How to Read an IMDb Rating Without Being Misled
Vote counts, weighting, brigading and survivorship: the numbers behind the number, and the three checks that tell you whether a rating is worth trusting.
Read article →Independent film & TV review site. We publish reviews and information only. Nothing on this site is downloadable and we host no software or install files. Read our disclaimer
Rotten Tomatoes vs IMDb vs Letterboxd
Three rating sites, three completely different answers. What each one is actually measuring, where each is gameable, and how to read them together to pick something worth watching.
Look up the same film on three sites and you’ll get three different verdicts: 94% fresh, a 6.1 from users, and three and a half stars from a stranger in a trench coat. None of them is lying to you. Each one is answering a question you didn’t realise you were asking, and the fastest way to stop being misled by rating sites is to learn which question each number is actually answering.
Rotten Tomatoes: A Yes/No Vote in a Percentage Costume
The Tomatometer is not an average of scores. It’s the share of counted critic reviews that were positive, which means a film can sit at 95% without a single critic loving it. If ninety-five critics give it a mild three stars out of five and five critics hate it, that’s a 95%: a number that reads like triumph and actually means broad, tepid agreement. This is the single most common misreading of any rating on the internet.
The percentage is also unstable at low sample sizes. A film with twelve early reviews can swing twenty points across a single afternoon as more critics file. The number that genuinely tells you how much critics liked something is the average rating, usually shown in smaller type next to the percentage, something like “6.4 out of 10”. When the percentage is high and that average is low, you are looking at a film everyone tolerated.
Divided from the critics’ score is the audience score, and the split between the two is arguably the most useful single signal on the site. Critics high, audience low usually means a demanding or confrontational film. Audience high, critics low usually means crowd-pleasing genre comfort, or an organised review-bombing campaign, which is why it’s worth checking whether the audience score arrived in a single angry afternoon.
IMDb: A Weighted Crowd Average, Skewed Towards Genre Fans
An IMDb rating is a weighted average of user votes on a 1-10 scale. The weighting matters: IMDb deliberately suppresses suspicious voting patterns, so the displayed rating is not the raw arithmetic mean and cannot be reverse-engineered by hand. The crowd here is the watching public: global, self-selecting and heavily skewed towards genre and internet-active audiences. There is no critic component at all, which makes it excellent at one question and useless at another. It tells you whether a broad audience enjoyed something. It tells you approximately nothing about whether it is good.
Two biases bite hard. The first is brigading around release: highly anticipated genre films and franchise entries attract organised voting in both directions, and scores often drift for months afterwards. The second is survivorship on older titles. Nobody logs a forgotten 1974 thriller to give it two stars; they log it because they sought it out. Deep-catalogue films therefore settle a point or so higher than they deserve, while brand-new releases are judged in the heat of a marketing campaign.
Vote count is the honesty test IMDb gives you for free and most people ignore. A 7.9 built on 400,000 votes is a genuine signal. A 7.9 built on 900 votes, most of them cast in the first weekend, is a press kit. Nothing under a few thousand votes should influence a decision on its own.
One more quirk: a 6.1 on IMDb is not a 6.1 anywhere else. IMDb’s effective working range is roughly 5 to 8.5, and anything above an 8 is genuinely unusual, because a large, diverse crowd rarely agrees that strongly on anything.
Letterboxd: Enthusiasts Grading in Public
Letterboxd uses five stars in half-star steps, and the people casting those stars are self-selected film enthusiasts. That single sentence explains almost all of its behaviour. Users rate things publicly, often with a written review their followers read, which makes every score a small act of taste-signalling. Nobody logs three films on a Sunday to look basic.
The result is a scale with a personality. Arthouse, foreign-language, cult and genuinely strange films inflate, because the audience is exactly the audience for them. Mainstream blockbusters tend to sit lower than their IMDb equivalents, not because they’re worse but because the people rating them care less about that kind of film. And because you only rate what you chose to watch, there’s a permanent selection bias: the person giving a horror film 4 stars already knew they liked horror.
It reads a whole point lower than IMDb on a like-for-like basis, and the middle is much narrower. On Letterboxd, 3.5 is a solid recommendation, 3.0 is a shrug, and anything above 4.0 means the film-devoted have adopted it. A 2.5 is not a bad film so much as a film that disappointed people who wanted to like it.
Where Letterboxd genuinely outperforms the other two is not the aggregate at all. It’s the social layer: what your friends logged, which lists a title appears on, and the written reviews. A film sitting at 3.9 with thirty argumentative reviews is more informative than the same film at 3.9 with none.
Where Each One Lies to You
Laid side by side, the three systems are measuring three different things with three different crowds. Once you see it as a table, the disagreements stop looking like contradictions.
| Site | What the number measures | Who is voting | Main bias | Best for |
|---|---|---|---|---|
| Rotten Tomatoes | Percentage of critic reviews counted as positive | Professional critics | Measures consensus, not enthusiasm; a 3/5 counts as fresh | “Is this agreeable to most people?” |
| IMDb | Weighted average of user votes on 1-10 | Self-selected global public | Brigading, tiny samples, survivorship on older titles | “Will a general audience enjoy this?” |
| Letterboxd | Mean of self-selected half-star ratings | Film enthusiasts, skewing young and English-speaking | Only interested people watch and rate; taste-signalling inflates the arthouse | “Is this interesting to people who care about film?” |
A 95% is not a 9.5
The most dangerous number in film discourse is a Rotten Tomatoes percentage, because it looks like a score out of 100 but behaves like a headcount. A film at 95% fresh and 6.2 average is a film almost every critic found acceptable and almost none found special. Meanwhile a 6.1 on IMDb and a 6.1 on Letterboxd describe two different films: the first was mildly enjoyed by a general crowd, the second genuinely disappointed a room full of enthusiasts.
How to Combine All Three Into a Decision
Read them as a panel, in a fixed order, and the noise drops away. Start with the Rotten Tomatoes average rating rather than the percentage. That 6.4 out of 10 tells you the intensity of critical feeling, which is the thing you actually want. Then move to IMDb for audience reception, checking the vote count before you trust the figure. Then check Letterboxd, and pay closest attention to the gap between it and IMDb: if Letterboxd is a full point higher, you’re looking at a film for enthusiasts; if it’s lower, the film-devoted passed on it.
The gaps are more informative than any individual number. Critics high and audience low means admirable and difficult: a great pick if you like being challenged and a bad one if you want to switch your brain off. Audience high and critics low means genre comfort or a campaign, and the vote pattern tells you which. When all three agree, you can stop researching.
A few working thresholds save a lot of scrolling. Ignore any score built on a few hundred votes, whatever the site. Treat a fresh percentage above 90 as “broadly acceptable” rather than “excellent”. Treat an IMDb rating between 6 and 7 as normal for anything genre-shaped. And treat Letterboxd’s 3.0-3.5 band as a perfectly good evening rather than faint praise. For consistency’s sake we also publish our own methodology-based score for the titles we review, not because our number is more correct, but because a fixed rubric means a 7 from us always means the same thing.
What These Scores Are Good For, and Where They All Fail
✓Why they’re worth using
- Thousands of opinions beat one, even a well-argued one
- Each measures a different crowd, so together they triangulate
- They’re free, instant, and work for titles nobody has reviewed
- The gap between them is itself useful information
✗Where all three mislead
- None of them knows your taste, your mood, or your evening
- All can be gamed by organised voting in either direction
- Sample size is usually invisible at a glance
- One decimal of precision hides enormous variance
The deeper limitation is that every one of these numbers is an average, and you have never once been an average viewer. A film that half the audience adores and half finds unwatchable produces a bland middle score that describes nobody’s experience. That is not a reason to ignore the scores; it’s a reason to read the disagreements, which are where the real information lives.
The Verdict
Pick one? IMDb · the widest crowd, with a built-in brigading counterweight
IMDb is the least wrong single source, and that’s the whole bar
Forced to keep just one, take IMDb: the sample is the largest, the weighting blunts organised voting, and the vote count is visible so you can judge how much to trust any given figure. Rotten Tomatoes tells you about consensus and Letterboxd tells you about enthusiasts, but IMDb tells you what the widest possible audience actually thought. Best of all, though, is two numbers: IMDb for the crowd and the Rotten Tomatoes audience score for the gap: if those two disagree sharply, that’s the film telling you what kind of film it is.
Ratings FAQ
Can a film really be 95% on Rotten Tomatoes and still be mediocre?
Yes, easily. The percentage counts how many critics were positive, not how positive they were. Check the average rating shown beside it. A high percentage with a mid-6 average means broad agreement that it’s fine, not that it’s great.
Why does the same film score 7.4 on IMDb and 3.2 on Letterboxd?
Different scales and different crowds. IMDb is a 1-10 vote from a broad public audience; Letterboxd is half-star ratings from film enthusiasts. The two scales don’t map onto each other, so compare each site’s score to its own average, not across sites.
Which of the three is easiest to game?
Audience scores are the softest target, because a coordinated campaign can move them before enough ordinary viewers have voted. IMDb’s weighting absorbs some of this, and Letterboxd’s smaller, more invested user base tends to notice and push back.
Should I trust a high score I don’t personally agree with?
Trust it as information, not as instruction. A high score means a lot of people got something out of it, which is real data. Whether it’s your something is a question only the review text and your own taste can answer.
Where does your own score fit alongside these?
Alongside them, not above them. Our number comes from a fixed six-category rubric, which makes it consistent rather than correct. Use it as a stable baseline when the public aggregates disagree with each other.