Ratings
Why a 7.8 with 40 Votes Isn't the Same as a 7.8 with 40,000
Updated Sep 17, 2026
Two titles both show a 7.8. One has forty ratings. The other has forty thousand. Most apps show you the score and hide the sample size that makes one of those numbers trustworthy and the other one noise — and once you know to look for it, you'll never read a rating the same way again.
A score is an average, and averages lie with small samples
Rating sites like TMDB store two numbers behind every score: the average itself, and the count of votes it's built from. A brand-new release with forty ratings can swing from 6.5 to 8.0 on the strength of ten more people voting — one coordinated review-bomb, one enthusiastic fanbase turning out early, or just plain small-sample luck. A title with tens of thousands of votes barely moves when a hundred more come in. The average is the same kind of number in both cases; how much it deserves your trust is not.
Why a niche or brand-new title is the case that breaks this
This isn't a hypothetical edge case — it's most of the long tail of any real catalog. A film that just left festivals, a foreign series with a small but devoted audience, an old show nobody's rated in years: these are exactly the titles most likely to carry a rating built on a handful of votes, and exactly the titles where a misleadingly high or low number does the most damage. A 9.2 from six votes tells you six people loved it. It does not tell you what you'll think.
What a system that takes this seriously actually does
Two houses to keep in order:
•
Don't confuse "no rating" with "unrated." A title that hasn't been scored yet at all is a different situation than a title that has plenty of votes and a mediocre score — collapsing them into the same "0" or blank display makes an unrated title look bad rather than simply unmeasured.
•
Don't let a token handful of votes carry the same visual weight as a mature sample. Some systems solve this with a minimum-votes floor: below a threshold, a title just doesn't get treated as "rated" for ranking purposes, however its raw average looks in isolation.
CouchReel applies exactly that kind of floor internally: a title with fewer than 20 votes behind it is written into the catalog as unrated rather than displaying a raw average built on a handful of opinions, and unrated titles simply don't get boosted or buried by recommendation ranking — they sit outside that math rather than skewing it. It's a small, quiet rule, but it's the difference between a recommendation engine that occasionally chases a six-vote fluke and one that doesn't.
The habit worth keeping
Next time a score looks surprisingly high or low for a title you've heard little about, check whether the app shows a vote count next to it. If it does, a small number next to a strong opinion is worth exactly what it sounds like: one strong opinion, not a consensus.