How we score

The HeartGuard score

One number from 0 to 100 for how suitable a film or series is for Christian viewers. It combines what's on screen with the message of the story, and it's a starting point for your family's decision, not a verdict on anyone's conscience ().

1. Content meters

Every title is rated from 0 (none) to 5 (extreme) on eight kinds of content. Some weigh more than others in the score:

Meter	Weight
Sex and nudity	×1.3
Blasphemy	×1.2
Violence	×1.0
Drugs and alcohol	×1.0
Attitudes	×1.0
Language	×0.8
Relationships	×0.8
Occult	×0.8
Frightening scenes	×0.6

Blasphemy is kept separate from general language, because misusing God's name matters differently to Christian families. The occult meter (witchcraft, spells, séances, fortune-telling, satanic themes) is added to the average without adding to what it is divided by, so a title without occult content scores exactly as it would without the meter. Faith itself isn't a meter: how a story treats it shows in the worldview score.

2. Content concern

The meters are combined into one content concern from 0 to 5:

concern = 0.6 × weighted average + 0.4 × worst meter

The weighted average reflects the film as a whole. The worst meter counts for 40% so that one serious problem isn't averaged away: a film that is clean except for explicit sex should not look nearly clean.

3. The score

score = 100 − 16 × concern + 8 × worldview

Each point of content concern takes 16 points off. The worldview (−2 to +2) adds or removes up to 16 points, so the same content scores higher when the story treats it as a dead end and lower when it glamorises it. The result is rounded and kept between 0 and 100.

Verdicts

80–100Recommended
60–79Watch with care
35–59Think it through
15–34Strong caution
0–14Not recommended

4. Hard limits

Explicit sex (the sex and nudity meter at 5) caps the score at 30, however strong the rest of the film is.
Pornographic content or sexualised minors make a title automatically Not recommended.
We always flag suicide (including implied or portrayed as heroic), sexual content involving minors and child abuse testimony, whatever the score.

Series

Every episode of a series gets its own analysis and score. The series score is not their average, which would let a run of gentle episodes hide a harder one. It is the score of its harder episodes: the hardest one in a short series, and in a longer one the hardest 10%, so a single unusual episode doesn’t decide it. Explicit sex or a hard flag in any episode still caps the series, as above. Each meter shows the series’ typical high point: the level that only the strongest 10% of episodes reach or pass, so you know what you’ll meet somewhere in it. Those high points can come from different episodes, which is why the score comes from the episodes themselves and not from the meters. Each series page also shows the range of its episodes’ scores and points out the episodes that stand out: those flagged, with a red verdict, or 15 or more points below the series score.

Worked examples

The community vote

Signed-in viewers say whether a title is appropriate for Christian viewers, and can add a comment explaining why. We show the share who say it is, next to our score but not blended into it: one counts people, the other measures content, and the number of votes is always shown beside it. From 10 votes, when the share and our score differ by more than 30 points, we show “Community sees this differently” on the title and take another look at our analysis. Comments that break the rules are hidden.

Where the meters come from

An AI review drafts the meters from what is known about the title: the plot, reviews and reporting that independent websites agree on, and its own knowledge. Some earlier analyses also counted language, blasphemy, slurs and drug references in the dialogue and read them in context (a prayer is not an exclamation); each title’s “How we analysed this” says which. Each title page says whether a human reviewer has checked it yet. Dialogue misses most of what is only seen, so visual content is marked as unverified until a person has checked it. If you think we got a title wrong, use “Was this analysis accurate?” on its page.