Data methodology

• Marketplace: eBay US.
• Currency: USD.
• Sellers: located in the United States. No ship-to ZIP in this version, so shipping is national as listed.
• Price statistics use active fixed-price listings only.
• Displayed cost = item asking price + listed shipping. Taxes are excluded.
• Each page states the retrieval time and the number of listings analyzed.

A current bid is neither an asking price nor an expected sale price. Auctions are shown on their own, ordered by end time, and are never mixed into the median, percentiles or histogram.

Completed-sale data is served by eBay's Marketplace Insights API, which is restricted. We do not have programmatic access to it, so we do not compute sold medians or sold-price history. Instead each card page links out to eBay's own sold listings, where you can inspect completed sales directly.

The distribution chart uses conventional equal-width dollar buckets with P20, median and P80 markers drawn on top. Equal-population quintile bars are avoided deliberately — five bars of identical height would tell you nothing. The axis is bounded near the 95th percentile so a single extreme listing cannot flatten the chart, but those listings stay counted in the final bucket and in every statistic. Expensive listings are never silently dropped.

A listing is included when all of the following hold:

• It is in eBay's Pokémon Individual Cards category.
• Its title contains the card name and, where we have one, the card number.
• For a graded view, the title states the selected grading company and grade. For the raw view, the title claims no grade at all.

A listing is excluded when it is a lot, bundle, "pick your card", mystery item, proxy, custom, reprint, sticker, empty slab or accessory, or when it is a different card, number or grade.

A listing is set aside as uncertain when the grade cannot be established from the listing. Uncertain listings are counted and shown, but never included in the statistics. A technically perfect median over wrongly matched cards is worthless.

eBay reports its own result totals as an indicator only, and one listing can contain more than one card. We therefore say "matching active listings", not "cards for sale".

The comparison table covers one grading company at a time, plus raw. That keeps the number of eBay API calls per page flat. The "PSA 10 asking premium" is the PSA 10 median ask divided by the PSA 9 median ask, minus one. The "share of graded listings" is PSA 10 listings divided by all graded listings in the table. Both are statistics about current supply and asking prices — not gem rates, not liquidity, not predicted prices.

Responses are cached per card, grader and grade so repeat visits do not hammer eBay: roughly 30 minutes for fixed-price statistics, 10 minutes for the auction board, and an hour for the grade comparison. The "Refresh current data" button on each card page forces a live pull.

Population figures come from PSA's own public population reports and are dated on the page. The gem rate calculator is arithmetic you control: PSA 10s divided by total graded. We link to the PSA report so you can verify every input.

Matching is the whole product, so we take reports seriously. Use the "Report incorrect listing" button on any card page, or the corrections page.