What the term actually describes
Temporal hollowing is the loss of fat and, sometimes, temporalis muscle volume at the temples — the region between the eyes and the hairline, at the sides of the forehead. What it looks like from the outside:
- More hollowing — a sunken, indented temple that shows more of the underlying skull contour, reading as a thinner, more skeletal upper face.
- Less hollowing — a smoother, fuller temple line with no visible indentation.
It's a depth and volume trait that lives in the soft tissue itself — driven by aging, weight loss, and genetics — not a distance between two points on the face's outline, which is the category every ratio Mogalytics computes falls into.
Why no landmark-geometry tool can read it
Face-landmark models place points on the visible surface of the face and compute distances or angles between those points. That approach works for a width, a height, or an angle — it does not work for temporal hollowing, because:
- Fat-pad and muscle volume changes how the surface between the hairline and the brow curves, not where any landmark itself sits.
- Two faces can share the exact same bitemporal-to-bizygomatic width ratio and still carry very different amounts of temple hollowing underneath.
- Reading depth this way generally needs a 3D scan or a side-lit still that reveals shadow contour — a flat set of 2D landmark coordinates has no depth channel to draw on.
This is a different limitation from the angle and distance problems covered in the full breakdown of why face-rating scores swing between scans — those are about a measurable geometry ratio getting distorted by pose; this is about a trait that isn't geometry-ratio-shaped to begin with.
Where this fits next to what Mogalytics actually reads
Mogalytics does not compute or display a temporal hollowing number, and it can't — the trait falls outside what any landmark-geometry ratio measures, and Mogalytics doesn't track the hairline or temple as a landmark at all. The closest existing page by region is:
- Bitemporal width — a width proportion (temple width divided by cheekbone width) built from the same general region, but it's a distance between two points, not a measure of fat or muscle depth.
A wide or narrow bitemporal reading says nothing about how hollow the temple looks — a wide-templed face can be either hollow or full, and so can a narrow-templed one. Mogalytics makes no claim of parity between the two.
Related reads
If temple or upper-face structure is what you're auditing, these are the closest reads among what this site covers:
- Bitemporal width — the closest camera-read analog by region, a width proportion rather than a volume read.
- Buccal fat & hollow cheeks — the same "soft-tissue volume, not a ratio" limitation, one region down at the lower cheek.
- Ogee curve — another contour trait shaped partly by soft-tissue depth rather than frontal landmark geometry.
- How does Omoggle's scoring work? — the publicly described mechanics this term shows up in.