I have seen the same Pokémon card sale appear on more than one website. At a glance, it looks like two pieces of evidence: one result on a marketplace and another in a price tracker. But if the tracker imported the marketplace transaction, there was only one sale.
That distinction matters a lot when you are trying to value a scarce card.
Three numbers that should never be mixed casually
Asking price: What somebody wants. It can stay listed for months.
Completed transaction: What a buyer and seller apparently agreed to, although some platforms obscure accepted offers or other details.
Automated market estimate: A model or summary based on data the service has access to. Useful for orientation, but not an executable offer.
Those numbers answer different questions. I would never average them into one “market value.”
A small example
Imagine you're researching a PSA 10 card and find these records:
| Source | Displayed result | What to check |
|---|---|---|
| Marketplace completed listings | $1,250 on October 9 | Exact card, sale format, shipping |
| Price tracking site | $1,250 on October 9 | Could be the same marketplace sale |
| Grading-service sales history | $1,250 on October 9 | Could also reference that transaction |
| Another auction | $1,100 on October 18 | Possibly a separate comparison |
You have four records, but perhaps only two independent transactions.
That does not make the data useless. It means you should group records by likely underlying sale. Compare dates, photos, listing titles and any visible identifying details. For graded cards, certification identifiers can help you check for resales, but keep your own collection identifiers private when sharing research publicly.
What if the prices disagree?
First check whether you are looking at the same printing. A Japanese promo, English set card, World Championships deck card and later reprint can have remarkably similar artwork. Our set-code guide helps with that first pass.
Then look at condition, grade, auction timing and shipping terms. A raw card described as mint is not equivalent to a PSA 10. A fixed-price sale may reflect a buyer who wanted the card immediately; a poorly timed auction may not.
If you only find two trustworthy comps, say so. A range with an explanation is more credible than an average with two decimal places.
My research note format
I would keep one row per *independent sale*, then attach links to every site reporting it. Add a column for “possible duplicate or resale” and a short explanation. This makes it easier to revisit the research months later without counting the same transaction twice.
The goal isn't to make the number bigger or smaller. It's to understand how much real market evidence you actually have.