Does Amtrak misrepresent lateness

[bergw]

I'm on the 42 today and it was about 45 minutes late leaving Pittsburgh, as Amtrak stated

"Train 42 is delayed departing Pittsburgh (PGH) due to mandatory crew delays"

I assume they mean mandatory rest, but also how does that happen at the departure station of a once a day route?

Every stop so far has been 40-45 minutes late as you'd expect, but their tracker just keeps showing it as arriving in time. I know it can get moving in the corridor but that much? I also noticed this on the meteor last year where the website would show the stops as on time or even a little early but we'd pull in 30 minutes late. Just seems almost deceptive, or like based on hope or some bs. The closer you go to arrival the more accurate it was but still wildly off until almost the minute of arrival

I do understand there are a lot of factors at play here, but it seems to me the default position should be to update all the stops to a 45 minutes delay and then update them to more accurate times as the next stop is reached. If you made up time sweet, update all he stops to the new delay interval

 

[Tom C]

Some of that is because of padding in the schedule. With the padding if 42 left Pittsburgh 45 down but had zero issues and got through it’s station stops promptly the rest of the way it would indeed arrive at the time stated.

 

[AmtrakBlue]

I'm on the 42 today and it was about 45 minutes late leaving Pittsburgh, as Amtrak stated

"Train 42 is delayed departing Pittsburgh (PGH) due to mandatory crew delays"

I assume they mean mandatory rest, but also how does that happen at the departure station of a once a day route?

Every stop so far has been 40-45 minutes late as you'd expect, but their tracker just keeps showing it as arriving in time. I know it can get moving in the corridor but that much? I also noticed this on the meteor last year where the website would show the stops as on time or even a little early but we'd pull in 30 minutes late. Just seems almost deceptive, or like based on hope or some bs. The closer you go to arrival the more accurate it was but still wildly off until almost the minute of arrival

I do understand there are a lot of factors at play here, but it seems to me the default position should be to update all the stops to a 45 minutes delay and then update them to more accurate times as the next stop is reached. If you made up time sweet, update all he stops to the new delay interval

It's a computer algorithm that estimates times for stops several stations up the line from the last update.

 

[bergw]

Seems likes it a pretty bad algorithm in that case. Especially given the first stop after Pittsburgh was listed as on time even as the train had departed pit late. Definitely erring on the side of them making up time. I get it, it just feels kinda manipulative.

I'm in no way upset by this mind you, late trains in America are part of the process. Just kinda wish I had a better idea if I shoulda rebooked on an earlier keystone to get to PHL. Too late now anyway

Thanks for the responses, appreciate the info

 

[Trollopian]

Some of that is because of padding in the schedule. With the padding if 42 left Pittsburgh 45 down but had zero issues and got through it’s station stops promptly the rest of the way it would indeed arrive at the time stated.

This. Here's an example of padding on just one smallish and wee-hours segment of the Floridian. Snipped from the timetable at https://www.railpassengers.org/site/assets/files/20928/floridian.pdf:



Do a little math and you can see that, eastbound, there's 20 minutes of "padding" between Alliance and Pittsburgh (i.e., the schedule allows 2 hours rather than 1 hour 40 minutes). Westbound, there's an astonishing 33 minutes of padding between Connellsville and Pittsburgh, which explains why the Cap (I stubbornly continue to call it that) often chugs into Pittsburgh a half-hour early; and 15 minutes of padding between Elyria and Toledo.

This is just one segment of a much longer route, but it shows why you'd be doing a disservice to passengers by automatically adding (say) 45 minutes to every stop further down the line if there's a delay at one. It depends on the reason.

(As to where Amtrak pads schedules, and why, I'll leave that to the bigger brains. But I hypothesize that there's almost always padding somewhere in stretches between major cities.)

 

[ThePun]

See my Trip Report thread.

Not moved in six hours on the Empire Builder.

 

[zephyr17]

(As to where Amtrak pads schedules, and why, I'll leave that to the bigger brains. But I hypothesize that there's almost always padding somewhere in stretches between major cities.)

On the Builder there is pad (officially called "recovery time") westbound at St. Paul, Minot, Shelby, Whitefish, Spokane, Wenatchee (only about 5 minutes) and Seattle.

 

[jis]

The recovery times tend to be near division points since the handover time at such points are officially recorded and reported.

 

[Acela150]

I'm on the 42 today and it was about 45 minutes late leaving Pittsburgh, as Amtrak stated

"Train 42 is delayed departing Pittsburgh (PGH) due to mandatory crew delays"

I assume they mean mandatory rest, but also how does that happen at the departure station of a once a day route?

Correct. Train 43 was delayed due to a Freight Train Tresspasser strike. The train arrived hours late and the crew which is based out of Harrisburg had to turn the train and then get their mandated rest.

The only crew base in Pittsburgh is for Engineers to Cumberland. So there is no one to recrew the train for an ontime departure in this case.

 

[acelafan]

I'm on the 42 today and it was about 45 minutes late leaving Pittsburgh, as Amtrak stated

"Train 42 is delayed departing Pittsburgh (PGH) due to mandatory crew delays"

I assume they mean mandatory rest, but also how does that happen at the departure station of a once a day route?

Every stop so far has been 40-45 minutes late as you'd expect, but their tracker just keeps showing it as arriving in time. I know it can get moving in the corridor but that much? I also noticed this on the meteor last year where the website would show the stops as on time or even a little early but we'd pull in 30 minutes late. Just seems almost deceptive, or like based on hope or some bs. The closer you go to arrival the more accurate it was but still wildly off until almost the minute of arrival

I do understand there are a lot of factors at play here, but it seems to me the default position should be to update all the stops to a 45 minutes delay and then update them to more accurate times as the next stop is reached. If you made up time sweet, update all he stops to the new delay interval

I agree that the algorithm for late trains making up time is almost always too optimistic. But it can happen, especially on the longer routes with lots of recovery time, for example the #7 Empire Builder that departed Seattle on 12/24.

I don't think Amtrak has the optimistic times set to be misleading on purpose, but it can set unrealistic expectations for the passengers which is not good.

https://asm.transitdocs.com/train/2024/12/24/A/7

 

[blueman271]

I would say yes, Amtrak does misrepresent delays. I’ll give an example. At the end of October I rode Acela 2258 from DC to New Haven and were scheduled to depart at 415pm. I was waiting in the lounge and as 415pm came and went the train was still listed as on time which it obviously wasn’t. Furthermore, the lounge attendant insisted the train would leave on-time, even as that became impossible. Granted this is only one example but I’m sure others that have ridden Amtrak extensively have similar stories.

To be fair, Amtrak isn’t the only transportation company that does this. AA is famous for their rolling delays, telling you every five minutes that your flight is 5 minutes late while you can see mechanics welding an engine on. The welding part is hyperbole but the rolling delays are not unfortunately. JetBlue is also notorious for misrepresenting delays. They will swear your flight will leave on time while the incoming aircraft has just landed and departure time is 10 minutes away. So Amtrak is in no way unique when it comes to misrepresenting delays.

 

[west point]

It appears to this poster that when a train leaves a station late that the best recorded enroute times are used for the next station arrivals. Example: Train A from X - Y scheduled for 40 minutes. However sometimes it only takes 30 minutes. So that projection will be used. Times continue on down the line until an on time projection occurs. This is especially evident on the crescent leaving Anniston going to ATL. Today left ATN :03 late expected ATL :30 early.

EDIT did arrive :27 early

 

[TrackWalker]

I believe Amtrak refers to it as artistic license.

 

[railiner]

The problem with projecting current lateness of a train to future station stops, is that some passengers may count on that lateness in going to catch their train, and if it makes up time, it will either have to hold for that projected time, or risk missing some passengers .

 

[MARC Rider]

This morning, 111 left New York 12 minutes behind schedule, but got into Wilmington 4 minutes early. Here's hoping they keep up the good work and get to Baltimore on time.

 

[AmtrakBlue]

This morning, 111 left New York 12 minutes behind schedule, but got into Wilmington 4 minutes early. Here's hoping they keep up the good work and get to Baltimore on time.

I'm sure that's because there's a lot less traffic on the corridor today.