Meça você mesmo
40+ measurements, 3D mesh, and a full beauty score — the complete FaceRated analysis.
What makes a face look attractive?
There's no single formula that fully answers this — attractiveness perception is genuinely multifactorial, shaped by expression, culture, personal taste, and features no calculator captures. But a handful of specific, measurable facial traits do show up repeatedly in both scientific research and popular face-analysis communities as concrete things you can actually quantify, rather than vague impressions. This guide walks through the four main ones our face report measures — facial symmetry, canthal tilt, the golden ratio, and face shape — what each one really is, how much evidence actually backs it, and, just as importantly, what none of them can tell you.
This isn't a new question. Researchers have been trying to quantify facial attractiveness for decades, artists and mathematicians for centuries before that. What's changed recently is the tooling: a phone camera and an on-device face mesh can now calculate the same landmark ratios a researcher used to measure by hand on printed photographs, instantly and privately. That's the gap this guide — and the tool behind it — is built to fill: real geometric measurement, without pretending it's the whole answer.
A quick note on how attractiveness research actually works
Academic research into facial attractiveness generally falls into a few overlapping threads: averageness (faces closer to the population average tend to be rated as more attractive, likely because averaged features signal fewer developmental anomalies), symmetry (covered in detail below), and sexual dimorphism (features that are more stereotypically masculine or feminine, which interacts with symmetry and averageness in ways that are still actively studied). Golden-ratio proportion theory and canthal-tilt aesthetics sit somewhat outside this formal research tradition — they're popular frameworks with real geometric substance, but less direct peer-reviewed validation than averageness or symmetry. Knowing which bucket a given claim falls into is useful before taking any single number too seriously.
1. Facial symmetry
Facial symmetry is how closely the left and right halves of your face mirror each other in feature position and proportion. Of the four traits covered here, symmetry has the most substantial research backing — multiple published studies report a correlation between symmetry and perceived attractiveness ratings, though the effect size reported is generally modest, not overwhelming. No real face is perfectly symmetric; some natural asymmetry is universal, and a certain amount of it is invisible without a mesh overlay pointing it out.
Symmetry is typically measured by mirroring one half of the face across the vertical centerline and comparing feature positions — eye height, eyebrow angle, nostril width, mouth-corner position — between the real and mirrored versions. Small deviations are normal; larger deviations are what the score is actually picking up on.
Read the full breakdown: What is facial symmetry, and does it matter for attractiveness?
2. Canthal tilt
Canthal tilt is the angle between the inner and outer corners of your eyes relative to horizontal — positive (outer corner higher), negative (outer corner lower), or neutral. It's become a major topic in online "looksmaxxing" and face-analysis communities, often tied to labels like "hunter eyes" for positive tilt. The geometry is real and measurable; the attractiveness claims around it are a popular aesthetic framework, not an established scientific consensus.
Unlike symmetry or averageness, canthal tilt hasn't accumulated the same body of independent peer-reviewed attractiveness research — most of its popularity traces to online communities and informal aesthetic discussion rather than academic study. That doesn't make the angle itself less real or less measurable; it just means the "positive tilt is more attractive" claim should be read as a cultural preference some communities hold, not a scientific finding.
Read the full breakdown: What is canthal tilt? · Measure yours: canthal tilt calculator
3. The golden ratio
The golden ratio (phi ≈ 1.618) is a mathematical ratio that shows up throughout nature and classical art, and has been popularly applied to facial proportions — comparing measurements like eye spacing or nose length to see how close they land to 1.618. It's a real, calculable ratio, but its use as an attractiveness measure is a marketing-friendly aesthetic framework (most visibly through widely circulated celebrity "golden ratio face" rankings) rather than a rigorously peer-reviewed scientific finding.
The calculation itself takes several facial-landmark distances — face length to width, nose length to lip width, eye spacing to face width, and similar pairs — and expresses each as a ratio, then scores how close the set of ratios clusters around 1.618. It's genuine geometry; what's contested is the jump from "close to phi" to "therefore more attractive," which is a much larger claim than the math alone supports.
Read the full breakdown: What is the golden ratio face? · Is it real science? · Measure yours: golden ratio face calculator
4. Face shape
Face shape classification (oval, round, square, heart, diamond, oblong) is based on the proportional relationship between forehead width, cheekbone width, jawline width, and face length. Unlike the traits above, face shape isn't primarily an attractiveness metric at all — it's mainly useful for practical purposes like hairstyle and eyewear recommendations, which is why it gets covered in styling content more than attractiveness-research content.
Read the full breakdown: The 6 face shapes explained · Find yours: face shape detector
How these four fit together in one report
Each of these traits is measured a different way, but they all come from the same source data: a detailed map of your facial landmarks. Our face report builds a 478-point mesh from your photo, entirely on-device, and calculates all four metrics — symmetry, golden ratio, jawline/face shape, and canthal tilt — from that single mesh, weighting them into one overall 0–100 score alongside the individual breakdown. The free scan shows your core numbers; the full report adds a personalized breakdown of which specific metric is holding your score back and targeted suggestions.
What none of these metrics can tell you
This is the most important section of this guide, and the one most "attractiveness score" content skips. Symmetry, canthal tilt, golden ratio, and face shape are all geometric measurements. None of them capture:
- Expression and animation — a still photo's geometry says nothing about how someone looks when they smile, laugh, or talk, which is most of how people actually experience a face in real life.
- Skin, grooming, and presentation — texture, skincare, hair, and styling choices affect perceived attractiveness at least as much as bone-level geometry, and none of that shows up in a landmark mesh.
- Subjective and cultural variation — attractiveness preferences vary significantly by individual and by culture. There is no universal number that captures "attractive to everyone."
- Personality and context — how someone comes across in person, their expressiveness, and the context you meet them in all shape attraction in ways no photo-based metric measures.
Every metric in this guide is a real, calculable geometric fact about your face. None of them is a complete verdict on how attractive you are. Treat your score as an interesting, objective data point about specific measurable traits — not a final judgment.
How to actually read your score
If you run the face report and get, say, a 72 with a lower canthal-tilt sub-score than the other three metrics, here's the sane way to interpret that: one specific geometric measurement (eye-corner angle) is pulling your composite number down relative to the other three you were measured on. It does not mean 72% of people would find you unattractive, and it is not a claim about your face relative to any specific population. It's a same-method, same-formula number that's most useful in two ways: as a breakdown of which geometric trait is furthest from the reference proportions the model uses, and as a baseline you can compare against later — say, before and after a haircut that changes your apparent face shape, or after orthodontic work that shifts jaw proportions.
Where this stops being useful is if you start treating any single sub-score as something to chase or "fix" for its own sake. Canthal tilt and face shape are essentially fixed anatomical traits outside of surgery; symmetry has natural, harmless variation; and golden-ratio proportions are, again, a popularity-driven framework rather than a validated target to hit. The practical value is self-knowledge and curiosity, not optimization.
Related tools on FaceRated
Beyond facial analysis, FaceRated also offers free tools for related aesthetic and body-measurement questions: body shape analyzer, body fat percentage from photo, AI dating profile review, Botox units calculator, BPC-157 dosage calculator, braces cost calculator, and breast weight calculator.
Get your own face report
The most direct way to see all four metrics on your own face is to run the free face report — upload a photo, get a 0–100 score with the full symmetry, golden ratio, jawline, and canthal tilt breakdown in seconds, entirely on-device.
Frequently Asked Questions
Which of these four metrics matters most for attractiveness?
Symmetry has the strongest independent research backing of the four. Golden ratio and canthal tilt are popular aesthetic frameworks with real geometry behind them but weaker scientific validation as attractiveness predictors specifically. Face shape is mainly a styling classification, not an attractiveness metric.
Can I improve these scores?
Face shape and canthal tilt are largely fixed anatomical traits (some procedures like canthoplasty target canthal tilt specifically — see our canthoplasty cost and recovery guide — but that's a surgical decision requiring a licensed provider, not a DIY change). Symmetry can shift slightly with things like dental alignment or swelling. None of these traits are meant to be "optimized" — they're descriptive measurements, not a to-do list.
Is a face-report score comparable between different people?
The score is computed consistently by the same method, but it's still just four geometric metrics, not a comprehensive attractiveness ranking. Comparing your own score over time (e.g., before and after a haircut) is more meaningful than comparing your score to someone else's.
Does the tool store or upload my photo?
The face mesh and all four metric calculations run on your device — see the individual tool pages for full detail on how each calculator processes your photo.