The game is simple: you see a station name, you click where you think it is on a blank map. No tube lines, no labels. The players are self-selecting and likely more transport-aware than average, so this is not a representative sample of London. But with nearly 135,000 data points across all 272 stations, the patterns are consistent enough to be worth examining. Before the data, see how much you already know.
Test yourself
Most accurately located station?
Oxford Circus. 300 m average error - the most accurately placed station in the game. Players know that intersection by heart.
Least accurately located station?
Elm Park. 13.7 km average error. District line, Zone 6. Most players click somewhere near Stratford and call it a day.
Which pair gets swapped the most?
Euston Square and Warren Street. 1,234 swaps - the most confused pair in the dataset. They are 250 m apart on Euston Road.
Highest miss rate?
Alperton. 82.5% strike rate. Piccadilly line, Zone 4. Almost nobody gets within 2 km.
Chesham is in which zone?
Zone 9. It is in Buckinghamshire - the furthest station from central London on the entire network. Most players guess Zone 3-4.
Most confidently wrong station?
Finsbury Park. 282 fast, confident guesses that land 3+ km away. Players click near the Finsbury area by Old Street. The station is in Haringey.
What confidence looks like
Oxford Circus: 1,373 guesses, 300 m average error. Players click within a short walk of the actual entrance. Every red dot below is a guess.
Waterloo (311 m), Leicester Square (334 m), Covent Garden (347 m), Piccadilly Circus (359 m) - all within a few streets. These are places people navigate by foot, not by tube map.
What guesswork looks like
Elm Park. District line, Zone 6. Average guess: nearly 14 km away. Over three quarters of guesses are complete misses.
Chorleywood (12.3 km avg error), Eastcote (10.7 km), Moor Park (10.6 km) - outer stations the tube map compresses into a tidy diagram. Chorleywood is in Hertfordshire. Most players place it somewhere in north London.
The stations people swap
Some stations get swapped so consistently the data looks like a mirror. These are the pairs where players click on one station when asked about the other.
Euston Square and Warren Street - 1,234 swaps
The most confused pair in the entire dataset. 863 Warren Street guesses land on Euston Square. 371 go the other way. They are 250 m apart on Euston Road and players treat them as interchangeable.
Leicester Square and Piccadilly Circus - 266 swaps
Two of the most famous stations in London, 400 m apart. 159 Piccadilly Circus guesses land on Leicester Square. 107 go the other way. Players know the West End - they just cannot separate the two.
The full leaderboard
| Pair | Total swaps |
|---|---|
| Euston Square / Warren Street | 1,234 |
| Bank / Cannon Street | 1,092 |
| Bayswater / Queensway | 720 |
| Great Portland St / Regent's Park | 718 |
| Baker Street / Regent's Park | 380 |
| Euston / Euston Square | 345 |
| Euston / Warren Street | 328 |
| Edgware Road (Bak) / Edgware Road (Circle) | 312 |
| Monument / Bank | 272 |
| Leicester Sq / Piccadilly Circus | 266 |
| Covent Garden / Leicester Square | 257 |
| Temple / Blackfriars | 245 |
| Mansion House / Cannon Street | 247 |
| Embankment / Westminster | 240 |
The leaderboard suggests a few recurring triggers for confusion. Shared names account for many of the top pairs (Euston/Euston Square, the two Edgware Roads). Shared fame plays a role too - Leicester Square and Piccadilly Circus are both landmarks, but players blur their exact positions. And the Thames corridor stations from Temple to Bank consistently get shuffled along the river.
Confidently wrong
Some stations produce fast guesses - under 5 seconds - with terrible accuracy. Players are not uncertain. They are certain, and they are wrong.
Names that reference a place trigger fast, confident clicks in roughly the right area. But "roughly right" and "correct" can be separated by several kilometres.
Mobile vs desktop
Desktop players average 1,741 m error. Mobile: 2,038 m. Tablet: 2,120 m - a 17% penalty for smaller screens. Whether that reflects reduced precision from touch input or a different player demographic is unclear from the data alone.
The zone cliff
Average error increases nearly ninefold from Zone 1 to Zone 3-4. Players know the centre. Beyond that, guesswork takes over.
Strike rates tell a similar story, with one wrinkle. Zone 3-4 has the highest miss rate at 58.3%, but Zone 5+ actually drops back to 50.9%. Why? Likely because some well-known outer stations - Brixton, Heathrow, Epping, Cockfosters - are familiar enough to anchor guesses in the right area, even if the surrounding stations are unknown. The outer zones are not uniformly blank. They are patchy: a few landmarks in a sea of guesswork.
What predicts accuracy?
Zone matters, but it is not the only factor. Major interchanges - stations where multiple lines meet - average 602 m error, compared to 2,141 m for other stations. Being a transport hub makes a station a landmark, regardless of zone.
The overall distribution is heavily skewed. 56.5% of all guesses land within 200 m. Another 8% land between 500 m and 1 km. But 5.1% of guesses are over 10 km off, and those outliers drag the average up to 1,919 m despite a median of just 98 m. Most players are reasonably good at most stations. A small number of outer-zone stations generate enormous errors that distort the headline figure.
Readers who want to test themselves can try Mind the Map here.
- All data from Mind the Map on tubenotifications.co.uk. 134,895 guesses across 6,288 games played between 27 February and 2 April 2026.
- 272 London Underground stations. All station positions from TfL open data.
- "Strike" defined as a guess landing more than 2 km from the actual station location.
- "Confident wrong" defined as guesses submitted in under 5 seconds with an error exceeding 1 km.
- "Confusion pair" proximity defined as a guess landing within 400 m of the other station in the pair.
- Device type inferred from screen width at time of play (mobile: under 768 px, tablet: 768-1024 px, desktop: over 1024 px).
Tube Alerter publishes data-led analysis on London transport, alongside disruption alerts for commuters on the Underground, Overground, DLR and Elizabeth line.
How Deep Is the Tube?
FOI data on every line, the deepest platforms, and what London's depths reveal about its geology.
Why Is It So Hot on the Tube?
Temperature data, tunnel physics, and what is actually being done about it.
The Tube Map: What It Hides
How to read it, walking shortcuts it conceals, and every version TfL publishes.
Play Mind the Map
Pin stations on a blank map. No tube lines, no labels. Just you vs London.
This article is regularly updated as new data comes in from Mind the Map. Send us a tip if you spot something we should include.