How are the routes selected? For example, in a commute from Roosevelt Ave in Queens to 42nd St (6th Ave), both the F and the 7 are direct routes. Which is being used? For another example, from Metrotech in downtown Brooklyn to Roosevelt Ave in Queens, the F and the R are direct, single-train routes, but to my experience, the consistently fastest way to go is the A/C to the E.

I was surprised with this article, very pleasantly so. As I explored it, the data seemed to be more and more at odds with reality until i found in the very small print: " The estimates are based on commutes that start and end between 6 a.m. and 10 a.m. on weekdays...." Readers need to know this right up front. The effect of burying the detail suggests that service is far better than it actually is. This is almost a commercial for how good the subway has gotten. There is a fix, however, to the query to allow readers / users to get much better information: Start End Day of Week Start Time The addition of the last two parameters will be an eyeopener for all but the subway honchos who know the ugly truth. Including "you cannot get there from here" during stations closures for work.

Interesting analysis and visualization. I have some comments and questions about this though. Firstly: what is the definition of a typical day? It differs from the median day, so is it the average day? Secondly: if you're willing to be late every other day, that would mean that you are late 50% of the time. The median is 50% of the observations, so the travel time for every other day should be the same as the median.

This is just not accurate. I take the A train, and I would LOVE if my trip was as short as this app says. My average trip is actually 20 minutes longer. Read the fine print. It says the data was supplied by the MTA, and anything that looked inconsistent was thrown out. It says if a train 'disappeared' ( whats that mean - taken out of service? moved to a different line ?) that trip wasn't counted. If a train seemed to go in the 'wrong direction' it wasn't counted. Come on. Anyone who's taken the A train has noticed magically disappearing trains and trains going backwards. Anyone who's watched the arrival time clock has seen the A train arrival time go from 20 minutes, to 9 minutes, to 7 minutes, then go back up to 12 minutes, then unexplicably STAY at 12 minutes for god knows how long. And after 33 minutes of waiting the train finally arrives. Tell me - where did that Train-In-12-Minutes go? Did THAT train go backwards? Was it counted in the statistics? When the MTA shorts the Rockaway line by sending multiple trains to Lefferts instead, does that count as on-time? Its definitely NOT considered on time by anyone who is stuck riding the Rockaway branches. (and that stupid shuttle just makes the trip longer with more waiting. that shuttle is NOT a substitute for an actual A train)

This article is impressive but misses a big point. Predictable service is not literally the goal; customers do not want reliably slow service. We want a FAST trip! I wish folks would stop using "reliability" to describe transit because speed is a more accurate goal. Quicker wait times, faster trains... this is reliably FAST. And if some trips are faster than others, we shouldn't make those fast trips slower just to be more predictable; we should speed everything else up too! People don't like saying speed because they think higher speeds are less safe. Not necessarily true. Increased safety ALLOWS for increased speed -- modern signals, better subway cars, etc. Plus slower service means people abandoning the subway and walking or driving, where they are more likely to get hurt than on the train. Please advocate for speed and build public awareness and pressure about the right priorities! This is supposed to be a rapid transit system, and it was designed for speed. Separated from traffic and built right beneath the surface for quick access, with express tracks and running 24/7. Fight for speed and make our city more efficient, livable, and sustainable!

How unpredictable is your New York Subway commute? Compared to the predictability of the subway trains in Paris, NYC is totally unpredictable. I've ridden hundreds of time on the Paris subway. Trains are always within 2 minutes of their published times. They have to be because the trains are scheduled at about 5 minutes or less apart during rush hour. and 10 minutes apart the rest of the time. Perhaps, NYC should hire the RAPT (Paris subway and bus system) to fix its mess.

I use the subway almost every weekend. Since the article reports that service has gotten a "little better," it is evidence that the governor should continue to let Mr. Byford manage the system rather than manipulate the MTA to take away his authority. Mr. Byford's plan to update the signaling seems reasonable to me and I think we should give him a chance. In my opinion, the subway is the most egalitarian way to transport people around the city and deserves our every effort to make it as good as possible.

A very well made presentation, but it's the really, really bad days that you remember, and all you need is a couple a month to form an opinion which no clever graphic will dislodge.

This data is confusing to me because it does not point out express trains in the morning that significantly cut the commute time on the A train from Rockaway Blvd. to Laffayette Blvd. (Brooklyn TEcchnical High School) or West 4th Street (NYU) 60 minutes. Other than that the commute time are similar to what I experienced in the late 70's early 80's. This only applies to the A train.

Even the "median time" enumerated in this tool is misleading. My commute is listed as 35 minutes, however, I've covered the trip in ~20/25 minutes during non-peak hours. Other cities do not have such WIDE discrepancies between peak and non-peak travel times (London, Paris, Berlin etc.). This tool would be well served to note the difference between a trip at 9am vs 2pm in order to highlight the MTA's inability to deliver a consistent, reliable level of service to its riders.

The NYC subway is amazing. I never have any problem and I've been riding everyday for 16 years. All the negative hype lately is pure politics.

Cool tool. Noticed it only does morning commutes, and takes the time from 6am to 10am. I'm assuming that trains run better 6 to 7am and 9 to 10 am than between 7 and 9. How about narrowing it down to an hour or half hour either side of your normal travel time? What about the evening ride, which I found to be longer than the morning commute? When I put my ride in, its showing the average is about 6 minutes shorter than what I experience, on most days. And the longest rides happen in the evening which can be 2x to 3x what the longest rides are in the morning.

I am surprised not to see any time-of-day variable included. But in any case, this is only one measure of what frustrates NYers about the subway. Another is the constant "Planned Work" that interferes unexpectedly with weekend, evening, and even mid-week travel.

I chose a station I knew was closed from January to mid June 2019. It told me my subway commute from that station had improved including a period when the station was closed. There is more noise in the data. Also--it's the trip home that is usually the worst part of the travel!

Great analysis and probably more informative to outsiders than to NYC strap hangers. For them, it just confirms what they well know although they could enhance the data further with their anecdotal experiences traveling the NYC subway. I'm sure it's the same way for other commuters who drive through LA or Seattle.

I cry foul on your data. Directly from the flawed beast you are trying to interview. All of this is hand selected and for all of the data points that paint a pretty picture, 6am to 10 am is far too restrictive. Who works 6a to 10a in NYC? Why not then flip this and show us the return home trip form 6p to 10p? When any schedule work is being done, all travel information is culled because the MTA knows it will sour their points. But, Sour is our level of service. I want to see all the Quantitative data from the MTA, nothing culled, and lets us see what this plots.

I left NYC (Queens) for good in 1972. Loved the subway. Could go anywhere in NYC, by myself or with a friend, from 1958 (age 11) onward. Favorite trip was Union Tpk. to 50th St (E or F) to go see the Rangers play. Also liked going to Yankee Stadium, Polo Grounds and Shea. Plus Union Tpke. or 53rd and Lex -------> Cortlandt Street-Radio Row. I enjoyed playing with this tool

One point this upshot article misses about the frustration of subway travel is not the unpredictable nature of how long the travel time between two stops will be, but rather the wait time for the train to arrive, even when there are no obvious causes for delay such as sick passengers or signal problems. I take the R train every day, and even during a single rush hour period, some trains will be spaced about 5-8 minutes apart and others more like 12, with no apparent rhyme or reason to it. So in order to be on time anywhere, I have to always allot 12 minutes just for waiting time. 12 minutes between trains during rush hour is just not appropriate for a train line that is the sole artery to much of Brooklyn!

We just returned from a trip to London and the tube's efficiency was a beautiful thing to behold. We used it every day for more than a week, and never waited for more than 2 minutes for the train we wanted. Announcements were helpful and clear, at times they even suggested faster routes we hadn't considered. It wasn't until the last day of our trip that we encountered multiple delays on 2 different lines (unwell passenger and fire, so forces beyond control). I hope Andy Byford can work his transport magic in NYC.

I don't get to be late for my job, but thanks for those 40 seconds. It's a sort of typically American result, almost imperceptible improvement is being presented as progress. Sure, in the literal sense of the word, there is progress being made. This subway system is a joke.

What if I want to never be late? Or maybe 2 times a year? Leave 2 hours early?