7 ms·

Using Countdown Clock Data to Understand the New York City Subway

151 points by lil_tee 8y ago

anthonybsd 8y ago
>The L train is also the only line that uses modern >communications-based train control (CBTC), which allows >trains to operate in a more automated fashion. It's not "more automated fashion", it's fully automated, LIDAR and all - they are fully autonomous. L trains have operators because of unions throwing a fit and most of the times they sit there twiddling their thumbs.
- eumoria 8y ago
  I don't mind paying a man to sit at the head of the train even if they're autonomous. I also don't believe making all the lines CBTC are going to solve the core issues of breakdowns outside of an ancient signaling system. Some of the lines in NYC are still using R32 cars[1]. They were built in 1965 that is ludicrous. Granted, they were refurbished in the late 80s but that's still 30 years ago. That said most of the lines are running "newer" stock (either from the late 80s[2] or mid-2000s[3]) but the focus on signaling, while at the core of many delays, is a band-aid on a system that is fundamentally broken by way of funding EVERYTHING appropriately. While I do appreciate the wifi in stations I completely don't care about it if the train doesn't arrive. While I do appreciate the "wait time" clocks they're meaningless when they say "5 minutes" for 15 minutes then switch to DELAY. You might as well not have a sign at all. There are fundamental bureaucratic issues going on causing many of these problems and technology is not a root solution. It's a patch. You won't be able to argue with me that one of the busiest transport system in one of the richest cities on the planet cannot manage to fund serious upgrades. Also wary of politicians blaming the unions like they're the cause of everything. This is an ongoing tactic that makes sure that nothing continues to happen. They aren't the problem directly even if there's a lot of dead weight. They're, again, a symptom of higher level management problems. [1] https://en.wikipedia.org/wiki/R32/A_(New_York_City_Subway_car) https://en.wikipedia.org/wiki/R32/A_(New_York_City_Subway_ca... [2] https://en.wikipedia.org/wiki/R62A_(New_York_City_Subway_car) https://en.wikipedia.org/wiki/R62A_(New_York_City_Subway_car... [3] https://en.wikipedia.org/wiki/R160_(New_York_City_Subway_car) https://en.wikipedia.org/wiki/R160_(New_York_City_Subway_car...
  - blattimwind 8y ago
    These all look like they're made in the 70s, including the supposedly modern R160.
  - rayiner 8y ago
    It's not a question of whether "one of the richest cities on the planet" can "fund serious upgrades." It's how much money are New Yorkers willing to throw into a system where it takes several times as much money to do the same thing as other countries? https://www.nytimes.com/2017/12/28/nyregion/new-york-subway-construction-costs.html https://www.nytimes.com/2017/12/28/nyregion/new-york-subway-.... New York is rich, but so is London. Based on that wealth, there is a certain amount of money New Yorkers are willing to pay for subway service. If it costs twice as much to do the same thing, half as much service will get purchased. (Indeed, it's worse. Crappier subway service reduces demand, which decreases the money available to fund it.) Costs are the fundamental problem with the MTA (and D.C.'s WMATA), and those are union problems, not management problems. If it cost Lenovo twice as much to build the same laptop as Acer, Lenovo would simply cease to exist. Transit systems generally cannot go bankrupt for political reasons, but they can decline to a state of government-funded life support where nobody uses them except people who have no other choice. (See, e.g., the transit systems in most cities outside NYC/Chicago/DC/Boston). Consider: > France’s unions are powerful, but Mr. Probst said they did not control project staffing. Isabelle Brochard of RATP, a state-owned company that operates the Paris Metro and is coordinating the Line 14 project, estimated there were 200 total workers on the job, each earning $60 per hour. The Second Avenue subway project employed about 700 workers, many making double that (although that included health insurance). This is a large-scale problem in the U.S. Our public services suck, which means that they turn into safety nets instead of something that are broadly used by the population. In turn, people have limited willingness to fund them (because people naturally are less willing to spend money on safety nets versus something they also use and benefit from). Unions aren't the only reason for our public services sucking, but to the extent they drive costs out of alignment with what is the case in other countries, they're a big part of the problem. If you can buy less service with the same amount of investment, that's a problem. One of the things that's happened in the last 30 years that nobody talks about is that Europe became far more market oriented, and their unions adapted. Anti-union rhetoric in the U.S. yielded a very different result, with most private unions dying out, and the public sector unions that survived remaining a bulwark of the "old way" of doing things.
- fredley 8y ago
  We have this system on a few lines in London now. The only job now for the train 'drivers' is to open and close the doors and the platform. https://en.wikipedia.org/wiki/Automation_of_the_London_Underground https://en.wikipedia.org/wiki/Automation_of_the_London_Under...
  - Piskvorrr 8y ago
    ...if the train is on happy path. And believe me, there are some really ugly unhappy paths where underground passenger rail is concerned; we definitely have no technology autonomous enough for that. https://en.wikipedia.org/wiki/List_of_London_Underground_accidents https://en.wikipedia.org/wiki/List_of_London_Underground_acc...
    - Symbiote 8y ago
      A system so safe, that almost all of the accidents in the last 50 years say "no serious injuries" or even "no injuries". One driver was killed in 1968, another in 1984, someone broken a leg in 2003, and a passenger in 2004 dragged along a platform. (The accident in 1975 was with a mainline rail train, not a metro train, I'm not sure why it's on the list.)
      - gsnedders 8y ago
        The line to Moorgate was owned by the Metropolitan Railway, which became part of the amalgamated London Passenger Transport Board, using the Underground brand, and branded as part of the Northern Line. The line up to Holloway Junction (where it joins the ECML) was owned and maintained as part of the Underground, operated by Underground drivers and Underground owned trains, and it was only transferred near the end of 1975 to British Rail. Perhaps most tellingly, it was London Underground 1938 Stock involved in the 1975 crash.
        Symbiote 8y ago
        > Perhaps most tellingly, it was London Underground 1938 Stock involved in the 1975 crash. Thanks, that's something I'd not realized. I'd assumed the situation was more like the current London Overground system.
- RossM 8y ago
  I feel it's worth reminding that there's a difference between autonomous trains, driverless trains and unattended trains. The L train is considered Grade of Automation 2 - it starts and stops itself, though it has someone in the driver's seat who controls the doors and gives control to the ATO system. Driverless (GoA3) trains move that role to a train attendant who controls the doors from within the train. Unattended (GoA4) trains function without needing an on-board operator at all. The driver probably is just sitting with a hand near the emergency stop button most of the time, but they're still required to pay attention. I imagine it's harder work now that they don't have to necessarily concentrate to operate the train. I live in London, where a lot of tube lines are already (or soon to be) GoA2, with the introduction of a part-GoA2 mainline service this year (Thameslink). I don't think we'll see a GoA 4 tube for many years; aside from the obvious union rows, you've got safety concerns (primarily platform-edge-doors, which can't be retrofitted to cramped stations), and there's mixed public opinion on the idea of being stuck on a broken down train in a tunnel without on-train staff. Then again, people don't seem to acknowledge that the airport terminal transfers are unattended GoA4.
  - fredley 8y ago
    The DLR is GoA 4. It wasn't exactly smooth sailing at first: https://www.youtube.com/watch?v=mwruoagYq0U https://www.youtube.com/watch?v=mwruoagYq0U
    - Symbiote 8y ago
      The DLR is level 3. The video is from 1988 (good find!), but the only accidents ever to have occurred were under manual override. The Copenhagen Metro is an example of a level 4 underground system. It's fairly new, so the tunnels all have an adjacent walkway which can be used in an emergency. https://en.wikipedia.org/wiki/Docklands_Light_Railway#Accidents_and_incidents https://en.wikipedia.org/wiki/Docklands_Light_Railway#Accide...
  - onion2k 8y ago
    Does the DLR count as a GoA3 train? There isn't a driver sitting at the front.
- gruez 8y ago
  >L trains have operators because of unions throwing a fit and most of the times they sit there twiddling their thumbs. rentiers, if you will
- gowld 8y ago
  > most of the times they sit there twiddling their thumbs. Most of the times? What would happen if operators were twiddling their thumbs (or absent) the rest of the times? Uber's self-driving car also has an operator "twiddling their thumbs most of the times", but I wouldn't want to eliminate the operator.
  - chimeracoder 8y ago
    > Most of the times? Well, they're not always twiddling their thumbs. Every time the train comes in to the station, they have to open the window and point up at the zebra-striped sign. That's not literally twiddling your thumbs, but I'll leave it as an exercise to the reader to determine whether that serves any purpose on a fully-automated train that couldn't be eliminated with no loss of functionality, safety, or convenience.
  - mseebach 8y ago
    > Uber's self-driving car also has an operator "twiddling their thumbs most of the times", but I wouldn't want to eliminate the operator. Eliminating the driver is literally the whole point of self driving car research, at least as pursued by Uber and Waymo.
- tomarr 8y ago
  Well this isn't correct either. It's not fully automated in the sense that doors are still operator controlled. LIDAR is completely unrelated and is used for asset management rather than signalling.
- ggm 8y ago
  Dismissing Unions assumes you really know all the reasons Unions might object. You go directly to 'taaaking our jooooobs' without even reflecting once on the other concerns. BTW, if people don't have jobs what is the economy? we were told a few years back that with the decline of manufacturing the jobs were in the service industry. Service, like y'know, staff on the underground? Maybe, the path out here is to keep them working and paid, but have them do other things? Unions are not bad. Sloppy critique of unions because 'story' is bad.
turdnagel 8y ago
Bless Todd W. Schneider and his in depth mathematical analyses of travel times in NYC. See also his epic Uber/Taxi post: http://toddwschneider.com/posts/analyzing-1-1-billion-nyc-taxi-and-uber-trips-with-a-vengeance/#update-2017 http://toddwschneider.com/posts/analyzing-1-1-billion-nyc-ta...
- macintux 8y ago
  I found it interesting to contemplate how much harder it would be for someone to do this level of analysis and draw conclusions from it if they weren't physically proximate. I love remote work, and I hope to sustain it for the rest of my career, but there is something to be said for direct exposure to the systems you study. (It also makes me appreciate all the more how hard it must be for archaeologists et al to glean details of the past from fragmentary evidence of a world long gone.)
puzzle 8y ago
He's completely right that the data is messy and sometimes makes no sense at all. Even a passenger waiting for a train, if paying attention, will notice trains appear, disappear and "jump around" on the countdown clocks.
Alex3917 8y ago
I notice the numbered trains are all coming out ahead of the lettered trains. Is this taking into account the bug in the MTA data where the countdown clocks for the lettered trains weren't accurate at the terminal stations until a week or so ago?
exegete 8y ago
The author does briefly mention this, but the countdown clocks are not every accurate. I often see a wait time of "3 min" and then it's there in the next 10 seconds. Or you can see the opposite where it just hangs on "3 min" for a minute or two and was really 5 minutes away. In reality it should just report at its accuracy, e.g., "<5 min" "5-10 min", "10-15 min", etc. PS, why do people stick their necks over the tracks to look if a train is coming?
- prophesi 8y ago
  Yeah, I've heard of one vendor needing to use both the real-time and static-time data in such a way to determine how off-schedule the trains are, and using that information to discard the unreliable real-time data bits.
- graedus 8y ago
  > PS, why do people stick their necks over the tracks to look if a train is coming? You can often see further down the tunnel from an angle close to the tracks than you can standing behind the yellow strip. If someone ahead of you on the platform is already doing this, then they're probably blocking your view down the tunnel, so you have to lean to see past them. It seems pretty silly, but sometimes it's useful, for example, in determining whether the express or local will come first, and you can then switch platforms at the last minute if it makes sense for you. If countdown timers were reliable and accurate, fewer people would feel compelled to lean out.
- jhj 8y ago
  Also, many days the system is outright not running at all, at least as far as displaying arrival times in the station, even on the L. This seems to be the case many weekends for some reason.
- tialaramex 8y ago
  Accurate countdown clocks are hard, they by definition involve predicting the future which is widely accepted to be a Hard Problem :) More specifically there are two major problems one of which is expensive to solve, and the other is impossible to solve. The expensive one is providing and making use of lots of sensors so that we can have the best possible evidence on which to base our prediction. Most obviously we need to sense where the train is - the more accurately the better. But it's also valuable to know if the train is moving, and even if the previous platform is full of passengers who will board it before it can leave, or empty. The estimates may prove wrong because of something as dramatic as a fault, or as trivial as somebody's coat trapped in the door. London's system has always seemed fine to me, but it's noticeable that buses (for which London also provides countdown clocks) have much poorer accuracy, because as well as the vagaries of passengers they must contend with variable traffic. It only takes one idiot trying to reverse a lorry onto a major street to add 3-4 minutes delay to your bus and of course if you can't see the lorry you have no idea that's where the time went. I don't agree that it makes sense to reduce the displayed precision as you suggest. The problem isn't with precision, it's accuracy, and you can't really fix that by reducing precision.
erikig 8y ago
Kudos to MTA NYC - the raw data to perform analysis is available here: http://web.mta.info/developers/MTA-Subway-Time-historical-data.html http://web.mta.info/developers/MTA-Subway-Time-historical-da... The details on how the countdown clocks work is available here: https://www.amny.com/transit/subway-countdown-clock-complaints-1.15351692 https://www.amny.com/transit/subway-countdown-clock-complain... GTFS Format: https://developers.google.com/transit/gtfs-realtime/ https://developers.google.com/transit/gtfs-realtime/
- 1wheel 8y ago
  That feed archive haven't been updated in years; you have to scrape and save the feed yourself or scrounge together different sources to look back in time.
xhrpost 8y ago
Moving due to the L tunnel shutdown next year, I've been asking if any other train comes nearly as often (I rarely wait more than 5 minutes). This is a very timely list, thanks!
hendersoon 8y ago
This is interesting analysis, but it's based on NYC MTA countdown clock data which is just rampantly inaccurate on lines other than the L. It is regularly, customarily, up to 4 minutes off in either direction, in my experience. Since trains are supposed to be spaced out 4-5 min or so, that means the clock is only really useful to tell if there's a huge delay.

151 points by lil_tee 8y ago

>The L train is also the only line that uses modern >communications-based train control (CBTC), which allows >trains to operate in a more automated fashion. It's not "more automated fashion", it's fully automated, LIDAR and all - they are fully autonomous. L trains have operators because of unions throwing a fit and most of the times they sit there twiddling their thumbs.

eumoria 8y ago
I don't mind paying a man to sit at the head of the train even if they're autonomous. I also don't believe making all the lines CBTC are going to solve the core issues of breakdowns outside of an ancient signaling system. Some of the lines in NYC are still using R32 cars[1]. They were built in 1965 that is ludicrous. Granted, they were refurbished in the late 80s but that's still 30 years ago. That said most of the lines are running "newer" stock (either from the late 80s[2] or mid-2000s[3]) but the focus on signaling, while at the core of many delays, is a band-aid on a system that is fundamentally broken by way of funding EVERYTHING appropriately. While I do appreciate the wifi in stations I completely don't care about it if the train doesn't arrive. While I do appreciate the "wait time" clocks they're meaningless when they say "5 minutes" for 15 minutes then switch to DELAY. You might as well not have a sign at all. There are fundamental bureaucratic issues going on causing many of these problems and technology is not a root solution. It's a patch. You won't be able to argue with me that one of the busiest transport system in one of the richest cities on the planet cannot manage to fund serious upgrades. Also wary of politicians blaming the unions like they're the cause of everything. This is an ongoing tactic that makes sure that nothing continues to happen. They aren't the problem directly even if there's a lot of dead weight. They're, again, a symptom of higher level management problems. [1] https://en.wikipedia.org/wiki/R32/A_(New_York_City_Subway_car) https://en.wikipedia.org/wiki/R32/A_(New_York_City_Subway_ca... [2] https://en.wikipedia.org/wiki/R62A_(New_York_City_Subway_car) https://en.wikipedia.org/wiki/R62A_(New_York_City_Subway_car... [3] https://en.wikipedia.org/wiki/R160_(New_York_City_Subway_car) https://en.wikipedia.org/wiki/R160_(New_York_City_Subway_car...
- blattimwind 8y ago
  These all look like they're made in the 70s, including the supposedly modern R160.
- rayiner 8y ago
  It's not a question of whether "one of the richest cities on the planet" can "fund serious upgrades." It's how much money are New Yorkers willing to throw into a system where it takes several times as much money to do the same thing as other countries? https://www.nytimes.com/2017/12/28/nyregion/new-york-subway-construction-costs.html https://www.nytimes.com/2017/12/28/nyregion/new-york-subway-.... New York is rich, but so is London. Based on that wealth, there is a certain amount of money New Yorkers are willing to pay for subway service. If it costs twice as much to do the same thing, half as much service will get purchased. (Indeed, it's worse. Crappier subway service reduces demand, which decreases the money available to fund it.) Costs are the fundamental problem with the MTA (and D.C.'s WMATA), and those are union problems, not management problems. If it cost Lenovo twice as much to build the same laptop as Acer, Lenovo would simply cease to exist. Transit systems generally cannot go bankrupt for political reasons, but they can decline to a state of government-funded life support where nobody uses them except people who have no other choice. (See, e.g., the transit systems in most cities outside NYC/Chicago/DC/Boston). Consider: > France’s unions are powerful, but Mr. Probst said they did not control project staffing. Isabelle Brochard of RATP, a state-owned company that operates the Paris Metro and is coordinating the Line 14 project, estimated there were 200 total workers on the job, each earning $60 per hour. The Second Avenue subway project employed about 700 workers, many making double that (although that included health insurance). This is a large-scale problem in the U.S. Our public services suck, which means that they turn into safety nets instead of something that are broadly used by the population. In turn, people have limited willingness to fund them (because people naturally are less willing to spend money on safety nets versus something they also use and benefit from). Unions aren't the only reason for our public services sucking, but to the extent they drive costs out of alignment with what is the case in other countries, they're a big part of the problem. If you can buy less service with the same amount of investment, that's a problem. One of the things that's happened in the last 30 years that nobody talks about is that Europe became far more market oriented, and their unions adapted. Anti-union rhetoric in the U.S. yielded a very different result, with most private unions dying out, and the public sector unions that survived remaining a bulwark of the "old way" of doing things.
fredley 8y ago
We have this system on a few lines in London now. The only job now for the train 'drivers' is to open and close the doors and the platform. https://en.wikipedia.org/wiki/Automation_of_the_London_Underground https://en.wikipedia.org/wiki/Automation_of_the_London_Under...
- Piskvorrr 8y ago
  ...if the train is on happy path. And believe me, there are some really ugly unhappy paths where underground passenger rail is concerned; we definitely have no technology autonomous enough for that. https://en.wikipedia.org/wiki/List_of_London_Underground_accidents https://en.wikipedia.org/wiki/List_of_London_Underground_acc...
  - Symbiote 8y ago
    A system so safe, that almost all of the accidents in the last 50 years say "no serious injuries" or even "no injuries". One driver was killed in 1968, another in 1984, someone broken a leg in 2003, and a passenger in 2004 dragged along a platform. (The accident in 1975 was with a mainline rail train, not a metro train, I'm not sure why it's on the list.)
    - gsnedders 8y ago
      The line to Moorgate was owned by the Metropolitan Railway, which became part of the amalgamated London Passenger Transport Board, using the Underground brand, and branded as part of the Northern Line. The line up to Holloway Junction (where it joins the ECML) was owned and maintained as part of the Underground, operated by Underground drivers and Underground owned trains, and it was only transferred near the end of 1975 to British Rail. Perhaps most tellingly, it was London Underground 1938 Stock involved in the 1975 crash.
      - Symbiote 8y ago
        > Perhaps most tellingly, it was London Underground 1938 Stock involved in the 1975 crash. Thanks, that's something I'd not realized. I'd assumed the situation was more like the current London Overground system.
RossM 8y ago
I feel it's worth reminding that there's a difference between autonomous trains, driverless trains and unattended trains. The L train is considered Grade of Automation 2 - it starts and stops itself, though it has someone in the driver's seat who controls the doors and gives control to the ATO system. Driverless (GoA3) trains move that role to a train attendant who controls the doors from within the train. Unattended (GoA4) trains function without needing an on-board operator at all. The driver probably is just sitting with a hand near the emergency stop button most of the time, but they're still required to pay attention. I imagine it's harder work now that they don't have to necessarily concentrate to operate the train. I live in London, where a lot of tube lines are already (or soon to be) GoA2, with the introduction of a part-GoA2 mainline service this year (Thameslink). I don't think we'll see a GoA 4 tube for many years; aside from the obvious union rows, you've got safety concerns (primarily platform-edge-doors, which can't be retrofitted to cramped stations), and there's mixed public opinion on the idea of being stuck on a broken down train in a tunnel without on-train staff. Then again, people don't seem to acknowledge that the airport terminal transfers are unattended GoA4.
- fredley 8y ago
  The DLR is GoA 4. It wasn't exactly smooth sailing at first: https://www.youtube.com/watch?v=mwruoagYq0U https://www.youtube.com/watch?v=mwruoagYq0U
  - Symbiote 8y ago
    The DLR is level 3. The video is from 1988 (good find!), but the only accidents ever to have occurred were under manual override. The Copenhagen Metro is an example of a level 4 underground system. It's fairly new, so the tunnels all have an adjacent walkway which can be used in an emergency. https://en.wikipedia.org/wiki/Docklands_Light_Railway#Accidents_and_incidents https://en.wikipedia.org/wiki/Docklands_Light_Railway#Accide...
- onion2k 8y ago
  Does the DLR count as a GoA3 train? There isn't a driver sitting at the front.
gruez 8y ago
>L trains have operators because of unions throwing a fit and most of the times they sit there twiddling their thumbs. rentiers, if you will
gowld 8y ago
> most of the times they sit there twiddling their thumbs. Most of the times? What would happen if operators were twiddling their thumbs (or absent) the rest of the times? Uber's self-driving car also has an operator "twiddling their thumbs most of the times", but I wouldn't want to eliminate the operator.
- chimeracoder 8y ago
  > Most of the times? Well, they're not always twiddling their thumbs. Every time the train comes in to the station, they have to open the window and point up at the zebra-striped sign. That's not literally twiddling your thumbs, but I'll leave it as an exercise to the reader to determine whether that serves any purpose on a fully-automated train that couldn't be eliminated with no loss of functionality, safety, or convenience.
- mseebach 8y ago
  > Uber's self-driving car also has an operator "twiddling their thumbs most of the times", but I wouldn't want to eliminate the operator. Eliminating the driver is literally the whole point of self driving car research, at least as pursued by Uber and Waymo.
tomarr 8y ago
Well this isn't correct either. It's not fully automated in the sense that doors are still operator controlled. LIDAR is completely unrelated and is used for asset management rather than signalling.
ggm 8y ago
Dismissing Unions assumes you really know all the reasons Unions might object. You go directly to 'taaaking our jooooobs' without even reflecting once on the other concerns. BTW, if people don't have jobs what is the economy? we were told a few years back that with the decline of manufacturing the jobs were in the service industry. Service, like y'know, staff on the underground? Maybe, the path out here is to keep them working and paid, but have them do other things? Unions are not bad. Sloppy critique of unions because 'story' is bad.

turdnagel 8y ago

Bless Todd W. Schneider and his in depth mathematical analyses of travel times in NYC. See also his epic Uber/Taxi post: http://toddwschneider.com/posts/analyzing-1-1-billion-nyc-taxi-and-uber-trips-with-a-vengeance/#update-2017 http://toddwschneider.com/posts/analyzing-1-1-billion-nyc-ta...

macintux 8y ago
I found it interesting to contemplate how much harder it would be for someone to do this level of analysis and draw conclusions from it if they weren't physically proximate. I love remote work, and I hope to sustain it for the rest of my career, but there is something to be said for direct exposure to the systems you study. (It also makes me appreciate all the more how hard it must be for archaeologists et al to glean details of the past from fragmentary evidence of a world long gone.)

puzzle 8y ago

He's completely right that the data is messy and sometimes makes no sense at all. Even a passenger waiting for a train, if paying attention, will notice trains appear, disappear and "jump around" on the countdown clocks.

Alex3917 8y ago

I notice the numbered trains are all coming out ahead of the lettered trains. Is this taking into account the bug in the MTA data where the countdown clocks for the lettered trains weren't accurate at the terminal stations until a week or so ago?

exegete 8y ago

The author does briefly mention this, but the countdown clocks are not every accurate. I often see a wait time of "3 min" and then it's there in the next 10 seconds. Or you can see the opposite where it just hangs on "3 min" for a minute or two and was really 5 minutes away. In reality it should just report at its accuracy, e.g., "<5 min" "5-10 min", "10-15 min", etc. PS, why do people stick their necks over the tracks to look if a train is coming?

prophesi 8y ago
Yeah, I've heard of one vendor needing to use both the real-time and static-time data in such a way to determine how off-schedule the trains are, and using that information to discard the unreliable real-time data bits.
graedus 8y ago
> PS, why do people stick their necks over the tracks to look if a train is coming? You can often see further down the tunnel from an angle close to the tracks than you can standing behind the yellow strip. If someone ahead of you on the platform is already doing this, then they're probably blocking your view down the tunnel, so you have to lean to see past them. It seems pretty silly, but sometimes it's useful, for example, in determining whether the express or local will come first, and you can then switch platforms at the last minute if it makes sense for you. If countdown timers were reliable and accurate, fewer people would feel compelled to lean out.
jhj 8y ago
Also, many days the system is outright not running at all, at least as far as displaying arrival times in the station, even on the L. This seems to be the case many weekends for some reason.
tialaramex 8y ago
Accurate countdown clocks are hard, they by definition involve predicting the future which is widely accepted to be a Hard Problem :) More specifically there are two major problems one of which is expensive to solve, and the other is impossible to solve. The expensive one is providing and making use of lots of sensors so that we can have the best possible evidence on which to base our prediction. Most obviously we need to sense where the train is - the more accurately the better. But it's also valuable to know if the train is moving, and even if the previous platform is full of passengers who will board it before it can leave, or empty. The estimates may prove wrong because of something as dramatic as a fault, or as trivial as somebody's coat trapped in the door. London's system has always seemed fine to me, but it's noticeable that buses (for which London also provides countdown clocks) have much poorer accuracy, because as well as the vagaries of passengers they must contend with variable traffic. It only takes one idiot trying to reverse a lorry onto a major street to add 3-4 minutes delay to your bus and of course if you can't see the lorry you have no idea that's where the time went. I don't agree that it makes sense to reduce the displayed precision as you suggest. The problem isn't with precision, it's accuracy, and you can't really fix that by reducing precision.

erikig 8y ago

Kudos to MTA NYC - the raw data to perform analysis is available here: http://web.mta.info/developers/MTA-Subway-Time-historical-data.html http://web.mta.info/developers/MTA-Subway-Time-historical-da... The details on how the countdown clocks work is available here: https://www.amny.com/transit/subway-countdown-clock-complaints-1.15351692 https://www.amny.com/transit/subway-countdown-clock-complain... GTFS Format: https://developers.google.com/transit/gtfs-realtime/ https://developers.google.com/transit/gtfs-realtime/

1wheel 8y ago
That feed archive haven't been updated in years; you have to scrape and save the feed yourself or scrounge together different sources to look back in time.

xhrpost 8y ago

Moving due to the L tunnel shutdown next year, I've been asking if any other train comes nearly as often (I rarely wait more than 5 minutes). This is a very timely list, thanks!

hendersoon 8y ago

This is interesting analysis, but it's based on NYC MTA countdown clock data which is just rampantly inaccurate on lines other than the L. It is regularly, customarily, up to 4 minutes off in either direction, in my experience. Since trains are supposed to be spaced out 4-5 min or so, that means the clock is only really useful to tell if there's a huge delay.