Barista Magazine is your home for the worldwide community of coffee and the people who make it.
Issue link: http://baristamagazine.epubxp.com/i/118056
"I t's like trying to score blue cheese on a cheddar form. How are you going to score blue cheese on a cheddar form?" So posed Jason Long—the problem facing any cupper with a washed standard and an alternately processed coffee on the cupping table. But Jason, co-owner of Café Imports, was not asking this rhetorically. He wanted an answer and was raising the issue as a project. How could we formally acknowledge that non-washed-process coffees are not defective simply because scoring standards tailored to their washed counterparts tend to define them as such? Over the course of the last year, I've followed and then taken the lead in developing distinct standards for the cup scoring of variously processed coffees. Today we use five separate scoring standards in our green cupping at Café Imports: Washed, natural, wet hulled, Brazil, and decaf. What follows is a description of how these standards came about, the thinking behind them, and how we now use them in our daily analysis of green coffee. Rational Numbers Allow me to frame the situation. Arriving at a green cupping, one receives a clipboard with a scoring form (or a tablet with the same). The form is broken down by attributes, including but not limited to: cleanliness, balance, sweetness, acidity, mouthfeel, flavor, aftertaste, and some sort of overall score. Scoring categories are on a 1–10 or 1–8 scale, and add up to a score between zero and 100 points. While the scoring method of putting coffee in your mouth and doing your best to note the experience remains subjective, a proper cupping lab should have a working standard delineating how those experiences are rated. An 84 needs to be tied to a definition in order to mean anything as 84. When people visit us to cup, I'm curious if I see them scoring without having looked over the standards that we have posted. This is not a my-way-or-the-highway thing. Most people can taste and then describe their results just fine. However, after having worked on our coffee scoring standards for the last year, I often wonder what exactly their numerical scores are referencing. I like to assume these folks are scoring on the Specialty Coffee Association of America's (SCAA) scale. The SCAA describes Good, Very Good, Excellent, and Outstanding for attribute scores, respectively, in the 6, 7, 8, and 9-point bands. On the 100-point scale, sub-80 coffees are categorized as Below Specialty Quality. In the Specialty spectrum, there are three categories: 80–84.99 is Very Good; 85–89.99 is Excellent; and 90+ is Outstanding. However, it often seems that people are just scoring from 80–90 on an "I like this good-better-best" scale, including 76 for Bad, 90–92 for Incredible, and keeping 93+ absolutely off limits. Mysterious Mathematics So it's tough. We all have preferences (and conceits). We cup together. We talk about our scores. Referring them to a defined standard, we start to come into a state of calibration. Things are going well...and then someone tosses something wet hulled (WH) onto the end of a Yirg 1 table. It's not ideal, but the realities of workload occasionally dictate such things. This WH has less sugar, and is lower both in quantity and quality of acidity. Instead of floral and apricot flavors, we taste green pepper and earth, while the clean and juicy viscosity of the Y1s is replaced by oily weight. Across the board the WH is looking pretty bad. It happens like this: We start seeing ranges. One person says it's a 68, and defends it as above. Someone else says 76, explaining that they didn't want to kill the thing just because it wasn't a Yirg. A third cupper says 84, claiming it was pretty solid attribute by attribute, for a wet-hulled coffee. We add it up and get a 76-point average, which happens to be equal to the person who essentially said that they preferred not scoring the coffee—arguably the least meaningful of the three. Next thing you know, there's a room full of cuppers constantly figuring "for a ___" into their scores. And again the question arises: where are the numbers coming from? What do they mean? We don't have "for a ___" defined on our grading scale. In practice, the implicit "for a ___" works well enough when the cupping panel is small, and very consistently cupping together from one day to the next. The "for a ___" develops organically, and is shared. The dynamic changes considerably when those few individuals working closely grow into a larger group, working disparately with less time for tacit calibration. Standard Aphorisms of the Dedicated Cupping Lab If 68 and 84 can both be accurate scores for a single coffee, then we have a decision to make. We need either to define our cuppers (So-and-So hates earthy, kills ferment, or loves everything, etc.), or better define the terms of our scoring. While 68/84 is an exaggerated example, we finally encountered enough of this here at Café Imports to seek a more stable solution than "So-and-So always ____." We sought the expert consultation of Paul Songer. For a very cold week in January of 2012, Paul worked with Jamin Haddox (who was then managing me in our quality control www.baristamagazine.com 85