How We Source Our Data, and Why We Publish the Lineage

Most Pokémon TCG market sites publish prices without telling you where they came from. PriceCharting scrapes eBay. Card Ladder scrapes English-market secondary sales. pokeca-chart aggregates Japanese domestic retailer data into ranking tables and explains nothing about the lineage. Snkrdunk Magazine writes opinion pieces without citing the trade prints they draw from. The standard practice in this category is to publish the number and assume readers will trust it.

We are not going to do that. This article documents every data source PokemonMarketWatch reads, every price column we publish, and the methodology choices that bridge them. If you cannot evaluate where our numbers come from, you cannot meaningfully disagree with them. That problem is solvable. This is the solution.

What we publish, and where each number comes from

PMW publishes three categories of numerical signal: Japanese-domestic retail prices, international-facing Japanese retailer prices, and US secondary marketplace prices. The three are structurally different and we never blend them silently.

Japanese-domestic retail prices are what a shopkeeper in Akihabara, Osaka, or Tohoku quotes a Japanese customer in JPY. These are the most predictive of the four for understanding sealed-product and singles trajectory, because they are the prices that the Japanese collector base actually transacts at. We pull these from:

  • Yuyu-tei (yuyu-tei.jp) for modern singles and sealed sell-side ask prices. Yuyu-tei is one of Japan's largest TCG shop chains. Their per-set listing pages publish sell-side asks plus stock status (在庫 ◯/×). We scrape daily.
  • Mandarake via the ekizo subdomain (ekizo.mandarake.co.jp) for vintage 1996-2003 sealed and singles, plus modern stock at their secondary retail locations. Mandarake is the canonical Japanese used-goods chain for vintage collectibles, with per-store inventory rolled into the central catalog. We capture pre-tax and tax-inclusive (税込) prices separately.
  • Surugaya (suruga-ya.jp) for new and used product, distinguishing condition explicitly. Surugaya is one of the few major JP retailers that consistently flags new vs used.
  • Card Rush via the pokeca-chart price API (more on that below).

International-facing Japanese retailer prices are what the same Japanese inventory costs once it is offered to non-Japanese buyers. The gap between this and the domestic price is the proxy-ecosystem premium, and it exists because The Pokémon Company Japan restricts its distributors from selling internationally. We pull these from:

  • Hareruya2 (hareruya2.com), a Shopify-backed English storefront that ships internationally and prices in JPY.
  • AmiAmi (amiami.com), the canonical preorder authority for new Japanese sealed product with international shipping support.

US secondary marketplace prices are what US sellers ask for the same Japanese product, in USD. We pull these from:

  • TCGCSV (tcgcsv.com), a free public mirror of TCGplayer's pricing data. We cross-reference Japanese category listings (categoryId 85) to capture the US-market ask for Japanese-named product.

The pokeca-chart breakthrough, and what it unlocked

In early May 2026 we reverse-engineered the backend API behind pokeca-chart, the most-cited Japanese-language price aggregator. Their per-card pages route through /{set-code}-{card-number}-{total}/ URL slugs, but the price chart itself loads from a separate PHP endpoint at /ch/php/get-chart-data.php. Their obfuscated JavaScript revealed two source columns under the hood:

  • price_01: Card Rush ask price (constant SHOP_ID_RUSH = 0x8)
  • price_02: Snkrdunk listing price (constant SHOP_ID_SNDN = 0x9)
  • price_03: a third source not labeled in their JavaScript
  • volume: daily listing-count signal across all tracked sources

This API returns the full historical time-series for each card, not just today's value. One scraper effectively unlocked three sources of price data plus listing volume plus historical depth, all on a clean JSON endpoint that requires no HTML parsing.

We publish this lineage because pokeca-chart does not. Their site presents their rankings as if they were primary observations. They are not. They are derived from Card Rush and Snkrdunk feeds with a third source layered in. Knowing this changes how you should weight their numbers in your own decisions, particularly when Card Rush is having a buyback campaign or Snkrdunk is running a promotion that distorts the data.

What we capture beyond prices

Price data alone is necessary but not sufficient for predicting Japanese market trajectory. We also capture supply events and competitor signals.

Supply events come from two sources. The first is the official Pokémon Company Japan press-release feed at pokemon-card.com/info/. Every official announcement, every Special Box reprint, every Pokémon Center lottery, every price-tier change posts here with a six-digit ID. We poll this daily and classify each entry by keyword into the categories that matter for market timing: price_hike_jp, reprint_announced, lottery_drop, set_release, id_verification, event. The May 2026 pack-price hike (announcement 005376) and the more recent My Number Card identity verification rollout (announcement 005450, May 22) both surfaced here.

The second supply-event source is a curated set of specialized Japanese Twitter accounts that operate as restock and lottery monitors. The market in Japan reacts to these accounts within minutes, often before the official site publishes the same news. We ingest from accounts like @pokeca_new_card (new release announcements), @pokecenobserver (Pokémon Center Online state changes, a real-time supply bot), and @poke_times (official Pokémon news mirror). We use TwitterAPI.io as our ingestion layer, paying per-tweet read rather than the official $200/month X API tier, which would not be cost-justified for our volume.

Competitor signals come from monitoring the editorial output of every English-language site that covers Japanese Pokémon market. Snkrdunk Magazine publishes the closest editorial product to PMW. We poll their RSS feed daily and filter for Pokémon-tagged articles. We do not reproduce their content. We use the metadata (publication date, headline, URL) to understand what topics Snkrdunk is covering and which JP-market angles still need English-language analysis.

Methodology choices we publish

Several editorial calls follow from this source mix. We document them rather than burying them, because each one is a choice that affects what conclusions you can draw from our numbers.

We never blend domestic JP and international-facing JP prices into a single number. Domestic Yuyu-tei and Mandarake prices answer the question "what does this card cost a Japanese collector." Hareruya2 and AmiAmi prices answer the question "what does this card cost an international buyer of Japanese inventory." Those are different questions. The gap between them is the proxy-ecosystem premium, and we treat that gap as signal worth tracking, not noise to average out.

We capture multiple condition tiers separately. Mandarake distinguishes new and used. Surugaya distinguishes new and used. Pokémon Card 151 boxes at the December 2024 reprint launch sold at one price, and used-but-mint copies from the original 2023 print run sold at a meaningfully different price. We do not collapse these into a "Mandarake price for 151" composite.

We treat C2C sale-comp data as a separate price type from shopkeeper-ask data. magi.camp gives us what individual Japanese collectors actually list cards for, with PSA-graded representation that is the truest live signal for graded singles trajectory. Yuyu-tei gives us what a professional retailer asks. The two diverge in characteristic ways. PSA10 grades on magi will sit above the corresponding Yuyu-tei sealed-pack expected-value calculation when launch hype peaks, and dip below it during reprint troughs. The divergence itself is signal.

We do not publish predictions as facts. When we forecast a US-side price response from a JP-side event, the forecast is explicitly flagged. The May 2026 JP pack-price hike implies a US-side response within four to six months based on the 2022-to-2023 precedent. That is a prediction, not a measurement. We say so.

When sources disagree, we report the disagreement rather than picking a side. If Card Rush is asking ¥156,500 for Mega Darkrai ex M5 118/081 and Snkrdunk is asking ¥178,888 for the same card on the same day, that 14% gap matters more than the average of the two would suggest. We publish both numbers.

The FX handling that matters

Every JPY-denominated price in PMW carries an explicit FX timestamp. We pull the JPY/USD rate daily from a public exchange-rate feed and store the rate alongside each price snapshot in our database. When we display a USD equivalent next to a JPY price, we are showing you what that price would have been in USD at the rate that was live when we captured it, not at today's rate. The yen's been on a multi-year decline against the dollar. Failing to disclose the FX-at-snapshot is one of the most common ways US-facing Japanese-market sites mislead their readers about the actual cost trajectory of products.

What we deliberately do not publish

We do not publish closed-auction realized prices systematically, because we do not yet have a reliable Japanese closed-auction ingest. Yahoo Auctions Japan is feasible via third-party scraper actors at roughly $8-30 per month for 500-listing-per-day coverage, but the data is comp-depth rather than real-time signal, and we have not yet completed thirty days of baseline pricing across our live sources to evaluate whether the marginal cost is justified. This is on the roadmap for the v1.5 build phase.

We do not publish grading-population data, because the Japanese grading services (ARS, TCG Slab Japan) do not publish their populations the way PSA does. PMW will reference grading service data when it appears on a marketplace listing, but we do not have an authoritative population feed to draw from.

We do not publish print-run estimates as primary data. Japanese analysts use community-driven heuristics for this (pull-rate logging, lottery allocation back-calculation, sealed-price forensics). When we cite a print-run estimate, we name the methodology and the source, and we flag the confidence interval.

Why publishing the lineage is the strategy

The Pokémon TCG market intelligence category is full of sites that look authoritative and refuse to show their work. We chose a different approach for one specific reason. If you are deciding whether to allocate capital to Japanese sealed product, the question that matters is not "what does the number say." The question that matters is "what does the number measure, and how confident should I be in its interpretation."

We can give you better numbers than the competition. We can also tell you exactly what those numbers are and are not. The first is table stakes. The second is the durable advantage. The list of sources above is not a marketing pitch. It is the operating manual.

If you have a question about any specific source, methodology choice, or pricing convention documented here, the contact email is on our About page. We answer methodology questions in writing and we update this article when our pipeline changes. The current version is dated. The next version will say what we changed and when.

PokemonMarketWatch reads the Japanese market in public, with all of its primary sources named. That is the product.

Last updated: May 22, 2026.