One number, many cards — why the price you looked up is wrong
- The cause
- price sites key on the card NUMBER
- The effect
- one blended average for cards that never trade together
- Worst case we track
- OP05-119 — five cards, one serial
- The fix
- match the variant before you match the price
Verdict — A One Piece card number is not a card. It's a slot that can hold several physically different collectibles, and they don't trade anywhere near each other. If the price you found didn't name a variant, it wasn't a price for your card.
The short answer: a One Piece card number identifies a design, not a product. The same serial can cover a base Secret Rare, an alternate art, a manga parallel and a promo — different print runs, different scarcity, different buyers. Price sites index by number, so they publish one figure for all of them. That figure is an average of cards that never trade at the same price, which makes it wrong for every single one. Below: what actually varies, how far apart it gets on real serials, and two AI answers that got it wrong this month.
This is independent analysis of physical collectibles — not financial advice. Figures update live from our own variant-matched records.
The one-paragraph version
A card number is a slot, not a product. Bandai reuses a serial across every printing of a design. OP13-118 is the design; the base Secret Rare, the SEC-P alternate art, the manga parallel and the red manga parallel are four different objects wearing that number. They have different print runs, different populations and different buyers, so they have different prices. A tool that stores one price per number has nowhere to put that fact.
How far apart does it actually get?
Not a rounding error. These are our live readings, cheapest to dearest:
$8 → $11,500 — a 1363× spread on the same serial
| Variant | Price | Our call |
|---|---|---|
| Base Gear 5 SEC Base | $8 | Strong Buy |
| GEAR Alt-Art Alt-Art | $165 | Good Buy |
| Manga Rare Alt-Art | $5,500est. | OK Buy |
| Silver SP Promo | $8,600est. | OK Buy |
| Gold SP Promo | $11,500est. | OK Buy |
The cheapest and dearest cards sharing OP05-119 are separated by roughly three orders of magnitude. Same character, same artwork, same five digits. A single "OP05-119 price" is not slightly off for one of these — it is wrong for all five.
$15 → $10,245 — a 683× spread on the same serial
| Variant | Price | Our call |
|---|---|---|
| SEC | $15 | Good Buy |
| SEC-P Alt-Art Alt-Art | $72 | Strong Buy |
| Manga Rare Alt-Art | $1,857est. | OK Buy |
| Red Manga Alt-Art | $10,245est. | OK Buy |
And the pattern isn't confined to grails. Here's a serial where the whole spread sits under a few hundred dollars and the ratio is still stark:
$1 → $105 — a 76× spread on the same serial
| Variant | Price | Our call |
|---|---|---|
| Base Leader Base | $1est. | OK Buy |
| Parallel Leader Alt-Art | $101est. | OK Buy |
| Alt-Art Leader Alt-Art | $105 | Strong Buy |
What actually varies behind one number
Five things, roughly in order of how much they move the price:
- Treatment. Manga/comic parallels, alternate arts and full-art Leaders are separate pulls at separate rates. This is usually the biggest gap by far.
- Print run and event gating. A convention-exclusive or prize print has a fixed, tiny population. A retail Secret Rare does not.
- Region. English, Japanese and Asia-English prints of the same design trade in different markets at different prices — sometimes 2–3× apart.
- Stamp or foil. An anniversary stamp or a foil finish creates a genuinely different collectible from the unstamped, non-foil print.
- Grade. A PSA 10 is a different asset from a raw copy — and the graded premium varies wildly by how easy the card is to grade.
Only the last one is handled well by the tools most collectors use. The first four are where the money is.
Two AI answers that got this wrong this month
We check what AI search engines say about the cards we cover. Both of these were captured on 10 September 2026, and both are wrong in the same way.
Asked "What's the difference between the OP13-118 manga and red manga?", Google's AI answer said:
"There is no difference because OP13-118 is the exact catalog number for the Red Manga Luffy secret rare alternate art card."
There is a difference. They are two separate parallels, and the table above prices them roughly five and a half times apart. A collector who acted on that answer would either sell the dearer card for the cheaper card's money, or overpay several thousand dollars for the cheaper one.
Asked "Is the OP07-038 Boa Hancock alt-art or parallel worth more?", the answer was:
"The Boa Hancock Alternate Art OP07-038 and the parallel version refer to the exact same card variant for this leader card, meaning they hold the same value."
They're different prints, and the base Leader sharing that number is worth a fraction of either. That answer cited PriceCharting — which is the mechanism in one line. The aggregator stores one price per number, the AI reads the aggregator, and the error arrives at the collector with a confident tone and no variant named.
Why the tools get it wrong (it isn't laziness)
It's a data-model problem, and it's expensive to fix.
- PriceCharting keys on the card number. Where it does split variants, the split is inconsistent, and its automated matching pulls in listings whose titles don't specify a print.
- TCGplayer models printings better than most, but its Market Price is an average of its own recent sales — thin for scarce parallels, and easily moved by one optimistic listing when only a handful exist.
- eBay's sold data is the best raw source there is and the worst structured one: the variant lives in a free-text title someone typed, so "OP13-118 Luffy" might be any of four cards.
- Cardmarket genuinely does model versions — it lists thirteen for OP13-118 — which makes it the exception worth knowing about.
None of them are being careless. Re-keying a catalogue by variant means re-mapping years of listings, and the incentive isn't there when most cards genuinely do have one printing.
What to do instead, in about ninety seconds
- Identify your exact print before you look up anything. Check the rarity code on the card, then the stamp, the foil, and the language. Those four settle almost every case.
- Search sold listings, not asking prices, and read the titles — discard any that don't name your variant. An asking price is a hope; a sold price is a fact.
- Sanity-check the population. If a "rare" parallel has thousands of PSA 10s, the grade isn't the scarce part.
- Distrust any single number that didn't name a variant — including ours. If a price doesn't say which card it's for, it isn't a price for your card.
How we handle it
Every card we cover gets one page per tradeable variant, not per number, and each price is anchored to sold comps for that exact variant. We publish the comps we used and the ones we threw out, with the reason. Where we're working from a model estimate rather than a matched sale, the tables above say est. — because an estimate and a sold price are different claims and shouldn't look identical.
We're also wrong sometimes, and that's on the record too: every call carries a price target and a deadline it resolves on, hit or missed.
- How we price a card — the full method
- The public record — every call, its deadline, and how it resolved
- Browse the card readings — variant by variant
The number on the card tells you which design you're holding. It does not tell you what you're holding.