IMDb vs Rotten Tomatoes: which score should you trust?

A movie sits at 94% on Rotten Tomatoes and 6.4 on IMDb, and the comments underneath fill up with people calling critics out of touch, or calling audiences philistines. Everyone is arguing about numbers that do not mean what they assume.

The two scores are not two opinions on the same question. They are answers to different questions, and once you know which question each one is answering, a wide gap between them tells you more about a movie than either number does alone.

The Tomatometer counts heads

This is the part most people get wrong, so it is worth being blunt about it. The Tomatometer is not an average rating. It is the share of approved critics whose review was, on balance, positive. Every review gets sorted into one of two buckets, fresh or rotten, and the percentage is how many landed in the first bucket.

A five-star rave and a shrugging three-star notice count identically. Both are fresh. Both add the same amount to the number.

Follow that through and something strange falls out. Imagine a movie where every single critic files a mild, faintly bored six-out-of-ten. Nobody would put it on a year-end list. Nobody would watch it twice. It scores 100%. Meanwhile a movie that half the critics call the best thing they have seen in years, and the other half call self-indulgent rubbish, lands at 50% and looks like a failure.

So the Tomatometer measures agreement. A high score means critics converged, not that they were excited. That is genuinely useful information, it is just not the information people read off it.

IMDb averages enthusiasm, from whoever showed up

IMDb runs the other way. Every user picks a number from one to ten and the site reports a weighted average, with the weighting there to blunt organised vote campaigns. Enthusiasm survives the maths. A ten and a six are not the same input.

The catch is who votes. Rating a movie on IMDb takes deliberate effort, and the people who bother skew towards the ones who felt something, in either direction. A movie that arrives to broad indifference collects fewer ratings, and the ones it does collect come disproportionately from people with a reason to log in.

You can see the shape of this in the distributions. Genre movies with devoted audiences carry a visible cluster of tens. Movies that annoyed people carry a matching spike at one. The average sits somewhere between two groups that never agreed, describing a middle position nobody actually holds.

Reading the gap

Once you stop treating the two numbers as rivals, the distance between them becomes the interesting signal.

High Tomatometer, middling IMDb. Critics agreed it was competent. Audiences were not moved. This pattern turns up constantly around well-made dramas that do everything correctly and leave no residue. It also turns up around movies whose pleasures are formal rather than emotional, where the craft is legible to someone watching professionally and invisible to someone watching at eleven at night.

Low Tomatometer, high IMDb. Critics split, and the people the movie was aimed at loved it. Horror lives here permanently. So does most science fiction with a cult, most action cinema with a stylistic signature, and nearly everything that was reviewed as one thing and found its audience as another. A 55% here is often more promising than an 85% elsewhere, because the low number is measuring disagreement and the disagreement is the point.

Both high. Usually safe. Also usually a movie you have already heard of, which limits how much the score is telling you.

Both low. Believe it.

Metacritic is doing a third thing

Worth knowing, since it turns up next to the other two. Metacritic normalises each critic’s review to a 0 to 100 score and averages those, weighting some publications more heavily than others. It keeps the intensity that Rotten Tomatoes throws away and it stays with critics rather than users.

That makes it the most informative of the three for how much critics liked something, and it also makes it the one with the narrowest range in practice. Almost everything lands between 40 and 80, so a 90 means considerably more than a 90% Tomatometer does.

None of this tells you whether you will like it

Here is the limitation that no aggregate solves. All three scores answer the question “did people in general rate this highly”. None of them answer “will this land for me, tonight”.

Those come apart more often than the numbers suggest. Half the movies worth your evening are ones the population average is lukewarm about, because the population average includes everyone who was never the audience. Aggregate ratings are a floor, useful for avoiding something actively broken. They are poor at surfacing the movie you will still be thinking about on Thursday.

That is the job Filmatic is built for. We hold IMDb and Rotten Tomatoes as filters, because a hard floor is genuinely useful when you are tired and do not want to gamble. The pick itself comes from your own swipes, and the Filmatic Score is our editors’ view rather than a crowd average, which means it can back something divisive instead of regressing to whatever offended nobody.

The short version

Rotten Tomatoes tells you whether critics agreed. IMDb tells you how strongly people who bothered to vote felt. Metacritic tells you how much critics liked it. A movie that is high on one and low on another is not a contradiction to resolve, it is usually a movie with a specific audience, and the only remaining question is whether that audience is you.