Drinks are commonly rated on a scale that runs to a hundred, and almost every published score sits in a narrow band near the top. The scale is doing something other than what its range implies, and the compression is the reason scores move prices.
The scale was borrowed from school marking
A hundred-point scale is intuitive to anyone educated in a system that graded work out of a hundred, which is why it spread faster than the twenty-point scales used in professional circles.
Readers already know what a mark in the nineties means and what one in the seventies means, so no explanation of the scale is needed.
That intuition is precisely what makes it misleading, because the working range in drinks is far narrower than in a school exam.
Almost nothing is scored below the eighties
Faulty or unpleasant products are usually not reviewed at all, since a publication has limited space and readers want recommendations.
The practical result is a scale that runs from roughly eighty to a hundred, with most commercially serious bottles clustered in a band of about a dozen points.
A gap of two points therefore represents a meaningful difference in that compressed band, while looking trivial on a scale that nominally runs to a hundred.
Small differences carry large commercial effects
Retailers use score thresholds in their buying and shelf-talker decisions, and distributors use them to decide what to carry at all.
A bottle crossing a round-number threshold gains visibility that a bottle a point below does not, which makes the boundary far more consequential than the underlying difference.
Producers are aware of this, and the pressure it creates on style is one of the standing criticisms of numerical scoring.
A score is one taster on one occasion
Scores are typically assigned in a flight of many samples, under conditions unlike those in which the drink will be consumed.
Bottle variation, storage history and the state of the taster's palate all move the figure, and none of them is visible in the published number.
Re-tasting the same wine months later commonly produces a different score from the same reviewer, which is a known feature rather than a scandal.
What the number is genuinely useful for
A score compresses a verdict into something sortable, which is its real function in a market with far more products than any buyer can taste.
It works best as a filter rather than as a ranking, separating the bottles worth considering from those that are not.
The written note beside the score carries the information that actually predicts whether a particular drinker will like the bottle, and it is the part most readers skip.