Hair in the way

Bone-pressed length means the ruler stops on bone. Hair at the base can stop it earlier than that, and because hair does not sit the same way twice, it stops it a different amount each time.

Why this is worse than it sounds

A single millimetre of inconsistency would be tolerable against the rest of the error budget in a self-measurement. Hair does not cost you a consistent millimetre - it costs you a variable amount, because how much it compresses under the ruler's edge depends on how it happens to be lying, how recently it was trimmed, and how hard you press. Pressing hard is already the largest source of difference between a careless and a careful reading; a mat of hair between the ruler and the bone adds a second, harder-to-control version of the same problem on top of it. The result is a length that moves between sessions for a reason that has nothing to do with the thing you are trying to measure.

What to actually do

Part the hair away from the midline before you place the ruler, rather than pressing through it. If it is thick enough that parting does not reliably clear a path to skin, trim the area at the base short - this is a measurement accommodation, not a grooming recommendation, and the site takes no view on hair beyond what it does to a reading. Then press as usual and confirm the ruler has stopped against bone rather than against compressed hair, which you can generally feel as a softer, less definite stop.

Where this sits in the method

This note covers only the hair itself. Finding the landmark in the first place is a separate step and worth reading if the stop itself is the part that is unclear, and how hard to press once you have found it is the piece that governs the fat pad rather than hair. Treat hair as one more variable to remove before the pressure question even applies, the same way you would clear a surface before placing a rigid ruler down for any other length reading.

It is a small thing on its own. It is also exactly the kind of small thing that, left uncontrolled, produces the "my number keeps changing" complaint that the rest of this site's error-bar material exists to explain - most of it is not the ruler, and this particular piece of it is not the ruler either.

None of this applies to a subjective judgement, which is worth being clear about since the two get lumped together online. Rate Cock is scoring a photograph, not pressing a ruler to bone, so hair at the landmark has no equivalent problem there - it is a different kind of number with a different set of failure modes, worked through by an image model rather than a technique. A numeric rating or a human judge's opinion can still be thrown off by plenty, just not by this.