Facial Proportions, Thirds and FWHR: Definitions, Evidence and Limits
Proportion measures compare one part of the face to another. They sound exact. In a frontal selfie they are approximations.
Where the definitions come from
Clinical anthropometry measures facial dimensions between landmarks located on the person, such as the hairline, nasion, subnasale and menton. Farkas (1994) is the standard bibliographic reference for head and face anthropometry. We have not used it to substantiate any Selfbaus landmark definition. The classic "facial thirds" assume a physical landmark you can locate.
What a frontal mesh offers instead
Selfbaus 0.2.0 measures between mesh points, divided by iris-center distance. Face width runs between oval points 234 and 454; face height between oval top point 10 and chin point 152. Three vertical segments are measured: 168 (between the brows) to 10, nasal base (point 2) to 168, and chin (152) to nasal base (2). Point 10 is the model's face-oval top, not the hairline, so the upper segment is not a clinical upper third. Point 6, not 2, sits at the nasal bridge; point 2 is the nasal base. Width points are contour estimates, not bony cheekbone width. We report these as mesh-defined proportions, named by their indices, and do not call them facial thirds.
FWHR
Facial width-to-height ratio (FWHR) compares face width to a height measure. Weston, Friday and Liò (2007) reported that the height of the upper face, between lip and brow, develops differently in males and females at puberty, independent of body size. Carré and McCormick (2008) built on a sex difference in facial metrics and reported, in three studies, associations between FWHR and trait dominance and aggression. A meta-analysis found a small but significant positive relationship between men's FWHR and aggression, and notes that other work failed to replicate the link (Haselhuhn et al., 2015). These are group-level associations in specific samples, measured with defined landmarks. They are not a basis for judging an individual, and they say nothing about attractiveness. Our mesh width and height use different landmarks and a casual capture, so we cannot transfer those findings to them.
What we do and do not say
Selfbaus does not present an FWHR value or interpret width against height. The three other display-only ratios it shows (eye spacing, mouth width, nasal contour width) are described in the iris-distance article. We report each ratio as measured in this photo, with the definition named. We do not give a "good" range, a percentile or an ideal, because we have no validated reference population for this method.
References
- Farkas, L. G. (1994). Anthropometry of the Head and Face. Raven Press. Bibliographic context only; not read for this article.
- Weston, E. M., Friday, A. E., & Liò, P. (2007). Biometric evidence that sexual selection has shaped the hominin face. PLoS ONE. https://journals.plos.org/plosone/article?id=10.1371%2Fjournal.pone.0000710
- Carré, J. M., & McCormick, C. M. (2008). In your face: facial metrics predict aggressive behaviour. Proc. R. Soc. B. https://doi.org/10.1098/rspb.2008.0873
- Haselhuhn, M. P., et al. (2015). Men's facial width-to-height ratio predicts aggression: a meta-analysis. PLoS ONE. https://journals.plos.org/plosone/article?id=10.1371%2Fjournal.pone.0122637