From a master’s word to a number two people can check
For most of Chinese tea’s documented history, quality was decided by one expert’s pronouncement — a buyer tasted a dozen lots and picked the best against a mental map built over decades. The method produced tasters whose names still carry weight in Hangzhou and Fuding, but it did not travel: a verdict formed over thirty years in one cupping room could not be handed to a new grader in an afternoon. Scoring rubrics exist to close that gap, trading a little of the master’s nuance for a number a second taster can reproduce.
The reference point is GB/T 23776, China’s national methodology for the sensory evaluation of tea, finalised in 2009 and revised in 2018. It defines a 100-point scale, but the split is not fixed across categories — it shifts with what each style is built to show. For green tea the weighting runs appearance 25, liquor colour 10, aroma 25, taste 30, infused leaf 10; oolong pushes aroma and taste to 30 points each; dark tea lifts taste as high as 35, reflecting its role as a post-fermented, mouth-coating brew. The companion article 100-point vs 20-point rubric — when each is right works through these weightings category by category, including the mid-point clustering that erases a rubric’s discrimination if a panel is not calibrated against reference samples first. Trade buyers, meanwhile, often reach for a faster 20-point card — four criteria at up to five points each — built for a go/no-go decision in the minutes between one sample and the next, not for a decimal-place argument over a competition medal.
Tea.academy examinations add a third layer, adapting the national standard for a training and certification setting rather than a factory floor. The article The official tea.academy rubric explained, criterion-by-criterion walks through its six weighted dimensions — dry leaf, aroma across a hot-warm-cold sequence, liquor colour, taste and mouthfeel, infused leaf, and a closing aftertaste-and-harmony score — built to make candidates name what a passing palate detects, not just feel it. Chen Hui Yi, who helped write that rubric, puts the difference in teaching terms: a 100-point card rewards a taster who can defend a decimal point; a 20-point card rewards one who can make a decision before the next sample arrives. Neither replaces the other — a buyer moving between a Fuding white-tea market and a formal cupping room needs both fluencies, and switching between them without noticing which one you are holding is the most common scoring error this topic exists to prevent.
None of this replaces a trained palate — it gives that palate a shared vocabulary. A shēng pu-erh scored by a panel in Hangzhou and the same lot re-scored at a Guangzhou cupping room should land within a point or two of each other, provided both panels calibrated against the same reference set first — the discipline covered in this site’s calibration protocol topic. A rubric is a reduction, never the tea itself. What it buys is a conversation two tasters, a thousand kilometres apart, can actually have.