How We Score Films and TV

Every score on this site comes out of the same six-category rubric, applied the same way to a four-hour prestige drama and a ninety-minute streaming comedy. No deciding a number first and working backwards, no nudging it up because a lead performance was charming, no rewriting history when the consensus goes the other way. A score you can’t audit isn’t a score, it’s a mood with a decimal point. So here is the whole machine, weights and all.

The Six Categories We Grade

We split a title into six things that can be judged separately, because most bad reviewing comes from collapsing six separate judgements into one impression and calling it taste. The six are writing and structure, direction and craft, performance, originality and ambition, execution of intent, and rewatch value. Six is deliberate. Fewer and the score is just a feeling wearing a jacket; more and the categories start overlapping until the same quality gets counted three times over.

Writing and structure means the script: whether scenes build, whether dialogue sounds like people, whether the story earns its turns or simply announces them. Direction and craft covers everything the camera and the edit do: framing, pacing, sound design, production design, the difference between a sequence that breathes and one that just sits there. Performance is the leads and the ensemble together; a brilliant central turn surrounded by cardboard is not a performance triumph, it’s a rescue mission.

Originality and ambition asks how hard the thing is trying and how familiar the ground is. Execution of intent is separate from all of that on purpose: a film that sets out to be a breezy heist comedy and delivers exactly that should not be marked down for refusing to be a Bergman picture. Rewatch value is the least weighted column precisely because it’s the most unfair one. A devastating one-and-done drama shouldn’t lose points for doing its job too well.

The Weights Behind the Number

Each category is graded 0-10 to one decimal, independently, before anyone sees a total. A running total has gravity, and graders drift towards numbers they’ve already written down. The six sub-scores are then combined using fixed weights that add up to 100%. Those weights haven’t changed since we started publishing sub-scores, and if they ever do, we’ll note it and leave old scores exactly as they were printed.

Writing carries the heaviest weight because almost every other good thing in a film or a season is downstream of it. You can photograph a bad script beautifully; you can’t act your way out of a scene with no reason to exist. Ambition is weighted below the three craft columns because we want to reward difficult things that work, not difficult things that merely exist. Plenty of very ambitious titles score badly here.

Three Scores, Three Different Journeys

Here is the part that matters: a 6.0 is not “60% of a great film”, and it is not a 7.0 that lost some points for being forgettable. It is a specific shape. Compare three real score profiles from our own review archive.

The 6.0 is professionally made and completely unambitious: a by-the-numbers thriller that hits every beat you expect and skips every risk it could have taken. Its execution score is its highest number, and that’s the point: it did what it set out to do, and what it set out to do wasn’t much. Craft can’t save a 4.5 in originality, which is why this profile lands at a middling 6.0 despite having nothing broken in it.

The 8.0 is the shape of a very good production with one real ceiling. Writing and direction are both strong, execution is near-perfect, the acting is genuinely good, but originality sits at 6.5: you have seen this story before, and its best scenes are the ones least like the rest of it. That single soft column is the entire difference between an 8.0 and a low 9. The 9.5 is different in kind rather than degree: it is above 9 in five of six columns, and its weakest score would still be the best line on most other titles’ sheets. That is not an accident. To climb from a 9.0 to a 9.5 you have to gain almost everywhere at once, which is exactly why so few reviews ever get there.

“Good” Is Not the Same as “Good For You”

The number describes how well made something is. It does not describe how much you, specifically, will enjoy it. A 9.2 horror film is a 9.2 for a person who has not willingly watched a horror film since 2011, and they will turn it off after twenty minutes. The score is a description, not a promise, and treating it as a promise is how people end up watching technically flawless films they resent.

This is also why a score is not a recommendation. A recommendation needs two things a rubric can never have: knowledge of the viewer and knowledge of the evening. The same 8.4 drama is the perfect pick for a long Sunday and a terrible pick after a fourteen-hour day, and no decimal point is going to tell you which one you’re having.

The practical fix is to use the score as a sorting tool and the sub-scores as the decision. If what you love about television is dialogue, read the writing column and ignore the total. If you watch for a great central performance, the performance line is your number. That’s the whole reason we publish the breakdown instead of a single opaque figure: you can disagree with us about one axis without throwing out the entire review.

Why a Rubric Helps, and Where Rubrics Mislead

We accept all four of those weaknesses and publish anyway, because the alternative (a vibe score) has the same problems plus no way to check our work. If you want to know why two films both landed on 7.6, the sheets will tell you. Often they got there for opposite reasons, and that difference is more useful than the shared number.

Ties, Rounding, and What We Refuse to Measure

When two titles land on the same reported score, we break the tie in a fixed order: the unrounded totals first, then execution of intent, then the more ambitious title, and finally rewatch value. That order is written down so nobody invents a tie-breaker at one in the morning that happens to favour the thing they liked more. On rounding: we report one decimal using standard rounding, and nothing rounds up to a 10. A 10 currently has to be earned with a raw score north of 9.95 across the board, and no title has managed it yet. A perfect score that is impossible to reach is more honest than one we hand out every awards season.

There is a longer list of things the rubric deliberately refuses to measure. It does not measure cultural importance: a landmark title that changed the industry can still score a 7.0, because influence is a fact about history and quality is a fact about the work. It does not measure whether a title suits your mood, your age, or your household. Parental controls and age ratings are a separate question with a separate guide. It does not measure the streaming service around it: bitrate, app quality and buffering are a platform issue, not a film issue, and we score them elsewhere.

The score also ignores awards, box office, and how loudly the internet argued about it. It genuinely does not care whether the characters are likeable, only whether they are written and played as people. And a score is not a popularity contest in either direction: a 4.0 that nobody watched is still a 4.0, and a title with a passionate fanbase does not get a bonus for the passion.

The Bottom Line

Methodology FAQ