Two experienced reviewers can taste the same bottle and reach opposite conclusions. Some of that is preference, but a substantial part is physiological: they are not receiving the same information.
Detection thresholds vary between people by large margins
The concentration at which a compound becomes noticeable differs enormously between individuals, and for some compounds a portion of the population cannot detect them at all.
Cork taint is the clearest case, since the threshold spread means one taster can find a bottle obviously musty while another finds it merely dull.
Neither is mistaken. The compound is present, and one nose is registering it while the other is not.
Specific anosmias are common and usually unnoticed
Being unable to smell one particular compound while having otherwise normal perception is widespread, and most people never discover which ones they miss.
A reviewer with a blind spot for a green pepper compound will consistently find wines fruitier than colleagues do, without any way of knowing why.
Professional panels manage this by using several tasters and treating a single dissenting note as information rather than as error.
Bitterness sensitivity divides the population
Sensitivity to certain bitter compounds is strongly genetic, and people at the sensitive end experience tannin, hops and bitter botanicals far more intensely.
That single difference shifts a taster's whole preference profile toward or away from bitter styles, and it shows up as a consistent bias across hundreds of reviews.
Reading a reviewer over time reveals the pattern more reliably than any individual score does.
Vocabulary diverges even when perception agrees
Tasting notes are descriptions by analogy, and the reference points come from each taster's own experience of food and place.
Two people can perceive an identical aroma and reach for different fruits to describe it, which reads as disagreement and is not.
Structured tasting vocabularies exist to reduce this, and they work within a trained panel while doing little for published reviews written for a general audience.
What follows for reading reviews
Consistency matters more than authority. A reviewer whose descriptions have matched your own experience before is a better guide than one with a larger reputation.
Disagreement between two credible reviewers usually indicates a drink sitting near a threshold, which often means it is genuinely divisive rather than that one of them is wrong.
The most useful reviews say what the drink is like rather than how good it is, because the description survives the difference in palates and the verdict does not.