PSLChad, Chadlite or HTN? Understand the PSL scale.
PSL is a community face-rating framework commonly discussed on a 1–8 scale. Explore the tier names, what people mean by them, and how they differ from a measured photo result.
The seven-tier ladder
High Tier Normie / HTN
A community band described as having clear contours and cohesive facial proportions.
See where your score falls
Use your photo to get a PSL score and explore the proportions behind it.
Get my PSL reference scoreSee a full sample reportOur four measured dimensions
Harmony
Evaluates how facial features work together as a cohesive whole, including facial thirds, eye spacing, width-to-height ratio (FWHR), and midface compactness.
Dimorphism
Dimorphism uses the measured widths of the jaw and chin to describe their relative shape.
Angularity
Evaluates the sharpness of facial bone structure, including jaw angle definition, cheekbone contours, and outer eye corner canthal tilt.
Miscellaneous
Miscellaneous combines additional geometric features, including bilateral balance and eye and mouth proportions. It is not a standalone symmetry score.
How our personal reference score is calculated
AttractiveNet supplies the overall PSL score from a 350 × 350 RGB input prepared by your browser. Our Cloudflare service runs AttractiveNet and MediaPipe; the browser uses the returned data to calculate and display Harmony, Dimorphism, Angularity and Miscellaneous.
- PSL score
S = clip(1 + 1.75 × (AttractiveNet − 1), 1, 8)- Geometry adjustment
δᵢ = 9 × rᵢ- Anchored dimension
Dᵢ = clip(S + δᵢ − 0.5 × Σ(wⱼ × δⱼ), 1, 8)- Dimension weights
w = [0.35, 0.20, 0.25, 0.20]
The 16 measurements enter fixed feature curves. Their weighted contributions form each group's relative value r. The four dimensions are anchored to the overall score, with room for strengths and weaker areas; their average is not forced to equal PSL. Measurement explanations, highlights and styling priorities use these same weighted curves.
Green segments mark ranges with a positive contribution in the current model's frozen feature curve, not a population-wide ideal range. The marker is your photo measurement; its spread models point-placement sensitivity, not a confidence interval for lighting or pose.
Each displayed contribution is 9 × feature-curve value × consistency weight. Contributions sum within their group before the shared weighted anchoring correction and 1–8 clipping. Highlights and priorities use these same contributions. Reference segments are limited to the model's recorded feature range.
A valid AttractiveNet result supplies PSL; all 16 valid geometry readings complete the four dimensions. The formula library shows each measurement, its positive-contribution segments and the shared scoring method.
Can styling change your PSL score?
Hair, grooming, frames, and posture (softmaxxing) directly enhance visual proportions and overall aesthetic presentation. Strategic haircuts frame facial width, and proper lighting highlights bone structure definition.