrive almost at the same time as the previous run, causing bunching (though this can be remedied by artificially slowing the run down).

    2. They obscure on-time performance problems because they measure adherence to a scheduled interval rather than wait time for the average rider

    3. They confuse riders

    The solution is to simply change the scheduled headway to be the average headway that the equipment/labor commitment allows, and advertise it as such. This will allow Muni to focus on headway/frequency rather than "schedule", which is really what riders care about.

  • Mario Tanev

    I should also mention, that as far as I understand, the synchronization between N and J is not to ensure timed transfers, but instead to regulate the traffic into Duboce tunnel. But as mentioned before, it doesn’t really work, due to inability to adhere to the schedule.

  • Mario Tanev

    We should also examine what happens in the fifth run. Due to the fourth run being faster than scheduled, there is now a wider gap between the fourth and fifth run than 12 minutes (this is a feedback loop). That’s very familiar to riders near the end of the line (say 30th and Church) who observe 2 J trains arriving together, at 20 minute intervals.

  • SteveS

    “Increased parking enforcement, Muni signal priority, stop consolidation, and faster boarding, including ticket machines at J-Church stops, remain among Haley’s top priorities for strategies yet to be used.”

    Well, Muni certainly does have a lot of good strategies that it isn’t using. Too bad it doesn’t have any good strategies that it is using.

  • guest

    @ Alex. Let’s not worry or comprehend that if there’s a probably further down the line that the streetcars cannot go any farther than the switchback. Embrace the fact that a single CAR hitting a bus/train can put 50-100+ people out and waiting for the next vehicle to arrive and a delay of an hour or more for the thousands of riders just trying to get home. We blame it on MUNI being slow when we’re stuck at a traffic light, when there is congestion from too many cars, or the fact that we have a demographic with varying needs. We’d save time if people would let people off the bus before others got on, and if people didn’t think that blocking the doors of trains in the subway just saved them 2 minutes while delaying everyone else on the train and the train behind them by 20 seconds per stop per day.

    MUNI is a social creature. It’s the craziest thing if you’ve ever been involved in the politics of moving a bus stop. Each has it’s own support group and they’d raise hell if you tried to take it away from them. Even at the cost of the greater good (your fellow riders…) we need it to be as close to our house as possible so save our stop but get rid of the next one that “I” never use but someone else does. We fight too much about what WE want and not what is good for everyone. Can we be a little selfless and help make it work for everyone? I thought we were supposed to be a progressive city; one that had a heart to care for others, and the understanding that if we all give a little, we could gain a lot.

  • Michael Smith

    Ah, the politics behind San Francisco transit problems. The supervisors demand that Muni institute fewer turn backs yet do not provide the support for reducing the problems that cause the underlying problems in the first place. We need to deal with the systemic issues, not hold yet another hearing. So to the board of supervisors: provide the money for maintaining the vehicles, the money for fixing the Muni underground system (fixing the tracks, the train control system, etc), the support for reducing traffic on transit streets, the willingness to push drivers and their supervisors to make sure that drivers really do start their trips when they are supposed to.

    Sure it is good politics to castigate Haley, but if the Supervisors are not part of the solution then they are part of the problem.

  • david vartanoff

    The first step is dynamic resequencing of the outbound trains so that no one ever has to wait more than one cycle thus relieving crowding both on the platform and the trains. Bonus shorter dwells, more efficient trips.

    Second, on all of the surface segments, gut the 4 way stop signs that are not at revenue stop intersections. LRV priority traffic signals w/adequate time for cross traffic will speed up trips. All of this is within MTA’s remit, no excuse for inaction.

    Maybe we need a penaly clause in top (mis)management contracts of $50 an unscheduled turnback. That might wake them up.

    Ticket machines are nice, but 2/3 of riders already
    have either a pass or a transfer, so not a criticality.

  • SteveS

    @guest There’s definitely a lot of blame to go around and NIMBY’s opposing stop consolidation are definitely one part of the problem, but a relative small part at least.

    But why shouldn’t we blame it on the MTA when a transit vehicle is stuck at a traffic light? The voters specifically restructured the MTA so that one agency would be in charge of both traffic and transit specifically so they could do things like implement signal priority. Likewise when a vehicle is slowed by double parking, this is just another part of the MTA not functioning.

    So when the MTA says there is inadequate traffic enforcement and signal priority still needs to be implemented, there is no one to blame but the internal priorities of the MTA. With things like enforcement it’s not even a matter of money, just a matter of will and priorities.

  • Alex

    Let’s focus on car and transit vehicle collisions? Really? REALLY? How often do those happen? Now, how often do we have a 1/4 mile+ bottleneck in the subway? Correct answer: maybe once or twice daily at MOST versus every single inbound K/L/M run (potentially every K/L/M/J/N run).

    The problems with using turnbacks as a method of improving service are twofold:

    1.) The MTA turns these things back with no rhyme or reason. When inbound trains are backing up in the avenues, perhaps they should run outbound trains a bit further.

    2.) It further reduces service in the sticks.

    Additionally, where is Haley on the complete lack of communication to MTA riders? When a BART train derailed recently, they were using Twitter (and e-mail, and text messages) to disseminate information to riders a few times an hour. There were pictures and details. When an SFMTA train derailed near Castro station, the MTA spent a few hours trying to pass off the problems as a ‘medical emergency’ before owning up to what really happened and occasionally tossing the riders a bone (or a botched shuttle bus).

    Informed riders can take alternate routes (altho the MTA is doing its best to make using BART as cost-ineffective as possible). Uninformed riders stand around at the non-shelters, in the rain, getting angrier.

    Mike: Nothing that dribbled out of Haley’s mouth was a solution. Nada. In fact the statistics he quoted are a.) bullshit b.) utterly meaningless and c.) probably fabricated. On what level is it inappropriate to excoriate Haley?

    What he quoted was a percentage of runs for a whole year that got turned back early. That’s crap. Pure and simple C-R-A-P. What that glosses over is that when MUNI has a bad day, it’s typically a /very/ bad day. I wanna see average percentage of runs turned back early by day (including standard deviation).

    Are they keeping track of unscheduled turnbacks? Because that’s one very easy way to make these numbers look good (despite riders knowing otherwise). It’s just like putting the trains into cutout mode to enter West Portal outbound. While it was against official policy, drivers were given a wink and a nod because putting the trains in cutout mode was the right thing to do to maintain service. If drivers aren’t being disciplined, and unscheduled turnbacks aren’t being monitored, Haley’s stats are 100% useless.

    Are they breaking it down by time of day? While there’s emphasis on short runs during peak hours, those are the most benign of the short runs. Peak service theoretically means more trains. More trains mean that if one is turned back early, there will still be some outbound service.

    “Switchbacks on the 14-mission declining from 180 in January to 116 in March up to the 25th, most of which are done just after peak hours, said Haley.”

    And there you have it. When there is a minimum of service to begin with, that’s when the MTA decides to gut service. That IS problematic because it means instead of service 15-20 minutes, it’s every 30-40 minutes (or more). Cuts during peak service would still leave much shorter headways (theoretically).

  • guest

    @ SteveS. I just wanted to point out the issue of near side stops at intersections with traffic lights and how people love to complain about them but dont appreciate when the system works right. (ie: we never give props to MTA when they get it right). Nope, not saying anything about their past performance and yes definitely MTA IS RESPONSIBLE for getting it done and it definitely needs to get done.

    Side note, MTA should really give access to people willing to analyze the NextBus information they collect. It’d give EVERYONE the factual information we all desire. It’ll clue us in on real dwell times, actual headways, when/where bunching occurs. All that and more!

    I still think it would be awesome if supervisors will go beyond calling hearings and push for plans and timelines for implementing changes and then report back (and that mainstream media needs cover the whole story as it evolves.. and not just big grand opening ceremonies… hence building anew is significantly sexier than fixing what we have..)

  • Nick

    Re: Enforcement of Double Parking

    I know PCO’s have no enforcement authority but they can still be used to give motorists a verbal warning to not double park or drive in transit lanes.

    I saw it yesterday. A PCO got in the face of a motorist who was blatantly trying to subvert the 10th and Market turn. They were my hero for the day.

  • guest

    re: alex. haha. No one said focus but SOMETIMES there are reasons beyond control. real annoying reasons why things dont work out the way they are supposed to. Like IF our vehicles were properly maintained, they wouldn’t break down as often. AND let’s not jump back in time and complain about MUNI botching the specs or this or that because that really wont help fix anything right now. AND for the other people who want to fire everyone and start over… ummm yea.. really? That doesn’t FIX what we have.

    Definitely subway delays are king, and infrastructure and technology needs to be improved. All for that 100%. I was always wondering if they’re using the exact same system in the new central subway. I mean the system has to be similar to say the least since we’re using the same LRV’s but hopefully it’ll be at least somewhat more reliable with only the single route using it.

    AND yes SIR, I am also YEARNING for better use of the technology we already have. Via NextBus signage, text message, or 511 alerts, ALL incidents should be provided to customers when they inquire about the specific line and major incidents should be posted systemwide to inform riders to use alternatives. But there are still gaps in the system. Think of the stops without nextbus, without a shelter, people without cell phones (seniors, kids…).

    Let’s PUSH our officials for the changes we want. Let’s stop cursing MUNI management and work with them to achieve the reforms we’re always talking about amongst ourselves. You may think this is all too idealistic or that it’s done before, but really, stop, think, try it again. If all the people riding streetcars say HEY, this isn’t good enough and we want you to fix subway delays NOW, you’ll see change. We need to stop being so scatterbrained and accusatory. Us saying the MTA should be doing this or that isn’t helping us, so if we’re all so gungho about fixing things, let’s create a detailed, step by step plan. Go out and really observe if removing xyz stop sign will make a difference and then instead of saying we need to remove stop signs, that “MTA, you need to remove xyz stop sign because it is impacting operations and does not serve to benefit the community”.

    If all our fellow citizens were ideal, we wouldn’t need to rely on stop signs to slow cars or enforcement of this and that. But we dont live in an ideal world. Just one where we the people cause problems for other people.

    Let’s use technology and use the powers of MTA.. and let’s fix MUNI!! one route at a time. (and please, let’s not get distracted by naysayers that say it cant be done or it wont work. Try It!) EACH route is different and the problem is.. there is no simple, easy, or standard solution to fix things and hence it has not been done.

  • Alex

    I understand that we need to cooperate, and in fact I think the lack of a proper riders’ advocacy group is really screwing over MUNI users/victims. But that doesn’t change the fact that we’re paying millions of dollars a year to these folks, and we shouldn’t have to do their job for them.

    Around 3 this afternoon I watched four inbound and two outbound trains go by. One of the inbound trains was a two-car pair that I had watched meander down to Sunset for a short run. The first two inbound trains I saw were < 1000ft together (the minimum following distance). That's not a technological solution, that's a drivers need to use some common sense and management needs to enforce the common sense adhere to headways mentality. So, no, when Haley tries to blame riders as a first resort I say tar and feather the guy.

    If you want information, the NextBus data isn't terribly fine grained but it is publicly available. OTOH if you do something wild like compare 3-4 year old daily service reports on rescuemuni's site to the castrated ones we get today… yes. The MTA needs to be more transparent. But the reason it's not is the same reason for Haley's blame the riders attitude and the MTA's presentation of useless statistics: the folks sucking down six figure salaries at the MTA want to make themselves look as good as possible.

  • Alex

    Sorry, ‘technological solution’ should read ‘technological problem’. And it’s certainly not a problem that removing stop signs or putting in freaking TVMs would solve. Haley is, as usual, skirting the issue of operational chaos.

  • Sean H

    I question MUNI on-time performance (OTP) measure. This is either done with manual checkers or GPS technology. Both techniques leave a ton of room for error and data manipulation. For manual checkers, they just check the terminus usually without checking midroute OTP. The drivers often know about the OTP checkers via the radio, so they change their driving behavior.

    With GPS technology, it often is down so that data is taken out of the analysis that could have indicated worse OTP. Also, in many systems every bus doesnt have the location technology, so there is plenty of room for data manipulation when you interpolate the non-equipped bus data.

    Most riders dont care if the bus is on time at the end of the route, they probably got off miles before- and late.

  • doogiehowser

    Too bad — Muni needs all the money for a Subway to Nowhere.

  • david vartanoff

    @Alex. deed Haley IS a problem. It is he who changed the Dailies from useful to useless. All the stop elimination crap is meaningless if there are no Js )or which ever) for 20 min while 2 each of the others go by.
    BTW, as to rider action, a couple decades ago in New York, A train riders got fed up w/ being short turned at 168th St and one night simply blocked the cab and non violently communicated that the motorman should procede to 207th as the train was signed for. He did.

    A couple of those would get the attention of both drivers and (mis)management.

  • Would a daily report like this one, made for Monday’s NextBus data (right-click on the image to zoom), which plots where each train was at each time of day, be useful for identifying systematic problems? Or is it too much information?

    (Bunching and missed runs jump out pretty well, but it’s not so clear *why* the problems are happening.)

  • david vartanoff

    not sure how to use that representation. what I miss from the older version of the Daily Reports was the detail of which routes got cheated by how many runs. That would provide a Supe with ammunition on behalf of constituents stranded by not outs. Does Next Bus detail turnbacks?

  • david, Right to Left on the graph is distance traveled. Top to bottom is time elapsed.

    You can see the tunnels because the line goes up but not over which means that the LRV is stopped (time continues and line goes down, but distance is not traveled so line doesn’t move to the right).

    It is hard to see, but does show how much the tunnels slow the overall performance. Thanks Eric, awesome stuff as always.

  • Thanks, Mike.

    David, can you link to an example of a report that has the information/format you are looking for? It’s probably possible to replicate most of it.

    I don’t think NextBus has any explicit indication of turnbacks, but you can often tell when they happen from looking at the plot. For example, look at the L Taraval, the lines sloping downward toward the left (outbound). At about 9:00am, 10:30am, 1:00pm, 1:30pm, 2:30pm, 3:00pm, 3:30pm, and then a whole bunch of them together at about 6:30pm, you get a vehicle that makes it out of the subway, gets about halfway through the surface line, and then turns back toward the east instead of continuing on to the end.

    P.S. Does anybody know why, in the latest GTFS schedules, at least some Muni routes are now represented as loops from a central point on the line instead of end-to-end routes? For example, most J runs are shown as loops from Church and Liberty to one end and back, and most N runs are shown as loops from between Van Ness and Church to an end and back. Are they actually dispatched that way now? I didn’t actually look at all the routes so I don’t know if they were all switched to be scheduled this way.

  • david vartanoff

    Eric, picked at random from 6 months before Haley gutted the system
    http://www.sfmta.com/cms/rdlymuni/uploaded/Service%20De
    livery%20Daily%20Report%20Aug%2005%202009.pdf

    note the bold NO 74X missed–remember the totally unused Culture Bus?

  • Michael

    Has anyone had a chance to watch the full hearing? Another story mentioned that Haley said they couldn’t run longer trains because of hills and tight turns, but that doesn’t add up since the N often goes home at night on the J tracks, and the J even has a second deadhead car sometimes. I can see how boarding might be a challenge at some stops, but that seems more fixable. Does Haley explain this in further detail?

  • Aaron Bialick

    Michael –

    I asked Haley about making physical changes to allow two-car J trains, and he said it was practically impossible – or, whatever could be done would require enormous capital. Neither he nor Paul Rose articulated clearly where the issues are that prevent two cars, but Haley did mention 18th Street and alluded to boarding issues (he isn’t the clearest speaker, and it was off the record). But I did think about the N factor afterwards as well – it’s something I’d like to follow up on. The expense of sending out more operators to run more single J trains was a limitation he definitely pointed out.

  • Alex

    Eric: That’s a pretty visualization, but it lacks enough information for me. NextBus is also kinda a poor source for that because the GPS is sometimes wildly inaccurate (I’ve seen vehicles going in excess of 100mph reported, as well as trains well off of their tracks, etc). They’re also, AFAIK, relatively unaware of short runs. They do collect the run #, which often goes to shit when the driver turns the vehicle around — but this isn’t publicly available. They also collect the direction ‘tag’, and that can sometimes be used to guesstimate if it’s been turned around (as on the streetcar.org ‘live on the f’ page) — but that’s dependent upon the driver to do the right thing.

    What’s also missing, and what I think are perhaps more important, are the whys. The old daily service reports showed WHY things were delayed, they showed every single incident. The current ones omit as much as possible.

    What I ended up doing at one point was simply overlaying (and blurring a bit) the GPS data onto a map of SF.

    The red represents vehicles moving slower than 7mph or so. The more opaque the red, the more vehicles in that area were going ‘slow’, the more translucent colors mean fewer vehicles. For a single route it wouldn’t be that hard to make a similar map just showing the level of service. Combining routes becomes more difficult because some vehicles report more than others. Likewise gaps in AT&T’s cellular coverage can make for some interesting gaps on the map (try plotting the 108 for instance).

  • Alex
  • Isn’t it 21st Street that keeps the J from having longer trains? The train can’t be longer than the intersection is wide because there is no room for boarding platforms on the private right of way.

    Thanks for the report, David. There is a lot of information there that wouldn’t be available from NextBus, but NextBus should be enough to know how many vehicles were active on each line, how many trips they made, whether each trip went the full distance, and how long the trips took.

  • EL

    I find Avalos displeasure with the #14 amusing. Instead of just holding a hearing on the matter, how about using his political power to support/legislate:

    1) A double-fine zone on Mission for double-parked vehicles,
    2) Prohibit parking on certain stretches of Mission where the lanes are so narrow, that Muni uses both lanes, or known choke points like Geneva or Randall,
    3) Prohibit turns (that block through traffic) or private vehicles traveling through along Mission,
    4) Create transit only lanes,

    I give a bit more slack to Wiener regarding the J-Church because the cost of rail infrastructure is so high. However, items 1 and 3 above could apply to Church Street as well.

  • Alex

    Eric: Should, but there are lots of reasons that the GPS data is inappropriate for this. What happens when you have an LRV shown as being in front of West Portal for 20+ minutes? Do you count that as a broken down train? A turnback? What happens when it shows up in the tunnel at Montgomery 25 minutes later?

    You can bet your ass that the MTA /is/ collecting data on which vehicles are making short runs. They need to make it publicly available. Reverse engineering is a poor solution to the opacity at the SFMTA.

    As for two car trains on the J. First, you don’t have enough vehicles for two car trains on the busiest routes (N and L — I’m still seeing one car trains on the N FFS), let alone the least used route (J) with fewer riders than the F. Second, yah, some of the stops are ill-suited for two car use.

  • SteveS

    @EL #2, #3 and #4 are all things that the MTA is already empowered to do without needing any action from the board of supervisors.

  • Alex

    Aaron: Cool. So which routes would you gut to provide two car service on the J? Take a look at the spare ratio in the daily service reports and tell me you honestly think that two car service on the J can be provided without gutting other lines.

  • Agreed, if they are collecting the short turn information, they should publish it. I still think it is worthwhile to do independent verification of the reports where possible, and to calculate what the impacts are on the typical wait times at different locations.

    And no, I don’t know what to do about discontinuities in the data except to acknowledge that a vehicle’s location is sometimes unknown for a while. I think the data is generally good enough to be able to draw reasonable conclusions from it, though.

  • Alex

    Eric, look at the /old/ DSRs on rescuemuni.org (google them) and tell me you could duplicate that (easily) with the publicly available NextBus info. They had info on just how unreliable the ATCS was that’s been hidden for a few years now. Hell NextBus is collecting way more info than they’re making public, and I’m sure the MTA is fine with that.

    As far as your visualization goes, I’d want to see some sort of x-axis labels like embarcadero, west portal, 19th, sunset, 46th, zoo for the L.

    Focusing on penny ante crap like stop removal to address a structural problem with how the MTA is run is just insane.

  • Michael

    @Alex: Peak hour LRV availability is a critical issue, but getting two-car trains on the J-Church ROW still might make a lot of sense. That’s because there are already plenty of LRVs available off-peak, and the J isn’t only crowded during the peak. LRVs that are going out of service during the midday and evenings could instead be used for two-car runs on the J. For instance, those Mission High students might arrive at school during the A.M. peak, but they leave during midday, on crowded one-car trains. We’ve all seen packed J-Church cars headed to and from Dolores Park on the weekend, too, and there is no shortage of LRVs on Saturday and Sunday.

    So, in the near term, the J-Church could run two-car trains off-peak without cannibalizing other lines. In the long run, we can plan to buy more LRVs and rehabilitate the ~10 that are currently out of service (at a cost of about $18 million).

    @Eric Fischer: 21st Street might be one of the challenge spots, though between the sidewalks and the intersection, that’s a pretty long potential boarding area. Anyone know what the MTA considers to be the problem stops? I’m not sure that major capital expenditures would make much sense, but I’m also having trouble finding spots that would actually require that.

  • Alex, no, I can’t replicate the old daily reports. Yes, it would be great if they would start publishing them like that again.

    But more constructively: I talked to the SFMTA data people a couple of weeks ago at Transportation Camp, where they were *soliciting suggestions* for things that people would like to see in their public data feeds that is not there now. I mentioned run numbers in NextBus and they seemed receptive to adding that. Send them mail at sfmtadata@sfmta.com and tell them what else you would like to have, and maybe they can get it for you.

  • EL

    SteveS wrote: “@EL #2, #3 and #4 are all things that the MTA is already empowered to do without needing any action from the board of supervisors.”

    On paper, yes because MTA Board is independent of the Board of Supervisors. But don’t be naive. In practice, the reality is no. If the MTA Board tried to legislate it and the objections come from the locals, it would take supervisor support to override them. That’s one reason why only a hearing is held to “express concern”, but no hard line position is taken.

    For example, when Sunday and nighttime parking meters came up, Newsom left it to the local Supervisors. How many actually stepped up to the plate to bat?

  • Alex

    Eric: Perhaps they should start by restoring the quality of data that they /were/ publishing instead of trying to push for something fancier with NextBus. If they need someone to suggest that they should push for a release of all the data that they’re paying NextBus for… well… geez. Do they need someone to tell them that water falling from the sky is rain?

    How about ensuring that all of the F-line vehicles have NextBus equipment installed? Apparently they can do the cable cars (with no electrical systems to speak of) but can’t do some of the older F-line trams.

    If they want a real suggestion it would be to publish ALL of the info included in the old-style DSRs and publish it in a machine readable format so that someone else can pretty it up and munch on it.

  • Jon

    Looks like the problem stops are 21st and 22nd, both outbound. Inbound has a lengthy platform at 21st and a stop before it enters the right-of-way (south side of the intersection) at 22nd.

    So the solution is easy- in the name of stop consolidation, remove both stops at 21st, and move the outbound stop at 22nd to the south side of the intersection. You still have only a two block distance between stops (350m/1150ft), which should be standard for this section of the line.

  • Alex

    Jon: Cool, so what does that do the bottleneck at the outbound Castro departure… which backs up traffic to Embarcadero, which the outbound J’s will get stuck in? The reason the J sucks is because the whole metro sucks, and none of these ridiculous hearings have addressed that.

    Stop consolidation will not fix operational issues. And, FFS, if turnbacks were as rare as Haley made out, they could probably do away with the supervisor at Sunset and Taraval whose ass is permanently glued to the seat in his truck.

  • Jon

    Hey, I was just saying how to fix the problem of not being able to run two car trains on the J line. Sure, the whole metro sucks and there are plenty of other problems to be fixed.

  • EL

    Obviously Jon, you haven’t been here in a while. The outbound stop at 22nd was moved over a year ago. And have you noticed that huge hill (which the J goes around rather than over), which is probably why the stops are so closely spaced ehre.

  • Jon

    Admittedly I was using Google Maps for that piece of information. I live less than a mile from that section of the J-Church but rarely have reason to use it.

    I’m aware that there is a hill, and that is why stop spacing is closer than on flatter sections of the line, but it’s still too close. It’s closer than most European cities have their bus stops, nevermind light rail.

  • pbear

    Switchbacks are MUNI’s current J-CHURCH line problem? Are they kidding? Pardon my rant, but, MUNI remains the longest running transit comedy act in the bay area and Mr. Nathaniel Ford its current, lead comedian. MUNI has run at 76.8 since January? Highly unlikely in real terms and more likey because the line has recently come under considerable scrutiny and MUNI/union are seeking to cover their proverbial bottoms. I’ve lived on the J-CHURCH line for 30 years and been through 2 systemwide reviews (including J-CHURCH line) done by MUNI declaring itself safe and offering acceptable service. While there have been rare times of exceptional service along J-CHURCH line and others, most the time at least J-CHURCH service is approaching the abysmal level.

  • Chobakasan

    In regards to the J Church: a) why is there only 1 car, when most other lines have 2 cars? b) has there been a census evaluation of the J Church line done in the last 10-15 years to reflect new residents/demand for the J Church? c) 76% on time reliability is a joke. anyone who takes J Church for more than 2 days in a row, knows this is not true.

  • Ken

    Despite living a block off Church out near Ceasar Chavez, I’ve avoided Muni for the past year or so since it’s generally more predictable and/or faster to walk over to 24th and Mission and grab BART to get downtown.

    Having given Muni a second chance over the past few weeks, however, in four or five trips inbound from Church and 27th I’ve found the Nextbus predictions to be pretty much 0% accurate. That is, the trains seem to show a pretty predictable 5-7 min after Nextbus shows them as “Arriving”.

    Do people actually find they’re able to use the J-Church on a specific timetable? That is, always count on catching the 8:10am train, for example?

  • DC

    San Francisco will NEVER be a world-class city without world-class transit. I am amazed by 40 minute J Church waits in the evening; “Switch problems” caused a 45 minute ride from Powell to Castro today.

ALSO ON STREETSBLOG

J-Church Line Could Be the First to Get All-Door Boarding

|
The J-Church has emerged as the top candidate for Muni to test all-door boarding as a way to speed up service on its busiest and least reliable lines. San Francisco Municipal Transportation Agency (SFMTA) Transit Director John Haley said today that it is being evaluated for a pilot program that could lead to a system-wide change […]

Supes Avalos, Wiener Clash on Equitable Spending Strategies for Muni

|
Supervisors John Avalos and Scott Wiener are sparring over how new revenue for transit should be spent to benefit the Muni riders who need it most. With tax measures proposed for the 2014 ballot that could significantly increase transportation funds, Avalos introduced a charter amendment yesterday that would “require the city to prioritize investments to address existing […]

Mayor, Eight Supervisors Promise to Ride Muni Every Day Until June 22

|
The SF Transit Riders Union’s challenge to ride Muni for 22 days kicked off yesterday with late sign-ons from Mayor Ed Lee and Supervisors London Breed and Mark Farrell, who had initially declined to commit. Supervisors Katy Tang and Malia Cohen still declined, and Supervisor Norman Yee has not confirmed a pledge since he tweeted a selfie on Muni after […]