How a face with East Asian ancestry scores across 17 structural metrics. Descriptive anthropometry, not a hierarchy.
Beauty is multi-ethnic. Published cross-cultural research finds no single ethnicity scores higher in aggregate. This page describes the metric distribution, not a ranking.
17 metrics · Farkas international norms · Free · No signup
Free score · $14.99 unlocks population-appropriate percentile report
Most legacy face-scoring tools were calibrated on Caucasian-dominant datasets, which means the eye aspect ratio, nasal index, and midface projection thresholds were set against one population's distribution. When a monolid or folded East Asian eye is scored against a Caucasian eye-aperture norm, the resulting percentile reads as a low score on a metric where the face is actually average for its population. Choe et al. (2004) on Korean craniofacial norms and Le et al. (2002) on Vietnamese norms both document distributions that diverge meaningfully from Western reference datasets on a handful of specific metrics.
The fix is not to weight populations differently in the composite. The fix is to show both percentiles. A useful East Asian face report shows you where you sit against the universal distribution (cross-population, descriptive) and against the population-appropriate distribution (Farkas international atlas plus Korean and Japanese datasets where validated norms exist). A metric where you sit at the 35th universal percentile but the 65th population-appropriate percentile is not a weak metric; it is a metric where the universal norm was the wrong reference.
The 17 structural metrics themselves are universal geometry. Facial thirds, fifths, FWHR, canthal tilt, jawline ratio, and the rest are defined the same way regardless of population. What changes is the reference distribution they get compared against. The free composite uses the universal distribution; the $14.99 paid report adds the population-appropriate one.
Korean and Japanese norms (Choe 2004; Wang et al. 2011) document a tendency toward a wider bizygomatic relative to bigonial measurement compared to European norms, contributing to the perception of higher cheekbones. Wide individual variance.
International atlases (Farkas et al. 2005) record slightly flatter midface projection on average in East Asian populations compared to European samples. Descriptive, not evaluative; flatter and projected both register as neutral on the composite when measured against population-appropriate norms.
Nasal width to nasal height ratio sits in a distinct sub-range. The score treats this as population-appropriate rather than as a deviation from a Caucasian norm.
Monolid and double-lid eyes register slightly different eye aspect ratios. Population-appropriate norms (Choe 2004) prevent the Caucasian eye-aperture mean from being treated as a universal target.
Average lower-face height relative to total facial height clusters slightly shorter in published East Asian norms vs European samples. Carried as descriptive context in the metric layer.
Darker, denser eyebrow hair and a typically lower hairline shift how the brow-to-eye distance metric reads compared to lighter-brow populations. The score treats this as structural signal rather than as a styling artifact.
A single composite forces a face into one reference distribution. For most users that distribution is implicitly Caucasian because the underlying training data was. The result is a percentile that systematically misreads any face whose population norms differ meaningfully from the training set. We avoid this by carrying both percentiles in the paid report.
The universal percentile answers: where do you sit relative to faces of all populations? The population-appropriate percentile answers: where do you sit relative to faces with similar ancestry? Both are descriptive. Neither is a verdict. The gap between them is itself diagnostic; a large gap usually points to a metric where the universal distribution is dominated by one population's mean.
Free score is the headline. Population-appropriate percentile is the plan.
The $14.99 Looksmax Report scores all 17 metrics with both universal and population-appropriate percentiles, identifies your two weakest metrics, and writes a soft-tissue-first plan.
Free, instant, private. 17 metrics with population-appropriate percentile context in the paid report.
17 metrics · Farkas international norms · Photos auto-deleted
All free. All private. All instant.
Is your smile genuine or forced?
How close are your proportions to φ?
AI attractiveness analysis
Rate my face 1–10
How attractive am I?
How symmetrical is your face?
Which photo gets more matches?
Best photo for LinkedIn
Your glow-up score