Why the Average Face Feels Attractive — The Power of the Prototype Over the Golden Ratio

What the phrase "perfect proportions" misses

Scroll through beauty content and you keep meeting the same sentences. "A face is ideally divided into thirds vertically and fifths horizontally," "A face close to the golden ratio of 1.618 is beautiful." It sounds as if beauty has a single correct coordinate you could verify with a ruler. Yet the answer psychology has kept arriving at, after decades of measuring attractiveness, is far humbler and, honestly, almost anticlimactically surprising.

That answer is 'the average.' An average face — made by computationally overlaying many people's faces — tends to look more attractive than most of the individual faces that went into it. It is not that there is one perfect-ratio ideal, but that a blurry center point formed by countless faces is what actually draws the eye.

Before we begin, let me draw one line. This piece is a story about mind and culture — about how attractiveness is perceived — and not an attempt to rank whose face is better or to enshrine any particular face as the correct one. Statistical tendencies in attractiveness and the worth of a person sit on entirely different planes; here we look only at the former, and only for fun.

Langlois's composite-face experiment — what 'average' really means

The starting point for this view is the 1990 study by Langlois and Roggman, 'Attractive Faces Are Only Average.' The two digitally aligned photographs of many people's faces and averaged them pixel by pixel into 'composite faces.' Intriguingly, as the number of overlaid faces grew — 2, 4, 8, 16, 32 — the composites were rated increasingly attractive, usually scoring higher than any of the individual faces used to make them.

Here is a point that is easy to misread. The 'average' at play does not mean 'plain' or 'common.' In the process of taking a mathematical average, individual deviations — blemishes, left-right asymmetries, unusually prominent contours — cancel one another out and disappear. What remains is a smooth-skinned, evenly balanced 'central face' that the group's faces share. It is attractive not because it is plain, but because it is a clean form with the noise of individuality wiped away, so it settles easily on the eye.

The shock of this result was that it reversed the direction. People had long thought of a beautiful face as an 'exception that specially departs from the ordinary,' yet the experiment said, on the contrary, that beauty sits in the middle of many faces — at the group's center.

Prototype and processing fluency — why the average feels easy

So why, of all things, does a face close to the average feel good? Rhodes's 2006 review, 'The Evolutionary Psychology of Facial Beauty,' lays two strands of explanation side by side. One is evolutionary; the other is a cognitive account resting on how our minds process information.

The evolutionary explanation comes at it from a different angle. It hypothesizes that averageness or symmetry may signal developmental stability — a sign of 'healthy genes' — so that we evolved to be drawn to such faces. But Rhodes herself notes that this link is not as solid as one might think. The correlation between attractiveness and actual health, even where it exists, is generally weak, and much of the same preference can be explained by the cognitive account alone. So today's researchers tend to see it less as one side being correct and more as evolutionary pressures and perceptual habits working together.

The key words of the cognitive account are 'prototype' and 'processing fluency.' Over a lifetime we see countless faces and build in our minds an average prototype — a sense of 'this is roughly what a face looks like.' Because the average face is closest to this prototype, the brain recognizes it smoothly, without effort. And the ease we feel when processing is this smooth, we often mistake for the liking of 'I find this face appealing.'

Familiarity layers on top of this. Like the mere-exposure effect — an old psychology finding that we tend, for no clear reason, to like what we have seen often — the average face resembles the sum of the countless faces we have encountered all our lives, so it feels somehow familiar even at first sight. It means attractiveness stands not on some mysterious exception but on the rather practical senses of familiarity and ease.

Symmetry and sexual dimorphism — what else besides the average

Of course, averageness alone does not explain all of attractiveness. Rhodes's review, and the 2011 overview 'Facial Attractiveness' by Little, Jones and DeBruine, list left-right symmetry and sexual dimorphism (masculine and feminine features) alongside averageness as leading cues of attractiveness.

Symmetry shows a mild correlation with attractiveness across many studies. But there is a trap here too: 'perfect symmetry,' made by mirroring one half of a face onto the other, often ends up looking somehow artificial and cold. What actually feels attractive is not mathematical perfection but symmetry within a natural range. Slight left-right differences are not a flaw but the default of a living face.

Sexual dimorphism calls for more caution. The finding that feminine features are generally favored in women's faces is relatively consistent, but the preference for masculinity in men's faces splits by observer and context. Under some conditions a softer impression inspires more trust; under others a sharper impression is preferred. It shows that a cue of attractiveness is itself not a fixed formula but a tendency that rides on circumstance.

Rereading the golden-ratio myth

From here we can take another look at the staples of Korean beauty content — 'the golden ratio 1.618' and 'the ideal thirds-and-fifths grid.' Such expressions are useful as handy rules of thumb for gauging a face's balance by eye. The problem is when they are talked about as if they were physical laws that generate beauty.

The claim that the golden ratio of 1.618 is the secret of a beautiful face does not, in fact, fit well with where attractiveness research points. The strong axes of attractiveness that experiments have repeatedly confirmed were not a particular numeric ratio but 'closeness to the group's average,' 'symmetry,' and 'familiarity.' What is more, follow-up studies show that the most attractive face may sit not at the exact average but at a point slightly shifted from the average in a particular direction — for example, a face averaged only from faces already rated attractive — further deflating the single-magic-ratio hypothesis.

In other words, the golden ratio is less the cause of beauty than a story attached after the fact to explain a face that already feels harmonious. The promise that laying down a ruler and hitting 1.618 will make you beautiful is not something research guarantees.

The 'consensus on attractiveness' is not a ranking of people

The large 2000 meta-analysis by Langlois and colleagues, 'Maxims or Myths of Beauty?', checked several common assumptions against data. One was the maxim that 'beauty is in the eye of the beholder,' yet in reality people's judgments of what is attractive agreed considerably more than expected — both within cultures and across them. Even infants too young to speak gazed longer at faces that adults had rated attractive.

Such 'consensus' is easy to misunderstand. That judgments agree means our perceptual systems are biased in a similar direction — it never means that some face is objectively finer or that its owner is a better person. The same meta-analysis also noted that attractiveness affects actual social treatment, but this should be read as revealing a bias baked into our society — the 'appearance halo' — not the excellence of beauty.

So using these studies as a tool to line faces up by score is to get the direction entirely wrong. Statistical tendencies in attractiveness speak only of group-level perceptual bias; they judge nothing about a person's individuality, ability, character, or the worth they hold. Between the verdict 'attractive' and the judgment 'a good person' there is, in fact, no bridge at all. The two must be kept firmly separate.

Setting down the ratio chart

As I put this subject in order, I felt the conclusion that lightened my heart was, of all things, 'the average is beautiful.' If beauty is not an unreachable perfect coordinate but a familiar middle formed by countless ordinary faces, then it is a quality that already lives, in good measure, in anyone's face.

So the ratio chart that beauty content holds out is enough to grip lightly, as just one of many lenses for understanding a face. Attractiveness is not an exam graded by numbers but something closer to a perceptual illusion produced by familiarity and ease. And this piece is, to the end, a story about culture and mind that peers at the workings of that illusion for fun — never a message telling you or anyone to fix or to judge a face. Rather than counting how many points your face scores in front of a ratio chart, wondering why a certain face feels easy on the eye is far more enjoyable — and closer, too, to what the research actually says.