What we are building
tea.degree exists to give Chinese tea a working evaluative grammar. Wine has had the WSET systematic approach since 1969 and the UC Davis 20-point scale since 1959. Coffee has the SCA cupping form and the Counter Culture flavor wheel. Chinese tea, despite a documented sensory tradition reaching back to Lu Yu’s Chá Jīng (茶经, ca. 760 CE), still mostly relies on informal vocabularies that do not survive translation between a Wuyi cliff tea producer, a Saint Petersburg buyer, and a Berlin sommelier. That gap is what this site addresses.
Our scope is narrow on purpose. We cover Chinese tea only — the six categories codified in GB/T 30766-2014 (green, yellow, white, oolong, black/hóngchá, dark/hēichá), plus re-pressed forms and aged stock. We do not assess matcha, masala chai, Ceylon or Assam orthodox, or Kenyan CTC except as comparative reference points when a Chinese counterpart needs context. A Lapsang Souchong from Tongmu village (桐木村) sits inside our scope; an Indian smoked black does not.
The instrument itself has three components. The sensory wheel is a 16+64 segment structure — sixteen primary descriptors (floral, fruity, roasted, mineral, vegetal, marine, woody, dairy, spiced, honeyed, medicinal, earthy, smoky, sweet-grain, animalic, off-notes) subdividing into sixty-four specific terms with Chinese, English and Russian glosses. The ten-axis scoring rubric covers dry leaf appearance, wet leaf, liquor colour, aroma intensity, aroma complexity, flavour intensity, flavour complexity, texture (口感), aftertaste (回甘 and 喉韵), and overall harmony — each scored 1-10 with anchor descriptions, summing to a 100-point reference that maps cleanly to the GB/T 23776-2018 sensory evaluation procedure used in Chinese export grading.
We are not a shop. We do not rank teas commercially, we do not accept sample fees, and we do not publish ratings of named products without the producer’s prior knowledge. What we publish are tools: the wheel as an open vocabulary, the rubric as a scoring form, blind-tasting protocols, a self-calibration test, and a comparison radar for overlaying three to five teas on the same ten axes. If the instrument is useful, the assessments people write with it will be useful. That is the whole proposition.
How the project started
tea.degree began in 2022 as an internal scoring sheet at Teamotea’s Saint Petersburg cupping room — a single A4 page taped above the bench, used to record arrivals from Yunnan, Fujian and Anhui. The sheet had ten columns and no wheel; tasters wrote descriptors in whatever language came first, which produced unusable archives. A 2021 cupping of seventeen Wǔyí Yán Chá (武夷岩茶) samples generated three different spellings of ‘mineral’ in three different alphabets and no agreed vocabulary for the yán yùn (岩韵) — the rock-rhyme aftertaste that defines the category.
The rebuild took eighteen months. We mapped the existing Chinese sensory lexicon from GB/T 14487-2017 (tea sensory terms standard), cross-referenced the 2016 World Tea Flavour Wheel from the University of California Davis Coffee Center adaptation, and conducted thirty-one interviews with producers in Anxi, Wuyishan, Jingmai, Yiwu and Fenghuang between November 2022 and August 2023. Every term on the published wheel has a documented origin: a standard, an interview, or a peer-reviewed paper. Terms that exist only in marketing copy were excluded.
Editorial principles
Three rules govern what gets published. First, every claim about a tea’s sensory profile is tied to a recorded session — date, water source and temperature, vessel volume, leaf weight, infusion times, and the tasters present. We use the gōngfū protocol from GB/T 23776-2018 (5 g leaf, 110 ml gàiwǎn, 100 °C for most categories, 85 °C for green and yellow) as default, and we record any deviation.
Second, descriptors are checked against the vocabulary library. A taster who writes ‘leather’ must accept the library’s anchor — the cured-hide note documented in aged shú pǔ’ěr from the 1990s Menghai stock, distinguished from ‘animalic’ (a separate segment of the wheel covering wet-fur and barnyard notes more common in mishandled shēng). If a new descriptor is needed, it goes through review before entering the library.
Third, we publish the rubric scores with the prose. Numbers without notes are unfalsifiable; notes without numbers are unverifiable. Both, together, can be argued with — which is the point. Disagreements about a score are how the instrument improves.
Sourcing and sample handling
Samples assessed on tea.degree come from three channels: producers who send material directly for a published evaluation (their identity disclosed at the top of the entry), blind samples drawn from Teamotea’s commercial inventory by an operations staff member not involved in the evaluation, and self-purchased reference lots used for calibration. We do not accept payment in exchange for assessment.
Storage matters as much as sourcing. Reference samples are held in food-grade aluminium-foil bags with one-way valves, in a dedicated room at 18-22 °C and below 60 % relative humidity, away from spice and coffee storage. Aged pǔ’ěr and liù bǎo reference lots are kept separately in unglazed clay jars in a Saint Petersburg cellar at controlled humidity, with provenance documents from the original purchase. The room is logged; samples are dated; nothing tasted is older than its label claims without a note explaining the discrepancy.
Transparency and corrections
Every scoring entry on tea.degree carries the taster’s identifier, the session date, and the protocol used. When we get something wrong — a misidentified cultivar, a misread harvest date, a descriptor that survey readers find unrecoverable in their own cup — we correct in place and log the change at the foot of the page with the original wording preserved. The correction log is public.
We do not silently update assessments to track a producer’s commercial trajectory. A 2023 score for a 2023 spring Bái Háo Yín Zhēn (白毫银针) stands as a 2023 score; if we re-taste the same lot in 2026, that becomes a separate entry with its own date, because the tea is no longer the same tea. Storage is part of identity for white, dark and aged categories — a fact the rubric explicitly accommodates through the wet-leaf and aftertaste axes.