Revolutions per mile, and why your odometer drifts

Revolutions per mile is the hinge between a tire and everything electronic on the vehicle. Count how many rotations one mile of road costs you, and — because a car's instruments have no other input — you have the exchange rate between physical distance and everything the dashboard believes.

Getting it from a size code takes one division. A mile is 63,360 inches; divide that by the tire's circumference in inches and you have it. A 225/65R17 measures 28.52 in across, giving a circumference of 89.6 in and 707.3 revolutions per mile.

And that figure is wrong. Not by much, but it is wrong, always, on every tire, in a predictable direction — and the reason it is wrong is more useful than the number itself.

Why the calculated figure and the tested figure disagree

The arithmetic above uses the tire's free diameter: its shape hanging in the air, unloaded, inflated, unstressed. Under a vehicle, the sidewall in contact with the road deflects. The axle therefore sits closer to the ground than half the free diameter, and the tire covers slightly less distance per revolution than its circumference suggests.

Less distance per revolution means more revolutions per mile. So a manufacturer's published figure — which comes from a rolling test, not a formula — generally sits above the geometric one, typically by a few percent, varying with construction, inflation pressure and load.

There is a second-order effect pulling the other way at speed: centrifugal force grows the tread band slightly, which increases the effective radius. It is smaller than the deflection term and does not cancel it.

None of this is a defect in the calculation. It is the difference between a nominal geometry and a loaded structure, and no formula that takes only a size code as input can know your inflation pressure, your corner weight, or how stiff your particular casing is.

Why the disagreement does not ruin the comparison

Here is the part that matters, and the reason a calculator built on geometry is still worth trusting for the questions people actually ask.

Almost nothing you want to know is an absolute revolutions-per-mile figure. What you want is a ratio: how much did this change when I swapped tires? Speedometer error is a ratio. Odometer drift is a ratio. The shift in engine speed at a cruise is a ratio. Effective final drive is a ratio.

Write the real figure as the geometric one multiplied by some deflection factor. When you divide the new tire's figure by the old tire's figure, the two deflection factors divide as well — and if the two tires are broadly similar in construction, running at similar pressures, under the same corner weight, those factors are close to equal and they very nearly cancel. What survives is the ratio of the diameters, which is exactly what the geometry gives you.

That is a real and load-bearing claim, so here is where it breaks down. Compare a soft passenger casing at 33 psi against a stiff light-truck casing at 65 psi, and the two deflection factors are not equal. The cancellation is partial, and a residual error rides along in the answer. The same applies when one tire is carrying a much heavier corner than the other, or when a size change is accompanied by a change of load range.

So: trust the ratio between two comparable tires. Treat the absolute figure as a good approximation that a manufacturer's tested number should be preferred to whenever you have one. The reasoning is set out in more formal terms on overall diameter and rolling radius and the revolutions per mile reference.

The tire moves more than the size change does

An uncomfortable comparison, and the strongest argument for keeping the absolute figure at arm's length.

Going up one width step from 225/65R17 to 235/65R17 takes the diameter from 28.52 in to 29.03 in — +0.51 in, or 1.79% — and revolutions per mile from 707.3 to 694.8.

Now consider the same tire simply wearing out. A passenger tire leaves the factory with something on the order of eight millimetres of usable tread and is worn out at about a millimetre and a half of it. Losing that depth from the radius removes roughly half an inch from the overall diameter, top and bottom combined.

Read those two paragraphs again. The diameter change from wearing one tire out is comparable to the diameter change from going up a size. Your revolutions per mile is not a constant that a size change disturbs; it is a slowly moving quantity that a size change nudges. Anyone quoting the figure to a tenth for a tire with 20,000 miles on it is quoting precision the object does not have.

It also means the drift below is not perfectly linear across a set's life. It steepens as the tread goes.

The odometer is an integrator, and it never resets

The speedometer's error is instantaneous: look away and it is gone. The odometer adds its error to a running total that follows the vehicle for the rest of its existence.

At the 1.79% of the example above, every 100 miles the instrument reports is 101.8 miles of road actually covered. Over 36,000 indicated miles, roughly 36,650 real ones have passed under the tires — about 650 the vehicle has no record of. Over 60,000 indicated, it is close to 61,100, and the uncounted gap is over a thousand miles.

Note the direction: a taller tire makes the vehicle under-count. It presents as lower-mileage than it is.

Where the uncounted miles turn into something real

• Resale. Mileage is one of the two or three numbers a valuation actually keys on, and the odometer is the only evidence anyone consults. A vehicle that spent four years on tires one size up carries a permanently understated figure, and it is the next owner who inherits the difference.
• Service history. Intervals recorded against an undercounting odometer arrive later in real distance than the schedule intends, every time, cumulatively.
• Reimbursed mileage. Anyone claiming a per-mile rate is claiming against the instrument, not against the road, and is quietly under-claiming for the life of the tires.
• Distance-based warranty and lease terms. These are written against the instrument too, so under-counting works in the owner's favour here — the one place the drift pays.

None of these is dramatic. All of them are one-directional, which is what makes them worth knowing about. An error that averages out is noise; an error with a sign is a bias.

Getting the number for your own tires

Two codes into the comparison tool returns both revolutions-per-mile figures, the diameter difference and the percentage, and the converter gives the full derived set for a single size, circumference included.

Then apply the caveat this whole article has been building to. The percentage between two similar tires is solid. The two absolute figures are geometric, they sit slightly below what a rolling test would produce, and they describe a brand-new tire that stopped being that tire the first week you drove it.