How accurate are we?
Every price here is a completed sale — what somebody actually paid — never a seller's asking price. That is the whole product. This page is the honest audit of it: how much of the game we can price, how fresh those numbers are, and what we deliberately refuse to print.
How much of the game do we actually price? the denominator first, not the flattering number
Of the 257 cards in our catalogue, 126 (49%) carry a published price. Only 47 (18%) rest on a deep enough sample — 4 or more completed sales — to be shown as a plain number. Another 79 sit on a thin sample of 1–3 sales and are shown only with that qualifier attached. The remaining 131 have no verified sold at all and say exactly that: “No verified solds — tracking live.”
A price site that quotes a number for every card in a game this young is either using asking prices or filling gaps with a model. We do neither, and the cost is visible right here in the denominator. Browse every card we track, including the ones still waiting for their first sale.
How fresh is the number? recency is a separate axis from sample depth — we never fold them into one score
This snapshot was built 2026-08-03 from completed sales on eBay US, Mercari JP, Yahoo! Auctions JP. The sweep runs daily. If a build is ever more than 26 hours old — one daily cycle plus slack — every page carries a stale banner at the top saying so, rather than quietly serving you yesterday's market as though it were today's.
On each card we show recency as its own badge: FRESH means the newest real sale is within 14 days, STALE means the freshest sale we have is older than that, and RECENCY UNKNOWN means the marketplace never told us when the sale happened — so we do not claim it is recent. A card can be deeply sampled and stale at the same time; you get to see both facts.
What we refuse to print, and why the part no asking-price site can copy
- A thin price, bare. 79 cards rest on 1–3 sales. They never appear as a plain number anywhere on this site — not in a table, not in search, not in a share image, not in our structured data. They appear as “~$X — rough number, only a few sales”, or not at all.
- A number for a card nobody has sold. 131 cards have no verified completed sale. We could interpolate one from similar cards in a minute. We don't; they say “tracking live”.
- A median we don't trust. 17 price series across 16 cards are flagged suspect — the sales are so widely dispersed that a median would be a fiction. Those are marked on the page and withheld from the weekly report rather than dressed up as a price.
- A price move we can't back. 34 series held enough dated, deduplicated sales in both windows to be eligible for a price move this week; 1 moved enough (2%+) to be worth printing. The rest are not "flat" — they are simply not a story we can back.
- Asking prices, ever, as a headline. A card listed at $200 that never sells at $200 is worth what it last sold for. Active listings are hope; we price history.
- Numbers we don't have the data for. Graded (PSA) values, pack expected-value, and a JP↔US arbitrage spread are simply absent from this site — not estimated, not "coming soon" with a placeholder figure. They appear when the underlying sold data exists and not one day before.
Each of the refusals above is enforced in code, not by editorial discipline: the build fails rather than render a bare thin price. Every number that survives all of it links back to the individual sold listings behind it — click any dot on a card's price chart to open the receipt.
How often are we wrong? the honest answer today is “we can't tell you yet”
The number you actually want is: when we print $475, what does the next completed sale go for? We are not publishing an error rate until we can measure it properly — hand-verifying a random sample of published prices against their source sales, and tracking how our number compares to the next real sale. Publishing a flattering figure off a handful of cards would be exactly the kind of number this page exists to argue against.
What you can do in the meantime is check us directly, which is the whole point of the receipts: every price on this site links to the completed listings it was computed from. If our median disagrees with what you see, the evidence is one click away.
The full record
- Methodology — how a raw sold listing becomes a price: bootleg filtering, outlier fences, spike quarantine, dedup, retraction voiding, currency conversion at a dated rate.
- Alert-tool diagnostics — the graded performance of a private alert tool the site's operator runs. It has nothing to do with the prices here; it is published because we don't delete our own record.
- Analyst log — that tool's raw output, unedited, including the calls that aged badly. Mostly commentary on asking prices, which is not what this site is about.
- Market overview — the shape of what we can price, and the tracked-value floor built only from confident prices.
Tell us we're wrong the fastest correction loop this site has
If a price here disagrees with what you are seeing in the market, say so: contact@palworldprices.com. Point at the card and, if you can, at a sale we missed or one we should not have counted — that is a concrete fix, and corrections get published in the changelog like everything else.
The same address reaches the operator of this site for rights-holder inquiries, including anything to do with card images, which are shown here under the publisher's fan-content policy. There is no form and no intermediary — it goes to a person.
Coverage figures on this page are computed from the same snapshot every price on the site is built from — they cannot drift from what the pages show. Prices are informational estimates from public completed-sale data, not financial advice.