Reviews measure expectation as much as quality

A rating is the distance between what somebody was promised and what turned up.

Two people receive identical items. One rates it highly, the other poorly, and neither is lying. The difference is what they were led to expect. Enthusiastic descriptions raise expectations, and raised expectations are harder to meet, so an item can be rated down for being merely good. Modest descriptions do the opposite. This is why ratings should never be compared across items described in very different registers, and why a plainly described product often scores better than a more capable one that was oversold. The rating is measuring the promise as much as the object.

It also means the text of a review is more useful than its score. Read for what the person expected and what they got, and you can often work out whether the gap would exist for you. Somebody disappointed that a compact item is compact has told you the size is as stated. Somebody delighted that a simple item is simple has told you the same thing from the other direction. Both reviews are informative once you stop reading the number and start reading the mismatch between what was hoped for and what arrived.

There are also patterns worth noticing in how reviews arrive. Batches appearing close together, unusually similar phrasing, reviews that read like a description rather than an experience, or notes that a review was written after the writer was asked or given something. None of that proves anything about any seller, and it is not an accusation to make. It is simply a reason to weight those entries lightly and give more attention to reviews that describe an ordinary week of use in ordinary language. Detail is the tell worth trusting: people who have really lived with something tend to mention small, specific, slightly boring things.