How we evaluate
Methodology
Every claim on Yuri Matcha carries a source and a label saying how we know it. This page documents where the data comes from, what each label means, and how much of the catalogue each one covers.
We catalogue 574 blends. 73 carry scores across all five flavour dimensions, and 12 of those come from tasting we did ourselves. The other 61 are built from producer documentation and published tasting notes, and every page says which. We would rather publish the coverage than imply we have tasted everything.
Scoring Framework
Five dimensions, scored 1–10. 73 of 574 blends carry scores.
Scored blends receive ratings across five flavor dimensions as integers from 1 to 10, where 5–6 is average for the category rather than bad. Scores come from one of two places and each blend says which: a tasting we conducted, or published notes from the producer or a documented third party. Blends without sufficient evidence are left unscored rather than estimated, which is currently 501 of 574.
Umami旨味
Measures the savory, broth-like depth that distinguishes high-grade ceremonial matcha. High umami (8–10) indicates significant shade-growing time, careful stone-milling of high-quality tencha, and ideal first-harvest timing. Low umami (1–3) is typical of culinary-grade or latte-focused blends.
- 1–3: Minimal, culinary-grade
- 5–6: Moderate, everyday drinking
- 8–10: Deep, ceremonial-grade
Bitterness苦味
Measures astringency and the intensity of catechin-driven bitterness. Note that bitterness is not a defect — it is a stylistic axis. High-bitterness matcha (7–9) suits drinkers who prefer a bracing, coffee-like character. Lower bitterness (2–4) is ideal for beginners or latte use.
- 1–3: Mild, beginner-friendly
- 5–6: Balanced, slight edge
- 7–10: Bold, bracing astringency
Sweetness甘味
Measures natural sweetness — not from additives, but from amino acids (particularly L-theanine) and the absence of harsh astringency. Sweet matcha (7–9) is often perceived as creamy or vegetal. This score is inversely related to bitterness for most blends.
- 1–3: Dry, mineral, austere
- 5–6: Faintly sweet
- 8–10: Notably sweet, creamy
Bodyボディ
Measures texture, mouthfeel, and the weight of the liquor. High body (8–10) indicates a dense, coating texture suitable for koicha. Low body (2–4) produces a lighter, more tea-like texture well-suited to usucha or matcha lattes.
- 1–3: Thin, tea-like
- 5–6: Medium, versatile
- 8–10: Dense, coating, koicha-suitable
Color色
Measures the vibrancy of the prepared liquor and dry powder. High color (8–10) indicates intense, vivid green — a signal of quality nitrogen-rich tencha and careful milling. Low color (2–4) indicates yellower or duller powder, common in lower-grade or older stock.
- 1–3: Dull yellow-green
- 5–6: Mid-green, standard
- 8–10: Vivid, deep green
Tasting Protocol
Standardized preparation used for all in-house tastings.
To minimize variables and enable score comparability across blends, all in-house tastings follow this protocol. Deviations (e.g., different water temperatures or preparation styles) are noted explicitly in the tasting data for that blend.
| Parameter | Standard Value |
|---|---|
| Amount | 2 g |
| Water volume | 60–70 ml (usucha) |
| Water temperature | 70–80 °C |
| Whisk type | Bamboo chasen, 80–100 tines |
| Sifting | Always |
| Bowl (chawan) | Ceramic, wide-mouth |
| Foam quality | Rated 1–5 |
| Evaluation window | 1–5 min after preparation |
| Water source | Filtered, low-mineral |
Tasting notes include: foam quality, astringency window (how quickly bitterness develops), finish length, and suitability assessments for usucha, koicha, and matcha latte preparation. All tasting data is stored in the database and visible on blend pages where available.
Evidence Standards
What counts as verified, researched, and tasted.
Tasted
The blend was prepared and evaluated in person using our standard tasting protocol. All five flavor scores are present, tasting-protocol data is recorded, and at least one detailed review exists in the database.
- Direct purchase and preparation
- Standardized tasting protocol followed
- Flavor scores assigned from tasting session
- Tasting data recorded (grams, temp, whisk, foam, astringency)
Verified
All factual claims have been cross-checked against at least one primary source (official brand documentation, verified retailer listing, or peer-reviewed publication). Flavor scores may come from verified third-party tasting notes with methodology parity.
- Origin, cultivar, and harvest season confirmed from primary source
- Price and availability confirmed from official retailer
- Source URLs logged with access dates
- Confidence rating assigned per claim
Researched
Factual data has been collected from secondary sources (press coverage, aggregator sites, distributor listings) but not verified against a primary source. Flavor scores are either absent or marked as estimates.
- Core facts collected (name, origin, price range)
- At least one source URL logged
- No confirmed primary-source verification
- Scores absent or clearly marked as estimates
Draft
A page stub exists with basic identifying information but has not yet been researched or verified. Draft pages are indexable but display a prominent warning and are excluded from recommendation algorithms.
- Entity exists in database
- Slug and name confirmed
- Research not yet completed
“Not Enough Evidence to Rate” Policy
Yuri Matcha does not assign placeholder scores or speculative ratings. When evidence is insufficient, scores are left blank and a visible notice is shown on the page. This applies when:
- We have not tasted the blend and cannot locate a verified third-party tasting note
- The blend has been reformulated and existing reviews predate the change
- The blend is discontinued and current stock cannot be confirmed
- Conflicting sources exist and cannot be resolved with available evidence
We believe an honest “unknown” is more useful than a fabricated score. Pages with insufficient evidence are still published because factual data (origin, price, certifications) has value even without flavor scores.
Frequently Asked Questions
How does Yuri Matcha score matcha blends?
Each blend is scored across five dimensions on a 1–10 scale: umami (savory depth), bitterness (astringency intensity), sweetness (natural sweetness), body (texture and weight), and color (vibrancy of green). Of 574 catalogued blends, 73 carry scores across all five dimensions. 12 of those come from tasting we conducted; the rest are built from producer documentation and published tasting notes, and every blend page states which. Blends without sufficient evidence are left unscored.
What does 'verified' mean on a blend page?
'Verified' means all factual claims on the page have been cross-checked against at least one primary source (official brand documentation, verified retailer listings, or peer-reviewed tasting notes). The verification date and method are displayed in the Trust Block on every entity page.
What does 'tasted' mean?
'Tasted' means the blend was prepared and evaluated in person using our standardized tasting protocol. Tasted blends receive full flavor scores and tasting-protocol data (grams, water temp, whisk type, sifting). Blends marked 'researched' have verified factual data but no in-person tasting.
How often are blend pages updated?
Pages are re-verified when prices change, availability changes, a new harvest season begins, or a blend is discontinued. The last-verified date and any changes are recorded in the changelog visible on each blend page.