A single score often misleads. What averages hide, critic versus audience scores, how ratings shift over time, manipulation and clues beyond the score.
A film rated 7 where almost everyone gave it 7 and a film averaging 7 from a mix of 10s and 4s are completely different. The first is reliably decent; the second divides people. The second could become a favourite or a waste of an evening. Where the distribution is visible, reading it alongside the average helps. If a service does not show the distribution, reading a few of the highest and lowest reviews serves as a substitute.
Critics weigh construction: direction, structure, context. Audiences weigh enjoyment and satisfaction. That is why some films show a wide gap between the two. The question is not which is right but which matches what you want tonight. High critic and low audience scores often point to an experimental or slow film; the reverse often points to reliable genre entertainment. It is a tendency, not a rule.
Immediately after release, the people who were waiting for the film watch and rate it first, so scores often start high. As general audiences arrive, scores can settle downward. Some films go the other way, undervalued at first and reassessed later. An opening-week score is still an unsettled number. Ratings for older films are often given by people who sought them out because they are already regarded as classics, so the same number means something different from a new release's score.
Organised campaigns to boost or sink a specific title exist. A large number of ratings before release, or an unnatural concentration of only 1s and 10s, is worth treating with suspicion. In those cases, reading a few actual reviews is faster than studying the number. Reviews that are extremely short and repeat near-identical sentences are another warning sign.
The most useful approach is finding people whose taste overlaps yours. Once you know a reviewer or publication that rated several of your favourites similarly, their verdict predicts your reaction far better than a global average. Keeping different references for different genres also works. Keeping a record of the scores you give the films you watch lets you start seeing, over time, the patterns in when expectation and reality diverged.
There are reasons the same film scores differently on different services. A ten-point scale and a five-point scale do not convert proportionally, and some services tally only 'good' or 'bad' and show the share of positive verdicts. In that case 90% means 'most people thought it was fine' rather than 'outstanding'. The user base differs too: sites where heavy film-watchers gather tend to score harshly, and those dominated by general audiences tend to be generous. So rather than comparing scores across services directly, it is more accurate to compare a film against other films on the same service.
When a rating is ambiguous, other information tells you more than the number. In particular, checking whether the reasons given by low scorers would bother you too speeds up the decision considerably.
Looking at a rating in the following order makes you less likely to be pushed around by it. First check whether enough people have rated it, then whether the distribution is concentrated or split. If it is split, read a few low-score reviews to see whether their reasons apply to your taste. Finally, look at the gap between critics and audiences, and decide based on whether tonight you want craft or light entertainment. Once it becomes habit, the whole process takes a few minutes.