Everyone rates things now — a tap on a star, a thumbs up, a score muttered to a friend on the walk out of the theater. Almost nobody rates things well. Scores get handed out in the emotional flush of the final scene, wobble wildly with mood, inflate year after year, and mean nothing when revisited later. Meanwhile, the aggregate scores we consult before watching measure the average feelings of strangers whose taste may overlap yours by half. There is a fix for both problems, and it is the same fix: build a small, honest rating system of your own. Done right, it takes a minute per title and quietly becomes the most useful entertainment tool you own — a map of your own taste, drawn by the only critic whose verdict actually predicts your enjoyment.

Why Bother Rating Anything at All

The skeptic's question deserves a real answer, because rating for its own sake is indeed pointless. The payoff of a personal system is threefold. It sharpens watching itself: knowing you will render a verdict makes you notice what is working and why, the difference between eating and tasting. It builds a usable memory: two years from now, when someone asks whether that series was worth it, your log answers instantly and honestly, where unaided memory recalls only a vague warmth. And it trains prediction: after fifty rated titles, patterns in what you loved and what merely impressed you become visible, and you start choosing better before you press play. The system pays for itself in avoided mediocre evenings.

Choose a Scale You Can Actually Feel

The scale wars — five stars versus ten points versus letter grades versus tiers — have a boring, correct resolution: the best scale is the one whose every step you can feel the difference between. For most people that rules out anything past ten points. A hundred-point scale invites false precision; nobody can defend the difference between a seventy-one and a seventy-three, and undefendable distinctions rot a system fast.

  • Five stars with halves — ten meaningful rungs, familiar, hard to overthink. The sensible default.
  • Tiers — loved it, liked it, fine, disliked it. Blunt, fast, nearly immune to inflation; ideal for people who resent deliberating.
  • Ten points — workable if you write one sentence defining each number and actually obey your own definitions.
  • Two axes — one score for quality, one for enjoyment. More on why this pair earns its extra effort below.

Separate Craft From Joy

The single upgrade that most improves any rating system: score quality and enjoyment separately, because they are different things and one number forces a dishonest average. Everyone owns the experience of the impeccably made film that left them cold, and the ramshackle comedy they have rewatched five times. Collapse those into single scores and both become meaningless threes. Keep them apart and your log suddenly tells the truth: high-craft, low-joy titles reveal what you respect but do not love; low-craft, high-joy titles reveal your comfort zones and guilty pleasures, which are neither guilty nor trivial — they are data.

Over time the two-axis log answers the deepest taste question of all: when craft and joy diverge, which way do you lean? Knowing that about yourself outperforms every aggregate score ever computed.

Pick Criteria That Predict Your Own Satisfaction

Beneath the headline score, the criteria worth tracking are the ones that predict whether future-you will thank past-you for the recommendation. Three earn their place in almost any system. Rewatchability: would you happily experience this again — the strongest single signal of personal love, and notably absent from professional criticism. Staying power: does it linger a week later, or did it evaporate by morning? Some titles rate higher in memory than they did at the credits, and that drift is signal, not error. Recommendability: who, specifically, would you tell to watch this? A title you cannot match to a single friend may have impressed you without connecting to anyone.

Resist criteria bloat. Scoring seven sub-categories per film turns a pleasure into an audit, and audited hobbies get abandoned. One or two numbers plus a sentence beats a spreadsheet symphony you quit by spring.

Fight Score Inflation Before It Starts

Every personal scale drifts upward if unattended, for predictable reasons: you pre-select things you expect to like, recency flatters the just-finished, and rounding up feels generous while rounding down feels mean. Left alone, three years of ratings compress into a mush of fours and the scale dies. The countermeasures are simple. Define the middle of your scale as your honest average experience and use it without shame — average is not an insult; it is the definition of most evenings. Rate twice: once at the credits, once a week later, keeping the second. Anchor with exemplars — pick a personal benchmark title for each rung and compare new candidates against them rather than against the ceiling. And audit annually: if last year contains no low scores, you were either miraculously lucky or drifting, and you already know which.

One more inflation trap deserves naming: the social score. Ratings shared with friends or posted publicly drift upward under the gentle pressure of being seen, because harsh numbers feel like unkindness and enthusiasm earns replies. If you share your log, keep a private column too. The public number can be diplomatic; the private one must stay true, or the whole instrument bends.

Keep the Log Lightweight and Alive

The system only works if it survives contact with your actual habits, so build for laziness. Log immediately — one line, thirty seconds, on whatever surface you will really use: a notes app, a pocket notebook, a spreadsheet if that genuinely pleases you. The entry that matters is the sentence, not the score: one honest line about why — what it did to you, where it lost you — turns a number into a memory hook. Two years later the sentence is the part you will treasure, and it costs you less time than choosing tonight's snack did.

Then, occasionally, read the thing. A rating log is not an archive; it is a mirror. Skim it each season and ask what it shows: a genre you consistently overrate going in, a format that never misses, a slow migration of taste you had not noticed. That reading habit is where the log stops being record-keeping and starts being self-knowledge.

Your History Beats the Aggregate

None of this requires abandoning collective scores; they remain decent coarse filters against outright disasters. But understand what they are: an average of strangers, weighted toward people motivated enough to vote, measuring a consensus you may sit nowhere near. Your own log, once it holds fifty entries, is a better instrument for the only question that matters — will I like this? — because it was calibrated on the correct subject. The aggregate tells you what the crowd felt. The log tells you what you feel, how reliably, and under what conditions. Consult both; trust yours.

Final Thoughts

A personal rating system is a small machine with an outsized return: a felt scale, craft and joy scored separately, one or two criteria that predict your actual satisfaction, active defenses against inflation, and a one-line log you genuinely keep and occasionally reread. Build it and every evening of watching feeds the next one — each verdict sharpening the picker, each entry deepening the map. The stars were never the point. The point is becoming the world's leading expert on a subject nobody else can study: what you love, and why.