# The dials that carry a face and how far they travel

Which measurements make a face read male, female, older or from another population, ranked by how reliably they separate faces, with measured ranges by sex, a slider mapping that spends its codes where faces differ, and the data as JSON. Every figure attributed.

*Figure: Range strips on one scale from 48 to 64 cm, each with a mean tick, a solid band at plus or minus one standard deviation and a pale band at plus or minus two. Female 54.2 ± 2.4; Male 56.3 ± 2.6.* Head circumference by sex, adults.

## The answer

*Figure: Range strips on one scale from 115 to 160 mm, each with a mean tick, a solid band at plus or minus one standard deviation and a pale band at plus or minus two. Female 133.8 ± 5.6; Male 142.6 ± 6.2.* Cheekbone width by sex (ANSUR II, n≈6,068).

A character creator needs a few dials that carry most of how a face reads. The measurement literature says which ones they are, and how far each should travel.

- **Cheekbone width is the strongest sex dial.** Bizygomatic width is 133.8 ± 5.6 mm in women and 142.6 ± 6.2 mm in men (US Army, n≈6,068): d = 1.39, and still 1.07 after adjusting for body size.[1]
- **Jaw width against cheekbone width drives the silhouette.** Bigonial width runs 70-82% of bizygomatic width. High reads square, low reads heart-shaped, and the ratio decides whether the cheekbones stand out.[12]
- **Head shape sets width; the facial index sets length.** The cephalic index (head breadth ÷ head length) and the facial index (face length ÷ face width) vary independently, so they need separate dials.[10][11]
- **Regions beat the whole face.** The reliable sex differences sit in the jaw, brow and nose. The male glabella sits 1.60 mm further forward and 1.43 mm lower, and the orbital rim 1.90 mm further back.[6] Whole-face ratios are the weakest signals.
- **Faces age by site, not uniformly.** The orbit grows 15-20% in area by the seventh decade and the maxilla loses 8-15% of its height,[4][7] while the fat compartments deflate one by one.[8] Cartilage keeps its own clock, on [its own page](/research/cartilaginous-tissues).

## The obvious dials and where they break

*Figure: A scale from -0.4 to 0.4 d. Marks: US Army, d = -0.24 at -0.24; Meta-analysis, d = 0.11 at 0.11; US Army, d = +0.16 at 0.16; Athletes, d = 0.31 at 0.31. Effect size d, male minus female; the US Army sample gives either sign depending on how height is defined.* Sex differences in facial width-to-height ratio are small and inconsistent.

Three shortcuts look right and fail.

**fWHR as the masculinity dial.** Facial width-to-height ratio is the most studied "masculine face" measure. Its sex difference is d = 0.11 across 10,853 people.[14] It vanished in athletes once body size was controlled,[15] it flipped sign in the army sample depending on how height was measured,[1] and after about age 48 women have the larger ratio.[16] Use cheekbone width instead.

**The neoclassical canons as targets.** Four of the eight classical proportion rules differ significantly between ethnic groups. The most variable measures (forehead height, eye spacing and nose width) are the canons' core assumptions.[13] Use them as a grid to deviate from.

**A linear 8-bit slider.** A byte gives 256 steps. Mapped linearly over ±3 SD, a third of the codes go to the 4.6% of people beyond ±2 SD, where neighbouring values look the same.

## Take the dials and the data

*Figure: Two sliders sit above one bell-shaped pile of people, measured in standard deviations. On the top slider, labelled linear, 256 tick marks are spaced evenly from -3 to +3 SD. The 86 ticks over the thin tails beyond ±2 SD, a third of them, are grey and labelled a third of the codes, on 4.6% of people. On the bottom slider, labelled quantile, the same 256 ticks bunch into a solid block under the middle of the bell and spread out toward the tails, labelled each code: 1 in 256 people. None of them are grey. The end ticks sit at about ±2.9 SD.* The same 256 slider codes, spent two ways. A linear map gives a third of them to the 4.6% of people beyond ±2 SD; a quantile map gives every code an equal share of people.

Map each slider code through the inverse normal CDF, so every code is an equally likely quantile:

```ts
// code 0-255 -> a measure in mm or degrees; probit is the inverse normal CDF
const dial = (code: number, mean: number, sd: number) =>
  mean + sd * probit((code + 0.5) / 256)
```

Codes 0 and 255 land at about ±2.9 SD, so the reach matches a wide linear slider, with the resolution where faces actually differ. If the parameter can be a float, the constraint disappears.

*Figure: Seven range strips, one per measure, each with a female and a male row on one shared scale: the mean as a tick, the solid band at ±1 SD and the pale strip at ±2 SD. Bizygomatic width (Zy-Zy · face width · mm · US Army, n≈6,068): female mean 133.8, male mean 142.6. Bigonial width (Go-Go · jaw width · mm · Farkas adult): female mean 91.0, male mean 97.0. Gonial angle (ramus ↔ corpus · degrees · Iranian patients, n=149 · direction disputed (see the jaw page)): female mean 124.7°, male mean 122.2°. Facial width-to-height (fWHR · athletes n=3,481 · pooled SD · d≈0.31): female mean 1.92, male mean 1.98. Interpupillary distance (pupil to pupil · mm · Turkish adults, n=756): female mean 61.4, male mean 63.9. Head breadth (eu-eu · mm · UK range, SD derived): female mean 150, male mean 155. Head circumference (round the skull · cm · n≈2,508 adults): female mean 54.2 cm, male mean 56.3 cm.*

The solid band is ±1 SD (about 68% of adults); the pale strip runs to ±2 SD (about 95%). Treat the edges as reach markers, not realism gates: going past them is stylisation.

| Dial | Female | Male | Source |
|---|---|---|---|
| Bizygomatic width | 133.8 ± 5.6 mm | 142.6 ± 6.2 mm | US Army, n≈6,068[1] |
| Bigonial width | 91 ± 5.9 mm | 97 ± 5.8 mm | Farkas adult reference[2] |
| Head circumference | 54.2 ± 2.4 cm | 56.3 ± 2.6 cm | n≈2,508[9] |
| Interpupillary distance | 61.4 mm | 63.9 mm | Turkish adults, n=756[5] |
| Gonial angle | 124.7 ± 3.2° | 122.2 ± 3.9° | 149 Iranian patients; direction disputed[3] |
| fWHR | 1.92 | 1.98 | athletes, n=3,481; weak, see above[15] |

*The headline dials, by sex*

These numbers as data, with means, SDs, samples, sources and a verification status per measure: [parameters.json](/research/facial-character-parameters/parameters.json).

Every measure in full, by region:

- [Cheekbones and the width-to-height ratio](https://www.mackenziebowes.com/research/facial-character-parameters/cheekbones) (2026-10-02): Bizygomatic width by sex and population, what varies in the zygomatic region, and why the facial width-to-height ratio is a weak sex signal. Every figure attributed. Markdown: https://www.mackenziebowes.com/research/facial-character-parameters/cheekbones.md
- [How the face sits in the head](https://www.mackenziebowes.com/research/facial-character-parameters/face-in-head) (2026-10-02): The facial index, face width against head width, the neoclassical canons and how head shape bends the face, plus the landmarks every table is built from. Every figure attributed. Markdown: https://www.mackenziebowes.com/research/facial-character-parameters/face-in-head.md
- [The cranium and how the face ages](https://www.mackenziebowes.com/research/facial-character-parameters/cranium-and-ageing) (2026-10-02): Head size and shape by sex, the cephalic index, how the vault changes with age and its genetics, then site-specific bone resorption and the facial fat compartments. Every figure attributed. Markdown: https://www.mackenziebowes.com/research/facial-character-parameters/cranium-and-ageing.md
- [The genetics of face shape](https://www.mackenziebowes.com/research/facial-character-parameters/genetics) (2026-10-02): The largest facial GWAS, the loci with an assigned facial trait, and why face shape is a continuum to blend rather than a set of gene switches. Every figure attributed. Markdown: https://www.mackenziebowes.com/research/facial-character-parameters/genetics.md
- [The jaw and mandible](https://www.mackenziebowes.com/research/facial-character-parameters/jaw) (2026-10-02): Gonial angle, bigonial and bicondylar width by sex, which jaw measures discriminate sex reliably, and how the jaw changes with age and dentition. Every figure attributed. Markdown: https://www.mackenziebowes.com/research/facial-character-parameters/jaw.md
- [The nose, eyes and brow](https://www.mackenziebowes.com/research/facial-character-parameters/nose-eyes-brow) (2026-10-02): Nose length, width and profile angles by population, interpupillary distance by age and sex, and male-female differences at the brow and orbit. Every figure attributed. Markdown: https://www.mackenziebowes.com/research/facial-character-parameters/nose-eyes-brow.md

## What this says about how we work

*Figure: A list of 13 rows: Errors fixed; Genetics corrected; Blank cells filled; IPD corrected; Sections added; Head-shape links; Vault genetics added; Key-measure chart split by sex; ±2 SD tier added; Second pass (2026-10-02); Uncited figures removed; Kept from the draft; Sources. A green tick fills each box in turn.* The revision notes, one box each.

We take a craft problem to the measurement literature, check the headline numbers against their papers, cut every figure we can't cite, and ship the result as dials and data a character creator can load.

- **Rechecked against the papers:** the bizygomatic means, SDs and effect sizes (n=6,068), the IPD table (n=756), the gonial-angle means (n=149), and the athlete fWHR means and d.
- **Not rechecked:** the bigonial width (Farkas, via a 2021 table), the nasal index, head breadth, head circumference, and the lifespan study's overall fWHR mean.
- **Cut because they couldn't be cited:** a multicentre nose CT table, unsourced IPD population ranges, the nasal-index classes, and a dozen smaller figures.
- **Limits of the evidence:** most samples cover one population each, the army sample is young and two-thirds men, and the head-breadth SD is derived from a 5-95th percentile range.

That's the reading we'd give your character pipeline before you build sliders on guesses.

If you're building a character creator and your faces keep drifting toward the same few looks, I'd like to hear about it: mackenzie@mackenziebowes.com or [@_mackenziebowes](https://x.com/_mackenziebowes).

## Sources

1. Caton & Dixson (2022). *Bizygomatic width is highly sexually dimorphic when adjusting for allometry.* Biology Letters. <https://pmc.ncbi.nlm.nih.gov/articles/PMC9554718/>
2. Farkas adult craniofacial norms (gonial width M 97 ± 5.8 / F 91 ± 5.9 mm), as tabulated in Ogodescu et al. (2021), *Updating Standards of Facial Growth.* <https://pmc.ncbi.nlm.nih.gov/articles/PMC8156684>
3. Gender Determination Using Mandibular Angle Measurement (mandibular angle M 122.19 ± 3.89 vs F 124.72 ± 3.24). <https://brieflands.com/journals/amhsr/articles/114608>
4. Shaw & Kahn / Pessa: *Changes in the Facial Skeleton With Aging.* <https://pmc.ncbi.nlm.nih.gov/articles/PMC3404279/>
5. Evaluation of interpupillary distance in the Turkish population (n≈756, by age/sex). <https://pmc.ncbi.nlm.nih.gov/articles/PMC4529256/>
6. Quantitative Analysis of Male Versus Female Frontal Bone and Orbital Skeletal Morphology. <https://pmc.ncbi.nlm.nih.gov/articles/PMC12459598>
7. Facial bone aging: an update and literature review (orbital enlargement, maxillary height loss, mandibular angle increase). <https://www.sciencedirect.com/science/article/pii/S2352587826000185>
8. Rohrich & Pessa (2007). *The Fat Compartments of the Face.* Plast Reconstr Surg. <https://pubmed.ncbi.nlm.nih.gov/17519724>
9. Microcephaly measurement in adults (n=2,508; head circumference M 56.3 ± 2.6 / F 54.2 ± 2.4 cm; correlation with brain volume and height). <https://pmc.ncbi.nlm.nih.gov/articles/PMC9126574/>
10. Cephalic index - dolichocephalic / mesaticephalic / brachycephalic classification. <https://en.wikipedia.org/wiki/Cephalic_index>
11. An Anthropometric Study of the Facial Index (facial index classes; M 88.15 / F 85.74). <https://pmc.ncbi.nlm.nih.gov/articles/PMC13399996>
12. Cheekbone prominence and midface width ratios (bitemporal/bizygomatic 75-85%; bigonial/bizygomatic 70-82%). <https://facialharmonyai.com/blog/cheekbone-prominence-bizygomatic-ratio>
13. Historical Tools of Anthropometric Facial Assessment: the neoclassical canons (systematic review). *Aesthetic Surgery Journal* 2022. <https://academic.oup.com/asj/article/42/1/NP1/6369423>
14. Geniole, Denson, Dixson, Carré & McCormick (2015). *Evidence from Meta-Analyses of the Facial Width-to-Height Ratio as an Evolved Cue of Threat.* PLOS ONE (sex difference d = .11, n = 10,853; fWHR related to BMI, r = .31). <https://pmc.ncbi.nlm.nih.gov/articles/PMC4504483/>
15. Kramer (2015). *Facial Width-to-Height Ratio in a Large Sample of Commonwealth Games Athletes.* Evolutionary Psychology (n = 3,481; men 1.98, women 1.92, d = 0.31; difference not significant after controlling for body size). <https://pmc.ncbi.nlm.nih.gov/articles/PMC10430000/>
16. *Tracking sexual dimorphism of facial width-to-height ratio across the lifespan: implications for perceived aggressiveness.* Royal Society Open Science (2022; 17,607 passport images; male larger only in young adulthood, reversed after age 48). <https://doi.org/10.1098/rsos.211500>

---

First published 2026-10-02, updated 2026-10-18. Web page: https://www.mackenziebowes.com/research/facial-character-parameters Data as JSON: https://www.mackenziebowes.com/research/facial-character-parameters/parameters.json
