FAQ

What is this?

Bomb Hills reads public lidar elevation data and the public road network, drapes one over the other, and finds every stretch of road that goes meaningfully downhill. Then it simulates a rider going down each one and ranks them by how fast you would actually get.

How accurate are the grades?

The elevation comes from the USGS 3D Elevation Program, which is airborne lidar. Where full-resolution data exists it is one metre or better, with vertical accuracy around ten centimetres. That is far more precise than anything you could measure with a phone.

We measure grade over a rolling 50 metre window rather than between adjacent points. That matters more than it sounds: the raw data carries roughly 30 cm of vertical noise, and across a 10 m step that alone fakes a 3% grade, which is the same size as the signal we are hunting.

How fast will I actually go?

Probably less than the number shown, and that is on purpose. The simulation integrates the real physics down the real profile:

v dv/ds = g(sinθ − Crrcosθ) − (ρCdA / 2m)v²

It assumes an 80 kg rider in a tuck, starting from one push, with a rolling resistance of 0.012 for urethane on good asphalt and a drag area of 0.30 m². Sit up, weigh less, ride rough tarmac, or meet a headwind and you will be slower. Nobody rides in a perfect tuck the whole way down a residential street.

Why is a hill I know missing?

A run has to hold at least 3% for at least 120 metres and drop at least 4 metres to qualify. Below 3% you decelerate against rolling resistance rather than accelerating, so it is not really a hill. We also drop unpaved roads, parking aisles, driveways, and anything tagged private in OpenStreetMap.

What does Bomb Hills not know?

A great deal, and this is the important answer. It reads terrain. It does not read traffic volume, speed limits, surface condition, gravel, potholes, sightlines, driveways, parked cars, pedestrians, construction, road closures, whether a street is private, or whether skating it is legal where you are.

A hill at the top of the list can still be a terrible place to ride. Go and look at it in person, in daylight, before you ride it.

What does the runout field mean?

What happens after the bottom. We follow the road past the end of the descent, taking the straightest continuation, and classify what the terrain does. Uphill is the best outcome, because it slows you down for free. Keeps dropping means you carry speed into whatever comes next. Road ends means exactly that, and deserves a look on the satellite view before you commit.

Why does the 3D terrain look exaggerated?

Because it is, by six times out of the box. Stillwater has 89 metres of relief across 13 kilometres. Drawn honestly at 1:1 the 3D view is a flat black sheet and tells you nothing. Every number in the sidebar is the real one; only the shape you see is dramatised, and the slider goes back to 1x whenever you want the truth.

Is the satellite imagery current?

It is USGS NAIP, flown for agricultural survey at roughly 0.3 m. Refresh cycles vary by state and it can be a couple of years old. Useful for seeing whether a road is really paved and what is at the bottom of it. Not useful for last week's roadworks.

Why do I have to pass a check to see the map?

To keep bots from scraping the run data wholesale. It is Cloudflare Turnstile, it does not track you between sites, and it is not used for advertising. The rest of the site is open.

Can you do my city?

Check it yourself, free and with no account, on the cities page. Big cities are built in squares of about 22 km rather than all at once, so you pick the part you actually ride.

We measure the local gradient over 300 metre transects, not total relief across the square. That distinction matters: measured as relief, western Kansas scores higher than Stillwater, because relief over 22 km picks up the regional tilt of the plains rather than anything rideable. Measured properly, Houston and western Kansas both return zero probes at or above a 3% grade, while Stillwater returns nearly half.

If your square comes back flat, we will tell you so rather than take your money.

Where does the data come from?

Elevation from the USGS 3D Elevation Program, public domain. Roads from OpenStreetMap contributors under the Open Database License. Aerial imagery from USDA NAIP, public domain. Base map tiles from OpenFreeMap. None of it is scraped from anyone's private database.