Compute how rare a token is inside its collection — rank, top-% percentile and trait frequencies, with the working shown.
How rarity is scored#
For each trait a token carries, its rarity contribution is
N ÷ count — collection supply over how many tokens share that
value. A token’s score is the sum across its traits; ranks order scores
descending (ties share a rank) and top-% is rank ÷ supply.
- Relative, not absolute. Ranks only exist inside the dataset supplied — they say how rare a token is within this collection, never a market value.
- Incomplete data fails visibly. Tokens missing a trait type, or carrying values absent from the count table, are scored on what they carry and flagged — never silently “fixed”.
- Verified collections are governed. Records only appear below when sourced trait-count data has been verified against a public source with a verification date. Records that fail consistency checks (e.g. counts exceeding supply) are excluded rather than shown.
Verified collections#
Explore your own data#
Paste a collection record — supply, a
trait_counts table (trait type → value → token count) and the
tokens to rank. Everything runs in your browser; nothing is
sent anywhere.
Dataset: governed collections registry — 0 verified collection(s) at generation 2026-10-06. Engine is static and dependency-free; calculations identical to the documented method above.