Every score on this site can be recomputed by hand from numbers printed on the same page. That is the whole point of publishing a method rather than a promise.
Weighted mean, rounded half-up to a whole number. The weights sum to 100%.
99×0.45 + 98×0.20 + 99×0.15 + 93×0.10 + 81×0.10 = 96.40 → 96
Those five numbers are printed on Marchetti & Braithe’s page. So are everyone else’s. If a score on this site cannot be reproduced from the sub-scores beside it, that is a bug and we want to hear about it.
A 0–100 rating for a whole producer, not a single bottle. The weighted mean of the five sub-scores below.
45% of the Crux Score. Blind pours scored in the glass, and the heaviest single input.
20% of the Crux Score. Spread across vintages and bottlings. A high floor beats a high ceiling.
15% of the Crux Score. Pour size, staff knowledge, reservation friction, what it costs to sit down.
10% of the Crux Score. The score read against this producer’s own average bottle, not the region’s.
10% of the Crux Score. Published additions, estate-fruit percentage, ownership. Silence scores low.
The difference between this producer's Crux Score and the median of whatever list you are looking at. The median is printed in the column header.
A shared rank. =37 means this producer ties with others on the same score.
Every wine assessed without the label visible. Producers not blind-tasted throughout cannot reach the top band.
Tasters log against a producer ID. The billing table is a separate schema and is not exposed to the scoring tools.
There is no featured tier and no promoted slot. The order is the order.
We do not take money to remove or soften a review.
A claimed listing at $149/month — the ability to correct factual details such as hours and grapes. Everyone who pays is published.
Factual errors — hours, address, grape, ownership, estate percentage — are corrected on request from anyone, customer or not, and the page’s revision date changes. Scores are not corrected on request; they are re-cut when we taste again. The full dataset is published so a disagreement can be about numbers rather than vibes.
The Crux Score is a weighted mean of five sub-scores: wine quality 45%, consistency 20%, tasting room 15%, value for price 10%, transparency 10%, rounded to the nearest whole number. Every sub-score is printed on the producer's page, so any score on this site can be recomputed by hand from numbers we publish.
85 is the median for Paso Robles: half the appellation scores above it and half below. The scale is 0–100 and is not curved per region — a Paso Robles 85 and a future Napa 85 are meant to be the same wine-buying proposition.
No. 3 producers pay $149 a month for a claimed listing, all of them are published on the customer list with their scores, and none is in the top ten; the highest-ranked of them sits at =37. Tasters log against a producer ID; the billing table lives in a separate schema with no join path to scoring.
A producer whose wines were not all poured blind cannot reach the top band; the cap is applied mechanically and the disclosure block on the page says so. 157 of 187 producers here were blind-tasted throughout.
Producers on the same Crux Score share a rank, shown with a leading equals sign — =37 means this producer and others all score the same. The next rank skips accordingly.
Scores are re-cut when a new vintage is assessed. Every page carries the date it was last revised, and superseded scores are kept in the vintage record rather than overwritten.