Oval face shape, and why detectors say it too often
Longer than wide, widest at the cheekbones, tapering gently to a rounded chin with no hard corners anywhere. Every clause in that sentence rules something out rather than pointing something out. No hard corners, so not square. Longer than wide, so not round. Tapering gently, so not heart and not diamond. Everything is in the middle of its range. Oval is the shape defined by not being any of the others. That is worth knowing before you treat an oval result as a finding about your face.
A residual category catches whatever is left
Five of the six shapes are defined by something being unusual: a corner at the jaw, a face longer than it should be, a forehead narrower than the cheekbones, a chin that comes to a point. Oval is defined by none of those being true. So any face without a standout feature lands there, and most faces do not have a standout feature.
Of 43 distinct faces run through the detector this site ships, 15 came back oval. That is the most of any shape, and nearly four times the four that came back oblong. Our prototype for oval is, in effect, the average of all the others, so a face that is average in every direction is closest to it by construction.
What that means for you: an oval result carries less information than the other five. The other shapes tell you something specific was measured. Oval tells you nothing was.
Oval borders everything, which is the problem
Our detector writes a result as a pair, oval/round rather than one label, when the two closest shapes are within a set margin of each other. Eight of the 43 faces came back as pairs. Every one of those eight pairs contained oval: four with round, three with heart, one with diamond.
A shape that borders one neighbour is a category. A shape that borders all of them is a gap in the map. That is not a flattering thing for us to publish about our own classifier, but it follows directly from oval being the residual, and any tool built on the same six labels will have the same property whether or not it reports it.
One more sign of the same thing. For every other shape there is a single measurement that does most of the ruling out — the forehead for diamond, the jaw for square, length against width for round and for oblong. Oval is the only shape with no dominant cause: the forehead and the jaw each account for exactly 16 of the 43, a dead tie. Nothing is deciding oval, which is another way of saying oval is what is decided last. The full grid is on the face shape chart, and the neighbour it collides with most has a boundary story of its own.
Now measure your own
Reading a description and holding your own face up against it is guesswork, and the descriptions overlap on purpose, because real faces do. The tool measures the four lengths in the diagram plus the jaw angle, in your browser, and shows you how close you sit to every shape rather than picking one and hiding the rest.
Carried down from the section above: if this returns oval, read it as nothing standing out rather than as a match. The measurements underneath it are the part worth looking at.
Front-facing photo, even light, hair off the forehead
Your four measurements will appear here, so you can see which shape they sit closest to.
Oval face shape questions
What is an oval face shape?
+
Longer than it is wide, widest at the cheekbones, tapering gently to a rounded chin with no hard corners. Read that list again and notice what it is doing: every clause rules something out rather than pointing something out. No corner, so not square. Not as wide as it is long, so not round. Tapering gently, so not heart and not diamond. Oval is the shape defined by not being any of the others.
Why do face shape tools keep saying oval?
+
Because a residual category catches whatever is left, and most faces do not have a standout feature. In our own measurements, 15 of 43 distinct faces came back oval, more than any other shape and nearly four times the number that came back oblong. If a tool has to return one label and the face is unremarkable, oval is where it lands. That is a property of the taxonomy rather than a fault in any one tool, but it does mean an oval result carries less information than the other five.
Is an oval face shape rare?
+
The opposite, at least as tools report it, though nobody can give you a real prevalence figure because no peer-reviewed distribution exists for these categories. What we can tell you is that oval was the most common result across the 43 faces we have measured, and that this says as much about where our thresholds sit as about the faces.
Can a face be between oval and something else?
+
Constantly, and that is the most telling number on this page. Our detector writes a result as a pair when the two closest shapes are within a set margin. Eight of 43 faces came back as pairs, and every single one of those pairs included oval. A category that borders everything is doing less work than a category that borders one thing.