Welcome to rarity.studio!

Generative art rarity rankings and analytics



Verify with official sources before any action! The accuracy is limited, only traits provided in metadata are suported and reflected in the ratings.

Art Blocks (Upcoming mints)



Hashmasks



MoonCats + WrapID Helper + Mint Order



Two rarity measures - an explanation

This is the Opensea checkbox approach.
Rarity of an individual trait determines the overall ranking for the whole item.

We compute rarity within each trait and, then, for each item, we identify its 1st,2nd,... rarest trait. The ranking is produced by evaluating the rarity of item's "rarest traits." If there is a draw, we compare 2nd rarest trait and so on. Therefore, rarity comes from individual traits but is compared across them, e.g. palette rarity is compared to shading rarity.

This is not what we usually interpret as rarity in the NFT space, yet it is a more comprehensive measure.
This measure reflects rarity of all traits. However, humans are not able to construct a mental model that provides such an evaluation - there are just too many dimensions. Therefore, the resulting ranking may seem unnatural, yet it is more elaborate than the other one.

Here's a nice explanation by Adam.

Example: Compare Archetypes #432 and #304. Archetype #432 has Truce Palette. Ther are only 4 Truces and, thus, it ranks as 9th in the single trait rating. However, in other attributes, it is pretty common. So, in the statistical rating it ranks low - as 243rd. That is just behind Archetype #304 which does not stand out in any particular way. However, it outperforms #432 in combining its "other traits". The fact that they rank almost the same here means that, if the mint were to happen again (with the observed probabilities), you have almost the same likelihood of minting for either one. Of course, they can't be more different. #304 ranks as 578th in the single trait ranking - that's pretty bad.