Comment by JumpCrisscross

1 year ago

> so foreign to my American perspective - our public transport is more like, "maybe it'll be on time, probably not, you'll have no way of knowing, also screw you"

New York’s Metro-North and LIRR have 95%+ on-time rates [1]. (EDIT: On time is defined as less than 6 minutes late [2].)

[1] https://www.metro-magazine.com/10217862/metro-north-lirr-exp...

[2] https://www.osc.ny.gov/files/reports/pdf/report-9-2025.pdf

That seems... poor? A one in twenty trains late stat will basically mean every person will have to deal with a late train roughly every week?

I bet, however, it's not uniformly distributed and some lines are late more than others.

Hopefully that 95% is them being honest about the current state, while they push higher.

  • > I bet, however, it's not uniformly distributed and some lines are late more than others

    Sure. But "with a nearly perfect on-time performance of 99.3% on Hudson, Harlem, and New Haven Lines," there isn't much room to hide problems. (LIRR is 96.3%.)

    The point is we have systems that have been well built and well maintained. They just don't get coverage because they just work. (The LIRR is far from perfect, mind you. But it's apparently outperforming the DB. You have to get to some of the worst routes during peak conditions to get in the neighbourhood of DB's systemic numbers.)

  • That is a lot better than the 62% on-time statistic for Deutsche Bahn inter-city travel.

    I deal with late trains around two-thirds of my trips. Sometimes up to two hours of delay.

  • If true it would beat the 92.5% statistic for Swiss trains that was mentioned in the article.

My memory from a few months of commuting on LIRR is that on-time is within ten minutes of schedule.

CURRENT TIME: 9:45

NEXT TRAIN: 9:42

STATUS: ON TIME

  • In your example a train arriving at 9:45 scheduled to arrive at 9:42 would still be on-time in Switzerland.

BART's on-time definition is "within 5 minutes of scheduled arrival at final station". I recall in NYC's Metro that trains would frequently "go express" and start skipping stops at arbitrary moments. Do these systems do this as well and have that definition? I think I now understand why they do this stuff: they're trying to juice metrics because skipping stations speeds up the train a lot.

The correct comparison in this case is probably Amtrak, which has much lower on time rates.