The desk's vocabulary

Buy at retail — the model expects it to clear 2×.

The desk's one verdict ladder. Every other call on the site — the Dossier, the item pages, the scanner — normalises into COP, WATCH or PASS, so a pill means the same thing everywhere.

Could go either way — watch the sellout before you commit.

The desk's one verdict ladder. Every other call on the site — the Dossier, the item pages, the scanner — normalises into COP, WATCH or PASS, so a pill means the same thing everywhere.

Sits at or under retail — skip it.

The desk's one verdict ladder. Every other call on the site — the Dossier, the item pages, the scanner — normalises into COP, WATCH or PASS, so a pill means the same thing everywhere.

Projected resale divided by retail, from public pre-drop signals — how hard the desk expects it to flip.

Where this drop ranks against every drop the desk is watching, on public interest — search, video, retailer breadth. A percentile, not an absolute.

The model's probability this clears 2× retail.

How many public signals the desk could read for this drop — search, video, wiki, retailer breadth. Coverage, not confidence.

The indexed median of recent eBay sold listings for this exact product, with how many sales it rests on.

The modeled peak of the post-drop premium — sell before it, or hold past it knowingly.

The retail release date. Preorders may open earlier.

A product's release into retail — the day (and often the minute) it goes on sale.

On DropForecast: Every page on the site hangs off a drop date: the Board lists them, the Oracle forecasts them, Sell-By counts from them.

Buying before the drop, at retail, from a retailer or the manufacturer's own store.

On DropForecast: For hyped sets the allocation is decided in the preorder window, which is why the Key dates block lists the preorder date separately from the street date when the wire has found one.

The manufacturer's or retailer's list price at the drop — the denominator of every multiple on the site.

On DropForecast: The projected multiple is projected resale divided by retail; 1× means it sells for what it cost.

What the product actually sells for on the secondary market after the drop — eBay, StockX, GOAT, local.

On DropForecast: DropForecast projects resale before the drop and indexes it after, from completed public sales.

Comparable completed sales for the exact product — what buyers paid, not what sellers are asking.

On DropForecast: The site's valuations rest on sold comps, never on live asks; the Sold comps link on a card goes to eBay's completed listings for that product.

The middle price of recent sold comps — half the sales cleared above it, half below.

On DropForecast: Shown as Sold median $X · n sales wherever the comp index has this exact product (see the metric above).

Resale divided by retail. A 2× multiple means it resells for twice what it cost at the drop.

On DropForecast: The projected multiple is the desk's headline number; 2× is the bar the Win odds are measured against.

To buy at retail — successfully, at the drop. "I copped a pair" means you got them at retail.

On DropForecast: COP is the desk's strongest verdict: buy at retail because the model expects it to clear 2×.

To resell a product you bought at retail for a profit, usually soon after the drop.

On DropForecast: The whole site is about what will flip and how hard — the projected multiple is the flip, in one number.

To keep a product past the drop rather than sell into the launch premium, betting the price rises later.

On DropForecast: Sell-By tells you what holding past the modeled peak costs; a hold is a choice you make knowingly, not by default.

Rip: open the box for the cards inside. Seal: keep it sealed as an investment.

On DropForecast: The Rip or Hold block runs the box math — the expected value of the pulls against the sealed price — for the sets we have odds sheets for.

The sealed formats of a card product. Hobby boxes are the shop-channel format with the best hits; blasters are the mass-retail format; a case is a factory pack of hobby boxes (typically 6–12).

On DropForecast: One product, one page: the formats are grouped into a product family so a hobby box, a jumbo box and a case do not become three near-duplicate pages.

The category you flip in — cards, sneakers, sealed TCG, LEGO, collectibles. Most resellers work one or two lanes.

On DropForecast: Drop Lanes on the Desk sorts the next 30 days by lane; Sell-By's decay priors are set per lane.

The date by which a flip should be sold — the modeled peak of the post-drop premium, before it burns off.

On DropForecast: Each Sell-By page dates the peak, the sell-by, an honest band around it, and what holding 60 days past it costs.

The official date a product reaches shelves — the manufacturer's release date, as opposed to the preorder date.

On DropForecast: Street dates come from manufacturer and retailer calendars read directly, never regexed out of prose.

A second wave of retail supply after the initial sellout.

On DropForecast: A restock is the fastest way a premium dies; the item pages carry a restock-risk band and the wire types restock stories as their own event.

How much of a limited product each retailer or shop is given to sell — set by the manufacturer, decided before the drop.

On DropForecast: Allocation-tight releases raise the scarcity side of the forecast; the scarcity flags on an Oracle page say whether one is confirmed.

The moment retail supply is gone — when a product can only be bought on the secondary market.

On DropForecast: A WATCH verdict means watch the sellout before you commit: a fast sellout is the first real signal the projected multiple will hold.

A product that does not flip — sells at or under retail after the drop, and sits.

On DropForecast: PASS is the desk's word for a projected brick: sits at or under retail — skip it.