How the numbers are made

Shelves. We read the online catalogs of 441 used and retro game stores in the US and Canada every 6–48 hours: every listing, its price, condition and stock, and when it disappears.

Matching. Each listing is matched to one catalog game (title, platform, edition, region). Listings we are unsure about — an unverified edition, unclear condition, a placeholder price, damage noted in the title, a graded copy — are left out of every number.

Other stores. For a game in a given condition (loose, game + box, complete, sealed), we take each store's cheapest copy and compare yours with the median of the other stores' asks — never including your own. Sister sites of one shop count once. We compare only when at least 4 other stores carry the game, and leave out asks more than 4× off the others' median (almost always a mismatch or a lot).

No store is named. Other stores appear only as a group of 4 or more other owners (your own sister stores never count), with their prices rounded to the dollar; nearby comparisons widen the area until a group is that large.

This week's changes. One list per store, biggest dollar change first; the first 25 are the week's work and the rest are in the list and the CSV. A listing gets a raise when it is under 0.8× what 5+ other stores ask and under the cheaper quarter of their asks (suggested: their median), when it is under 85% of what the game sold for online (5+ sales in 90 days; suggested: 95% of that), or when the online price rose 10%+ in 3 months. It gets a lower when it is over 1.5× what other stores ask and above the pricier quarter, or over 1.5× what it sells for online. A store that prices above the market on purpose is judged against its own level: we only point out listings far above even that. A change must move the price by $3 and 10% or more; suggested prices are rounded to whole dollars (or .99 / .95 when your prices end that way). The dots show how sure we are: ●●○medium confidence 5–9 stores, ●●●high confidence 10 or more stores or online sales that agree.

Your price. Your median price is the median, over your compared listings, of your price divided by the other stores' median. Groups (nearby, state, network) show their typical store. Rank places you among every store with 20+ compared listings; below that we don't rank.

What stores pay. Some stores publish a buylist — what they pay in cash and store credit for each game and condition. The price check shows the median across the stores that list the game.

Sold online. Completed online sales of each game, cleaned by us: only the same game, platform and US release; sets, lots, graded copies, reproductions, imports and box / manual-only sales are left out, and the condition is read from the listing's title (sales that don't state it are left out of the figures). We show the median of the last 90 days (6 or 12 months only when sales are thin, and we say so), how many sales it comes from, and the 3-month trend. A sold figure that disagrees sharply with what stores ask (often a remake or another edition) is held back.

Offers at the counter. Your cash and credit percentages of the sold median when 5+ recent sales back it, otherwise of other stores' median ask; rounded down, and never more than 10% above what stores pay for it (their published buylists).

Leaving shelves. We also record when each listing goes from in stock to sold out at the stores we read; as that history builds, reports add how fast games sell at a given price.

Market prices last rebuilt: 2026-10-21 19:45 UTC.