How I score coffee | Cup of Record

Scores run from 1 to 5 in half-point steps, and every coffee is brewed the same way: one V60 recipe, tasted black. 4.5 and up is exceptional — buy again without hesitation. 4 is very good, a happy rebuy at the right price. 3 to 3.5 is a solid daily coffee, unremarkable in the cup. 2 to 2.5 is flawed: finish the bag, never rebuy. 1 to 1.5 is a struggle to finish. No coffee has scored a 5. That slot stays open deliberately.

The scale

The scale is anchored to purchase advice:

4.5 and up — exceptional; buy again without hesitation
4 — very good; happily rebuy at the right price
3 to 3.5 — solid daily coffee, unremarkable in the cup
2 to 2.5 — flawed; would finish the bag but never rebuy
1 to 1.5 — struggled to finish

The method

Each coffee is brewed on the same V60 recipe (25 g, written up in the brew guide) and tasted black, usually over ice. Fixed method, fixed dose, no milk to hide behind. When a coffee is brewed several times across the bag, the score reflects the best honest cup it produced.

The sub-scores

Four dimensions get their own 1–5 mark: aroma (dry and wet), acidity (quality, then intensity), body (weight and texture) and finish (length and cleanliness). The headline score is a judgement rather than an average, since one loud fault, a harsh finish for instance, can sink an otherwise pretty cup.

Price honesty

Every review lists the price per 100 g in US dollars at the time of purchase. Expensive coffees get no leniency and no penalty: the score describes the cup, and the price sits beside it so you can decide whether the two belong together.

Who is scoring

Every score on this site comes from one person: me, Barnaby Robson. I have logged every specialty coffee I have bought since 2021, most of them tasted in Southeast Asia, where the roasting scene is better than its international reputation suggests. One palate scoring everything is the point of the site. Aggregated ratings smooth away the very differences a serious drinker cares about, while a single reviewer with a fixed method gives you a consistent yardstick you can calibrate against your own taste.