The one idea that changes your coffee
Sour or bitter? You decide.
Almost every disappointing cup you have ever made was under-extracted or over-extracted — and the two failures taste nothing alike. This simulator gives you the dials that decide which one you get, including the one you do not actually get to set.
What extraction actually means
Coffee is a soluble solid. When water meets ground coffee it starts dissolving things out of it, and it does so in a stubborn order: fruit acids first, then sugars and balance, and finally bitter compounds. About 28% of a roasted bean will dissolve at all; the rest is just wet grounds.
So the whole game is knowing when to stop. Pull the water off too early and you get the acids without the sugars — that is sour. Leave it too long and you drag out the bitterness behind them. The window where the sugars have arrived but the bitterness has not is roughly 18–22% extraction, and every brewing method ever invented is a different trick for hitting it.
Why time isn't always yours
Here is the part most brewing guides skip. You cannot always change one dial on its own, because in half the brewing world the brew time is not something you set — it is something the coffee does to you.
| Family | What the water does | The clock | Grinding finer |
|---|---|---|---|
| Immersion French press, cupping, cold brew |
Sits with the coffee for the whole brew | Yours. You decide when to plunge | Extracts faster, same duration |
| Percolation V60, moka, espresso, drip |
Passes through the bed | The bed's. It takes as long as it takes | Extracts faster and drains slower |
Water forcing its way through packed coffee behaves like water through any packed bed: the finer the particles, the harder the squeeze. Halve the grind size and you roughly quadruple the drawdown time. This is why a barista grinds finer to slow a shot down — the grinder is the timer.
Which means grind is a far more violent lever in percolation than it looks: it raises extraction twice over, through surface area and through contact time at once. And it is exactly why the course calls infusion the more forgiving method — an immersion lets you rescue a bad grind by steeping longer. A percolation just fights back.
So the simulator has two modes. In immersion the brew time is a dial you control. In percolation that dial disappears, a dose dial takes its place, and the clock becomes a readout the model computes for you.
Most people say "bitter" when they mean "sour", then fix it in exactly the wrong direction. Sour is sharp and makes you wince at the sides of your tongue — like biting a lemon. Bitter is flat and dries the back of it — like dark chocolate. Sour means you extracted too little. Bitter means too much. Get this backwards and every adjustment you make will make the cup worse.
Dial in your cup
Start from a brewing method, then pull it apart. In an immersion the brew time is yours; in a percolation the bed takes it away from you and hands back a clock instead.
Drag any dial. The curve redraws for the settings you are on right now.
Seven things to try
Do not just brew "correctly". Break it on purpose — that is where the intuition comes from.
- Make the same coffee sour, then bitter. Load French press — an immersion, so the clock is yours. Pull the steep down to 1:00 and read the words in the cup panel. Now push it out to 8:00. Same coffee, same grind, opposite failures.
- Watch the grinder move the clock. Load V60 and grind finer, from 750 µm to 600. You never touch a time control — but the brew time climbs from 2:45 to 4:18 on its own, and extraction jumps from 20% to nearly 27%. That is grind hitting you twice.
- Run a real espresso dial-in. Load Espresso: 28 seconds, in the zone. Grind two notches finer and the shot slows to 32 seconds. Go coarser instead and it gushes through in 21 seconds, sour. This is the loop every barista runs every morning, and the grinder is the only dial they touch.
- Separate strength from extraction. Sitting in the zone, change only the ratio. Extraction barely moves; strength moves a lot. Confusing the two is why "my coffee is too weak, I'll brew it longer" ends in bitterness.
- Find out why moka pots run bitter. Load Moka. It lands just past the window before you touch anything — because the water arrives near 110 °C. Drop the temperature to 94 °C and see what a moka could taste like.
- Build a ristretto. From Espresso, pull the ratio to 1:1. Extraction collapses, strength goes through the roof, and the model tells you the cup ran out of water. That is the whole ristretto paradox in two numbers.
- Prove that immersion forgives. Take the same coarse grind through both modes. In immersion you can steep your way back into the zone. Switch to percolation and the bed simply hands the water back too fast — you have to fix the grind.
Where the presets come from
These are the starting points the simulator ships with — ordinary, defensible recipes rather than competition ones.
| Method | Family | Grind | Dose | Time | Ratio | Temp | Pressure |
|---|---|---|---|---|---|---|---|
| V60 | Percolation | Medium | 15 g | 2:45 | 1:16 | 94 °C | 1 bar |
| French press | Immersion | Coarse | 30 g | 4:00 | 1:16 | 94 °C | 1 bar |
| Aeropress | Immersion | Medium-fine | 15 g | 2:30 | 1:14 | 88 °C | 1 bar |
| Moka | Percolation | Fine | 17 g | 1:00 | 1:6 | 110 °C | 1.5 bar |
| Espresso | Percolation | Extra fine | 18 g | 0:28 | 1:2 | 93 °C | 9 bar |
You can add water to raise extraction — but you will thin the body doing it. If a cup is under-extracted, it is almost always better to grind finer and keep the ratio where it is. Fix taste with the grinder, fix strength with the scale.
None of this is repeatable by eye. A scoop of coffee can vary by 30% depending on how the beans settle, which is more than enough to move you out of the window. Weigh the coffee, weigh the water, and every dial above becomes something you can actually control.
Two models, stacked. Extraction is first-order kinetics: surface area, temperature, pressure and contact time set how fast coffee dissolves, and the available water sets the ceiling. Drawdown follows flow through a packed bed, which is why the clock scales with the square of the dose and inversely with the square of the particle size. Because every brewer's bed geometry differs, the drawdown is computed relative to each preset's known-good recipe rather than predicted from scratch — a moka basket and a V60 cone would never fit one constant.
Real coffee is messier still: bean density, roast level, water chemistry, channelling and how evenly your grinder performs all matter, and none of them are in here. Treat the relationships as real and the decimals as decoration.
That is the core of it. Once sour and bitter are two different things in your mouth rather than one vague "bad", you stop guessing — you taste the cup, name the failure, and turn the right dial.