How to know if you're attractive — and which of the signs hold up

Search this question and you get the same article fifteen times: a numbered list of signs. People hold eye contact a beat too long. Strangers do you favours. You get double takes. The lists are cheerful, they are ancient, and none of them can be checked — to know your double-take rate is high you would have to know everyone else's, and you never will.

There is a smaller, duller answer that can be checked. Part of a face is geometry, and geometry is measurable to a fraction of a degree. This is what that part covers, what the research says it is worth, and where it stops.

Why the signs listicles cannot answer the question

A sign is useful when its absence would tell you something. Take the most common one on those lists: strangers smile at you. Suppose they do. Is that a high rate? You have no baseline, no control group, and no access to what happens when you are not in the room. The claim cannot fail, which is exactly why it is comfortable to read and useless to act on.

The other structural problem is the observer. Attractiveness is a rating other people produce, and you are the only person on earth who cannot produce it about your own face — you have looked at it too many times, and mostly in a mirror, which is not the version anyone else sees.

The four things about a face that are actually measurable

From a single straight-on photo, four families of number can be extracted reliably. Everything else people list is either depth information a front photo does not carry, or an opinion.

• Symmetry. How closely the left and right sides mirror each other across matched landmark pairs. Reported as a percentage, and measured here. Nobody scores 100, and the normal range is much narrower than people assume.
• Proportion. The vertical thirds and the horizontal fifths, plus how each ratio compares to phi. The golden ratio calculator reports them individually rather than as one composite number, and the full guide to facial proportions explains where the rules came from.
• Shape classification. Which named category your outline and features fall into — face shape, eye shape, nose type, lip shape. These are labels, not scores; no shape is ranked above another anywhere on this site.
• Angles. Canthal tilt in degrees is the clearest example: a real angle, measured per eye, that people constantly guess at from photographs and constantly get wrong.

Photos, mirrors, and why the same face reads differently

This is the single most common version of the question, and it has an answer that has nothing to do with how attractive anyone is.

1. 1You know the flipped version. A mirror shows your face reversed, and that is the version you have practised recognising for your whole life. A photograph shows the unflipped one — the one everyone else knows — and it reads as subtly, unplaceably wrong. Try flipping a photo of yourself horizontally; most people find the flipped copy immediately more familiar and more flattering, and it is the same face.
2. 2The lens is too close. A phone's front camera at arm's length sits far closer than a person talking to you does. Whatever is nearest the lens — usually the nose — is enlarged relative to everything else. This is a property of distance, not of the nose.
3. 3The light is wrong. Overhead light drops shadows under the brow and eyes; light from below does the opposite. A mirror is usually lit deliberately. A snapshot usually is not.
4. 4The expression is frozen. Faces are read in motion. A still frame catches one arbitrary moment out of thousands, and it is very often not a moment you would have chosen.

The popular signs, checked against the research

What actually moves how a face gets rated

Rhodes' 2006 review in the Annual Review of Psychology is the standard summary: averageness, symmetry and sexual dimorphism all contribute, with modest effect sizes, and none of them acts as the single lever the popular accounts describe. Two things follow from that, and they point in a more useful direction than a bone-structure audit.

First, the fixed geometry explains less of the variance than people assume. Second, the parts that are not fixed — skin condition, sleep, grooming, expression, posture, and how a photograph was lit and framed — move ratings measurably, and all of them are within reach. That is not a consolation prize. It is where the available gains are.

How to get a measured answer about your own face

Take one photo straight on at eye level, in even light, with a relaxed expression, and run it. The output is a set of numbers with the landmarks they came from — not a verdict.

• Attractiveness test — the composite score, and the four measurements it is built from.
• Am I pretty? — the same measurements, framed for the question people actually type.
• Rate my face — a rating out of ten that shows its arithmetic instead of hiding it.
• Face symmetry test — left against right, pair by pair.
• PSL scale — what that 1-10 ladder is, where it came from, and why the labels at the bottom of it are not used here.

What the measurements will not tell you

They will not tell you whether you are attractive, because that is not a property of a face — it is a rating produced by other people, in a context, with a history. Two faces with near-identical geometry get rated very differently depending on lighting, expression, grooming and who is doing the rating.

They also will not tell you anything about depth. Orbital depth, the profile of a nose, chin and lip projection — none of it is present in a front photo, and every page here says so where it applies rather than filling the gap with a plausible number.

And a number is a poor thing to argue with. If the question behind the search is really about how you feel when you look at a photograph of yourself, a symmetry percentage will not settle it — that one is worth talking to someone about rather than measuring harder.

If you are weighing an actual change, the simulations render one procedure on your own photo so you can see the difference instead of imagining it. It is a visualisation. Nothing on this site is medical advice, and no result here is a diagnosis.