Research and editorial policy

No in-house play-test ratings

We organise official specifications and third-party accounts of actual use. We do not invent first-hand experience or star ratings.

Review sample size

We include approved posts about actual use that match the specification and were published within the past 180 days, using Japan dates. Accounts are counted using stable source IDs. Matching display names across platforms are not treated as proof of the same person.

For each topic, we distinguish positive, negative, mixed, neutral, indeterminate and no mention. The positive share is positive mentions divided by all mentions, including indeterminate ones. A zero denominator is shown as unknown. Unknown authors and publication dates are counted separately.

Positive and negative opinions from the same counting unit are classified as mixed. We avoid strong trend claims with fewer than 10 units. This summarises voluntary posts and is not a representative survey of all buyers.

Comparing prices

We compare listings for the same SKU: new, single-item, in stock, tax included, with confirmed shipping and within the validity period. Points and member coupons are not deducted from the standard price. If the destination is unknown, only confirmed nationwide flat-rate shipping is included.

AI and sources

AI assists with classification and drafts. Counts and percentages are calculated by software; sources, figures and claims are checked before approval. If a source is withdrawn or usage permission expires, affected summaries and articles are reviewed.

Advertising policy

Affiliate availability and commission rates do not affect lowest-price rankings or review summaries. Affiliate links and advertisements are labelled.