Subway maps distort the reality on the ground for all kinds of reasons. What happens when we make decisions based on them?
London’s city center takes up about two percent of the city. On the Tube map, it looks four times as big.
So public transit maps are distorted, quite on purpose. All of them enlarge city centers. Many use a fixed distance between stations out in the boonies, even if, in reality, they’re spaced wildly differently. Curvy lines are made straight. Transfers are coded with dots, lines, and everything in between.
According to Zhan Guo, an assistant professor of urban planning and transportation policy at NYU Wagner, certain cities allow for more flight of fancy than others. San Francisco and New York have a lot of geographic markers, so passengers will only accept so much map distortion. New York’s grid system further discourages excessive futzing. In Chicago, the line is elevated, which leaves even less leeway. But in a place like London, with twisty streets, few geographical markers other than the Thames, and an underground system, you can pull a lot more over on people.
Not that transit map distortions are malicious. A lot of these simplifications are necessary for good design. If people could keep a geographically accurate map in their heads, we might use them more—but, as we’ve known for decades, we can only remember so many discrete bits of information at once (between five and nine, as per George Miller’s seminal 1956 paper, "The Magical Number Seven, Plus or Minus Two").
Schematic maps, by necessity, balance between detail and readability. “If you try to get everything in, it becomes meaningless, usually,” says Lance Wyman, who designed D.C.’s Metro map more than 30 years ago and is redesigning it to incorporate the new Silver Line. When Wyman, who also designed the Mexico City Metro map (as well as the logo for the 1968 Mexico Olympic Games), sits down to convert a wily transit system into something people can understand, he aims not just to make it readable, but to enable each viewer to describe it to a friend. The stations need to be sequenced correctly, and their intersections must be clear. Color-coding helps—but be mindful of the colorblind. Keep the map legible—which includes simple station names. (D.C.’s had gotten rather long, which, Wyman says, "is always a struggle," though WMATA did agree to shorten more than half a dozen station names for an updated map unveiled this summer.) Finally, give the map some character: "A big part of my work has always been to integrate the essence of the community," Wyman says. "From my own experience, when it works, everyone kind of sees that it works."
So, to play to our cognitive capacities and honor all these principles, there’s no choice but to distort maps. The thing is, we humans are always using heuristics to make decisions, public transit choices included. Everyone might know a map "works," but we don’t necessarily know how it's working on us.
When we plan out a route, we consider tons of factors: distance, travel time, number of transfers, walking time, waiting time, direction, reliability, level of crowding, operating hours, cost, seat availability, aesthetics, and more. So what happens when we make decisions based on inaccurate maps?
• • • • •
Zhan Guo has only one transit map displayed in his office, and it's lying on its back in his bookcase. It's surprising, perhaps, for a researcher who studies how we plot our travel routes. But it's nowhere near as surprising as his main finding: we trust transit maps even more than our own experience.
In an extensive study of the London Tube, published last year, Guo found that passengers tend to take routes that look shorter, even if they’re actually longer. This is no small difference: the map depiction is about two times more influential than lived experience. Passengers who knew the system well were less likely to be duped than first-time riders, but even they still regularly fell under the Tube map’s spell.
It makes intuitive that sense that we prefer transfers that look convenient, but convenience gets tricky when complicated transfers appear simpler than they actually are. In the London Underground, the Victoria and Oxford Circus stations are depicted as dots, while Baker St. and Bank/Monument are each two dots connected with lines. The first two involve relatively complex transfers, but by modeling the various alternatives a passenger on a given route must consider before making a transfer choice, Guo calculated that their codification on the map might attract 960 and 516 additional passengers, respectively, every workday. As for Baker St. and Bank/Monument? That dot-line-dot icon, which looks like it involves an underground trek, probably deters 216 and 147 respective passengers each day.
We scorn indirect routes, too. In a 2008 study of Santiago’s Metro map, a group of researchers in Chile found that we are not likely to pick a route with portions that veer significantly from our destination. And University of Florida professor Hartwig Hochmair found, in a study of road maps, that we prefer to take a straight road and then go through a squiggly bit at the end of car trip than to squiggle first and drive straight later. We are picky travelers.
Picky, but often inaccurate. Based on Guo’s findings, a New Yorker might think nothing of transferring at the map-dot that is Times Square—which involves winding up, down and around through an enormous station—but avoid at all costs the long line connecting the 14th St. 1/2/3 with the L. It’s a long underground walk, to be sure, but worse than Times Square? Unlikely.
Or take the enlarged-center syndrome. In Boston, the distance from Park St. to Downtown Crossing looks at least a mile long on the MBTA map (at left) but in reality, it’s easily walkable. If we blindly rely on the map, we’ll probably waste time.
Individual decisions bloom into system-wide effects. Center enlargement—a necessity if you want all the names and transfers to be legible—has a dual effect on mass behavior, Guo surmises.
"Those last one or two stops are generally in an urban center, so they tend to congest the system," he says. If passengers got out at Park St. and walked instead of transferring for one stop, it would clear out a ton of human traffic underground. The flip side: the apparently short distance from the outskirts to downtown areas could make city hubs seem much more accessible to many people who live far away.
Which leads to Guo’s big question: "Can we change the map in order to change people’s behavior?" If we believe maps over our own knowledge, and we do, the answer is likely yes. In a new study of D.C.’s Metrorail system, Guo is measuring the difference. He’s collaborated with Wyman to produce three variations on the Metro map, all of which increase the apparent length of the Blue and Orange Lines at the point that they cross the Potomac River. One increases to the west, one to the north, and one in both directions. Comparing reactions to these maps to the current one, and to a real geographic map, will help Guo better understand how both route length and directionality can factor into passengers' decisions.
Can Guo and Wyman encourage people, through design alone, to transfer to the Yellow Line over the river instead of staying on the Blue? This isn’t just a matter of intellectual interest: The Rosslyn tunnel is overcrowded, so rerouting some human traffic would make a difference. "Even if you can switch one or two percent of passengers from the Blue to the Yellow Line, that’s a big success, because the cost is zero," Guo says.
That is, altering infrastructure is expensive, and since many transit agencies, including WMATA, face big budget shortfalls, it’s often difficult, if not impossible. But changing a map—making this crowded line look longer and less convenient, or replacing that complex-looking transfer with a dot—could change usage patterns practically for free. Mapmakers could nudge us to not only use less crowded lines, but also get out and walk, transfer at less trafficked stations, or even use alternate transit systems.
Unless passengers catch on. Though if Guo's work has any merit at all, it's not likely we will.