The chart in the article

A percentile chart in a lifestyle article looks like data. Sometimes it is - redrawn cleanly from a real study. Often it is a self-report survey run by the outlet itself, or an older figure recycled without its source, presented with the same visual confidence either way.

The chart tells you nothing about its own provenance

A curve on a page carries no information about where its numbers came from unless the article states it. A chart built from a professionally measured sample like Veale et al. (2015) and a chart built from an anonymous online reader poll can look identical - same axis, same smooth curve, same percentile labels - while resting on data that differ enormously in what they are worth.

That is the whole problem with a redrawn chart: the redrawing removes the one piece of information that would let a reader judge it.

What to check before trusting one

Look for a named source, not just a number. A citation to a specific paper, with authors and a year, is checkable. "Studies show" or no attribution at all is not, and an article that omits its source is not necessarily wrong, but it has given the reader nothing to verify.

Look for the measurement method behind the figures. Self-reported figures run consistently higher than clinically measured ones, for reasons that have nothing to do with honesty and everything to do with method. A chart that does not say whether its input was measured or reported cannot be placed against a source like Veale with any confidence, because the two kinds of input do not produce comparable numbers.

Look at whether the curve is smoother than the stated sample size would justify. A hand-drawn or software-smoothed curve over a small or unstated sample is an illustration, not a plotted result, whatever it resembles. The same caution applies to a figure an article attributes to "AI analysis" of user photos - an image model cannot recover an absolute length from a picture at all, so a chart built on that kind of input is not a percentile of measurements in any sense this site would recognise.

Why this matters more here than for most numbers

This is a distribution with a narrow spread in the middle, which means small differences in where a chart places the median move a reader's apparent percentile by a real amount. A chart quietly built on an inflated self-report figure will place an accurate, carefully taken measurement lower on the curve than it belongs, which is the least useful direction for that error to run.

The fix is not to distrust every chart. It is to ask the same question this site asks of any figure: what was measured, by whom, and how - the same standard a trustworthy percentile calculator should meet, and one a magazine feature usually does not bother stating.

None of this is a substitute for knowing what you actually want from a number. A percentile answers a question about distance from a mean. It says nothing about how a photograph reads to a viewer, which is the separate question that a service like Rate Cock or a photo-based rating hub is built to answer, and that a human reviewer answers a third way again.