Traits and odds
Every Feline's page states the odds of what you are looking at. This page explains where those numbers come from, how to read them, and, at more length, where they do not apply. That last part is where collectors get hurt.
Frozen generator weights, not market math
When the recipe was frozen, every trait carried a weight: the probability the generator would draw it. Those weights are the odds published on each Feline's card, phrased the way a person reads them, for example "3.2 in every 1,000 felines".
They are evidence from the generator, recorded before the first Feline was drawn. They never change, no matter what mints, sells, or moves.
What the card shows
- Per-trait odds: the frozen weight of each trait on the plate.
- Setting probability: the odds of the combination, when every component's weight is on record.
- A tier word as a reading aid, derived from the frozen weights, never replacing the figure.
The one-in number is about the recipe, not the collection
The figure printed as "1 in 360,577,338" is the frozen generator's odds of drawing that exact combination of all ten traits: ten weights, multiplied together.
Read that with the supply in mind. The collection is at most 3,333 Felines. A one-in-360-million combination is therefore not a claim that 360 million cats exist and yours is the only one. It is a statement about a single roll of a ten-part recipe. Almost every Feline in a ten-trait collection carries a combination whose odds run into the millions, because that is what multiplying ten probabilities does.
What the figure is good for: comparing one setting with another on a fixed scale that was recorded before the first Feline was drawn and cannot be recomputed later.
What the figure is not good for: estimating scarcity, ranking Felines, or reasoning about price.
The tier word is shorthand, not a second opinion
Each confirmed Feline also carries a tier word, for people who do not want to read a nine-digit number. One thing about it matters before you use it to compare two Felines: the tier word is not a function of the one-in figure. Across the confirmed Felines published so far, the ranges the tiers cover overlap, so a Feline in one tier can carry a larger one-in figure than a Feline in a tier that sounds rarer.
The tier and the figure are two readings of the same frozen weights. The figure is the precise one.
| Tier | Felines | Lowest odds seen | Highest odds seen |
|---|---|---|---|
| Rare | 29 | 1 in 531,203 | 1 in 111,274,730 |
| Uncommon | 21 | 1 in 624,225 | 1 in 37,642,689 |
| Epic | 11 | 1 in 2,860,552 | 1 in 360,577,338 |
The ranges overlap. That is the point of the paragraph above this table, and it is why the figure is the precise reading and the word is the shorthand.
How this differs from marketplace rarity
Marketplace rarity is usually a share of whatever set a venue happens to list, recomputed as listings change. It answers "how common is this among what we can see today". The odds here answer a fixed question: "how likely was the generator to draw this, ever". Same word, different fact. The house publishes the fixed one because it is the one that cannot be gamed, drifted, or quietly recomputed.
What the house will not publish
- A rarity rank. A rank is a comparison, a comparison invites a price, and a price is a claim this product does not make.
- Live scarcity counts ("only 4 left with this hat"). The count of served felines moves as the mint fills and would tell you something different every week about an object that has not changed.
Checking the odds yourself
The recipe digest on every Feline's card commits the house to one exact recipe, weights included. The odds shown for a trait are identical on every Feline that carries it; any inconsistency would be visible across the public catalogue.