112 Coffee

How we measure an extraction

Version 1.0 · October 2026

Every extraction on this platform is compared with extractions from other people, on other machines, in other countries. That only works if we all measure the same way. This page is the shared method. It takes two minutes to read and needs only a scale and a timer.

1. What you need

  • A scale with 0.1 g resolution. Weigh the coffee and the drink, not the volume. Put the cup on the scale under the portafilter.
  • A timer. The machine's own timer is fine if it starts when you press the button.
  • Fresh coffee, rested at least 7 days after roasting for espresso.
  • Optional: a refractometer for TDS, and a water test (strips or a lab report).

2. The recipe: three numbers

Number How to measure it Unit
Dose Ground coffee in the basket, weighed after grinding. g, one decimal
Yield The drink in the cup, weighed when the flow stops. Crema included. g, one decimal
Time From the moment you press the button until you stop the shot. Pre-infusion is included. s
  • Stop the shot at the target yield, then wait for the last drops before reading the scale. Write down what the scale shows, not the target.
  • Do not count from the first drop. On this platform, time always starts at the button.
  • Ratio = yield ÷ dose. 18 g → 36 g is 1:2.

3. Coffee and grinder

  • Roast date and days off roast on the day of the shot.
  • Storage: sealed bag with valve, airtight container, or freezer.
  • Grinder brand and model, and the setting exactly as it reads on the dial. Settings are compared only between people with the same grinder.
  • Purge 2–3 g after a big change of setting; the grinder holds old coffee.
  • Basket size (for example 18 g or 20 g) and whether you use a puck screen.

4. Machine

  • Brand and model, group type, and brew temperature if the machine shows it.
  • Pressure: the pressure at the puck, normally 9 bar. A pump rated at 15 bar does not mean 15 bar on the coffee.
  • Flush the group for 2–3 seconds before each shot. Lock in the portafilter and start within 10 seconds, so the coffee does not cook on the hot group.

5. Water

Water is about 90% of an espresso, and what is dissolved in it changes extraction and taste. Logging it is optional.

If you do not enter your water, we assume our reference water, close to the Specialty Coffee Association (SCA) target:

Parameter Reference water Acceptable range (SCA)
Total dissolved solids (TDS) 150 mg/L 75–250 mg/L
Total hardness (as CaCO₃) 70 mg/L 17–85 mg/L calcium hardness
Total alkalinity (as CaCO₃) 40 mg/L around 40 mg/L
pH 7.0 6.5–7.5
Sodium 10 mg/L around 10 mg/L
Chlorine 0 0

How to enter your water:

  • Use a preset (reference water, soft, medium tap, hard tap) or type the numbers from a test strip or lab report.
  • Hardness and alkalinity matter more than TDS. If you only have one number, give hardness.
  • If your report gives calcium (Ca) and magnesium (Mg) in mg/L instead of hardness, enter them; we convert: hardness = 2.5 × Ca + 4.1 × Mg.
  • 1 °dH (German degree) = 17.8 mg/L as CaCO₃. 1 grain per gallon = 17.1 mg/L. 1 ppm = 1 mg/L.

What the simulator does with water (an estimate, not a measurement):

  • Harder water extracts a little faster.
  • Higher alkalinity buffers acidity: fruit and acidity taste muted, the cup can turn flat.
  • Very soft water with very low alkalinity tastes sharp and sour.

6. Tasting: one word

Taste the shot within one minute, after stirring it. Choose the word that fits best:

  • Sour: sharp, salty, thin, short finish, no sweetness. Usually under-extracted.
  • Balanced: pleasant acidity, sweetness, a gentle bitterness at the end.
  • Bitter: harsh, dry, astringent, a long unpleasant finish. Usually over-extracted.

Sour and bitter are easy to confuse. Astringency (a dry mouth) belongs to bitter. If in doubt, taste a second shot 3 seconds shorter and 3 seconds longer.

7. TDS and extraction yield (optional, refractometer)

  1. Stir the shot well, take a sample from the middle with a pipette.
  2. Let it cool to room temperature (or use a syringe filter if your refractometer needs it).
  3. Zero the refractometer with the same water you brew with.
  4. Read TDS (%). Take two readings; if they differ by more than 0.1, take a third.
  5. Extraction yield (%) = TDS × yield ÷ dose. Example: 9.5% × 36 g ÷ 18 g = 19%.

8. Automatic (bean-to-cup) machines

Bean-to-cup machines are welcome. Switch on Automatic machine in your profile or on the extraction form. Differences:

  • Pressure: the pump is rated 14–17 bar, but at the coffee the pressure is about 8–10 bar, as on a manual machine. You do not need to enter it.
  • Dose: usually 7–12 g, set by the strength setting. If you do not know it, choose "I don't know" and we assume 9 g. To estimate it, let a used puck dry for a day and weigh it: the dry puck weighs about 80% of the dose.
  • Yield: set in ml on the machine. Weigh the cup if you can; otherwise enter ml and we convert (1 ml ≈ 0.9 g with crema).
  • Grinder: write the step number. Change one step at a time and wait 2–3 cups before judging, because the grinder still holds the old coffee.
  • Time: from the button press until the flow stops, as for manual machines.

9. One shot, one entry

  • Log every shot you taste, also the bad ones. Bad shots teach the model where the edges are.
  • Change one thing at a time between shots.
  • Add a photo of the cup or the extraction if you like; it is visible to everyone.

10. Why this matters

When a barista in Tokyo and a roaster in Târgu Mureș log the same coffee with the same method, their numbers can be compared. The simulator learns from these comparisons and the roasters see how their coffee behaves on real machines.

Questions or suggestions about the method: write to us from your account page.

Sources for the water values: SCA Water Quality Handbook (summarised by coffeeparts.com.au and asc1inc.com).