How we count swear words

What is counted

• The text of the film's English subtitle file. Formatting tags and styling codes are removed first, so only the words on screen are read.
• Each subtitle line is checked against a word list. A word matches as a whole word, ignoring capital letters, together with a fixed set of common endings such as -s, -ed and -ing. Some entries also match inside a longer compound word. A list of exceptions stops ordinary words that look similar from being flagged.
• Every appearance counts. If a word is said ten times, it adds ten to the total. Nothing is de-duplicated.
• Each flagged word is filed under a category, and we keep the count for each word. On the film pages the words are masked, for example f***, except for religious exclamations, which are ordinary words.
• The word list and the filter profiles are the same ones the Family Filter TV app uses to mute, so the page counts what the app would act on.

The categories and the filter profiles

Flagged words fall into these categories. The table shows which categories each filter profile mutes.

Each film page also shows how many subtitle lines each profile would mute. That is a count of lines, not words: a line with three flagged words is one muted line but adds three to the word count, and a line is muted only if it holds a word in a category the profile covers.

The language level

The level shown on each page (clean, mild, moderate or strong) comes from the number of lines the filter profiles would mute, checked in this order:

• Clean: the strict filter would mute no lines.
• Mild: the mild filter would mute no lines and the strict filter at most 5.
• Moderate: the mild filter would mute 1 to 9 lines, or the strict filter 6 to 30.
• Strong: anything above that.

Where in the film

Each flagged word is placed in a ten-minute stretch of the film, by the start time of the subtitle line it is in. That is why the pages say “between minute 20 and 30” and never give an exact time for a word.

Which subtitles we use

We look for English subtitle files for the film. Where there is more than one, we use the file whose last line best matches the film's running time, and a file that is clearly timed for a different cut of the film is not used. Each film page shows the date it was last analysed.

The comparisons with “films we have analysed” use only films where the subtitle timing matched the running time closely and the subtitles have at least 400 lines.

Why our number can differ from other sites

• Bleeped or partly spoken words. We count words written in the subtitles. A word that is bleeped, muffled or left out of the subtitle file cannot be counted, even if a viewer hears it or another guide lists it.
• Different subtitle releases. Subtitle files for the same film are not identical. A different release, or a different cut of the film, can give a different number.
• What counts as a swear word. Our list has 7 categories, including religious exclamations and crude language alongside strong profanity. Other sites choose their own scope, so totals are not directly comparable. To compare like with like, look at the strong profanity figure, or at the F-word figure, which counts every flagged word that contains it.
• Counting every appearance. We count each time a word appears, not each scene it appears in.

What these numbers do not cover

They cover language only. They do not rate violence, nudity or frightening scenes, and we do not recommend an age for any film. For a worked example, see Uncut Gems (2019).