The 16 facial harmony metrics, explained
MogTracker scores a face on sixteen proportions. Each guide below explains what the measurement is, how it is derived from the landmark mesh, the reference value it is scored against, and what can actually change it.
Fourteen of the sixteen carry weight in the composite score. Two — eye separation ratio and hunter eyes — are shown for context only, because they overlap with metrics already counted. Reference values come from the Bomer and Heard reference faces, measured through the same pipeline that scores users rather than assigned by hand.
Eyes
- Canthal TiltThe angle between the inner and outer corner of your eye, and one of the two heaviest components of the overall score.12% of composite
- Eye Aspect RatioHow tall your eye opening is relative to its width — the difference between an almond eye and a round one.7% of composite
- Eye Spacing RatioThe width of one eye compared with the gap between your eyes — the classic "one eye-width apart" rule, measured properly.6% of composite
- Eye Separation RatioHow far apart your eyes sit relative to the full width of your face. Shown for context; it does not affect your score.Display only
- Hunter EyesA combined score for the deep-set, hooded hunter eye look, built from canthal tilt and eye aspect ratio. Shown for context — it does not affect your score.Display only
Nose
Jaw
Proportions
- Facial ThirdsWhether your face divides into three equal horizontal bands — jointly the heaviest metric in the score.12% of composite
- Lower Third ProportionsHow the lower third splits between your upper lip and the chin below it.5% of composite
- Facial FifthsWhether your face divides into five equal vertical columns, each about one eye wide.4% of composite
- Eye-to-Mouth RatioThe vertical distance from your eyes to your mouth, relative to overall face height.5% of composite
Lips
Symmetry
What each metric is worth, and how people actually score on it
Weight is what the metric contributes to the composite. The averages come from 123,813 scans, and they say something the weights alone do not: several heavily-weighted metrics barely separate anyone. If nearly every face lands in the top band of a metric, that metric adds close to a constant to everyone's score rather than distinguishing between them.
| Metric | Weight | Mean | Spread |
|---|---|---|---|
| Facial Thirds | 12% | 58.24 | Spread across bands |
| Canthal Tilt | 12% | 20.33 | Almost everyone scores bottom |
| Cheekbone-to-Jaw Ratio | 10% | 76.34 | Skewed high |
| Nose Width Ratio | 9% | 92.19 | Almost everyone scores top |
| Facial Width-to-Height Ratio | 8% | 85.3 | Almost everyone scores top |
| Eye Aspect Ratio | 7% | 69.25 | Skewed high |
| Lip Ratio | 7% | 46 | Spread across bands |
| Mouth-to-Nose Ratio | 6% | 81.15 | Skewed high |
| Facial Symmetry | 6% | 76.33 | Skewed high |
| Eye Spacing Ratio | 6% | 59.6 | Spread across bands |
| Eye-to-Mouth Ratio | 5% | 90.89 | Almost everyone scores top |
| Lower Third Proportions | 5% | 74.06 | Skewed high |
| Facial Fifths | 4% | 6.51 | Almost everyone scores bottom |
| Philtrum Length | 3% | 72.65 | Skewed high |
| Eye Separation Ratio | — | 59.55 | Spread across bands |
| Hunter Eyes | — | 37.47 | Spread across bands |
Facial fifths and canthal tilt are where most composites are lost, and nose width, eye-to-mouth placement and overall face proportions are where almost nobody loses anything. Those five between them carry 38% of the composite. Snapshot captured 2026-08-14.