LooksMeter Research

The evidence behind the measurements. Data studies, measurement standards and evidence reviews. Each document states its method, cites its sources and says what it cannot show.

Norms coverage
14 facial ratios
Sources cited
81
Faces in the norms dataset
1,404
Last updated
19 November 2026

Data studies

Our own measurements, reported with the sample, the method and the limits.

Data study · Version 1.2

Facial measurement norms: percentiles for 14 facial ratios across 1,404 faces

The median and spread of canthal tilt, fWHR, midface ratio, eye spacing and ten more measurements, taken from 1,404 front-facing photos. With the share of faces that actually meet each commonly quoted target.

· 10 min read

Data study · Version 1.0

Do AI models rate faces like humans? Inter-rater agreement, shared vs private taste and determinism

How much two human raters agree with each other, how geometry-plus-vision models compare against human panel consensus and why unanchored chat models drift across identical photos.

· 9 min read

Measurement standards

What each measurement is, how it is read and what can make it wrong.

Measurement standard · Version 1.1

Measurement standards: how LooksMeter defines and reads each facial measurement

The definition, unit, reading rules and known sources of error for every facial measurement LooksMeter reports. Written so a result can be checked, repeated and compared.

· 9 min read

Evidence reviews

What peer-reviewed studies report about the claims made for facial measurements.

Evidence review · Version 1.1

The golden ratio and facial attractiveness: what the evidence shows

Studies that tested 1.618 against rated attractiveness, the proportions they found instead and how often real faces land on the classical targets.

· 10 min read

Evidence review · Version 1.0

Camera distance and facial proportions: how a selfie changes what a face measures

What published studies report about perspective distortion at arm's length, which measurements it moves and how to take a photo that measures fairly.

· 8 min read

Evidence review · Version 1.1

Facial width-to-height ratio (fWHR): what it predicts and what it does not

The first findings, the two meta-analyses and the null results behind a ratio that has drawn a great deal of research attention.

· 9 min read

Evidence review · Version 1.0

Facial symmetry and attractiveness: how much it matters

What experiments, reviews and cohort data report about symmetry, attractiveness and health. With what measurement studies show about asymmetry in healthy faces.

· 9 min read

Evidence review · Version 1.0

Canthal tilt and periorbital proportions: what the evidence shows

What peer-reviewed studies report about medial vs lateral canthal tilt, palpebral fissure shape, brow-cheek balance and why sagittal head tilt shifts every photo reading.

· 9 min read

Evidence review · Version 1.0

Jawline, gonial angle and the lower facial third: what the evidence shows

Peer-reviewed measurements of preferred gonial angle, chin and mandibular prominence tolerances, lower-third taper and lip proportions across 3D surveys and clinical rating panels.

· 10 min read

Evidence review · Version 1.0

Facial averageness and attractiveness: what composite-face studies show

Why mathematically blended composites outscore most individual faces, how shape warping separated geometry from smooth skin and why top-decile faces still depart selectively from the mean.

· 9 min read

Evidence review · Version 1.0

The attractiveness halo effect and 100-millisecond impressions: what meta-analyses show

What 76-study and 919-effect-size meta-analyses actually find about appearance bias, why 100-millisecond snap ratings track untimed judgements and why different photos of one face vary as much as different people.

· 9 min read

Evidence review · Version 1.0

Sexual dimorphism and neoteny in human faces: why masculine and feminine preferences diverge

Why femininised female morphs win consistently across cultures while extreme masculinisation in male faces splits raters between perceived dominance and perceived warmth.

· 9 min read

Evidence review · Version 1.0

Facial adiposity, skin tone homogeneity and coloration: what controlled experiments find

How soft-tissue fullness and visible chromophore distribution move perceived age, health and attractiveness independently of underlying bone structure.

· 8 min read

Evidence review · Version 1.0

Side-profile convexity, Ricketts E-line and chin projection: what orthognathic panels show

Published layperson vs clinician tolerances for chin retrusion and protrusion, Ricketts' aesthetic plane across populations and why profile balance depends on nose and lip projection together.

· 9 min read

Evidence review · Version 1.0

Beard density, stubble length and cranial hair framing: what perception studies show

Why heavy stubble often outscores both clean-shaven and full beards on attractiveness, how full beards raise perceived age and status and why clean-shaved scalps outperform thinning coverage.

· 8 min read

Evidence review · Version 1.0

Smile arc, gingival display and buccal corridors: what orthodontic rating panels find

Published layperson thresholds for upper gum exposure, midline cant, smile-arc consonance and dark corner corridors compared with stricter orthodontist tolerances.

· 8 min read

Evidence review · Version 1.0

Waist-to-hip ratio, shoulder-to-waist V-taper and body composition: what rating studies show

How waist-to-hip ratio and body mass index interact in female physique ratings and why upper-body waist-to-chest taper accounts for the majority of variance in male physique studies.

· 8 min read

Editorial standards

How this research is done

Sources
We cite published work from peer-reviewed journals. Every reference is checked against Crossref before it is used: the DOI resolves and the title and authors match. Papers indexed in PubMed are checked there as well. Findings are summarised from the published abstract. Where a paper has no abstract we could obtain, we report only what its title states and say so.
Our own data
Figures from LooksMeter scans are aggregates. Each data study publishes how the sample was built, how repeat scans were handled and what the sample cannot show. No photo and no single person's measurement is published.
Evidence grades
Evidence reviews grade each claim. Consistent: several independent studies point the same way. Mixed: studies point in different directions. Limited: one or two studies, not yet repeated. Not supported: tested, with no effect found or one too small to matter.
Versions and corrections
Every document carries a version and a date. A material change raises the version. Corrections are welcome at [email protected].
What this is not
These documents are written by the LooksMeter team. They have not been peer reviewed or reviewed by a medical professional. The measurements they describe are not clinically validated and are not a diagnosis.

Continue