The Transit Map That Lies (And Is Better For It)
Pull up the map of any subway on earth - London, New York, Tokyo, Paris - and you are looking at a lie. The distances are wrong. The angles are wrong. Whole neighbourhoods are compressed or stretched, the river is a tidy zigzag, and stations that are miles apart sit a neat centimetre from each other. It is one of the most successful lies in the history of design, and it is a masterclass in the single most underrated concept-phase skill there is: deciding what to throw away.
The lie has an author. In 1931 an out-of-work engineering draughtsman named Harry Beck sketched, on his own time, a new diagram for the London Underground. Every previous Tube map had done the obvious, honest thing: it drew the lines where they actually were, over a real street map, which meant the crowded centre was an illegible knot and the far-flung suburbs sprawled off the edge. Beck's insight was heretical. A passenger underground, he reasoned, cannot see the city and does not care where they physically are. They need to know only three things: which line they are on, which stations come in what order, and where the lines cross. Everything else - geography itself - is noise.
So he deleted it. He straightened every line to verticals, horizontals and 45-degree diagonals, like the circuit diagrams he drew at his day job. He made the packed central stations bigger and the distant ones closer, evening out the spacing until the whole network was legible at a glance. He kept the topology - the connections and the order - and threw away the geometry. The result was not a map at all; it was a schematic. The transit authority thought it too radical and sat on it. When they finally printed it in 1933 it was an instant, permanent success, and every transit map in the world has been a variation on Beck's decision ever since.
Here is the concept-phase lesson, and it is uncomfortable. Beck's genius was not what he drew; it was what he chose not to. The hard, high-leverage decision was subtractive: identifying that geographic accuracy - the thing everyone assumed a map must have - was precisely the thing making the map worse. This is information design in its purest form, and it generalises to all of it: the core of data visualisation is not adding the right marks, it is removing the wrong ones. A design that tries to represent everything represents nothing usefully. Deciding what to omit IS the design.
The proof that this is a genuine, non-obvious decision - and not just "simplify" - is that it can be got wrong. In 1972 the great Massimo Vignelli applied the same logic to the New York City Subway map, producing a beautiful, pure diagram that made Beck's move even more aggressively. New Yorkers hated it. The reason is instructive: unlike London's riders, New York's spend a lot of time above ground, connecting the subway to the street grid, so for them the geographic lie broke something they actually needed. The city replaced it in 1979. Same abstraction, different context, opposite outcome - which tells you the decision about what to discard is not a style, it is a judgement about who the reader is and what they can afford to lose.
That is the whole thing. Every act of cartography, every dashboard, every diagram, every interface is a series of decisions about what to leave out, made early, before a single line is drawn - because once you have decided to keep geography, no amount of downstream polish will make the centre of the map legible. Abstraction is not a rendering step; it is the concept-phase act of choosing the one true structure and sacrificing everything that obscures it.
At Depix, this is the counterintuitive half of the thesis. We talk constantly about the early decision that carries everything - and the transit map is the reminder that the decision is very often a deletion. The most valuable thing a designer does is not add the feature; it is having the nerve, at the start, to throw away the thing everyone assumes is essential. Beck threw away the map and gave us something better. The next time a design feels cluttered, don't ask what to add. Ask what a braver concept phase would have chosen to leave out.
Sources:
