A marathon route is 42.195 kilometers of city. Run any direction from the starting line and you'll cover roughly the same area. But marathon organizers don't choose routes at random. They choose them to showcase.
Renninger (arXiv:2603.22183) analyzed 311 marathon routes across five continents, comparing the density of landmarks along the actual route against the density along alternative routes sharing identical start and finish points. The numbers: museums are 15.7 times denser near the marathon route than elsewhere in the city. Luxury brands are approximately 8.5 times more concentrated near the route than the city median.
The race course is a curatorial act. The same start and finish could be connected by routes through industrial districts, residential suburbs, or institutional corridors. The marathon organizer chooses the route that presents the city as its tourism board would. Monuments, cultural institutions, and commercial landmarks are overrepresented not because the city is uniformly rich with them, but because the route seeks them out.
This matters because marathon routes are increasingly used as proxies for urban analysis — tracking air quality, measuring pedestrian infrastructure, modeling crowd behavior. If the route systematically oversamples the city's showcase neighborhoods, any measurement taken along it inherits that selection bias. Air quality along the marathon route is air quality in the nicest parts of town.
The through-claim: marathon routes are selective portraits of cities, not representative samples. The 15.7x museum density isn't a statistical coincidence — it's evidence that the route is designed to present a city's best face, and any research using the route as a proxy for the city inherits this curation as bias.