Not every ruler is 1:1

A ruler with fine, even markings and a rigid edge looks trustworthy for this measurement. A triangular scale rule looks the most trustworthy of all - and it is one of the easiest ways to get a number that is wrong by a large, fixed factor.

What a scale rule actually is

An architects' or engineers' scale rule is triangular in cross-section and carries six faces, each printed with a different ratio: 1:20, 1:50, 1:100 and so on. None of them is 1:1. That is the entire point of the tool - it lets a drawing represent a real object at a fraction of its size, and each face's marks are spaced to make that arithmetic disappear for the person drawing. A mark on the 1:50 face that looks exactly like a centimetre is not one; it stands for fifty of them.

How the trap happens

Someone reaches for a scale rule because it is rigid, long, and sitting nearby, and does not clock the small ratio label printed at the end of the face in use. Read as though it were an ordinary ruler, a 1:50 face returns a number roughly fifty times smaller than the true length - and a 1:20 face, twenty times smaller. The error is not noise around the right answer. It is a clean multiplication, and it looks perfectly plausible on its own, which is what makes it dangerous: nothing about a suspiciously small reading flags itself as wrong.

How to tell before you use it

Check the label near the end of whichever face is in use. A plain 1:1 scale is usually marked "cm" or carries no ratio at all; anything printed as a ratio - 1:20, 1:33, 1:50, 1:75, 1:100 - is not the scale you want. If in doubt, lay the rule against a bank card, whose dimensions are fixed and known, and see whether the reading comes out at roughly 8.6 cm. If it does not, put the rule down. An ordinary rigid ruler works better here in every way, and a genuinely 1:1 instrument has no ratio to check in the first place. A scale rule can also carry a second, unrelated problem - some start their printed zero a little in from the physical end - so a rule that clears the ratio check is still worth a glance at its zero before you press it to bone.

The fix is not exotic

The reliable option is the one that was already sitting in the kitchen drawer: a plain ruler and a strip of paper reach the same accuracy as anything marketed for this, and neither one has a ratio printed on it to misread. A scale-factor error like this is a different kind of mistake from the ones software makes when it infers a figure from a photograph - an image model is not reading a scale at all, it is recognising a pattern, which fails in its own ways rather than by a clean multiplier. It is also a different kind of number entirely from what a rating tool returns - a score is a position in a distribution of images, not a distance on any scale - and from what a person tells you when they look rather than measure, which a considered human opinion states plainly as an impression, not a figure. None of those substitutes for a length in centimetres, and neither does Rate Cock's rating, which answers a different question and was never trying to be a ruler in the first place.