# Cupping (/en/docs/foundation/phase-5/chapter-26)
## Phenomenon [#phenomenon]
You go to a coffee shop's public cupping for the first time. A long table holds rows of small bowls — five per coffee — each with ground coffee in it. Someone with a scale and a kettle pours water bowl by bowl. When a timer goes off, everyone uses a spoon to push aside a layer of floating grounds on each bowl and leans in to smell.
A little later, everyone starts scooping up coffee and sucking it into their mouths with a loud "sssslurp", then spitting it into a cup.
You are a bit confused. These people brew meticulous pour-overs every day. So why, when it is time to "seriously evaluate a coffee", do they use the simplest method there is — coffee in a bowl, add water, wait?
## Variables [#variables]
To compare two coffees, the coffee has to be the only thing that differs. But the taste of a cup carries plenty of other things with it:
| Variable mixed in | How it gets in the way of the comparison |
| ----------------- | ----------------------------------------------------------------------------------------- |
| Who brews | How they pour, how much agitation — different for every person and every brew |
| Brewing method | Immersion or percolation, paper or metal — extraction and texture differ (Chapter 25) |
| Recipe | Ratio, grind, temperature, time |
| Roast | How dark, and how long ago |
| When you taste | Hot or cooled down |
| Who tastes | Knowing which coffee it is, what it cost or who roasted it shapes perception (Chapter 35) |
Cupping is a protocol for **holding those variables still**.
## Experiment [#experiment]
A thought experiment: you want to compare coffees A and B, and there are two ways to do it.
* Way one: you brew A as a pour-over, a friend brews B as a pour-over, and you swap and taste.
* Way two: both coffees get the same dose, same grind and same water, steeped in several identical bowls, started at the same time and tasted at the same time.
If you find B better, which way lets you say more confidently that "the coffee is better"?
A second thought experiment: five bowls of the same coffee.
Why five bowls instead of one? Should all five taste the same? If one tastes different, what does that tell you?
One last one: taste the same bowl just after it is poured and warm, and again when it has nearly cooled.
Will it taste the same? What would you miss by tasting at only one temperature?
## Observation [#observation]
In way one, "B is better" might mean B is better — or that your friend brews better, or that you used different recipes. The results are tangled together. Way two removes the brewer and the recipe, so whatever difference remains is more likely to come from the coffee itself.
The second experiment: five bowls are there to check **uniformity**. If one bowl tastes clearly different — say, musty or with an off-ferment note — the cupping records it as a uniformity or defect problem: a few bad beans in the lot may have landed in that one bowl. With a single bowl, you might wrongly condemn the coffee, or let it off.
The third experiment: the taste changes. One study had tasters evaluate the same coffee at different temperatures: at cooler temperatures, descriptions like sour and sweet came up more; at hotter temperatures, overall intensity, roasted flavor and bitterness stood out more. That is why cuppers taste repeatedly as the coffee cools.
## Model [#model]
### Cupping is a measuring instrument [#cupping-is-a-measuring-instrument]
Think of cupping as an instrument with one design goal: **make differences between coffees show, and make every other difference disappear**.
Taking the cupping protocol the Specialty Coffee Association used for many years as an example, it specifies roughly this:
| What the protocol holds still | How |
| ----------------------------- | ------------------------------------------------------------------------------------------------------------------- |
| Roast | Roasted within 24 hours of cupping, rested at least 8 hours; roast degree checked with a color meter |
| Ratio | 8.25 g of coffee per 150 mL of water, scaled to the bowl's volume |
| Grind | Slightly coarser than for paper-filter drip, ground just before cupping, water poured within 15 minutes of grinding |
| Water | Clean, odor-free, poured at about 93 °C |
| Time | The floating layer of grounds steeps for three to five minutes before being pushed aside (the "break") |
| Uniformity | At least five bowls per coffee |
| When to taste | Tasting starts once the coffee cools to about 70 °C, then repeats as it cools |
From 2024 the association replaced this protocol with a new assessment method, but the idea of holding variables still has not changed.
### Why immersion, of all things [#why-immersion-of-all-things]
Cupping uses immersion rather than a more refined pour-over, for the reason Chapter 28 set up: in immersion, time is set directly, grind does not drag the time along with it, and there is almost no flow to control. No pouring technique means no "who brews better" variable. One person can prepare dozens of bowls in a few minutes, and every bowl runs on the same logic.
The cost: there is no paper in a cupping bowl, so fines and oils stay in, and it tastes heavier and less clean than a pour-over. Cupping is not after the best-tasting cup. It is after **the fairest comparison**.
### Slurping helps the nose [#slurping-helps-the-nose]
That loud slurp looks theatrical, but its purpose is plain: pull the coffee in together with air, spread it across the mouth and push the aroma toward the back. As Chapter 35 explained, when you swallow and breathe out, aroma travels from the back of the mouth into the nose, and much of what we call "taste" comes from there. Slurping is a standard cupping move; as for exactly how much it strengthens perception, we did not find a study that measured it directly. Spitting is so that tasting dozens of bowls in a morning does not mean swallowing too much caffeine.
### Instruments have error too [#instruments-have-error-too]
Cupping holds many variables still, but one it cannot hold: **the taster**.
* One study had four certified cuppers and a trained descriptive sensory panel evaluate the same 13 coffees, and found little overlap between what the two methods reported; the flavor words the cuppers wrote also rarely agreed with one another. The researchers concluded that cupping and descriptive analysis give different information and cannot replace each other.
* The coffee trade has long relied on "expert cuppers" for grading, and their judgments rest largely on personal experience; standards also differ between institutions and companies.
So sensory science adds a few more safeguards: code the samples so tasters do not know which coffee is which (blind tasting), randomize the tasting order, and have several people score separately before combining. The association's new assessment method splits "describing what is in the cup" from "an overall impression of its quality" into two separate forms (Chapter 35) — itself an admission that scores carry human judgment.
## Transfer [#transfer]
You can take the cupping idea home, and not only for cupping:
| What you want to know | The cupping approach |
| ----------------------------------------------------- | ----------------------------------------------------------- |
| Which of two bags suits your taste better | Same ratio, same grind, one bowl each, taste side by side |
| Whether a bag has "one bad bean" in it | Steep several bowls and see whether only one differs |
| Whether the tasting notes on the bag are steering you | Have someone number and shuffle the bowls, then taste blind |
| How this coffee will taste in your pour-over | Cupping cannot answer that; brew it in your pour-over |
The last row matters: cupping tells you how coffees compare on this one standard instrument. It does not tell you how a coffee will taste in your pour-over, with your grinder, in your water.
## Misconceptions [#misconceptions]
**"A cupping score is objective and does not depend on who tastes."** Cupping is a judgment made by trained people under standardized conditions. The conditions are held still; the differences between people remain, and research has found that cuppers' descriptions do not agree well.
**"What you taste on the cupping table is how the coffee tastes in any brewer."** Cupping is unfiltered immersion, with a heavier texture. In a pour-over the same coffee will differ in cleanness, body and extraction.
**"Cupping is a ritual for professionals and pointless at home."** At its core, cupping is just "hold everything else still and compare one thing". Two bowls and a scale at home are enough.
## Real Lab [#real-lab]
A small cupping at home, about twenty minutes:
1. Pick two or three different coffees. Give each coffee two identical bowls or cups, each with about 8 g of grounds and 150 mL of water (what matters is that every bowl is the same). Grind a little coarser than your usual pour-over, with the same setting for every coffee.
2. Smell the dry grounds first and note your impressions.
3. Pour just-boiled water that has rested a moment into all the bowls at the same time, and time four minutes.
4. After four minutes, gently push the layer on top back with a spoon and smell up close. Then skim off the foam and floating grounds.
5. Wait a few minutes until it cools a little, then scoop and slurp. Rinse the spoon in clean water between bowls.
6. Taste another round a few minutes later, and again when the coffee is cooler. Note the changes each round.
If someone can help, ask them to number the bottoms of the bowls and shuffle them so you do not know which coffee is which. Reveal after tasting.
To go further, follow on with Chapter 35's sensory triangle test:
## Check Your Model [#check-your-model]
**Predict.** You and a friend each cup the same two coffees at home, using the same parameters. You find A sweeter; they find B sweeter. Before looking at any explanation, write down: does this mean cupping is useless? Which variables might not have been held still?
Not necessarily. Variables that may not have been held still include: different water, different grinders (the same "slightly coarse" makes different particles), coffee at different temperatures when tasted, and the fact that the two of you perceive sweetness and acidity differently to begin with. Cupping held the brewing still, not the people. Put both of you at the same table, at the same time, tasting the same bowls blind, and compare again.
**Diagnose.** Of a coffee's five bowls, four are clean and sweet and one is clearly musty. How should you record this coffee? What would have happened with only one bowl?
It suggests the lot contains a few defective beans that happened to land in that bowl. Record a uniformity problem and a defect, not just an average impression. With a single bowl, the result would have been down to luck: if it were that bowl, you would judge the whole lot harshly; if not, you would miss the defect entirely.
**Explain.** In your own words, explain to a friend why professionals evaluate coffee with the "crudest" method — immersion — rather than the most refined pour-over.
To evaluate coffee, the coffee needs to be the only thing that differs. A pour-over's result depends heavily on who brews it and how they pour, and those differences would leak into the comparison. In immersion the time is set directly and there is no flow to control, so dozens of bowls can be prepared the same way, fairly. The cost is fines and oils in the cup, so it is less clean than a pour-over — but cupping wants a fair comparison, not the best-tasting cup.