23 July 2010

Metrics That Matter

The future of Caltrain (if it doesn't go bankrupt first) is likely to hinge on the quality of the service provided. Even if they don't ride the train, all peninsula residents can still benefit from quality train service that vacuums traffic off the road. So how exactly does one define and measure Quality?

From the point of view of a Caltrain rider, quality can be defined in terms of just four Metrics That Matter.
  1. What is the average trip time between my origin and my destination?
  2. What is the best trip time between my origin and my destination?
  3. At my origin, how long is the average waiting time between trains that go to my destination?
  4. At my origin, what is the longest waiting time between trains that go to my destination?
The first two metrics measure trip time on board the train, and the next two can be used to measure waiting time on the platform, which is just as important, and often perceived as longer than it really is. Notice that all four metrics are measured in units of time, between a specific origin and destination. These metrics don't measure operating cost. Not ease of maintenance. Not reduced emissions or noise. Not electricity or diesel. Not seats per hour. Not consist utilization. All those other metrics are important to a train operator, but riders don't understand them, and frankly don't care. Riders care primarily about time.

Figuring the Metrics

The four metrics that matter are objective measures that are reasonably straightforward to extract from a timetable. A computer can churn through a timetable to extract the metrics for every possible origin & destination pair, by making a few simple assumptions about rider behaviors as shown in the graphic at left.

For example, we can crunch the current Caltrain timetable (with 90 trains per weekday and 5 trains per peak hour), with the result shown at right. For simplicity, the graphic shows a limited sample of ten stations; simply follow the blue lines to find the intersection of the desired origin and destination, and read off the four metrics. A more complete version of this table is provided below.

Effective Trip Time

The four metrics are useful to consider separately, but ultimately we'd like to compare entire timetables to determine which timetable is better. To do this, we need to consolidate the four metrics that matter into a single "effective trip time" metric for each origin and destination pair. Thus far, the four metrics required very few assumptions and could be quantified quite objectively. As we construct an effective trip time metric, things get a bit more subjective and debatable.

The effective trip time is not the trip time experienced by any particular passenger; rather, it is a single figure of merit that reflects a global average of trip times taken by all passengers between a given origin and destination.

If a passenger showed up randomly, the effective trip time would simply be the average trip time plus 50% of the average wait time. However, most passengers don't show up randomly. They tend to show up before their train, and they also tend to prefer faster express trips. Therefore, we can create a reasonable measure of effective trip time as the sum of:
  • 70% of the average trip time
  • 30% of the best trip time (to favor express service)
  • 20% of the mean wait between trains (far less than the random arrival figure of 50%)
  • 15% of the maximum service gap (to penalize very large gaps between trains)
The first two terms of this sum (70% + 30%) bias the on-board portion of the trip toward express service. The last two (adding up to 35% of headway if the timetable is regularly spaced) reflect the time waiting at the station, with most but not all passengers showing up some minutes before their train. The combination of the four metrics that matter into a single effective trip time allows us to represent the current Caltrain timetable as shown in the graphic at right.

Notice that the effective trip time metric is constructed so as to punish very large service gaps. Due to the speed / frequency trade-off that Caltrain currently must contend with, many smaller stops are severely under-served during rush hour, commonly with gaps of 40 minutes or more, to clear the tracks for Baby Bullets. This is reflected in the table: for example, Palo Alto to 4th & King (the highest traffic and best-served O&D pair) is covered in 47 minutes, but the similar distance between California Ave and 22nd Street (an under-served O&D pair) effectively takes 67 minutes.

Whether or not you agree with the exact weighting used to construct the effective trip time metric, the fact remains that with some optimally chosen weights, the metric is a valid one. The numbers can easily be recalculated using different weighting assumptions.

Measuring the Quality of an Entire Timetable

Now that we have a single number that describes the effective trip time between any given O&D pair, it's time to generalize the approach to encompass all O&D pairs and to construct a single figure of merit that captures the service performance of an entire timetable. Obviously, not all O&D pairs can be served optimally: any timetable is inherently a trade-off between minimizing trip times for most riders at the cost of longer trip times for some riders. Figuring out what works best thus requires ridership weighting.

Ridership weights can be derived from actual weekday ridership data, shown as blue bars in the chart at right (these weights have been scaled such that they add up to one). Unfortunately, actual ridership is not always an exact reflection of underlying travel demand. Some stations suffer from a vicious circle; they have poor ridership in large part because they are poorly served. A good example of such a station is California Avenue in Palo Alto, where average weekday ridership was 1225 riders in 2002 before service was cut to make way for the Baby Bullet, dropping to 891 riders in 2010. This 27% drop occurred at the same time as overall ridership increased by 19%, and is unlikely to have been caused by any shifts in employment or residential patterns in the rather thriving vicinity of this station.

The most desirable approach would be to implement a full-featured ridership model, of the sort that has recently generated so much controversy for the state-wide high-speed rail project. That is unfortunately beyond our means, so we will simply use direct ridership weighting, with some filling in where service currently isn't provided (e.g. Transbay or Atherton).

The ridership weight of an O&D pair is the product of origin ridership and destination ridership (a measure of how many people travel on that O&D pair) and is shown by the light blue circles in the figure at left. Big circles mean heavy ridership, small circles mean light ridership. An optimal timetable will seek to provide the best effective trip time for O&D pairs where a big circle represents heavy ridership.

To construct a single figure of merit for an entire timetable, we first need to come up with a new, ridership-weighted service score for each O&D pair. This score, where higher is better, is the product of origin and destination ridership (the size of the blue circle), divided by effective trip time. Dividing by effective trip time means that shorter trip times increase the score for that O&D pair. Now all that's left to do is to sum up all the O&D service scores, and presto, we've got ourselves a single figure of merit. This allows us to compare timetables and, provided that we agree on the method used to construct that figure of merit, to determine objectively which of two timetables is the better one.

Any disagreements about which is the best timetable can then be reduced to disagreements over the scoring method. Planners tend to fall in love with their favorite solution, so taking the discussion away from the solution and focusing instead on the scoring method allows a dispassionate debate that is less colored by subjective preferences. Indeed, without an agreed-upon scoring framework (clearly defined metrics), comparing timetables is a subjective and useless exercise.

The Great Timetable Shoot-Out

Enough with metrics, are we ready for some fun, or rather, as much fun as can be had with timetables? Let's put three different timetables through their paces, and see how they stack up in terms of service quality.

Contestant #1: today's 90-train-per-day, 5-train-per-hour Caltrain timetable, to which we will assign a score of 100 for the purpose of comparison.
Contestant #2: the official Caltrain 2025 timetable published in Appendix K of the Preliminary Alternatives Analysis. This 162-train-per-day timetable shows 10 trains per hour in each direction at peak times, with every train performing a skip-stop pattern to keep the entire end-to-end run under 65 minutes--about ten minutes slower than today's Baby Bullet.
Contestant #3: a plain vanilla takt-timetable featuring only 6 trains per hour in each direction at peak times. Four of those trains are all-stops locals (running at regular 15-minute intervals, with a 93-minute run from Transbay to Tamien) and two of them are expresses, which take advantage of the four-track peninsula rail corridor to overtake locals. One of those overtakes occurs at the middle of the line at the Redwood City station, where the local and the express stop simultaneously on opposite sides of the same platform and exchange passengers--at this stop only, we make an exception to the 3-minute transfer rule. This cross-platform transfer extends the benefits of express service to a much wider selection of O&D pairs than today's Baby Bullet.
While Contestant #2 takes first place, Contestant #3 is competing with a huge handicap of 40% fewer trains per hour, but comes in just 1% behind in service quality. That's right: there exists a SIX train per hour time table that provides nearly the same quality of service as Caltrain's TEN train per hour timetable. How can that possibly be?

The secret weapon: the mid-line overtake.

The Trickle Down Effect

With the peninsula being reconfigured to four tracks for high-speed rail, it would be a terrible shame not to take advantage of some of that new track capacity to run better and more efficient local service, providing measurable benefits to local peninsula communities. Better service just might be the sugar coating to make the bitter pill of high-speed rail go down a little bit easier in places like Palo Alto, Atherton, Belmont, or Burlingame. Otherwise, why even bother?
at

55 comments:

  1. Ze Ace23 July, 2010 23:22

    Nice analysis. It'd be nice to know just how sensitive your results are to the weighting you used. i.e. does changing the weight from 70-30 to 60-40 cause a serious change in the results. Assuming it doesn't these results look solid. I agree that a mid-line overtake would be awesome. In the ideal word we'd have cross platform transfers to HSR too!

    ReplyDelete
  2. Alon Levy24 July, 2010 13:48

    Clem, in addition to the four objective metrics you use, there should be a fifth: punctuality. When the travel time is not predictable, riders respond by budgeting extra time for travel, which lengthens their commute.

    You should also include a transfer penalty, instead of a one-transfer rule. I don't know what Caltrain uses in its ridership projections, but New York's MTA assumes passengers perceive time spent waiting or transferring as taking 1.75 times longer than time spent on the train.

    I'm not sure what the punctuality metric would do to this analysis; both your takt proposal and Caltrain's 10 tph proposal require high schedule adherence. But the transfer penalty would make your proposal look better, because zero-wait cross-platform transfers are zero-penalty, whereas the transfers Caltrain is planning on aren't.

    ReplyDelete
  3. Clem25 July, 2010 17:33

    @Ze Ace: here are a few variations of the weightings. In each case, A%-B%-C%-D% are the weightings for (A) mean trip time, (B) best trip time, (C) mean gap time, and (D) max gap time.

    The scores are given in the order Caltrain 2010 (always 100); Caltrain 2025; Overtake 6TPH.

    Base case (from blog post): 70%-30%-20%-15%
    100 147 145

    Random "show-and-go": 100%-0%-50%-0%
    100 151 143

    No gap penalty: 70%-30%-35%-0%
    100 144 139

    Bigger gap penalty: 70%-30%-10%-25%
    100 148 149

    Bigger express bias: 50%-50%-20%-15%
    100 143 141

    Trip time only: 70%-30%-0%-0%
    100 137 138

    As you can see it behaves more or less as you might expect, with Caltrain 2025 really pulling away only in the random "show and go" scenario where riders don't pay attention to the timetable and just grab whatever next train comes along. This is only realistic, of course, with very short headways-- and whether Caltrain will ever afford to operate 6 minute headways is another basic question of realism.

    The other thing I'd like to emphasize is that the overtake timetable picked for this comparison has not been optimized at all. The only thing I did is tune it so that an all-stops local has the same SF-SJ run time as Caltrain 2025. That ensures that we are comparing apples to apples in terms of train performance and dwell time assumptions.

    A free autographed Caltrain timetable pamphlet to anyone who submits a better timetable, in keeping with the station-to-station timings shown here!

    ReplyDelete
  4. crzwdjk25 July, 2010 21:23

    I think ultimately the weights have to be derived empirically, but the basic conclusions are right anyway. And I think an express SF-PA-MV-SJ train has great potential of capturing a lot of ridership. Or even zone expresses, skipping everything except Millbrae on the way to Palo Alto. Then again, the Limited trains get lots of ridership at stations like Burlingame and San Carlos, going in reverse-peak direction.

    ReplyDelete
  5. Richard Mlynarik25 July, 2010 23:16

    Clem mentioned as an aside: "whether Caltrain will ever afford to operate 6 minute headways is another basic question of realism."

    One thing left unsaid -- because we're focused on the most important customer "Metrics that Matter" and not transit operator minutiae -- is that to operate Caltrain's whacky and unprecedented 10tph random-stops random-limiteds, 2-SF-terminals 2-tracks-only 2025 schedule requires 33 trains and 33 crews at peaks, whereas the unoptimized 6tph overtakes scheme requires 20. (Both assuming 15 minute peak turnbacks.)

    Where will broke Caltrain manage to dig up the 65% more equipment and train operating employees? Why would anybody choose to pay for such profligacy and inefficiency? Who'd bet the house on a scheme that has been implemented nowhere else in the world? All in order to operate a worse service from a "Metrics that Matter" perspective?


    Arcady comments: "Limited trains get lots of ridership at stations like Burlingame and San Carlos, going in reverse-peak direction."

    This isn't as true today as it was before the "98 train" schedule (with dismal peak hour headways at many minor stations, and a strong emphasis on serving stations with parking lots based on northbound AM riders.) But it could be again.

    Note that the unoptimized 6pth with overtakes schedule presented has quite a lot of space in it for throwing in some extra trains. (Note 2tph express and 4tph local; there are obviously 2 unfilled fast slots every hour.) Now between Transbay and Redwood City (yeah, yeah...) some of that extra space will be consumed by high speed trains heading into and out of the Bay Area, but even so there's still scope to slip in a couple "peak of peak" supplemental trains an hour that make a set of strategic, commute-targeted stops. All this provided that the cost of operating an excess number of trains can be justified by extra riders gained.

    The most important thing for the most successful regional transportation plans is to start out with "good bones" in the form of a solid regular-headway Takt timetbable, something that riders and connecting transit up and down the line can rely upon at all times of day. Once that is solidly in place, some supplemental peak stuff can be considered: icing on the cake.

    At least that's the way professional, success-oriented organizations go about it.

    ReplyDelete
  6. mike25 July, 2010 23:21

    Am I overlooking the link to the ridership matrix in spreadsheet or text format? It's hard to construct and evaluate alternative timetables without that input data...

    ReplyDelete
  7. Clem26 July, 2010 20:53

    @mike: here are the ridership weights I used, based on 2009 average weekday boardings. To fast-forward them to 2025, I pegged SF Transbay at 8000 and jacked up Mission Bay (4th & King) and SJ to 5000. I also filled in Atherton and Broadway with typical small station numbers. Note that only the relative ridership weights matter to the final score, since the score is normalized to Caltrain 2010 = 100 points.

    SF Transbay 8000
    Mission Bay 5000
    22nd St 927
    Bayshore 153
    So. San Francisco 353
    San Bruno 458
    Millbrae 2724
    Broadway 250
    Burlingame 727
    San Mateo 1436
    Hayward Park 237
    Hillsdale 1941
    Belmont 457
    San Carlos 1006
    Redwood City 2187
    Atherton 400
    Menlo Park 1446
    Palo Alto 3962
    California Ave 901
    San Antonio 648
    Mountain View 3455
    Sunnyvale 1916
    Lawrence 636
    Santa Clara 741
    San Jose 5000
    Tamien 652

    Again, this ridership weighting is crude because it represents only the demand served by the existing service pattern.

    I have heard a criticism that a takt overtake timetable with an all-stops local massively over-serves smaller stations (15-minute service in Atherton!) at the expense of track capacity and faster trip times for higher-ridership O&D pairs. While that complaint sounds true on the face of it, the scores don't bear out that this is a show-stopper--even with the 2009 ridership weights, which severely under-estimate demand at the lower ridership stations.

    If you were to compensate for this distortion, by say taking the square root of the ridership weights as a crude way of "leveling" the actual 2009 ridership data to estimate the true underlying demand, the overtake takt timetable does even better because it serves the smaller stops extremely well and taps far more of the (currently untapped) demand that exists there. Under such square-root weighting, the timetable scores are: Caltrain 2010, 100; Caltrain 2025, 168; Overtake Takt, 168.

    Of course, nothing will replace a proper model of the true underlying travel demand, which may vary by time of day. That's another criticism that has been mentioned: the overtake takt timetable is perfectly symmetrical, and therefore does not adapt to any sort of peak / reverse-peak differentials.

    @Richard: The express trains in the overtake takt timetable would probably experience huge demand compared to the locals, and since they only run every half hour it is likely that they would be double-length EMUs. That requires about 25 EMU trainset (5 of them doubled up, for 20 consists). Granted, that's still only 75% the 33 consists required to operate the Caltrain 2025 10 TPH timetable.

    ReplyDelete
  8. Anonymous26 July, 2010 22:13

    What market are you trying to serve with the 6 tph schedule? It's slower than driving in traffic. Off peak service is pretty meek. Bus or BART is or will be faster from many of those cities.

    The elephant and 900 pound gorilla in the room is the huge funding gap for the Peninsula and High Speed Rail. What's the point of spending $2 billion on Transbay and $1 to $5 billion on upgrading Caltrain (not including new rolling stock or new operating funds) if High Speed Rail is $30 billion+ underfunded? I don't get it...I don't see the imaginary trains.

    Reality Check: All the CEQA lawyers and environmentalists in this state will never come to agreement on anything that is this big in scale since they basically hate the industrial revolution and want to extract as much money as possible from the infinite process since our economy doesn't actually make anything anymore.

    Where does this train go in the end? THE EMPEROR HAS NO CLOTHES!

    ReplyDelete
  9. James26 July, 2010 22:30

    The 'cross-platform transfer local to express' makes a lot of sense. Sounds like something CHSR should consider located at one of the high ridership origin/destination stops in the Central Valley. Cross-platform stops timed for the three different branches, SF, Sacto., and LA would make for a very well-served station. (Fresno?)


    The SF or Sacto. southbound cross-platform local/express could arrive some minutes before the LA north-bound cross-platform local/express so that SF south could transfer to Sacto north and v.v.

    O.T.: If BART is the modern system its designers thought it was, why could it not be adapted for cross-platform-local-express? Wouldn't BART benefit from such "Metrics That Matter" applied to some (or all) of the five branch lines? If "M.T.M" is as inherently as good as Clem's limited example shows, it sounds like it would be well worth investing in a few, correctly placed cross-platforms. A way to grow BART capacity (or at least efficiency) as the population grows. Even if BART reaches capacity at least the passengers could be moved with better efficiency. Or is BART hard wired for all-stops local until the Sun goes supernova?

    ReplyDelete
  10. James26 July, 2010 23:43

    @Anon 22:13
    $30 Billion? Heck the oil companies spill that much.

    ReplyDelete
  11. Richard Mlynarik27 July, 2010 00:13

    Anonymous said: "The elephant and 900 pound gorilla in the room is the huge funding gap for the Peninsula and High Speed Rail."

    Please note that a better Caltrain service, serving the residents of the San Francisco peninsula, can be quite independent of HSR.

    It could be complementary (or at the very least minimally incompatible) in many ways, and there could be a number of synergies in shared use of civil structures, tracks, signals and stations, but there's nothing other than rank institutional and professional inability that has stopped Caltrain planning for and working for and incrementally building what needs to be done for such better regional service independent of CHSRA, and having done so over the last couple of decades.

    If anybody at Caltrain were professional and skilled enough to have done any planning, we would have the mid-line overtake station and quadruple track needed for radically better service under construction right now. Instead taxpayers' money is being flushed down the drain in San Bruno. This has nothing to do with a real or hypothetical $30bn HSR funding shortfall; it's a shortfall of a quite different type.

    The main problem for Caltrain -- and for residents up and down the peninsula -- is that HSR as "planned" by CHSRA and as actively, counter-factually and politically promoted by Caltrain staff is all downside -- all costs and no benefits. Worse, more expensive regional service combined with non-stop airline surrogate intercity trains blasting through people's back yards to no local benefit. Lose-lose.

    ReplyDelete
  12. Adirondacker1280027 July, 2010 00:44

    intercity trains blasting through people's back yards to no local benefit.

    Since when are intercity trains traveling through suburbia at 125, something that happens all over the world, including US suburbs are rich and prosperous as those on the Peninsula, "blasting through people's backyards"? Along with those intercity trains sedately cruising through the suburbs at 125 there's going to be a nice grade separated electrified Caltrain sedately cruising through those suburbs which seems beneficial to me me.. but what do I know I used grade separated electric trains most workdays for decades...

    ReplyDelete
  13. Anonymous27 July, 2010 14:12

    @Adirondacker12800 two proofs using the power of *science*:

    No stops
    + high speed (it is high speed no argument no question thank you very much)
    + no HS service to us
    + worse Caltrain with HS than without it (This blog has shown why several times. Good research and education by the blog owner. Thank you!!!)
    = NIMBYS are on to some thing. (Also Caltrain riders, also any one in "midwest flyover land" between San Jose and San Francisco.)

    Another proof:

    No benefit
    - some costs
    - some impacts
    = bad deal.

    Why do people in the Adirondacks want bad deals for people in California? Do you hate us for our climate? Is it earthquake envy? Do you resent that our state government is even worse than New York's? Why???

    ReplyDelete
  14. Adirondacker1280027 July, 2010 15:20

    high speed (it is high speed no argument no question thank you very much)

    125 MPH was high speed rail back in the 1960s. Trains have been running through rich suburbs in the US and Japan since then. Lots of other places too since the 70s. The cows don't go dry, the chickens still lay, the low fat creamer at Starbucks doesn't curdle. Most places in the world where the trains only run at 125 it's a source of embarrassment that they are so slow.

    no HS service to us

    Except at SFO, wherever they decide the Mid Peninsula stop will be and San Jose. Which if they work it right, it would be really really really hard not to, will be across the platform from Caltrain.

    worse Caltrain with HS than without it

    Only if Californians make a series of really stupid decisions. How to avoid those problems has also been discussed here throughly.

    ReplyDelete
  15. Clem27 July, 2010 18:59

    What market are you trying to serve with the 6 tph schedule?

    Peninsula residents who like to get to work or wherever they need to go conveniently, comfortably, and quickly?

    Off peak service is pretty meek.

    The analysis presented here is for peak periods only. You may be alluding to the fact that my overtake schedule timetable shows no off-peak service at all. That's simply because it wasn't relevant to this analysis, although it would be relatively straightforward to add off-peak trains (e.g. an all-stops local every 30 minutes). That's not however what drives service quality, track capacity, platform capacity or fleet size.

    Bus or BART is or will be faster from many of those cities.

    The Caltrain bullets already achieve the highest average speeds of any public transit in the Bay Area (50 mph), certainly faster than rush-hour commutes on 101. The overtake timetable being contemplated here is even faster.

    If a BART logo slapped on those trains will make you feel better about the whole thing, here you go.

    ReplyDelete
  16. Jon28 July, 2010 12:37

    A good post, but surely to do a fair comparison you should include services from the CAHSR planned timetable in both the Caltrain 2025 timetable and your own takt-timetable. Passengers travelling from one HSR station to another HSR station will almost certainly take HSR rather than Caltrain. Also passengers are likely to take a combination of HSR and Caltrain if necessary to speed up their commutes.

    ReplyDelete
  17. Adirondacker1280028 July, 2010 12:49

    Passengers travelling from one HSR station to another HSR station will almost certainly take HSR rather than Caltrain.

    Assuming that HSR and Caltrain have the same fare structure and assuming that you don't need a reservation to get a ticket vaild on an HSR train.

    ReplyDelete
  18. Caelestor28 July, 2010 14:37

    Another fabulous timetable post.

    I've always wondered why no one ever thinks of building one strategic station with two platforms serving four tracks to allow for overtakes like Asia.

    ReplyDelete
  19. Jon28 July, 2010 15:12

    Assuming that HSR and Caltrain have the same fare structure and assuming that you don't need a reservation to get a ticket vaild on an HSR train.

    If the system does not have these things customers will be inconvenienced and possibly deterred from riding simply due to the ticketing system. How would that make sense?

    I would say that the ability to purchase walk-up tickets on
    HSR is more important than integrating Caltrain and HSR ticketing, although both are desirable.

    ReplyDelete
  20. Adirondacker1280028 July, 2010 15:33

    ...for the same reason Amtrak charges more to go from Martinez to Oakland than BART does.

    With a six train an hour scenario there will be a Caltrain train toddling in every ten minutes. Why would any one wait until the next HSR train comes in?

    Might be all peachy keen for the locals. The people going from San Francisco to points south of San Jose will probably take a dimmer view of it as they stand between San Francisco and San Jose while they wait for commuters to get get off the train.

    ReplyDelete
  21. Jon28 July, 2010 18:01

    With a six train an hour scenario there will be a Caltrain train toddling in every ten minutes. Why would any one wait until the next HSR train comes in?

    Because if it takes Caltrain 65 minutes to get from San Jose to San Francisco, and HSR does it in 30 minutes, it's worth waiting a little longer for the HSR if those are your departure and destination points.


    The people going from San Francisco to points south of San Jose will probably take a dimmer view of it as they stand between San Francisco and San Jose while they wait for commuters to get get off the train.

    Suppose that everyone who buys a ticket in advance gets an automatic seat reservation, while those who buy a walk-up fare don't. That means the people buying walk-up fares (i.e. commuters) would have to stand while the long-distance travelers get a seat.

    You would probably want to have one or two unreserved cars on each train so that people buying walk-up fares have a reasonable chance of getting a seat. Once those cars are full, the people without reservations would have to stand.

    ReplyDelete
  22. Adirondacker1280028 July, 2010 19:12

    Because if it takes Caltrain 65 minutes to get from San Jose to San Francisco,

    Caltrain isn't and won't be BART. They are going to have these things called "express trains" The "express trains" will be serving stations where HSR doesn't stop. And they will have travel times comparable to HSR.

    Suppose that everyone who buys a ticket in advance gets an automatic seat reservation, while those who buy a walk-up fare don't.

    How do you enforce that in a 30 minute ride? How does someone boarding in San Francisco with no seat reservation know that the empty seat isn't reserved for someone who will be boarding at SFO? How does the passenger with a seat reservation who boards at SFO know that the freeloader warming his seat doesn't have a reservation. How many conductors and assistant conductors are on board to keep the fist fights from breaking out?

    ReplyDelete
  23. Alon Levy28 July, 2010 20:49

    Assigned seats. If you have a ticket that says you get seat 59 on car 6, and seat 59 on car 6 is occupied by someone else, you show your ticket and the person moves.

    ReplyDelete
  24. Joey28 July, 2010 21:23

    @Adirondacker

    In Japan, certain traincars are all reserved seats and others are for those without reservations. Also, on some Swiss trains (ETR610s specifically), they have small monochrome LCDs above each seat listing the reservation and destination of each passenger

    ReplyDelete
  25. Peter28 July, 2010 21:25

    @ Joey

    I think the newer and refurbished ICE cars do the same.

    ReplyDelete
  26. Adirondacker1280028 July, 2010 22:01

    Assigned seats. If you have a ticket that says you get seat 59 on car 6, and seat 59 on car 6 is occupied by someone else, you show your ticket and the person moves.

    And if they don't? And how do you prevent them from sitting in unsold seats? or seats that are reserved for SFO to Fresno and therefore unoccupied between San Francisco and SFO. It will be a free for all needing lots and lots of conductors.

    ReplyDelete
  27. Peter28 July, 2010 22:11

    How many conductors are there on an ICE? On an IC train? They use the same system. It's really not that complicated.

    Just because Amtrak hasn't figured it out yet doesn't mean that it won't work.

    Someone who got on the train at Transbay will find a seat on the train. They're not about to get back off just because they didn't find a seat right away.

    ReplyDelete
  28. Adirondacker1280028 July, 2010 22:50

    How many conductors are there on an ICE?

    What's the fare premium on an ICE? What's the on board surcharge to upgrade your ticket or buy a ticket if you get caught trying to leech a ride before the conductor gets through the car?

    Someone who got on the train at Transbay will find a seat on the train.

    And unless Caltrain fares are much lower people will get on the HSR train. If 600 people get on a train with 500 seats something is gonna give. It's not gonna be pretty unless there's platoons of conductors to enforce it in the 20 minute ride to Palo Alto.

    ReplyDelete
  29. Clem28 July, 2010 22:55

    Passengers travelling from one HSR station to another HSR station will almost certainly take HSR rather than Caltrain.

    This is extremely unlikely. The private HSR operator is nearly certain to use a yield-managed fare structure, just the same as airlines. This pricing model is fundamentally different from subsidized public transportation like Caltrain or BART, where fares are kept artificially below the point of maximum profit in order to boost ridership (and relieve rush hour highway congestion). For HSR, the bum that fills the seat is the bum that pays the most, and you can be sure that won't be a peninsula commuter.

    The only way to integrate the fare structures would be to have Caltrain or BART pay the HSR operator to accept commuters... those very same highest-yielding commuters that they lose in this deal. Hell will freeze over first.

    This is one of the reasons (among several) why using HSR as an ersatz Caltrain express service will never work in practice, and why confining Caltrain to two tracks is a huge, HUGE step backwards for peninsula commuters.

    Repeat with me: Any train, any track, any platform!

    ReplyDelete
  30. Jon28 July, 2010 23:52

    Alon Levy, Joey and Peter have already made the points I was going to make regarding assigned seats. LCD displays above the seats saying whether they are reserved or not works in the rest of the world without fist fights breaking out. I have been riding such trains for years and have never seen any drama greater than someone politely pointing out to someone else that they are in their seat. In reality even that rarely happens, as the trains are not usually full enough for the issue to occur.

    HSR rail is not that different to conventional rail, and not that similar to air travel. It costs much less to add an extra car to a train, high speed or otherwise, than it does to add the same amount of seats to an airplane. For this reason, and for reasons of comfort, HSR operators will towards having more seats that are necessary, where as airlines must aim for zero empty seats in order to make the flight cost effective.

    Repeat with me: Any train, any track, any platform!

    I completely agree, and agree that Caltrain express service should be preserved. But I also think anyone travelling between two HSR stations will want to do it in the shortest amount of time possible, even if it means spending a few dollars more on a high-speed service. I'm not sure why that is such a controversial point.

    ReplyDelete
  31. Alon Levy29 July, 2010 00:52

    Clem, at least one of the potential private operators of CAHSR, JR East, doesn't do any yield management. You can look up Shinkansen fares in a table. And in principle they're the same as on the slow trains; in practice, there's an unavoidable seat surcharge, as all Shinkansen passengers must have a seat.

    And Adirondacker, the ICE has a fare premium, but the IC doesn't. The TGV shows up on TER schedules as an express train, and costs the same as the slow trains.

    What needs tons of assistant conductors is the Amtrak first-come-first-served model, where people who book tickets together aren't guaranteed to find seats together. Large groups and Amtrak don't interact too well, not without conductors telling everyone exactly where to sit.

    ReplyDelete
  32. Peter29 July, 2010 07:29

    @ adirondacker 12800

    I gave the example of the ICE simply because they have the new displays for the seat reservations. Get on any IC or EC train and you'll find the same seat reservation system works just fine with little sheets of paper. It's really not as difficult or as controversial as you think it is.

    I don't believe the long distance trains in Germany have more than two conductors.

    ReplyDelete
  33. crzwdjk29 July, 2010 09:57

    How useful are IC trains for commuting though? How far out is the first stop from the city, and how big is the commuting zone? It seems to me that European cities have shorter commuting distances from their suburbs, and certainly nothing like the 45 miles from SJ to SF, or even the 35 miles from Mountain View.

    ReplyDelete
  34. Alon Levy29 July, 2010 10:06

    It doesn't matter how far it is. What matters is how long the trip takes.

    ReplyDelete
  35. Peter29 July, 2010 12:39

    @ Arcady

    My argument was not about reservations on commute operations. You won't find seat reservations for RB or RE trains in Germany, for example. My argument was simply that a seat reservation system is not the Anti-Christ that some are making it out to be. It works quite well, actually. Hell, you don't have people fighting over seats on airplanes when they have an assigned seat, do you?

    ReplyDelete
  36. Adirondacker1280029 July, 2010 14:05

    Peter, the fantasy that that they are drooling over is that you show up at Transbay at 5:13 instead of your regular time of 5:16. Instead of using your monthly Caltrain ticket to get on the Caltrain 5:18 express to San Jose you just dart onto the 5:15 to Los Angeles and everyone, happy as clams, gets a seat and no one wants to get on at SFO and take the train to Fresno has to maneuver their luggage around the standees for Palo Alto.

    ReplyDelete
  37. Peter29 July, 2010 14:34

    @ adirondacker12800

    In that case, I agree that is a little far-fetched.

    ReplyDelete
  38. Alon Levy29 July, 2010 17:18

    Okay, Adirondacker, so you just completely ignored the parts about reserved and unreserved cars.

    ReplyDelete
  39. Adirondacker1280029 July, 2010 17:44

    Who is going to be in the unreserved cars between San Jose and Los Angeles?

    ReplyDelete
  40. John Bacon29 July, 2010 17:51

    Commuter use of HSR reserved seat trains: In order to prevent delays during peak traffic periods the first objective should be to prevent commuters from interfering with reserved seat long distance passengers. The second objective should be to maximize HSR system yield during off peak hours by allowing commuters to reserve seats less than 5 minutes before boarding. Commuters could use POS touch screen ticket upgrade displays near each HSR train door where unreserved seats are still available. Platform edge screens combined with turnstiles opposite each train door would only admit reserved seat boarding pass holders. Global displays would tell reserved seat holders which train door to use and inform hopeful commuters which boarding area is closest to the most not-yet-reserved HSR seats.

    ReplyDelete
  41. crzwdjk29 July, 2010 23:13

    But will it work with Clipper? This is a serious question, by the way. If it doesn't, it definitely won't be interchangeable with Caltrain.

    ReplyDelete
  42. thatbruce30 July, 2010 09:49

    @John Bacon: Platform edge screens combined with turnstiles opposite each train door would only admit reserved seat boarding pass holders.

    No. That will cut the passenger throughput in and out of the train tremendously, leading to increased dwell times and longer run times for the trains.

    Your problem is that you've spread the turnstiles out linearly and coupled them directly to the train admitting passengers, meaning that if one door has a problem admitting passengers (whether mechanical failure or the tourist unfamilar with the system), the train needs to wait for the problem to be resolved before it can depart. Any attendants that you have on the platform to deal with any such problems will need to be in good shape, as good 'ol Murphy indicates that problems will occur at each end of the train concurrently.

    Keeping the platform access control decoupled from the trains departure means that if there is a problem with one or more turnstiles, it doesn't interfere with the trains themselves. Keeping the turnstiles all in a small number of locations also means that the turnstile attendant is more available to help with issues relating to the turnstiles not functioning (or the tourist who doesn't know how they work etc).

    ReplyDelete
  43. Adirondacker1280030 July, 2010 10:41

    Thatbruce, how do you separate the HSR passengers inside the turnstiles for the express track from the passengers inside the turnstiles for the Caltrain express? How much wider does the platform need to be to accommodate the fare control sturcture running the length of the platform? What happens when there's a special event and they want to have a extra HSR train serve a Caltrain station that normally doesn't have HSR service?

    ReplyDelete
  44. Richard Mlynarik30 July, 2010 11:08

    Turnstiles? What are you trying to build? Some sort of low-throughput and cost-plus-plus-plus-plus-plus security apparatus with transportation an afterthought?

    Can anybody tell me where to find the turnstiles, fare gates, security fencing, paid areas, unpaid areas, internal passport inspection zones (GOVERNMENT ISSUED ID REQUIRED TO TRAVEL), security queues, X-ray machines, paid mezzanine levels, dedicated secured platforms allocated only to specific trains, airside shopping, landside shopping, "sterile platforms", etc in this spanky-new facility? The architects must have overlooked something really basic somehow. Quick, send over MTC's Clipper Card consultant$ and the Transbay Terminal World Class designer$ ASAP to set them straight!

    ReplyDelete
  45. thatbruce30 July, 2010 13:30

    @Adirondacker12800: how do you separate the HSR passengers inside the turnstiles for the express track from the passengers inside the turnstiles for the Caltrain express?

    On the shared platforms, your platform access control (turnstiles, ticket inspector etc) makes sure that only passengers with tickets have access, irrespective of which service the ticket is for. You rely on the onboard train staff to catch anyone who boards the incorrect service.

    On the dedicated platforms, your platform access control only permits those with tickets for the services that are going to use that platform. ( Eurostar is a good example to use for dedicated platforms, as the UK's status re Schengen requires that passports are checked. Other European HSRs operating completely within the Schengen area do not have any requirement for passport control, and most of the time, do not run off dedicated platforms).



    What happens when there's a special event and they want to have a extra HSR train serve a Caltrain station that normally doesn't have HSR service?

    You allow CAHSR ticket holders access to the normally-Caltrain-only platform for that time period, assuming that the access control mechanism at the Caltrain station recognises the CAHSR tickets.

    @Richard: Here, I'll add an
    airport to the train-station/transportation center/shopping centre mix (besides, I haven't visited the new Berlin Hbf). The access control to the underground platforms, which provide all-stops, fast, express (unreserved and reserved such as IC) and HSR services (infrequent ICE, Thalys, Fyra), is the yellow pole beside the entrance to the escalator (and those shiny poles to stop airport baggage carts going down to the platform), which is a timestamping machine for the national zone-based ticket. Most of the tickets used there don't require interaction with the machine; you just need to have a valid ticket on you in order to be on the platform.

    Being based at an airport, all of the extra security things are present, if you are catching a plane. If not, you can wander around the unsecured areas as much as you like (loitering near the mid-east and US checkin desks will attract attention though).

    I'm not saying that you need access control such as turnstiles, I'm saying that if you have them, you don't put them on the platform itself as that will slow down passengers entering/leaving the trains. Nor do you need platforms dedicated all the time to a given service, as experience from Europe has shown that it is possible for multiple services with differing fare structures to use the same set of platforms without too much drama.

    ReplyDelete
  46. thatbruce30 July, 2010 13:31

    @Adirondacker12800: how do you separate the HSR passengers inside the turnstiles for the express track from the passengers inside the turnstiles for the Caltrain express?

    On the shared platforms, your platform access control (turnstiles, ticket inspector etc) makes sure that only passengers with tickets have access, irrespective of which service the ticket is for. You rely on the onboard train staff to catch anyone who boards the incorrect service.

    On the dedicated platforms, your platform access control only permits those with tickets for the services that are going to use that platform. ( Eurostar is a good example to use for dedicated platforms, as the UK's status re Schengen requires that passports are checked. Other European HSRs operating completely within the Schengen area do not have any requirement for passport control, and most of the time, do not run off dedicated platforms).



    What happens when there's a special event and they want to have a extra HSR train serve a Caltrain station that normally doesn't have HSR service?

    You allow CAHSR ticket holders access to the normally-Caltrain-only platform for that time period, assuming that the access control mechanism at the Caltrain station recognises the CAHSR tickets.

    ReplyDelete
  47. thatbruce30 July, 2010 13:33

    Dear Blogger; if you are going to complain about a post being too long and imply that the post needs to be split up, don't post the post.

    ReplyDelete
  48. thatbruce30 July, 2010 13:48

    Speaking of multiple services using the same platform, here is an example of a station platform where two services have differing platform heights.

    The high platforms in the background are used by the metro system, and the low platforms in the foreground are used by the tram system.

    ( The metro train pictured there also switches from third-rail to overhead, as well as differing platform widths between the metro and tram systems; a youtube search for sneltrain zuid shows the changeover. )

    ReplyDelete
  49. Adirondacker1280030 July, 2010 14:42

    On the shared platforms, your platform access control (turnstiles, ticket inspector etc) makes sure that only passengers with tickets have access, irrespective of which service the ticket is for. You rely on the onboard train staff to catch anyone who boards the incorrect service.

    Then either the elaborate turnstiles are superfluous or the on board staff is superfluous. Unless your aim is to needless annoy passengers and have twice as much staff as you need. Lets not forget the company that provides and services the superfluous turnstiles.

    On the dedicated platforms, your platform access control only permits those with tickets for the services that are going to use that platform. ( Eurostar is a good example to use for dedicated platforms,

    Ah yes pick different platform heights and loading gauges. Why not make Caltrain third rail to make it even more like BART and as slow as BART?

    The reason Eurostrar has dedicated platforms is that it has different platform heights and loading gauges compared to the legacy systems. There is no legacy system in California to speak of. What little there is will be altered beyond recognition by the time the first HSR train comes through.

    A rational system won't have dedicated platforms because rational people plan for scheduled maintenance and unexpected problems. Unless your aim is to single track when there's a problem instead of going to three tracks out of four when there's a problem.

    You allow CAHSR ticket holders access to the normally-Caltrain-only platform for that time period, assuming that the access control mechanism at the Caltrain station recognises the CAHSR tickets.


    If they have conductors on the trains why do they need turnstiles at the non-HSR stations? Or if they have strung turnstiles all up and down the line why do they need conductors?

    ReplyDelete
  50. Joey30 July, 2010 17:14

    No. The Eurostar has separate platforms because the UK requires all entering passengers to be passport and security checked. True, the Eurostars conform to narrow British loading gauges, but loading gauges and platform heights are frequently not very synced between the train and the platform in Europe (at least the places I've been).

    ReplyDelete
  51. thatbruce30 July, 2010 17:41

    @Adirondacker12800 (and Richard): You probably want to re-read the comment at 29 July, 2010 17:51 from John Bacon, which is where the references to turnstiles entered the conversation, and redirect your ire accordingly.

    As I said earlier, if you have platform access controls, be they turnstiles or whatever, and you don't want them to have any impact on the dwell times of trains at the platforms, you have the platform access controls at the entrance to the platforms, not on the platforms themselves.

    ReplyDelete
  52. crzwdjk30 July, 2010 22:06

    Richard, I suspect you'd be in some trouble with the Polizei if they found out that you were travelling without your Ausweis. Your beloved Germany not only requires a national ID to travel, it requires it for just about anything else and people must carry it with them at all times. I know, this is nitpicking your over the top rant, but those are much more effective when they're actually accurate :)

    ReplyDelete
  53. Alon Levy30 July, 2010 23:04

    people must carry it with them at all times.

    Wrong. In Germany, all people must have ID, but they are not obliged to have it on them. If challenged by a police officer, they can say, "It's at home," unlike in some other countries (e.g. Israel), where they'd get fined.

    ReplyDelete
  54. David M16 July, 2012 09:54

    This analysis ignores my segment of the commute: the South County extension. Service to/from Gilroy prove an embarrassing shortcoming in the Caltrain schedule.

    ReplyDelete
    Replies
    1. Clem05 August, 2012 10:12

      The demographics clearly indicate otherwise: the Gilroy branch has negligible population, negligible jobs, and demonstrates negligible ridership. Even if you increased frequency, there just isn't enough of a market there.

      Delete