How it works

Everything that decides a number on this site is written down here.

The Cinemini score

You rate from 1 to 10. Your rating counts toward a drama’s score when:

• you have marked it Finished, or Watching or Dropped at episode 5 or later (for a shorter drama, its last episode);
• your account is at least 24 hours old;
• your account is not held for review by our manipulation checks.

Ratings that do not count yet still show on your own page, and start counting as soon as they qualify.

The score is a weighted average that keeps a handful of votes from producing a 10: score = v/(v+10) × R + 10/(v+10) × C, where R is the drama’s average of counted ratings, v how many there are, and C the average of all counted ratings on the site (6.5 until there are enough). With fewer than 5 counted ratings no score is shown, and the drama does not appear in the rankings.

If accounts younger than 7 days give at least 10 of a drama’s counted ratings and more than half of them, their ratings are set aside and the page says so, until a person has looked.

Trope tags

Tags start as nominations from three sources, and each one says which:

• AI-nominated: an AI model read the app’s own description and picked tropes from our list. We checked it against 100 dramas by hand: its first pick was clearly right 94 times in 100 and its second 81, its later picks much less often, so we only nominate a later pick when the app’s own tags agree.
• From app tags: the app’s labels mapped to our list, only for the few mappings that proved reliable.
• Suggested by viewers: anyone signed in can suggest a trope from the list.

A tag is confirmed when it has at least 3 more agree votes than disagree votes and at least three quarters agree. It is hidden when disagree votes lead by 3. Trope filters on Find use confirmed and still-open nominations.

Reviews

Every review is written by a signed-in viewer. We never write reviews, never use AI to write them, and never pay for them. We do not hide reviews for being negative. We refuse spam (links, handles, contact details, promotions, slurs), and staff may hide a review that is spam, an unmarked spoiler, abuse, or that its writer asked us to remove; the writer sees the reason.

Where we use AI, and where we do not

• We use it to keep the database full and correct: reading new dramas from the apps every day, matching the same drama across apps, nominating tropes, and finding a drama from a description you type.
• We do not use it to write anything you read as an opinion: no AI reviews, no AI ratings, no AI write-ups of dramas.
• Anything AI produced and no person has confirmed is labelled as such.

Where the facts come from

Titles, episode counts, free episodes, cast lists and view counts come from each app’s public pages, read once a day. The cost to finish uses the app’s published coin price, which only ReelShort publishes. App descriptions are used inside our search index but not shown here: they are the apps’ text, not ours.