Face preference test: how your choices shape a result
This face preference test fits a small pairwise model to faces you selected and passed within each completed crowd.
Published · Independent guide and face preference test
What a face preference test records
The face preference test saves a pick as soon as you tap a portrait, then replaces that face in the same position. You can select multiple faces and undo the latest pick. Moving to the next round records the faces left on screen as passes. These server-owned records support private reopening and preference modelling; browser-supplied scores never authorize a paid feature or overwrite your choices.
How a face preference test compares your choices
This face preference test compares selected and passed portraits using reviewed tags. Each crowd contributes equally, so choosing more faces does not give one crowd greater influence. A regularized pairwise model looks for recurring differences in hair, expressions, glasses and contours. Different selections can therefore produce different summaries from the same collection; the labels remain tentative descriptions of your choices.
What a face preference test cannot conclude
A face preference test can describe the collection’s selections without claiming to explain all real-world attraction. A small correlated sample cannot establish why you made every choice. The report presents selected examples and tentative feature patterns. No population rating dataset is used, so we do not invent average attractiveness, rarity, hidden-gem percentiles or claims that another person would choose you back.
Why face preference test portraits need consistent framing
A face preference test is easier to interpret when backgrounds, framing and light are reasonably consistent. Otherwise a landscape, dramatic pose or unusual accessory can dominate the choice. Consistency reduces distraction; it does not remove every correlation between facial features and styling.
How a face preference test chooses later crowds
Later face preference test rounds mix model-ranked candidates with exploratory examples. The first 10 rounds are free. One Credit unlocks five more, up to 50 per test. Unseen portraits are prioritized; previously passed faces can return near the end of the collection. A paid five-round grant is recorded once and its rounds reopen without a second charge. More choices do not guarantee a higher prediction score.
Why a face preference test can show mixed patterns
A face preference test may connect several different details with your selections. Hair, glasses, contours and expressions can occur together. Inspect the actual examples before deciding which detail mattered to you; the model cannot disentangle every influence from a short session.
Sources & scope
For the original service, see MyType’s own explanation and current terms. Our test at mytypefyi.app has its own features, data handling and pricing. Features and prices at other services can change.
