How this is calculated
Every number on this map traces back to a public federal count. Nothing here is modelled from guesswork, and nowhere does the copy claim more precision than the source data has.
The AADT number
Every road carries a measured Average Annual Daily Traffic count (AADT) from the federal highway inventory — the number of vehicles, both directions combined, that cross a point on that road on an average day of the year.
From a daily count to a felt time
AADT is a whole-day, both-directions total, which is not how a rider experiences a road. The felt-time figure converts it to how often you’d expect to meet a vehicle during a daylight hour:
encounters per hour = AADT × 0.07
Two adjustments cancel on the way there. AADT counts both directions, which you would halve to get just your own; oncoming traffic arrives at closing speed, which you would double. Those cancel, which is why the formula above needs no factor of two anywhere in it.
The 0.07 peak fraction itself is measured, not assumed: across hundreds of continuous traffic-count stations, a rural weekday hour between 08:00 and 18:00 averages 7.33% of that day’s total traffic, and 0.07 lands exactly on hours 09:00, 10:00, 11:00 and 17:00. It is a daylight-hour figure — the quiet stretch either side of it, night included, is quieter still.
What counts as a road here
Paved rural roads under 1,000 vehicles a day, drawn only where the federal inventory positively records the surface as paved. A segment the inventory leaves blank is left off the map entirely rather than guessed at — so some quiet pavement is missing here, never gravel drawn as pavement.
That filter has a real cost, and it is not even across states. Surface type is recorded at full extent only on the National Highway System; off it, states report on a sample of their own low-volume rural network, and some states sample much more than others. Texas states a surface for 13.3% of its rural low-traffic segments — California reports 57%, Kansas 70%, Nebraska 88%, Washington 81%, while Colorado, Arizona, South Dakota and Oklahoma report 100%. Below 90% surveyed, the app says so directly in the state view: the rest of that state’s quiet roads are not shown because they were never surveyed, not because they don’t exist. Absent is not zero.
When to ride
The band ramp answers how quiet is this road. The when-to-ride panel answers and when — and the swing inside a single day turns out to be bigger than the whole band ramp: across rural continuous count stations, 5am on a Sunday carries about 0.8% of that day’s traffic and noon carries about 9.1%, an eleven-fold swing.
Each road’s when-to-ride block draws from one of three tiers:
| Tier | Share of segments | What it says |
|---|---|---|
| Station-direct | 0.45% | Measured on this exact road, by a federal counting station on it. |
| State regional | 96.5% | A regional pattern built from that state’s own continuous counters, none of them on this road. |
| Borrowed | 5.3% | That state has too few counters to build its own pattern, so a neighboring region’s shape is used and the panel says so. |
Because 99.5% of the map is an estimate rather than a direct measurement, the panel prints coarse felt ranges rather than a precise count outside a station-direct match, and never a finer breakdown than that — a false precision would misstate what the source data actually supports.
Data vintage
The road counts are the FHWA Highway Performance Monitoring System, 2018 Public Release — the last release consistent across all fifty states, and rural traffic volumes at this scale move slowly between releases.
Utah is the one exception worth naming. UDOT publishes newer, 2024 AADT figures, but without the companion fields (urban classification, functional class, surface type) this map needs to filter a rural paved road from everything else, so the 2018 release is what’s shown here for Utah too. Spot checks against the 2024 Utah numbers agree closely with what’s mapped: UT-261 reads 139–169 here against 160 in the 2024 data, UT-95 reads 186 against 207, and UT-12 reads 296 against 302. NV-375’s count of 200 also matches the “about 200 cars a day” figure independently cited for that road.
Source: FHWA Highway Performance Monitoring System and FHWA Travel Monitoring Analysis System. Both public domain, US Government Works.