18 December 2011

The Baby Bullet Effect

 For many years prior to 2004, including throughout the dot-com boom, Caltrain operated an all-stops a timetable with less stop-skipping (see February 2000 example) that fairly well revealed the underlying ridership demand at each station.  In 2004, that all changed with the advent of the Baby Bullet.  While the Bullet was a marketing triumph and remains a successful source of ridership and revenue, there was an under-reported flip-side to this new service: many small (and not-so-small) stations lost service.

Caltrain publishes annual ridership counts for each station, which can be distilled into a single table of historical counts of weekday boardings for each station reaching all the way back to 1992 (download 55kB Excel spreadsheet).  This data reveals interesting patterns.

The share of ridership at each station, which was fairly stable over several years leading up to the 2004 launch of the Baby Bullet, settled into a new pattern that has shown itself to be fairly stable in the years since 2004.  The change in each station's share is shown in the figure at right (also available as a 141kB PDF file), where 100% represents each station's average ridership share over the period 1999 - 2003, or the initial size of its slice of the ridership pie back in the pre-bullet days.  Following 2004, some slices got bigger, while other slices got smaller.  The entire pie also got a bit bigger, although that is not shown in this figure of the proportional trends for each station; ridership has only recently exceeded the 2001 peak.  The Baby Bullet Effect has divided stations into two groups: winners and losers.  Most of the losers were small and could justifiably be dispensed with.  Some were not, and are under-served to this day:
  • California Avenue in Palo Alto, 1376 weekday boardings in 2001, down to just 895 at last count
  • Lawrence in Santa Clara, 1309 weekday boardings in 2001, down to just 531 in 2011
  • Santa Clara, 1124 weekday boardings in 2001, down to just 656 in 2011
  • Burlingame, 985 weekday boardings in 2001, down to just 675 in 2011
  • Belmont, 892 weekday boardings in 2001, down to just 369 in 2011
  • San Bruno, 844 weekday boardings in 2001, down to just 403 in 2011
All of these places have more residents and jobs than implied by today's poor ridership, and are consistently under-served by Caltrain.

Planning for the Future

The future timetable plans revealed so far by Caltrain, including their notional electrification timetable and the timetables evaluated in the blended operations analysis, consist of an all-skip-stop service pattern as illustrated at left which "bakes in" the ridership pie slices as they exist today.  While the Baby Bullet is slated to be discontinued, its negative impact will live on at the places listed above, which will continue to be served by only two trains per hour (out of six).  Stakeholders at those stations should not allow this to happen.

As they plan their future operational concept, it is important that Caltrain base their stopping pattern on raw population and jobs data and not on the highly distorted ridership patterns induced by the Baby Bullet Effect.


  1. Some comments:
    - Electrification doesn't seem to speed up the baby bullet travel times so that other stops can be served. Something is wrong there.

    - Atherton, Broadway, and Hayward Park are still served. That makes no sense: Atherton ridership can be easily absorbed by Menlo Park or RWC, Broadway is too close to Millbrae and its high frequency, and Hayward Park is sandwiched between two major stops.

    - When the Gilroy shuttle is running, having the main line trains continuing to Tamien is useless.

    - The timetable is more taktlike, which is nice. Skip stop still makes no sense.

    Ultimately, the best course of action (besides electrification) is to construct that midline overtake and terminate some local trains at RWC. This way, all the riders south of RWC get express service, and stops north of RWC get more frequency w/o slowing down the Santa Clara county riders. Plus, it allows for future extension of the locals over the Dumbarton bridge. Sample schedule

  2. For various reasons, I've found myself spending a lot of time in the region between California Ave and San Antonio, when I am in town, so I've had some first-hand riding experience with this.

    Is California Ave under-served? Absolutely. But! I find myself using VTA 522 to get to Palo Alto to catch Baby Bullets. The California Ave station is poorly designed for pedestrian access, because the only underpass is in the parking lot. The Palo Alto one is somewhat better, despite the bus station transfer being slow.

    Presuming all this stuff could be fixed, would it even be feasible to serve these "under-served" stations without losing too much time on express service? Maybe it's just as well to have long stop spacing and fill in the gaps with bus transfers?

  3. Atherton still served? Please tell me when?

  4. One particularly egregious example I spotted recently is the Oracle shuttle. It's shown as running from either the Hillsdale or San Carlos stations to Ralston Ave along the street parallel to the line, and then down Ralston to Oracle's campus. Of course, Caltrain also has a station at Ralston Ave: Belmont. But it's so poorly served, with 1 local train per hour to northern stations and 0 tph in the peak to most stations in Santa Clara County, that it would make a very poor connection point for the shuttle.

    I've also noticed that Caltrain's poorest-performing train in the peak is the one that makes local stops between RWC and SF. Since they do still need local service, maybe it's time to use that spare capacity and save SamTrans some money by offering full fare integration in San Mateo County.

  5. It looks like 22nd is one of the Baby Bullet winners, and it's down to 1 tph in this schedule. Is 22nd due for more service in the reverse direction?

  6. @Morris: Atherton and Broadway are weekend-only for all-stops service.

    @Steve: Looks like 22nd street is winning largely because of "reverse" commuters: people who live in SF and work in the Peninsula. So look for that pattern.

  7. Matthew re California Avenue this is a prime example of why Caltrain and everybody remotely connected with it need to be terminated, with immense prejudice.

    Here's how it works

    At Caltrain, it's important that getting to from the train take as long as possible, ideally an extra four minutes or so, because BABY BULLET IS THE SOLUTION TO EVERYTHING and BABY BULLET SAVES TIME and so who cares if you have to detour a quarter mile out of your way and who cares if you have to wait an hour for a train to come and who cares if you have to tag and tag off every damned trip at at defense contract Cubic's "smart" card machines? BABY BULLET, BABY!

    So to "design", a station, what they do is that some some-cretin monkey at PCJPB, or an in-house expensive consultant with similarly skill, sits down at a CAD system draws up the stupidest, most rider-hostile, most ignorant, most expensive crazy thing that they can imagine. It's important that it be maximally inconvenient for riders, and it's important that it be completely incompatible with obvious needs (eg level boarding.)

    They then get some sub-cretin at the local city DPW to sign off on it. So it's What The City Wants.


    They then put it out to bid.

    They then spend $10 million of your tax dollars on "improving Caltrain" by making the service WORSE FOR RIDERS.

    The first time that anybody who isn't a construction contractor or agency staffer finds out how badly they've fucked over the public is when the projects is finished.

    They do this over and over and over. Every single time. Look at the fucking disaster of San Bruno -- spend hundreds of millions to make things WORSE, again.

    They can't be trusted with a single dollar. They are NEVER EVER EVER going to improve or modernize or upgrade the system. All they're going to do is piss money down holes.

    Caltrain needs to die, ASAP, because none of us will ever live see an acceptable level of acceptable quality transportation on the SF peninsula until these people are terminated.

  8. You see: "all negotiated with the city" makes it just peachy to waste $13 million dollars and screw your riders. And remember, former BART train driver Bob "CBOSS" Doty was supposedly the least dim bulb at Caltrain. Amazing.

    Subject: RE: Cal Ave: sad sad sad
    From: "Doty, Robert" <dotyr@samtrans.com>
    Date: Wed, 30 May 2007 08:23:41 -0700
    To: "Richard Mlynarik" <Mly@POBox.COM>

    And if the tunnel can't be smaller (I did reduce it to 9 x 16 already) and if the ADA access must be oriented towards to bus stop area and if I am not allowed to fit in a switchback then what? All of this has been argued.

    All negotiated with the city.

    Please don't go through the roof..... unless you plan on installing skylights. This business will drive you nuts.


    Robert L Doty
    Director, Rail Transportation,
    Engineering & Construction

    -----Original Message-----
    From: Richard Mlynarik
    Sent: Tuesday, May 29, 2007 5:30 PM
    To: Doty, Robert
    Subject: Re: Cal Ave: sad sad sad

    Doty, Robert wrote:

    >> Ah! But you live in the world of logic, ascetics and freedom. I live in the world of regulations, politics and "constraints". Can't say that I like it that much either.
    >> Bob.

    Not directed at you personally, but this makes me go through the roof. I've spent the last ten years telling myself that everything will be fine as soon as Jerry/Wally/Harold/Darrell stop *deliberately* messing things up, but with crap like this going down I don't see how things will ever change, and frankly I don't see any reason why you shouldn't just plan for the BART extension, since it is clear that "commuter rail technology" is a total, irredeemable basket case. What's supposed to be the advantage? Not speed -- we waste it. And not cost either, it increasingly seems. People should start lobbying against useless Caltrain capital funding!

    -- Richard.

    [... original message continues ...]

  9. [... original message continued ...]

    1. Just put the damned tunnel at the NORTH END OF THE STATION. The ramps head SOUTH.

    North is where EVERYBODY comes from. That's where actual pedestrian human beings and actual bus and shuttle riders and even (I know you hate us...) bike riders come from. Nobody, nobody NOBODY wants to walk half the length of the parking lot just to have to come back the same way they went -- especially not the slow walking elderly who can and do take the stairs.

    That way 95% of your riders save 4-5 minutes of transit time EVERY SINGLE DAY getting on and off the train. Jesus christ on a pogo stick. If this can't be done, FORGET IT.

    2. Every 6 inches of depth saved from the arbitrary Caltrain-engineering-standard top of tunnel to top of rail clearance and every 6 inches lopped off the arbitrary Caltrain-engineering-standard passenger tunnel height means SIX FEET less ADA disaster ramp. That all adds up really, really, really quickly -- about 5 seconds of walking time per six inches of unnecessary excavation.

    This stuff needs to shaved way, way down. Some "standard" that somebody decided on one afternoon becomes a real impediment when it adds and extra 50 feet or so of walking access time to half the trips from the station. Take a whole foot off the tunnel internal headroom (to 8ft floor-ceiling) _just for starters_.

    3. I see allowance (in terms of ADA Ramp Hell) for 8 feet more ramp (ie 8 inches more platform elevation.) What happens for 550mm ATOR (21.5") platforms -- where does the extra 7 feet of pain come from? Hello?

    4. This unspeakably wrongly-located tunnel is going to be there long after your blue squares and mini highs are dead and buried. Then what? One can't place 30+-year infrastructure for transient, interim, technically nonsensical 10-year stupidities. (I assume that some bozo decided that the underpass ramps had to start and end right at the Blue Squares. But when the squares are gone, all you're left with is a horribly inutile station ... forever.)

    5. It CAN be done right. I don't see how it's _that_ difficult, and I've looked really hard. If it can't be done right, don't waste my tax money on it, thank you.

    6. The lack of any tie-in to the eastern side of Alma? A complete and total scandal. And by clverly, cleverly positioning your tunnel right next to the diving merging auto ramp, your engineers have ensured that there never *can* be such a connection, outfall pipe or not outfall pipe (and there are ways around the pipe.)
    How did anybody in Palo Alto ever sign off on this stupidity ... or haven't they seen it yet?

    Summary: if BS is the only thing we can get for $13 MILLION dollars well, perhaps it is far better not to try at all until wiser heads prevail. Seriously, spend it on Burlingame or Santa Clara or something. It's not as if you haven't completely abandoned any semblance of useful rail service to Cal Ave already. (Stanford Industrial Park is still there, and the jobs are still there, but the hundreds of riders who I use to join boarding evening expresses a re all back on 280 and 101.)

  10. Thanks Richard. I can't remember how many times I had to help people find their way to the northbound platform. It was a lot.

  11. @Morris: yes, Atherton. If you of all people didn't know this, Caltrain is doing a poor job of selling the electrification project. One of the community benefits it provides is to resume or increase service to smaller stops during the rush, without loss to trip times, thanks to the quicker start/stop capability of electric trains.

    @Steve: it's a sample northbound AM schedule. The southbound AM (see Caltrain document) has 4 tph at 22nd.

    @Matthew: you've put your finger on a key issue, that trip times consist of more than a just start-to-stop train ride. Saving 60 seconds of access time is just as valuable as 60 seconds of trip time, or even more since riders perceive waiting time as "slower" than an equal amount of ride time. In all the service metrics discussed on this blog, waiting time (and its cousin, frequency) forms a key part of the equation.

    I am well acquainted with Cal Ave since that's near my job. I go out of my way to avoid it because of shitty service there; I use University instead, where I'm not screwed if I miss my train.

  12. @Clem:

    It isn't that CalTrain has done a poor job of selling electrification, the benefits of a $1.5 billion electrification project, simply don't come close to yielding an adequate public benefit.

    Richard's extensive comments on CalTrain in general tell the whole story.

    The whole idea of having HSR on the CalTrain ROW yields nothing but poorer service for commuters. Bullet train service by itself, as you have so elegantly illustrated here, has degraded local commuter service.

  13. A little bit off-topic, but I checked out the Santa Clara station, and they've actually managed to get it more or less right. The stairs are in reasonable locations on both platforms, the tunnel is just big enough to not feel claustrophobic, but the floor level is not gratuitously deep. There's a reasonable provision for continuing the tunnel to Brokaw, but that would be up to the city to do. And the west side is actually designed in such a way that they can install four tracks. They even remembered to install Clipper readers and ticket machines on the new platform. Also, it's interesting to note that the platform is curved.

  14. Anonymous, just wait to you see where they've planned -- planned! -- to place their designed-in-USA catenary support poles on the platform.

    Ca;train will always disappoint even the lowest expectation.

    PS Tens of millions later, it still isn't possible to get from the station to the eastern side of the tracks without a half mile detour. Another triumph.

    PPS Level boarding? What's that?

  15. @Morris: on the one hand, you have a point that 1.5 billion is a very hefty tab compared to equivalent international projects (e.g. Auckland NZ). On the other hand, that includes a fleet replacement which is overdue anyway and will be incurred regardless of electrification.

    But the most important point is this: your outrage would be better targeted at six billion being lined up to bring BART to San Jose. Caltrain is small potatoes compared to some of the other public-private wealth transfers that are currently in the pipeline.

  16. Wait, they rebuilt Santa Clara? I thought they were going to rebuild it as a joint BART/Caltrain/ACE/Amtrak station once BART finally makes it there. Will they be able to add that in, or will they tear everything down and start over?

    Also- "the west side is actually designed in such a way that they can install four tracks." Does that mean it's set up for a FFSS configuration? What happened to SFFS? Not that it really matters, as HSR is planning to shoot over it on an aerial...

  17. Jon,

    Why do something once and do it right when you can do it badly three, four or more times? Seventeenth times a charm!

    Doing over is what makes our rockin' world go 'round.

    Santa Clara Caltrain, Caltrain/BART Caltrain/BART/HSR: think of it as the Millbrae Caltrain Caltrain/BART Caltrain/BART/HSR of the south. Only even awesomer.

  18. Well, it's only a matter of 20 years or so until they finally get the money and build the BART subway to San Jose. I suppose we can just live with the holdout rule and no ACE stops until then. Keep in mind that this project helps real passengers on real trains right now, as opposed to the BART extension that still only exists in the fevered imaginings of Carl Guardino. And as far as the track configuration goes, it's going to be SFFS for the Caltrain tracks, just like at Lawrence, with platforms only on the local tracks, except the northbound platform also has the Coast line on the other side.

  19. Having been a former commuter between SF and Lawrence, it became really clear that service infrequency was only one of the problems. I'd say there are 5 major problems there in this order.
    1) Poor final mile transit (there is virtually none at Lawrence -- company shuttles are very limited, and at the SF end it's unpredictable(I had to plan to be at the station 20 minutes before departure to avoid missing my train -- thank you MUNI!))
    2) Poor caltrain reliability; I was more than 20 minutes late in either direction at least once a week -- the worst was nearly 3 hours late.
    3) Poor caltrain speed (even the baby "bullet" is still very slow)
    4) Poor caltrain schedules (what if I need to work late? or work early?)
    5) Poor caltrain frequency

    If #1 isn't fixed none of the rest will matter. If #2 isn't fixed none of the rest will matter either.

    Solving #1 requires cooperation with local transit agencies, and balanced funding of those agencies as part of a comprehensive regional plan. Collaboration with city planners and private developers for transit oriented development could also be positive.

    Solving #2 requires better preventative maintenance, at least 3 tracks for caltrain for the entire route, full grade separation, better contingency planning, and better infrastructure design to avoid silly problems like heat restrictions.

    #3 reduces desirability, but is tolerable if #1 and #2 are fixed. The ability to have a fast cross platform transfer to HSR from locals would be a way of dealing with it, but that has obvious problems.

    #5 is tolerable if #2 and #1 are fixed, with guaranteed connections for final mile transit, as long as there are at least 2 services an hour.

    #4 and #5 can both be assisted by running shorter, lighter rolling stock that only needs one crew member, and having modern train control.

    I eventually bailed and just started driving, and sincerely doubt a positive outcome will be made with caltrain as it is on the above issues.

    I think it's important that we consider the entire end to end trip effectiveness, rather than just caltrain. Most residences and workplaces are not walking distance from caltrain stations, so a comprehensive solution really needs to be created. HSR plans ought to be considered to be part of this.

  20. “Caltrain operated an all stops schedule.” This is an inaccurate assumption.

    Caltrain has been running some peak period express/limited stop service for decades. In the years prior to Baby Bullet, Caltrain did reduce/consolidate express/limited service due to budget cuts, etc.

    Considering pre Baby Bullet Caltrain service, the underlying ridership demand has never been fully realized. Additionally, due to the lack of decent off-peak and evening service, the true ridership demand at many stations has never been realized.

    Dispensing with so-called “small” stations is counterproductive.

    There are some advocates (and Caltrain staff) who defend the shitty schedule, using the claim that: Caltrain did a good job matching the service to the demand. Its actually the other way around, the so-called “demand” is reflective of the poor schedules. Stations with poor service will always have poor demand regardless of how many residents and jobs are in the vicinity of the station and Clem’s analysis pretty much shows that.

    The proposed future timetable still sucks during the mid-day and sucks even more in the evening and as you point out, some stations are still underserved. One train per hour during the peak hours is pretty much worthless. Even 2 trains per hour is no way to build strong ridership.

    I have been analyzing/studying Caltrain ridership, schedules, and fares for some 25 years and have found that ridership is sensitive to a number of factors, most notably increasing gas prices, 911 and the dot com bust, large fare increases, schedules, and more.

  21. Note that acceptable frequency for a transit service rather depends on trip length. The longer the trip, the more planning it generally requires and the lower the frequency that riders will consider acceptable. Caltrain seems insistent that the only correct way to use the system is to commute to and from SF, rather than do shorter intermediate trips, and that's how they justify their low frequency.

    On the reliability front, grade separation won't solve most of the problems. I've found that about half, or fewer, of Caltrain delays are caused by "trespasser" or grade crossing incidents, and the other half are due to equipment problems, mostly with the aging and unreliable diesel locomotives. The minor delays are due to a combination of lack of level boarding and poor operational practices and inconsistency by the train crews.

  22. @Jeff Carter, thanks for the correction. Do you happen to have a pre-bullet schedule handy?

  23. Clem, February 2000 Caltrain schedule for your nostalgic edification. I have many others. (Yes, that's a sad admission.)

  24. Interesting to note that the old schedule also had five trains per hour, and that that old end to end all-stop running time was also 1:31, though with stops at Atherton and Broadway, which takes 1:36 on weekends. Did they boost OTP with schedule padding, or did dwell times get worse, or operating practices get sloppier, or what? Also, the fastest end to end train was 1:09, not too bad compared to the current 57 minutes. And the Gilroy-to-SF market was served with a train that stopped at all the South Bay stops and then ran express, which is the only reasonable way of doing it.

  25. Yes anonymous, that's what the so-called "facts" and "evidence" and "history" appear to show .... but ... but ... but ... but ... you simply don't understand the Big Picture.

    It's way more complicated than amateur armchair bloggers mislead you to believe.

    You see, 6 whole trains per hour is impossible without CBOSS!

    6 trains per hour requires four vast new empty platform tracks at San Jose.

    Futuristic all-stops 91 minute travel time (for 50 whole miles) requires CBOSS! And electrification, using 15 year old Alstom trains running 10 cars long! And San Bruno grade separation! And a tunnel in Milbrae!

    Grade crossings require CBOSS!

    Freight trains require CBOSS!

    Caltrain wouldn't lie to us about all this, would they?

  26. @ Clem, Glad to be of help. Perhaps you can figure a way to post more pre-bullet schedules on this blog? It may provide for some interesting discussion.

    BTW, Caltrain likes to say that prior to Baby Bullet, they ran an all stops schedule, even referring to it as “trolley-like,” (whatever that is supposed to mean) and that the pre BB schedule was the reason for declining ridership, never mind 911, the dot-com bust, the week/moribund economy, and don’t forget the weekend shutdown.

    Then come Baby Bullet, yes it was a marketing coup, but there were a number of factors that led to the huge increases in ridership, BB being just one of them. For the first time, Caltrain ran a very effective marketing campaign and by their own admission, the largest in Caltrain history. Baby Bullet and then Caltrain reinvented (2005) ads appeared in local newspapers for months.

    Increasing gas prices are the primary reasons for Caltrain ridership increases, first due to damage to gulf oil facilities due to Hurricane Katrina, and then gas prices raising to $4.00/gallon. The return of weekend service also helped to boost ridership. Other factors leading to ridership increases include: improving economy/job increases, increased traffic congestion (which comes with the job increases), the BART connection at Millbrae, improved reliability/on-time performance of Caltrain, proof-of-payment most likely led to more fare revenues and therefore increased ridership. My arguments/criticisms of Caltrains dogma that BB being the primary reason for increased ridership really seemed to strike a nerve with some Caltrain staff personnel, as they really gave me the cold shoulder, which tells me that my arguments were really valid and makes it look as if they are not doing their homework. I would bring up these issues from time to time, over the years, at the monthly Caltrain/JPB board meetings looking for their response. I did quite a bit of research and presented numerous charts and spreadsheets of ridership and gas prices, and letters regarding the reasons for increased ridership, to Caltrain at the board meetings. I finally received somewhat patronizing response from Caltrain management to my arguments, trying to validate their position, it even got to the point where I was threatened with censure/no communication with staff, other than at the monthly Caltrain/JPB board meetings. They tried to put the lid on me so to speak. They wanted to focus on how to further improve Caltrain in the future and I should look forward to that.

  27. @ Anonymous said: “Interesting to note that the old schedule also had five trains per hour….”

    It gets better than that, if you look at the pre-Caltrans schedule, under Southern Pacific, they ran (express) trains on three minute headways. Five trains departing San Francisco in 12 minutes beginning at 5:14 pm, between 5:14 pm and 6:00 pm there were seven trains departing San Francisco. If you go back to the late 1950’s, SP ran eight trains on 3-minute headways beginning at 5:14 pm through 5:35 pm and a total of ten trains in less than one hour through 6:00 pm (and using Steam engines for Christ’s sake). Granted the service was different in those days, being that it was a true “commuter” service i.e. from the suburbs to the City in the morning and back to the suburbs in the evening. But we hear all this talk of *we can’t run more service till CBOSS, etc…. These (signal) upgrades will allow us to run up to six trains per hour in each direction. So what does this tell you?

    @ Anonymous: “Did they boost OTP with schedule padding, or did dwell times get worse, or operating practices get sloppier, or what?”

    All three are true, they boosted the schedule padding, dwell times got worse, and operating practices got sloppier… Dwell times are bad because of increased ridership and bicycle boarding, completely aggravated by the stupid single door, high steps, gallery cars. Many of us told Caltrans back in the early 1980’s that they should get the Bombardier cars as they were in the process of procuring the current fleet of gallery cars, oh but they wouldn’t listen, so now we have to suffer from overlong dwell times.

    @ Anonymous: “and that that old end to end all-stop running time was also 1:31, though with stops at Atherton and Broadway”

    It is 1:31 for northbound trains and 1:30 for southbound trains, serving 24 stations. The current schedule is 1:31 for northbound and southbound trains but now serving only 22 stations. Caltrain is definitely making backwards progress…

  28. @Jeff Carter: I too have been skeptical of the current diminutive projectile train schedules; particularly reverse commuter trains skipping the job rich Lawrence and Santa Clara stops near the end of their runs where few through passengers remain on the train to be annoyed by calling on additional stations. Clem’s post shows that by far the greatest neglected station ridership drop-off, 778, occurs at Lawrence. Two of the three Lawrence employer shuttles call on no other Caltrain stop. A Caltrain reverse commuter not willing to take the slowest train available to a nearby rail station is stuck as a long distance pedestrian−unless he has a bicycle which works well enough until he has to make a formal presentation on a rainy day. Express trains not stopping at the Santa Clara leaves the San Jose Airport or Golden Triangle commuter unable to take advantage of the ten-minute-headway-shuttle to the S. J. Airport and the First Street Light Rail.
    Most riders who formerly used the California Avenue and Belmont stations 1.6 and 1.7 miles respectively from the currently well served University Avenue and Hillsdale stations can easily switch to those busy stations by driving from home. Most of those former California and Belmont station riders are probably still riding Caltrain. However the BART Airport and Millbrae Extension trains both stopping near Caltrain’s San Bruno station provides frequent service direct to the most heavily patronized market Street Stations in the entire San Francisco Bay area. Severe permanent Caltrain ridership losses were inevitable whatever level service Caltrain might provide.
    The current rush-hour-express scheme blows away the ridership increase that should accrue if constant all-stop hourly service were provided. A passenger with convenient access to a particular station would have to memorize only that at a constant number of minutes past the hour service throughout the day and evening 7 days a week at that station would be provided. This is a particularly important service attribute in a travel market where Caltrain’s ridership is only 7% of transport provided by the parallel I280 and SR101 freeways−there are many potential occasional riders many of whom don’t have immediate access to a schedule.
    Here is one solution using current Caltrain rolling stock: Schedule all hourly local trains as one engine two Bombardier car trains. A careful performance evaluation indicates the acceleration, running, and deceleration mostly in dynamic mode, total run times would be 8.49 minutes faster than a 5 car local train between the Lawrence and Bayshore passing tracks. The fastest current express running times between those passing track segments is 42 minutes. Caltrain’s current mid-day local running time between Lawrence and Bayshore is 67 minutes. Subtracting 8 minutes running time to 59 minutes would leave a 17 minute running time closure between the fastest expresses and two car all-stop locals.
    This approach will work only if locals are not overcrowded and are rarely delayed by wheel chairs. Fifteen inches above current platform level loading areas for one door each side for all stations may be sufficient to speed the loading process with only one conductor. Requiring only one conductor on all local trains may save enough operating funds in order to pay for any extra express trains needed to fend off two-car-local-train overcrowding.

  29. The big innovation of the "Baby Bullet" schedule revamp was that most Caltrain crews were rostered -- the indignity! after decades of work practice -- to complete two whole passenger carrying revenue round trips per work day.

    A "work" day of driving, or worse, conductor-ing, a train from San Jose to San Francisco, then one to San Jose, then going home, became (mostly) a thing of the past.

    "Double doubles" -- two round trips, once reserved only for the lowest on the seniority pole -- became the norm.

    This huge step was, of course, was fought tooth and nail by TWU and BLE -- no human being can be expected to make 80 stops in one day (aside: bus drivers are super-human, as are train drivers driving cars to work) -- but nobody seems to have died or had their heads explode from the sheer effort ... so far.

    Everything else -- the dodgy train overtakes, the insanely unjustified and inutile quadruple and triple track construction, the sub-sixty-minute SJ-SF expresses -- is a side show to this.

    Having most Caltrain crews work more than a single round trip a day is the only improvement that has been made in the service in two decades. The trains are still overstaffed by a factor of two or more, but we make up for it by having them work more than 3 hours a day here in the 21st century.

    Sad, but true.

  30. It is my assumption but losing Gilroy extension ridership can be one of reason why ridership of Santa Clara and Lawrence were so reduced.

  31. Schedule all hourly local trains as one engine two Bombardier car trains........A careful performance evaluation indicates the acceleration, running, and deceleration mostly in dynamic mode, total run times would be 8.49 minutes faster than a 5 car local train between the Lawrence and Bayshore passing tracks. .......

    If you see Caltrain ridership report, express-local trains which stops all station north of Redwood City have poorest performance since 2005. (on the other hand, express-local train which stop south of Redwood city demonstrates excellent performance next to Baby Bullet.

    If Caltrain operate "Short" local train only north of Redwood city, 2 Car should be enought to handle the demand. Station spacing north of Redwood city to Milbrae are relatively shorter - advantage of short train (accerelation) is fully demonstrates.

  32. "It is my assumption but losing Gilroy extension ridership can be one of reason why ridership of Santa Clara and Lawrence were so reduced."

    South of SJ ridership has always pretty much irrelevant to everything -- except Caltrain's disastrous operating and capital budget that is. The percentage drops may be large, but the numbers are negligible anyway. (It's like when the Amtrak people go on about ridership skyrocketing up by 10% or whatever -- what's 110% of zero again? A rounding error.)

    Perhaps if trains ran "short" just to Capitol, where the population drops off, on UPRR-dominated track, note, and did it every 30 minutes then some ridership might develop after many years of haemorrhaging cash, then, maybe, maybe. But I doubt it. And who can afford to find out? (A: VTA!)

    That irrelevance aside ...

    The decimation of Sunnyvale, Lawrence and Santa Clara station ridership is, like that of California Avenue, entirely self-inflicted by Caltrain's staff.

    The businesses, or at rather the density of business activity, that was there in the early-mid 2000s is there and in fact higher, but the train service is infinitely worse -- "hour headway" equates to "doesn't exist, I'm driving, screw you!"

    There was a good deal of SF and northern San Mateo County to southern Santa Clara ridership back in the day (I was part of it at times), and there would be even more today (changing demographics of "dot-com" riders will pay to suffer by taking the train, but not suffer hour headways) except that the train service just isn't there.

    Ride a "reverse commute" train from SF some morning: they're ghost trains south of Mountain View because (pretty much) nobody actually wants to go to "downtown" San Jose, and the train doesn't make the stops where people do want to go.

    Sad, but true. And entirely self-inflicted. And of course they're immune to any change because BABY BULLET HAS SAVED CALTRAIN (except when their There Is No Alternative coming all-skip-stop Metra-in-the-1970s-style madness will replace it.)

    A astonishing but true contribution to this state of affairs was (and remains) the unshakeable religious conviction of former Rail Transformation Head Bob "CBOSS" Doty that full station parking lots (for park-and-riders heading northbound towards SF in the morning) mean that trains heading southbound carrying people to workplaces have no need to stop at those stations either. You can't make this stuff up! You'll also see this insanity with trains than express past Palo Alto (in either direction) or past 22nd (in the "reverse commute" direction) -- there's absolutely no possible justification for not stopping, except that the parking supply is deemed to be full, so too bad for work-bound shuttle riders and pedestrians and cyclists!

    I just don't know how to fix anything other than to kill off Caltrain and start from a blank slate. Nobody involved displays a single glimmer of normal human intelligence.

  33. @Richard Mlynarik: There are examples where rail transit labor costs have been sharply reduced in Chicago, where I grew up. The Suburban Chicago commuter train practice of having riders place their tickets in clip-holders near every seat eases the task of determining who hasn’t paid thus effectively increasing the speed and reducing uncertainty for conductors performing their ticket checking tasks. This approach would be particularly appropriate for Caltrain with their customers’ varied travel patterns. (Note: While a few Chicago transit practices have been helpful for present and future San Francisco area transit such as push-pull operation and FSSF on two extensive four track lines but other Chicago Suburban Transit practices such as gallery-cars and stub-end downtown terminals have become a nightmare for Caltrain. I wish those who travel to distant lands to see attractive, financially successful rail services would use similar discretion when upon returning to California in spite of a markedly different environment. They often say we should duplicate exactly the hardware and operating practices they have seen. For example the Swiss Railways 10% schedule pads are appropriate when running a frequent bi-directional service through numerous single track tunnels plus frequent sharp curves while maintaining timed transfers with connecting services. Curves sharply limit make-up time speed increases and off schedule running on single track sections delays traffic in both directions. These conditions do not exist on SF-SJ Rail except to a very limited extent while approaching their terminals. Considering the logic of the situation rather than the credibility of the source the CHSR Authority’s projected 1% schedule pad for the one intermediate stop SF to LA expresses run on a top priority basis is probably correct.)
    Back to crew efficiency: Enabling the head-end operator to activate the opening and closing of passenger car doors plus a 1.25’ X16’ elevator to rise 1.25’ in order to fill the gap between the 1.25’ high handicapped loading platform and a Bombardier car door would allow one employee to perform all train movement operations related to providing passenger service. Low cost reliable hardware needed to construct 20 square-foot 1.25’ lift distance elevators could be obtained from the ubiquitous garage door and automobile lifting industry suppliers.
    This approach could eliminate wheel chair delays and reduce running times. Also the minimum two crew member requirement needed for wheel chair boarding could be dropped. In addition the remaining conductor could get off an inbound train in order to meet his scheduled outbound run during system delays.
    Although running passenger trains with no conductor would be possible a wiser course would be to use one conductor per train unencumbered with operating duties to resume selling tickets on board. A proof-of-payment system for a frequent service for a population familiar with public transportation makes sense. But an infrequent service severely fining passengers not familiar with the system for making a mistake, as Caltrain does, will sharply discourage potential new customers. Implementing a proof of payment system for Caltrain looks like another example of importing a foreign system that is not appropriate for local conditions.

  34. Richard Mlynarik said...

    Since Bob Doty left the Caltrain, can Caltrain improve those mistakes?

  35. Re: "can Caltrain improve those mistakes?"

    I don't know. The signs aren't promising. And remember, Bob was by all accounts the least stupid of the lot. (True anecdote: a former Caltrain Chief Engineer was unfamilar with and perplexed by the term "throughput". You know, trains per hour. THe things that ruin what he fundamentally thought of as a static and trafficless freight railroad.)

    My information these days is third hand (very reliable, and independently confirmed, but third hand) so I can't possibly say.

    But given that they continue to do nothing right, at even the most trivial level, even on issues which have no downside to anybody, it's hard to be encouraged or to try to be encouraging.

    Another thing one has to try to understand is that Baby Bullet is the only "success" that most of the sad cases involved have ever seen or will ever see in their careers. Defend it to the death in every detail!

    And indeed more people are riding the train than a decade ago (though everybody acknowledges that Caltrain gave away so many free rides back before Proof of Payment that the difference from Dot-Com v1.0 peak isn't remotely like what the official ridership counts say). The agency is delivering more train-hours and train-km or service with fewer person-hours, which as I said above is the single biggest "innovation" of the last twenty years. They also worked out that fare elasticity was such that they could hugely jack up fares without shedding net revenue. (Passengers yes, revenue no.) So there's that. Success. Sort of. Can't argue with it, right?

    But when "armchair engineers" (or hugely experienced and professionally qualified foreigners for that matter) or even "Caltrain riders" suggest that a stop might be added here or things made more rational there -- let alone that the service should be run like analogous lines elsewhere in the world -- they fall back on clinging to Baby Bullet and The Push Model and You Can't Argue With Success and We're The Experts and Ridership Breaking All Records.

    (The "Push Model" FYI was Doty's and Bouchard's term for "people will drive from their homes in Sunnyvale to Mountain View and use the Mountain View parking lot if the train doesn't stop at Sunnyvale. Therefore we only need to stop trains at Sunnyvale to the extent that parking isn't available in Mountain View." This is "pushing" riders to wider-spaced stations, and running bizarre patterns of infrequent skip-stop semi-express trains to match parking demand.)

    You know, with all of this I really want to believe that things will improve. But such beliefs are entirely contrary to experience.

    As for me, it was San Bruno that finally made me finally go thermonuclear and stop pretending to myself that there was a single functioning neuron operating anywhere at 1250 San Carlos. (I should and could have given up years before, but I pointlessly kept Providing Constructive Input instead. Sucker!) Those people are beyond redemption.

  36. @ Richard Mlynarik

    How would you set up a decent Caltrain schedule, better than the one they now have?

    The current one has good points and lots of really bad points.

    What are your quick ideas to make Caltrain better given a clean slate?

  37. Dear anonymous: pretty much everything on this blog is how not to do Caltrain badly.

    Clem and I don't agree 100%, maybe 85%. That's not because we're twins or smarter than everybody, but mainly because we've seen how stuff works elsewhere, have an analytical bent, and don't see why we here shouldn't have stuff as nice as people elsewhere are allowed.

    See the top-right of this blog's home page "Corridor To Do List" for a blindingly obvious list of quite simple things that anyone non-stupid would do tomorrow to make our trivial local toy train set workbest and at lowest cost.

    So what to do:
    * All improvements delivered in accordance with an advertised strategic plan that with clearly defined stages delivers concrete service improvements to the public. Every single project exists solely in relation to the phased delivery of service, and not a dollar is spent that does not accord with it.
    * No FRA.
    * No freight for now. It can be reintroduced on CALTRAIN's terms later if necessary.
    * Adopt somebody or other's standards (I know Germany's best, but take you pick) and don't reinvent the wheel.
    * Full grade separation, introduced progressively. Nearly all separations are elevated tracks, leaving roads and pedestrian cut-throughs at grade, where they belong.
    * HS via Altamont. No HS trains screwing things up between where avoidable.
    * HS 100% integrated with Caltrain timetable and infrastructure. Shared platforms (at 550mm height; on this height Clem disagrees.)
    * TJPA Transbay blown up and killed without mercy. Workable, useful, cheaper, much higher capacity train station at that site instead.
    * Limited quadruple tracking, driven by service needs.
    * FSSF track configuration.
    * Modern single-level (double decker possible in the future) articulated high-acceleration level boarding rolling stock. Everybody knows I have it for Stadler FLIRTs, but Bombardier, Alstom, Siemens, CAF, Talgo will build similar things.
    * One Person Operation of trains.
    * Roving Proof of Payment fare inspection. Paper-based (not smartcard) fare media is perfectly adequate.
    * Caltrain timetables (30 minute or better regular "takt") completely integrated with bus service along the line, which operate on the same headways and are timed to allow best possible and shortest possible transfers at stations.
    * Caltrain fares completely integrated with local bus tickets. eg SamTrans pass works on Caltrain. Transit must be mode neutral and transfers must be without penalty or inconvenience.
    * There must be a few more, I'm just typing off the top of my head. If things are obvious to me, they ought to be obvious to anybody who looks around and travels or reads even a little.

    None of this stuff is novel in the slightest. It's just copying what works, and works insanely well, in Central Europe. Steal, don't reinvent!

    As for a specific timetable, the nicest thing I recall (it's been a while) can be see towards the end of the comments on this article. Search for "S-Bahn S.Mateo". That scenario features a Transbay-Redwood City local service shuttle, with all trains that run from SJ making fewer stops north of RWC.

    There are other variants which are about as good. Key is always to have timed, cross-platform, same-direction transfers between local and express trains (either at Hillsdale or, with more construction cost, Redwood City), and to operate a regular and predictable and simple small number (just "express" and "local" ideally) of easy to use service patterns.

    Summary: the "Corridor To Do List". A simple and obvious and also cheapest route out of the stupid present.

    Most simply: put customers first.

  38. A couple of random side-notes. First, about the Baby Bullets: everyone loves them, because they're light and spacious and have power outlets and tables so you can get some work done, and the ride is so much smoother too. Basically, the "everyone loves Baby Bullets" is at least about the newer, non-terrible Bombardier cars replacing the old, terrible gallery cars (even if the newest of them are only from 2001).

    Also, the Clipper mess is mostly not Caltrain's fault, because the MTC pretty much forced it on them whether they like it or not. I'm sure Caltrain would be okay with paper tickets, and probably wouldn't mind letting customers buy a monthly pass at a station. The one thing that Caltrain can and should be blamed for is not taking the opportunity to come up with a better fare structure, taking advatange of Clipper's capabilities. Then again, maybe Clipper's capabilities were designed by the MTC's consultants only to allow for Caltrain's existing fare structure.

    Finally, if anyone has the time to go to Caltrain board meetings, feel free to ask them to take a look at updating their timetable, especially in the reverse commute direction, because many of the constraints that were there when it was first made aren't there anymore. It's time to start taking advantage of eliminating the holdout rule at California Ave, Burlingame, and Santa Clara

  39. So I took a look at the AM Peak Passenger Activity numbers between 2002 and 2010. Here are the noticeable declines that I've found:

    Traditional Peak: Noticeably fewer people are getting on the train at Bayshore, Burlingame, Belmont, California Avenue, Lawrence, and Santa Clara. (San Jose and Sunnyvale both have over 1000 people getting on the train, probably because they're absorbing riders from Santa Clara and Lawrence respectively.) Fewer people are getting off the train at Bayshore, S. SF, Belmont, San Carlos, California Avenue, and San Antonio.

    Reverse peak: Fewer people are getting on the train at Bayshore, S. SF, and Belmont. Fewer people are getting off the train at Belmont, California Avenue, San Antonio, Sunnyvale, Lawrence, and Santa Clara.

    Here are the first things I would do in a revised schedule:
    - Belmont, Bayshore, Lawrence, and San Antonio need at least 2 tph in both directions. California Avenue and Santa Clara need at least one more tph.
    - For the reverse peak, more trains need to stop at Bayshore-San Bruno and Burlingame. The train needs to stop more at the stations between PA and SJ.

    In summary: the Baby Bullet work in the traditional peak because most riders are headed for SF, but inadequate for the reverse, because California Ave & Sunnyvale-Lawrence are just as important as SJ.

  40. Caltrain did not change the time table since October 2005. Aftet that, they did add/reduce night/midday train. However, there is no change in peak period time table.

    After 2005, Caltrain made several capital improvements. (1) Investment on Station construction: Eliminate Hold out rule. Burlingame, Hilsdale, California Avenue and soon, Santa Clara. (2) Add more signal in between to accomodate tigher train spacing. (3) CEMOF maintanance facility to improve train reliability and reduce maintenance time instead of sending engine to outside. (4)Installed Mini-High platform at more stations to allow easier wheel chair handing for Bombardia car.

    All of those imvestment should be connect with revenue increase, meaning improve time table. Caltrain can add more train in the peak period!

    Unfortunatelly,we did not see any return from those investment, except safety.

  41. Caltrain does run shuttles from Belmont to Hillsdale to make up for the 1 stop per hour. I can't say whether that's a win (short shuttle ride followed by connection with Baby Bullet) or a lose (time on shuttle).

  42. Shuttles from Belmont to Hillsdale should be changed into Hilsdale - Hayward park suttle. Then, Hayward park can be closed.
    Use the resource (caltrain claim that every stop cosume x Gallon of fuel) from Hayward park, add more train stop at Belmot, hopefully 2 train/h in peak period. Because Belmot have more potntial ridership than Hayward park.

  43. The problem with bus shuttles from one Caltrain station to another is that the Caltrain stations are suburban to begin with, with limited foot traffic to-train access, and limited surrounding walk-accessible employment. Rather than a way to get where people are going they're just another delay and another reason not to take the train;

    Suburbanites aren't going to drive to a station, pay to park, wait to take a shuttle bus to another station a mile or so away and wait again for a train. They're going to drive to the station at which the train stops. (Bob Doty's "push model" of the world, right there for you.)

    Few enough potential Caltrain passengers work close to, say,. Belmont or Hayward Park stations, and it's hard to imagine anybody putting up with getting of the train, taking a shuttle bus, and then having to walk further to get from the no-trains-train-station to their actual non-train-station destination.

    I'd pick Belmont over Hayward Park any day, but don't imagine for an instant that a half-assed bus shuttle to the inherently poorly-located Hayward Park station site is going to carry more than a dozen people a day. (Note: I don't believe buses or connecting transit are inherently "half-assed"; not if they're reliable, timed, and non-circuitously routed to meaningful destinations.)

    Fun fact to know: the rebuilt Hayward Park station and its empty-since-day-one parking lot were promoted by America's Finest Transportation Planning Professionals as (this is not a joke!) a way for motorists driving on Hwy 92 across the Bay or over the hills to take Caltrain. Mind-boggling. (Just consider for a few seconds what's involved in doing so and how poorly that compares to just staying in the private car the entire way.)

    (Way off topic, but the same sort of clowns "modelled" (aka fraudulently cranked the parameters to produce fraud-abetting ridership "predictions") that the San Bruno and SSF BART extension parking garages would be used by motorists who drove from Pacifica, parked, and took BART a couple miles to the airport!)

  44. Adirondacker1280017 January, 2012 15:26

    Suburbanites aren't going to drive to a station, pay to park, wait to take a shuttle bus to another station a mile or so away and wait again for a train. They're going to drive to the station at which the train stops.

    Assuming they can actually find parking at the station and the parking isn't the size of a football field ( or a Walmart ).
    Shuttle buses, generally referred to as station jitneys, are proving to be very popular with suburbanites around New York. Two varieties, one that circulates in neighborhoods so it skips the step where you get in the automobile and one where you park at a small remote parking lot and get dropped almost at trainside....cheaper than building parking garages and since you don't have to navigate the parking garage, some find it's "better". The older more well established ones supposedly break even.

  45. Just a quick comment: pretending all Google, Genentech, etc employees to switch to Caltrain would be crazy. Because:
    1) Caltrain train stops in the Silicon Valley are too far away from the workplaces. You still would need shuttles to take people from the Caltrain station to their workplace.
    2) Caltrain, in San Francisco, only goes to the Bayshore area (not many people working in the Valley live there) and to the downtown area. Meaning once they get to the Transbay terminal, they would probably still need to take Muni...
    Currently, big tech companies have buses literally directly going to any neighborhood in SF.
    Hence, many Google, Genentech, etc employees would have to switch from taking 1 direct shuttle to taking Muni+Caltrain+Shuttle from the station to their workplace. This will NEVER happen.
    Besides, how many thousands of people use those tech companies' "luxury white buses"? Shifting those flows to Caltrain could increase ridership by 1/3 overnight...can Caltrain accommodate all those extra passengers?