Where the data comes from
All aggregates come from Armoiria's opt-in closet-benchmark cohorts (in-app feature, off by default). Participants explicitly choose to contribute anonymized aggregates; no names, no photos, no garment-level rows leave the aggregate pipeline.
The k-anonymity gate
No statistic publishes from a cohort smaller than 100 participants. If a slice (by closet size, region, or season) falls under the gate, that slice is withheld entirely rather than rounded or fuzzed.
What is excluded
Free-text fields, photos, and any identifier are never part of the study. Wears are counted from explicit outfit logs only — no inferred or modeled wear counts enter an aggregate.
Wear counting
A wear is one logged outfit appearance of a garment on one day; repeat outfits on the same day count once per garment. The study year runs January 1 to December 31 in the participant's local timezone.
Reproducibility
Each edition archives its cohort size, gate status, and stat definitions alongside the numbers, so year-over-year comparisons can be audited against the same definitions.