Ride BananaGet there under your own steam

Riding the City

Learning the Back Streets Beats Learning the Main Roads

The road that gets you there is the one you already know from a car. The one you will enjoy is usually a street over, and nobody signposts it.

A cyclist rides alone on a sunlit empty street in Lazio, Italy, casting a long shadow.
Photograph by Cameron Knott via Pexels
Editorial note. Independent reporting and analysis. Nothing here is sponsored or paid for. How we work.

This is less a set of instructions about route knowledge than an argument, and it is worth saying so at the start.

The argument in brief

  • Main roads are direct because they were built for through traffic.
  • Parallel streets often cover the same distance with fewer stops.
  • Route knowledge accumulates faster if you deliberately vary it.

Why your mental map is a driver's map

Most people learn a city from behind a windscreen or from public transport, which teaches the arterial network and nothing else. That network exists to move through traffic efficiently, which is why it is direct, wide and unpleasant to ride along.

The residential streets between those arteries usually connect in ways that are invisible unless you have ridden them. A rider using a driver's mental map therefore takes the worst available route while believing it is the only one. Replacing that map is the single largest improvement most new commuters make, and it takes a few weeks rather than a few years.

What a good parallel route looks like

It runs roughly alongside the main road, typically one or two streets away, and crosses the arteries at controlled points. It has low traffic volumes, which matters more than low speed limits, because volume determines how often you interact with anything.

The bit that actually matters: it has few enough signals that you keep your momentum, which is what makes it feel faster than it measures. It usually involves a couple of awkward moments where the parallel street does not continue and you have to cross something. Those awkward moments are exactly why routing software does not find it and why local riders all know it.

Finding them deliberately

Riding a different variation of your commute once a week builds the map far faster than riding the same route daily for a year. Going home a different way is lower risk than experimenting on the morning ride, since arriving late matters less. Satellite imagery reveals cut-throughs, park paths and alley connections that street maps show poorly or not at all.

With a loaded rack, local cycling groups and colleagues who ride are the fastest source of all, and their knowledge is rarely written down anywhere. Heat maps built from riders' recorded activity show which streets riders actually use, which is different from which are designated.

The trade you are making

Back street routes are typically slightly longer in distance and frequently shorter in time because of the reduced stopping. They are quieter, which lowers the mental load and makes the ride restful rather than a task to be survived. They are often worse lit and less overlooked, which matters differently at eight in the morning and at eleven at night.

Surface quality on residential streets can be poorer, since maintenance budgets follow traffic volumes.

Different riders weigh those differently, and the right answer changes with the season and the time of day.

Building more than one route

A single known route is fragile, because roadworks, a closure or an incident removes your only option. Two or three variations mean a blocked street is an inconvenience rather than an improvisation in traffic.

The bit that actually matters: having a wet weather route, which avoids exposed sections and poor surfaces, is worth the effort in a rainy climate. A night route that stays on lit and overlooked streets is worth having even if you rarely use it. Riders who have several options describe their commute as reliable, which is almost entirely a function of having alternatives.

Where routing software helps and where it does not

Routing tools are good at finding designated infrastructure and at showing you a plausible line through an unfamiliar city. They are poor at judging how stressful a road feels, which is the variable riders actually care about. They also change their recommendations as data updates, which can send you somewhere unpleasant without warning.

Using them to generate candidates, and then riding the candidates to judge them, is the combination that works. After a few months the tool becomes a fallback for unfamiliar areas rather than the source of your daily route.

The takeaway

Go home a different way once a week, and within a month you will have a map no navigation app can give you.

Fix the small rattle. Every large repair began as one.

Questions readers ask

How do I find quiet cycling routes in a new city?

Satellite imagery for cut-throughs and park paths, rider heat maps for the streets people actually use, and local cyclists for the awkward connections. Then ride the candidates at a quiet time to judge them yourself.

Is a longer quiet route actually slower?

Often not, because urban journey time is dominated by stopping rather than by distance. A route with fewer signals and fewer conflicts frequently arrives at the same time and costs far less effort.

Riding the Cityroutesnavigationlocal knowledge
More in Riding the City
Joel Pereira
Workshop writer, Ride Banana

Joel writes the repair pieces and believes most bikes are one cable adjustment from feeling new.

Also by Joel Pereira