Help
FAQ
Why is the fastest route different from the one I know?
Chronométro follows a static route map and timing model. Ride times come from saved timetable data; waits and transfers are estimates. Direction and branches matter, so the return trip may have a different answer.
The game does not use live service patterns, disruptions, or real-time fast and semi-fast service decisions.
London: branch services currently share the line's estimated wait, rather than using a separate wait for every destination. Ordinary Underground changes use a fixed two-minute estimate; linked records within a larger station hub use three minutes. Real interchanges vary considerably, so these assumptions can change which route appears fastest.
How does it find the fastest route?
The game treats stations and connections as a weighted graph, then uses Dijkstra's algorithm to find the route with the lowest total time. This Veritasium video explains the basic idea.
Which lines are included?
| City | Included | Excluded |
|---|---|---|
| Paris | Metro, RER A–E, and trams T1–T14. | Buses, ORLYVAL, and CDG VAL. |
| London | All 11 Underground lines, Elizabeth line, and DLR. | London Overground, trams, buses, and national rail. |
| Chicago | Red, Blue, Brown, Green, Orange, Pink, Purple, and Yellow ‘L’ lines. | CTA buses and Metra. |
| Washington, DC | Red, Blue, Orange, Silver, Yellow, and Green Metrorail lines. | Buses, MARC, VRE, Amtrak, and other non-Metrorail services. |
| Boston | Red, Orange, Blue, Green B/C/D/E, and Mattapan. | Buses (including Silver Line), replacement shuttles, Commuter Rail, ferries, CapeFLYER, Amtrak, and non-MBTA services. |
| Berlin | U1–U9 and all 16 regular S-Bahn lines, including S15. | Temporary routes, diversions, discontinued S45, replacement buses, trams, ferries, regional and long-distance rail. |
| Madrid | Metro lines 1–12 and Ramal R, including MetroSur and outer sections. | Metro Ligero, Cercanías, buses, replacement services, and unopened extensions. |
| New York City | Subway routes 1–7, A–G, J/Z, L, M, N–W, and the 42 St, Franklin Av, and Rockaway shuttles. | Staten Island Railway, PATH, LIRR, Metro-North, buses, and ferries. |
Puzzles require a transfer and can use any of the included lines. Some start and end stations have geographic restrictions; these do not limit where your route can travel.
Which stations can be the start or finish?
Paris: all included Métro stations are eligible, including suburban stations. RER-only endpoints must be within the actual Paris municipal boundary, not the wider metropolitan area. Trams can be used along a route but do not supply endpoints on their own. The RER boundary is the Paris municipal polygon (commune 75056) from the French government’s administrative geography API, retrieved 4 September 2026. Source: data.gouv.fr, Contours administratifs (including IGN Admin Express), published under ODbL; the saved boundary is distributed unchanged under that licence.
London: all included Underground and DLR stations are eligible. Elizabeth-line-only endpoints must be within the actual Greater London boundary (the 32 boroughs and the City of London), not just the central City of London district. The Greater London Authority boundary was retrieved 4 September 2026 under the Open Government Licence v2 and converted from British National Grid to WGS84 without simplification. Contains National Statistics data © Crown copyright and database right 2015. Contains Ordnance Survey data © Crown copyright and database right 2015.
Chicago: every included ‘L’ parent station is eligible, with no geographic endpoint restriction.
Washington: every included Metrorail parent station is eligible, with no geographic endpoint restriction.
Boston: every included MBTA parent station is eligible, with no geographic endpoint restriction.
Berlin: all included U-Bahn stations are eligible. S-Bahn-only endpoints must be within the actual Berlin state boundary; Brandenburg stations remain available along routes. S-Bahn-only eligibility uses the ALKIS state boundary described in About.
Madrid: every included Metro station is eligible, including stations outside Madrid municipality, with no geographic endpoint restriction.
New York City: every included subway station complex is eligible, with no geographic endpoint restriction. Constituent platforms use the MTA's official complex definitions.
Geographic boundaries are checked against each station’s saved coordinates and apply only to start and finish eligibility, not to stations or services available along a route. Stations served by an unrestricted mode remain eligible. Other puzzle-selection rules still apply.
The game models normal service rather than temporary closures or diversions. Where available, official historical data restores missing stations and sections. Details of the sources and restorations are in About.
When is walking an option?
The source transit data lists walking connections between some stations. The game offers a walk only when one of those links exists, or when separate records belong to the same station hub.
Paris: walking links and equivalent station records follow the saved GTFS data and the game's configured transfer rules.
London: linked platforms count as one station at the start and finish of a puzzle. This is deliberate: you can begin on the appropriate platform within the named station and finish on arrival at that station. Changing between those platforms midway through a route still uses the modeled interchange time.
Chicago: CTA platform/direction stop records are collapsed only through their GTFS parent_station. No proximity-based walking transfers are invented.
Washington: Farragut Crossing is the only modeled walk between distinct stations, as a five-minute walk between Farragut North and Farragut West.
Boston: the Winter Street Concourse is the only modeled walk between distinct stations, as a five-minute walk between Park Street and Downtown Crossing.
Berlin: four connections between separate stations use a five-minute walk: Spandau–Rathaus Spandau, Charlottenburg–Wilmersdorfer Straße, Messe Nord/ZOB–Kaiserdamm, and Yorckstraße (Großgörschenstraße)–Yorckstraße. Changes within the same station, including U-Bahn/S-Bahn interchanges, use a three-minute estimate.
Madrid: platforms collapse only through explicit CRTM parent stations. Required changes within lines 7, 9, and 10 use the modeled interchange; no proximity links are invented.
New York City: platforms in an official MTA station complex count as the same puzzle station. Mid-route changes use positive GTFS transfer times where supplied and a three-minute fallback otherwise; no proximity links are invented.
How are waiting times estimated?
Paris: the game looks at weekday departures from 7:00–10:00 AM, finds the median gap between services, and uses half of that as the expected wait. When there is not enough data for a direction, it falls back to the line and then the transport mode.
London: Underground waits use the same half-median-gap method with saved weekday morning timetable departures. Elizabeth uses a fixed expected wait of 2½ minutes because direction-level headways were not available in the saved source snapshot. When an entire ride can be made on either Circle or H&C, their frequencies are combined to estimate the wait for the first suitable train. The benefit is not applied after their routes diverge.
Chicago: waits use half the median gap between scheduled weekday departures from 7:00–10:00 AM. If a direction has too little data, the model falls back to the route and then a three-minute ‘L’ wait.
Washington: waits use the same half-median-gap method with scheduled weekday Metrorail departures from 7:00–10:00 AM. Fallbacks proceed from direction to line and then to a four-minute Metrorail wait.
Boston: waits use half the median gap between scheduled departures from 7:00–10:00 AM on the representative weekday. They are calculated per complete service pattern, so a required Green branch does not receive the combined trunk frequency. Fallbacks proceed from direction to route and then to four minutes for rapid transit or five minutes for light rail.
Berlin: U-Bahn waits use half the median gap between scheduled weekday morning departures. S-Bahn waits use half the published regular interval. Independent services combine their frequencies only when they cover the entire selected ride, so branches and Ringbahn directions keep their own waits.
Madrid: waits use representative weekday frequency blocks where available, with a 3½-minute ordinary fallback. Required changes at Tres Olivos, Estadio Metropolitano, and Puerta de Arganda add a three-minute interchange and a fresh wait.
New York City: waits use half the median gap between scheduled weekday departures from 7:00–10:00 AM. If a direction has too little data, the model falls back to the route and then a four-minute Subway wait. Express, local, branch, and shuttle patterns remain distinct; 6/6 Express, 7/7 Express, F/F Express, and J/Z frequencies combine only when either service can complete the entire selected ride.
The model adds a wait for the first service you board. It does not add time for entering the departure station or reaching your preferred starting platform.
Does the site save anything?
Your route and scores are saved automatically in this browser using local storage, separately for each city and puzzle date. You can visit the rules, reload, or close the tab and return to your game. There are no accounts or cookies, and saved progress is not sent to a server or shared between devices. Clearing site data removes it. If browser storage is unavailable, you can still play, but progress will not be saved.