What's a normal TCGplayer refund rate?

This page doesn't have the numbers yet. We're saying that up front instead of padding it with a made-up percentage, because a benchmark you can't trust is worse than no benchmark at all.

What we can tell you is what we're measuring, why we're positioned to measure it, and the methodology we're committing to before a single result comes in. Writing the counting rules down first is what keeps a benchmark from quietly bending toward whichever number looks best once someone's staring at it.

Refund rates: everyone guesses, nobody compares

Every seller has a gut feeling about whether their refund rate is normal. Almost nobody has anything to compare it to. Search for a real TCGplayer refund or dispute rate benchmark, broken down by condition, game, or price, and you find policy documentation and forum anecdotes, not data. Breaking your own dispute rate down by set, condition, and price band covers how to see your own numbers today. It doesn't tell you whether those numbers are high, low, or ordinary for a store like yours, because nobody has published the comparison point yet.

Order history at scale is the missing ingredient

Hoard sits on both sides of the data this benchmark needs: a public TCG price and catalog layer, and our own sellers' order and refund history, anonymised, across live stores actually processing disputes day to day. Most tools in this category see one store at a time, if they see refund data at all. Seeing it across many stores, in a form that can be grouped by condition, game, and price band, is what makes a benchmark like this possible for us and not for a tool that only ever touches one seller's numbers.

Four cuts the benchmark will report

The study is not complete. This is the methodology as planned, not a set of findings:

• Average refund and dispute rate as a share of orders. Refunds and disputes divided by closed orders, across the seller base, as the headline figure.
• Broken down by condition tier. Near mint, lightly played, moderately played, and heavily played, to see whether disputes cluster at a particular grading boundary.
• Broken down by game. Magic, Pokemon, Yu-Gi-Oh, One Piece, Lorcana, and the rest of the catalog, since dispute behavior plausibly differs by game and by buyer base.
• Broken down by price band. Sub-$5, $5 to $25, and $25-plus, since a dispute on a cheap card and a dispute on an expensive one carry very different dollar weight even at the same rate.

All of it pulled from Hoard's own anonymised seller base, once there's enough of it to publish responsibly.

Too thin a sample to publish yet

A benchmark is only honest if it reflects the market, not a handful of the largest accounts on the platform. Right now, the pool of Hoard sellers with enough closed-order history to compute a stable refund rate is too small to publish a number from responsibly. That's not a stall tactic. A benchmark built on a thin, unrepresentative slice of stores would be worse than publishing nothing, because it would look authoritative while actually describing three or four large sellers instead of the market. This is a normal, expected stage for a benchmark like this, not a sign the study stalled.

Check back before you cite a number

We'll update this page directly once the seller base is large enough to compute a real number, rather than publish a separate announcement and leave this one stale. If you're citing refund or dispute rate data for anything public, this is the page to watch, not the one to cite yet.

In the meantime, breaking down your own dispute rate by set, condition, and price band is the version of this you can run today, on your own inventory, without waiting on the aggregate benchmark. And the TCGplayer refunds guide covers what actually lowers a refund rate in the first place, independent of what the eventual benchmark says is average.