How does the board find an opportunity?

It takes the other language’s Near Mint sales median, nets out a 13.25% selling fee, and lists a live Near Mint ask (delivered price) at least 20% under that. An ask under a third of it is usually not the same card and is refused.

How often is this updated?

The charts are rebuilt weekly from recorded sales (last run 2026-10-09, sales through 2026-10-09). The board is recomputed every 30 minutes from the asks already on file.

Language Arbitrage

Most cards are printed in Japanese and in English, and the two copies of the same card rarely sell for the same price. Every chart compares Near Mint sales of matched pairs, one Japanese card against its English counterpart; the board at the bottom finds live asks priced under the other side.

English ÷ Japanese1.20×

Same card, 95% range 1.11×–1.25×, over 898 pairs.

The cheap side todayJapanese

The Japanese copy is the cheaper one: the English copy of the same card sells for 20% more. See the board ↓

Widest Modern gap8.49×

Ditto V (Shining Fates Shiny Vault): the Japanese copy is cheaper · 70 EN / 9 JP sales.

Ratio distribution

How far apart are the English and Japanese prices of the same card?

Pooled 1.20×

← English is cheaperEnglish ÷ Japanese, log scaleJapanese is cheaper →

Japanese cheaper
English cheaper
Wide, left out of the pool
Pooled reading · 95% range

Off the axis: 25 under ×0.125, 19 over ×8.

Card pairs by what the English copy costs in Japanese copies; right of the dashed line, Japanese is cheaper. Hatched: 42 wide pairs, left out of the pooled reading.

Where the gap moves

Does the gap depend on the card’s era or its price?

By era

Vintagecut—4 pairs · cutcutVintage: 4 pairs, under the 15-pair floor — cut, not estimated
Classiccut—0 pairs · cut

Widest pairs

Which cards sit furthest apart, in each direction?

Japanese is cheaper

The English copy costs this many Japanese copies

1Ditto VShining Fates Shiny Vault · SV118/SV1228.49×Ditto V (Shining Fates Shiny Vault; Japanese S4a: Shiny Star V, Holo): the English copy's Near Mint median is 8.49× the Japanese copy's — EN $9.00 on 70 sales, JP $1.06 on 9EN $9.00 · n 70 / JP $1.06 · n 9JP listing ↗
2Eternatus VMAXShining Fates Shiny Vault · SV122/SV122

Coverage

How much of the Japanese catalogue stands behind the reading?

Japanese printings on file27,525
Linked to one English card· 3,896 more matched by position only, held back11,350
Card pairs over the sales floor· 5+ Near Mint sales on each side940
Pairs in the pooled reading898

12,279 printings cut · no confident English match
1,181 printings cut · no shared finish
9,623 pairs cut · under the sales floor

Language Arbitrage board

Which live copies are priced under the other language’s market right now?

All 0 Buy JP 0 Buy EN 0

All eras

Method & caveats

Charts rebuilt weekly from recorded Near Mint sales (last run 2026-10-09); a pair is one Japanese card matched to one English card on the same finish, and a reading needs at least 15 pairs. Pairs far outside the rest are drawn hatched and left out of the reading: that is where mismatches hide.

The board is recomputed every 30 minutes from asks already on file: a live ask against the other language’s sales median, after the selling fee. The full method below.

Questions

Is the Japanese or the English copy of a Pokémon card cheaper?

The Japanese copy is the cheaper one: the English copy of the same card sells for 20% more.

How big is the price gap between Japanese and English cards?

Across 898 matched card pairs, the English copy's Near Mint median is 1.20× the Japanese copy's (95% range 1.11×–1.25×) over the last 180 days.

Does the gap depend on the era or the card’s price?

By era, yes: Mid 0.81×, Modern 1.22×. By card price, yes: Under $10 1.13×, $10–$100 2.48×.

Which Modern card has the widest gap right now?

Ditto V (Shining Fates Shiny Vault): the English copy sells for 8.49× the Japanese one, on 70 English and 9 Japanese Near Mint sales.

How are Japanese and English cards matched?

A Japanese printing is linked to an English one only when the name matches exactly one card in the English set its Japanese set maps to and both are the same kind of card (Pokémon, Trainer or Energy), and the two copies are compared on the same finish (holo with holo, reverse with reverse). Each side needs 5 Near Mint recorded sales in 180 days.

What is a wide pair?

A pair whose price ratio sits far outside the middle of the pairs it is compared with (beyond 1.5 interquartile ranges, within its era group where that group is large enough). Mismatched cards and finish-label gaps live there, so wide pairs are drawn and listed but never pooled, never in the widest lists and never on the board.

See also: Condition Arbitrage — what LP, MP, HP and Damaged copies sell for against Near Mint, raw.

Method

What this measures

One language pair is one Japanese printing matched to one English printing, compared on the same finish: holo with holo, reverse holo with reverse holo, and non-holo only when neither printing also has a holo. Within a side the plain print wins (no stamp, first or unlimited edition); two plain candidates is ambiguous and the pair is cut. A Japanese printing is linked only when its name matches exactly one card in the English set its own set maps to (the English set holding most of its names, released near it) and the two agree on what the card is (Pokémon, Trainer or Energy); a match made by position alone is held back.

Prices

Each side’s price is the median of its recorded Near Mint raw sales on TCGplayer over the trailing 180 days. A side needs at least 5 sales, and one whose dearest sale is more than five times its own median is refused. Sales that state no condition never enter any number. Japanese sales on file begin in late May 2026, so the window is shorter in practice.

The ratio

Each pair gives one ratio, the English median over the Japanese median; above 1 the Japanese copy is the cheap side. The page’s reading is the median of those ratios across at least 15 pairs, with a 95% interval from 1,000 resamples of the pairs. The cheap side is named only when the whole interval sits on one side of parity.

Wide pairs

A pair whose ratio lies more than 1.5 interquartile ranges outside the middle half of the pairs it is screened with (its own era group where that group is large enough, every pair otherwise) is wide: drawn hatched and listed, but left out of the reading, the splits, the widest lists and the board. Mismatched cards and finish labels that differ between the two catalogues live in these tails, so the screen is a quality gate, not a trim of real markets.

Era and price

Whether the ratio differs by era group (from the English set) or by price tier (the geometric mean of the two medians, never one side’s own price) is a Kruskal–Wallis test on the log ratios, the two tests adjusted together (Benjamini–Hochberg). A split is drawn only at adjusted p under 0.05 with two groups of at least 15 pairs; a group under that floor is shown as cut, never estimated.