Data & Ranking Methodology

MangaKazoku is a discovery layer over third-party anime and manga metadata. Most factual catalogue fields originate with Jikan's MyAnimeList API or AniList; MangaKazoku normalizes those records and adds its own list activity, reviews, and discovery tools.

Rankings and “trending” lists

Homepage trends use the number of non-dropped watchlist or reading-list updates made during the previous seven days. When there is too little recent activity, the page says which fallback it uses: current-season rating for anime or catalogue recency for manga.

“Best of” pages are automated catalogue rankings. A title must match the stated genre or demographic and pass public-catalogue safety checks; eligible results are ordered by rating. They are not staff reviews or paid placements.

Recommendations

“Anime like” pages compare catalogue neighbours using overlapping genres, themes, and format. Catalogue score is a tie-breaker, not the sole recommendation signal. Each indexable recommendation page shows the matching evidence beneath the results.

Rating confidence

A precise score can be misleading when only a handful of people rated a title. MangaKazoku therefore hides catalogue scores and related rank claims unless at least 50 known source ratings support them. These upstream catalogue ratings are labelled separately from the average of reviews written by MangaKazoku members.

Freshness, limitations, and corrections

Imports run on a schedule, but upstream data can be delayed, incomplete, or wrong. Detail pages expose their latest known update time. MangaKazoku is not a primary or authoritative source, and availability can differ by country.

To report a reproducible catalogue or ranking problem, open an issue in the public project repository. Include the affected canonical URL and a link to the upstream record when possible. See also the about page and disclaimer.