How the scores are made

1. What Filtro is — and is not

Filtro Digital evaluates WORKS — movies, series, games, books, and music — across content and values axes. Scores are OPINION WITH EVIDENCE, never a factual verdict. Criticizing a work is not the same as making a factual claim about the people who made it.

We INFORM and RECOMMEND; we do not ban, we do not veto, and we are not an official ratings authority. The decision to watch is always yours. The goal is to help you find what matches your values — not to tell anyone what they may or may not watch.

2. The 14 axes

Each work is read across 14 axes in two groups. The 8 ALERT axes flag what part of the audience prefers to avoid; the 6 VALUE axes flag qualities part of the audience seeks out.

Alert — ideological (form the Woke Score): identity politics, historical revisionism, faith subversion, and anti-family content. Alert — advisory (content warnings): inappropriate sexualization, graphic violence, profanity, and drug glamorization.

Value (form the Valor Score): pro-family, faith and transcendence, virtue and heroism, Western values, artistic integrity, and artistic merit.

3. How a score is born: AI + community

The score is HYBRID and ADAPTIVE. First, AI makes an initial reading of each axis from the available information about the work. Then the community votes and refines that reading — and that is where the score matures.

Votes are weighted by reputation: new accounts count for less, to make manipulation harder. The more the community participates, the more the score reflects the collective reading and the less it depends on the AI. That is why scores CHANGE over time. Cases where AI and community diverge sharply are routed to human review.

From the per-axis readings we derive three summary scores: the Woke Score (0 to −100 — the more negative, the more ideological alert content), the Valor Score (0 to +100), and the FiltroScore (0 to 100, a "worth watching" reading combining content cleanliness, value, and the audience rating). Alongside the global score, once you fill in your values profile, we also show your personal Match.

4. The role of AI — and its limits

AI lets us read a large catalog consistently, but it is the STARTING POINT, never the final word. Every AI reading is put through the community — which votes and corrects — and sensitive cases go through human review. It is this hybrid design, not AI alone, that produces the score you see.

We work to a clear standard: no score should be a number without a reason. Where a score does not yet show its justification, we treat it as an editorial estimate — never as established fact — and we are actively expanding evidence coverage so that every critique carries the element of the work that supports it.

5. Evidence: why every critique should have a reason

Evidence is the OBSERVABLE ELEMENT OF THE WORK that supports a score — a scene, a theme, a treatment. It exists precisely so that a score is a reasoned judgment and not a bare number.

Our evidence describes what is in the work; it does not assert intentions or make accusations about named people. When you disagree with a reading, evidence is what allows an honest debate about the work itself.

6. How to contest an evaluation

If you think a score or a piece of evidence is wrong, there are two paths. Anyone can use the "Report" button on the work's page — the report enters our human-review queue. Parties responsible for a work (authors, creators, studios) who wish to respond have a dedicated Right of Reply channel, at en/right-of-reply.

When one of our errors is pointed out, we fix it. The rules for what is not allowed in community contributions are in the Moderation Policy, at en/moderation-policy.

7. Editorial independence

Filtro shows "where to watch" links, some of which are affiliate links — we may earn a commission when you subscribe or rent through them. That commission does NOT enter the score calculation and does not influence any evaluation. The editorial score and the commercial link are separate things, and we always mark which is which.

8. A living method

This methodology evolves. We revise the axes, the way AI and the community combine, and evidence coverage over time, and we record the last-updated date here. Questions or corrections are welcome through the channels above.