Methodology - TCG Mart London
Methodology

How we price Pokémon cards for the market you collect in.

Every value on this site — and every valuation in Cardfolio — is priced against the market it belongs to. A British collector should not be reading an American number with the currency symbol swapped, and an American collector should not be reading a British one. This page explains how we get there, how we tell you which prices to trust, and where the limits are. They are guide prices for tracking a collection — estimates, not quotes.

Why / The problem
Almost every Pokémon price database prices one market and leaves everyone else to do the conversion. Take an American number, turn it into pounds and call it a day, and your Base Set holos come out badly wrong — because the UK market doesn't behave like the US market. This engine began as a private fix for one collector's own spreadsheet. It now values 74,000+ cards, English and Japanese, across two markets — each refreshed every week, with the most-traded cards re-read from live market activity the same week.
· 01 HOW IT WORKS

Two reads on every price.

The market baseline refreshes weekly, right across the catalogue, so every card carries a current value in whichever market you are viewing. Then, the same week, a second read goes through the cards that trade most — the valuable, liquid singles — and re-prices them from live market activity, the highest-value cards first. Each price in the app carries its own "price as of" date.

Read 1

A market baseline for every card.

We start from recent-sale market data compiled by an independent third-party price guide. For the US market that data is the market, and it is read as reported. For the UK it is converted at the live exchange rate and then recalibrated — British supply is tighter and collectors here consistently pay more than a straight conversion would suggest, and vintage holos most of all. Either way, every card in the catalogue carries a current value in the market you are viewing.

Read 2

A live market read for the cards that matter most.

For the valuable, actively-traded singles — and every card someone actually owns — we go further than a baseline. We read the live market directly: thousands of real listings for that exact card, sifted down to the ones that genuinely match, and let real market activity set the number. Where that read is clear, consistent, and agrees with the baseline, it replaces it. Where it disagrees, the baseline holds and the read is discarded.

Condition

Anchored to a clean copy.

Ungraded values are anchored to Near Mint, with sensible adjustments for played grades. Graded cards are a different product on a different market, so they are valued separately — and a PSA 10 trades on an effectively global one. The same benchmark therefore serves both markets: shown as reported for the US, and converted at the live exchange rate for the UK, with no separate regional adjustment.

Filtering

The right card — and only the right card.

Wrong editions, reprints, bundles, graded slabs, custom proxies, damaged copies and sellers outside the market being priced are all sifted out before a listing is allowed to count. Getting this right — telling near-identical cards apart at scale — is most of the work, and most of the difference.

Stability

One noisy week can't whipsaw your collection.

Card markets are noisy — one odd sale or a quiet week can throw a single reading off. So a sudden, outsized swing in a card's value isn't published on the strength of one week alone: it's held at the last trusted figure until the next read confirms the move is real. Genuine shifts still come through — just a beat later, and without the whiplash. Your collection's value follows the market without lurching on its noise.

Calibration

Checked against reality, continually.

The engine is measured against live market activity in both markets and tuned as they move. The public collection on this site is valued by the same engine — it is the standing proof, on our own cards first.

· 02 CONFIDENCE

Every price tells you how far to trust it.

Not every card trades often enough to price with the same certainty — so we don't pretend otherwise. Each value carries a confidence indicator, shown as a coloured dot in the app, so you can see at a glance how much weight a number can bear.

High

Priced from your own market.

Set by a clear, consistent read of live market activity for clean, Near-Mint copies of that exact card — not inferred from somewhere else and converted. As close to the live market as a guide gets; lean on it.

Medium

A solid market guide.

A dependable, market-calibrated estimate — drawn from solid market data, or read from live activity where clean Near-Mint copies are too thin to confirm on their own. Not pinned to a single confirmed reading, but trustworthy for everyday tracking.

Low

A rough guide.

For a card that trades rarely, has thin data, or sits at the cheap end of the market, where even a run of genuine listings leaves too little room between them to pin a precise number down. Useful as a ballpark — the start of a conversation, not the last word.

The same card can shift bands as the market changes: a clean run of market activity moves it up; a quiet spell eases it back. Its price plays a part too — the cheaper end of the market is harder to pin to the pound however active the listings are, so it takes a stronger, more valuable card to earn the very top band. A card can also sit at different confidence in each market. The indicator always reflects today's evidence, not a label fixed once and forgotten.

The craft
Turning thousands of messy, real-world listings into one number you can trust is harder than it looks. The precise way we identify the right card, weigh each signal, and decide when a market read is strong enough to overrule the baseline is our own — refined continually against both live markets. The result is on every card; the recipe stays in the kitchen.
· 03 LIMITS

What these numbers are not.

A methodology you can't fault is one that admits its edges. Here are ours.

01 · Estimates

Not offers.

An estimate is what comparable copies have recently sold for — not what a buyer will hand you today. It isn't an appraisal, an insurance valuation, or a guaranteed resale price. Condition, photos, timing, and platform fees all move the real number.

02 · Thin data

Obscure cards wobble.

A card that sells twice a year has a noisy market — and, in a different way, so does one cheap enough that a pound or two of difference between listings is a large swing. Both carry a Low confidence dot — useful as a ballpark, but treat them with more caution than a liquid chase card.

03 · Cadence

Weekly, not real-time.

Both reads refresh every week: the baseline across every card, and the deeper market read across the valuable, most-traded cards right after it — highest-value first. It is not a live ticker — every price in Cardfolio shows the date it was last updated, so check it before you act on it.

04 · Condition

We assume a clean copy.

Ungraded estimates are calibrated to clean, collector-grade copies — the kind buyers actually bid on. A heavily played copy of the same card will trade below its estimate, sometimes well below.

05 · Big decisions

Check the comps yourself.

Before you buy, sell, or insure an expensive card, spend five minutes checking recent sold listings in your own market. Every card page in Cardfolio links straight to them, on the marketplace for the market you are viewing. Our estimate tells you where the market has been; live comps tell you where it is this week. Use both.

06 · Not advice

Collecting first.

Nothing here is financial advice. We built this to document collections honestly, not to predict markets or sell a dream.

· 04 PROOF

We use it on our own collection. Publicly.

The TCG Mart London collection — 2,000+ cards across 8 master sets — is valued by this exact engine and published openly, price history and all. If the methodology is wrong somewhere, it's wrong in public, on our own cards first. Then it gets fixed.