Skip to main content
Claude and other large models have deep knowledge of Magic: The Gathering and decent knowledge of Pokemon. Knowledge of newer games — Lorcana, One Piece, Riftbound, Flesh and Blood, Star Wars Unlimited, Digimon, Pokemon Japan — is thinner, mostly because a lot of these games didn’t exist yet or weren’t well covered when the model was trained. If you sell one of these and your assistant sounds confident but vague, that’s why.

​
Step 1: Recognize the pattern

If your assistant says something like “Lorcana commons typically run $0.10–0.50,” that’s a guess dressed up as a fact, not something backed by your data. It can reliably reason about your actual data for any game — your prices, your movers, your sales, your rules — but general market claims for these games are exactly where it’s most likely to be wrong or outdated.

​
Step 2: Ground it before you ask

Open a chat with a snapshot instead of a question:
“I sell mostly Lorcana. Here’s a snapshot of my store — pull my inventory value, my recent sales, and my current pricing rules for Lorcana before we talk.”
Let it run those reads first. Every answer after that is sharper because it’s reasoning from your real numbers instead of generic priors about the game.

​
Step 3: Ask your real question

Once it’s grounded, ask what you actually wanted:
“Which Lorcana cards am I underpricing relative to my own sales history?”
That’s a question about your data, which the assistant answers well regardless of how thin its general Lorcana knowledge is.

​
Pokemon specifically: graded cards and thin market data

Pokemon gets extra help here. When TCGplayer’s own market price is missing or zero, especially for graded slabs, Hoard falls back to PriceCharting data to fill the gap rather than leaving you with no number at all. So a “the market price is $0” answer for a Pokemon card usually means the underlying pricing already accounted for that, not that the card is worthless — ask your assistant which price source it used if the number looks off.

​
The habit worth keeping

This is the same grounding habit that helps with any cross-game question, not just newer games: let your assistant read before it reasons, and treat any unprompted claim about market norms as a guess until it’s backed by your own numbers.