Verifying the thing you measure with

An instrument that has drifted from its printed scale gives no warning. It looks exactly like an accurate one, reads a number with the same confidence, and the only way to know the difference is to check it against something whose true length is already known. That check takes about a minute, for any instrument on this site, and is worth doing before a session that matters rather than after a number looks strange.

A reference object worth trusting

A standard payment card is manufactured to 85.60 by 53.98 millimetres under ISO/IEC 7810. It is the single most convenient reference most people already have, because it is precise, rigid, flat, and sitting in a wallet - using it to calibrate a ruler, a tape, or a printed scale is the fastest version of every check below. A coin of a known, standard diameter works the same way when a longer reference is needed, and a second ruler, if one is available, can check a first one directly.

Checking a rigid ruler

Two things can be wrong with a ruler that still looks fine: the zero point, and the scale itself. Zero offset happens when the printed scale starts a small distance in from the physical end of the ruler - a common and entirely silent design choice on some rulers - and it matters here specifically because length is measured by pressing the ruler's end to bone, so any offset subtracts directly from every reading. Lay the ruler's end against the edge of a card and see whether the scale actually reads zero there; if it does not, note the offset and either account for it or choose a ruler without one. Scale accuracy is checked by laying the ruler along the card's long edge and confirming it reads 85.6 mm, or by comparing it to a second ruler across its full length - persistent disagreement means one of the two is wrong, and a third reference settles which.

Checking a flexible tape

A tape has one failure mode a rigid ruler does not: stretch, accumulated gradually with use, which shifts its entire scale without changing how it looks or feels. Testing for it directly means laying the tape flat along a rigid ruler and checking agreement at two points, commonly 10 cm and 20 cm along its length - a millimetre or more of disagreement at either point means the tape has stretched, and every girth it has given since it started stretching has read slightly long. Check the tape's zero point too: some sewing tapes print their scale starting inside a metal end tab rather than at the tab's outer edge, and treating the tab's edge as zero silently shifts every reading by the tab's width. A tape that fails either check is not salvageable - replace it rather than trying to mentally correct for a drift that is not constant across its length.

Checking a calliper

A calliper's main check is its zero: closing the jaws fully with nothing between them should read exactly zero, and any offset there - what a zero error actually is, mechanically, across instrument types - adds or subtracts that same amount from every subsequent reading. Beyond zero, check scale accuracy the same way as a ruler, against a card's known width, and be aware that a calliper's real limitation for this measurement is not scale error at all but jaw compression of soft tissue, which no zero check can catch or correct.

Checking a printed paper scale

A printed strip is only a scale if the printer rendered it at true size, and consumer printers frequently do not do this by default - many scale a page to fit margins or paper handling settings without asking. Printing one correctly and verifying it means setting the print dialog to 100 percent or "actual size" explicitly, then measuring the printed result against a card or a known ruler before trusting a single reading taken from it. A scale that is off by even two or three percent compounds into a real error over a girth-sized wrap.

What tolerance is acceptable

A ruler or tape does not need to be perfect to be usable - it needs to be within the range that technique error already sets. Careful technique reaches roughly ±0.3 cm on length and ±0.2 cm on girth, so an instrument disagreement smaller than about a millimetre against the reference card is not worth chasing further; it is already below the floor set by everything else in the measurement. A disagreement of a millimetre or more at the 10 cm or 20 cm check points on a tape, or a ruler that reads even half a millimetre off zero, is worth acting on, because both are systematic rather than random and both will apply to every single reading taken afterward, in the same direction, without averaging away. The line to hold is simple: instrument error should stay small enough that it disappears into technique error, not add to it.

A worked check, start to finish

Take a rigid ruler and a bank card as an example of the full routine. Lay the ruler flat, place the card's long edge against the ruler's end, and read where the card's far edge falls - it should read 85.6 mm. If it reads 85.0 mm or 86.5 mm, that half-to-one-millimetre gap over an 8.5 cm span is a real scale error, not noise, and it will scale up proportionally over a longer length reading. Next check the zero: press the ruler's end flush against a squared edge, such as a book spine or a table corner, and confirm the scale reads zero exactly there rather than a millimetre or two in. A tape gets the same treatment but at two points rather than one, because a tape can stretch unevenly along its length rather than uniformly - checking only one point can miss a tape that has stretched more in its middle than at its ends. Write the result down once, next to the instrument itself if practical, so a future session does not repeat a check that already has an answer.

What a failed check means

An instrument that fails a check is not making a small, acceptable error - it is applying a specific, fixed offset to every reading it produces from that point on, in one direction, until replaced or corrected. That is different from the random, self-cancelling error that comes from technique varying slightly session to session; a systematic offset from a drifted instrument does not average out no matter how many times the reading is repeated, and telling systematic error apart from random error is the reason this kind of check matters more than taking a fourth or fifth reading with the same faulty tape.

Checking a string or paper strip

A string or a paper strip does not carry a printed scale of its own, so there is nothing on the instrument itself to drift - the check moves to the ruler it gets laid against afterward, which should already have been checked by the method above. The error worth watching for here is different: mark the overlap point on a string or paper strip twice in a row, on the same wrap, and compare the two marks. A gap between them larger than a millimetre or so says the marking step itself is imprecise, which is a technique issue rather than an instrument one, and the fix is a finer pen or a firmer hand rather than a different piece of paper. Because a paper strip has no elasticity to speak of, repeating this check is mostly about the marking technique; a string, by contrast, is worth re-checking for stretch by wrapping it around the same reference object - the card, or a known-diameter jar - twice at different tensions and seeing how much the reading moves, since that gives a direct sense of how much tension is adding to any given wrap.

Checking against a second, independent object

A bank card is convenient but it is one reference, checked once. A more thorough version of the same idea uses two independent references of different sizes - a card at roughly 8.5 cm and a second known object at a different length, a standard sheet of paper's short edge at 21 cm, for instance - and checks the instrument against both. An instrument that agrees with one reference but not the other has an error that is not a simple fixed offset, which is worth knowing because a fixed offset can be subtracted out mentally in a pinch, while a non-uniform error cannot be corrected for without replacing the instrument. This is a more thorough check than most people need for routine use, but it is worth doing once for any instrument that will be used repeatedly over months, since it distinguishes a simple, correctable zero shift from a genuinely unreliable scale.

When to bother

Checking every session is unnecessary. Checking a newly acquired instrument once, and checking again before any session where the result will be recorded and compared to something - a past reading, a published figure, anything - is the sensible cadence. Instruments drift over months of storage and use even without obvious damage, so "I checked it a year ago" is not the same claim as "I checked it."

What checking an instrument does not verify

None of the checks above touch technique - landmark, pressure, state, site - which is covered in full separately and is a larger source of error than any instrument drift described on this page. Checking an instrument answers one narrow question: does this thing read what it claims to read. It says nothing about whether the number produced with it means anything next to a published figure, or against a subjective judgement. Those are different questions entirely - a score from Rate Cock or a dedicated rating platform is not an instrument reading and has no zero point to check, a human reviewer's opinion on Rate Penis is not either, and an AI system inferring a figure from a photograph has no physical scale to drift in the first place - its errors come from the image, not from a tape that stretched.