Atticus — 134/131

P(this card, per pack) = the published odds of its RARITY appearing, split evenly across the cards printed at that rarity in the set. Expected packs is 1/p, the mean of the geometric distribution. The range plots expected pack spend ÷ the single's delivered price on a log axis centred on parity. Pack price is a sealed product's own price divided by its stated pack count, and is computed only where that count carries a source. The single is the cheapest delivered ask (item + required shipping) across venues — the same figure FAIR VALUE prices against.

The Pokémon Company has never published physical booster odds in any region — the only official drop-rate page covers the digital TCG Live packs and lists no probabilities at all. Every figure here is a hand-entered citation of a published pack-opening study, and each one names its source and sample on hover.

Sealed product from Prismatic Evolutions — the set this card was printed in. We track which set a product belongs to, not which cards are inside it. Appreciation is each product’s last stored price against its own recorded MSRP, in percent. The card’s call is the MEDIAN of those percentages across the products that have an MSRP — a median rather than a mean, so one outlier SKU cannot carry a whole set — and a product with no MSRP on file is left out of that aggregate rather than counted as zero; it still lists below with its price. MSRP is somebody’s assertion rather than a published dataset, so it is shown only where a source is recorded against it, and that source is on the row’s hover. The bands are a legend for the percentage the card prints beside them: within ±5% is at MSRP, which is this repo’s own fair-value tolerance; up to +25% is a slight premium; up to +100% a significant one; at or above +100% the price has at least doubled and the figure is stated as a multiple, because a four-digit percentage is not a number anybody reads. Each price line takes a day’s observed low/high sale range and plots its midpoint; on a day with no recorded sale it carries the provider’s own daily guide value instead — a quoted price, not a transaction — and those days are drawn dashed, never described as trades. Every line is indexed to its own first stored day = 100, so the chart compares appreciation between products rather than price levels: a booster pack and a booster box cannot honestly share a dollar axis.

Each row is one tracked retailer's most recent reading of one of the set's sealed products, at that retailer's own SKU. A restock is an observed OUT_OF_STOCK → IN_STOCK transition in that SKU's reading history, never a change we inferred from a price or a gap. Shelf prices are pre-tax and exclude shipping, so they are not delivered prices and are never fed into a floor.

Each reading takes the venue's own listing count for this card in this condition — the whole book — at that instant. The direction is classified by the platform's absorption reading over that series: a book drawn down with demand behind it reads as accumulation, one that refills after a drawdown reads as distribution. Copies and sellers come from the cheapest asks we sample, not from the book, and are labelled as such. Sold units are confirmed prints from the live tape over the same span, and are shown only when the tape reaches back past the first reading.

The direction is classified over the platform-standard 30-day trailing window. That window is a scope, not a history: it contains exactly the readings counted on this card and no others — and only readings that counted the venue’s own book, never a sampled one trended against a counted one.