The old headline made it sound like Google had measured every train on Earth and produced the definitive ranking of the ten most crowded lines.
That is not what happened. Google analyzed optional crowdedness reports submitted by Maps users between October 2018 and June 2019 during peak commuting hours.
What Google actually measured
Google said its ranking was based on aggregated and anonymized reports from Maps users, focusing on peak commuting hours from 6 a.m. to 10 a.m. Buenos Aires and São Paulo dominated the top ten, with three lines each, while New York’s L train was the only U.S. line to appear.
To predict crowdedness before you board
Past riders contribute
Google introduced crowdedness predictions based partly on previous rider reports.
The goal is practical
Users can decide whether to board a packed vehicle or wait for another service.
The feature expanded
In 2021 Google said crowdedness predictions were expanding across more than 10,000 transit agencies in 100 countries.
Some systems provide richer information
Google has also experimented with live or car-level crowding information where transit agencies provide the data.
Cities and systems changed
Ridership patterns changed dramatically after 2019
Pandemic-era disruption alone makes pre-2020 commuting behavior a poor current benchmark.
Transit agencies changed service
Frequency, rolling stock and routes affect crowding.
The user base changed
Optional reports reflect the people using Google Maps and choosing to submit feedback.
Crowding changes by hour and direction
One line can be brutal at 8 a.m. and comfortable at noon.
A permanent global ranking needs a common methodology
Passenger counts, capacity and load factors need comparable data across systems.
A lesson in reading data headlines
When a list says “according to Google Maps,” ask what was counted, when, where and by whom. The methodology often changes the meaning more than the ranking does.
The 2019 Google list was real. Calling it the world’s permanent crowding table was the mistake.
It captured a moment in user-reported transit experience.
That is valuable data — as long as the timestamp stays attached.