A brew ratio is beverage yield divided by dose, written 1:n. The default n of 2.0 appears in basket specs, competition worksheets and home-dial-in guides for one practical reason: it lands extraction yield inside the 18–22 % window at normal strengths, on the widest range of roasts, without needing a refractometer to defend it.

Where the 1:2 benchmark comes from

The same ratio recurs across hardware generations because basket capacity and puck hydraulics scaled together — as dose rose, a proportional yield kept water contact and dissolved-solids percentages stable.

Dose → yield pairs that keep returning at 1:2
Context Dose Yield Ratio Why it holds
Traditional double basket 14 g 28 g 1:2.0 Small basket, short contact time, dark roast solubility
Modern precision basket 18 g 36 g 1:2.0 Deepest puck of the practical range; most common home default
Competition recipes 20 g 40 g 1:2.0 Higher dose keeps TDS up while yield scales with puck depth
21 g competition baskets 21 g 42 g 1:2.0 Geometry limit of ridgeless double baskets
Ristretto deviation 18 g 27 g 1:1.5 Deliberate departure — bought with body, paid for in extraction
Lungo deviation 18 g 45 g 1:2.5 Deliberate departure — bought with yield, paid for in strength

The extraction math of one bean, three ratios

The table below is one medium-roast bean pulled on the same grinder setting, only the programmed yield changed. TDS was measured on a digital refractometer; extraction yield follows EY = (yield × TDS) ÷ dose.

Logged example — 18 g dose, medium roast, same grind
Metric 1:1.5 — 27 g 1:2.0 — 36 g 1:2.5 — 45 g
Shot time 24 s 27 s 33 s
Measured TDS 11.5 % 10.0 % 8.9 %
Extraction yield 17.3 % 20.0 % 22.3 %
Window status EY low On target On target
Cup character Heavy, sweet, sour edge Balanced body and acidity Thinner, brighter, drier finish

Two numbers do all the work: at 1:2.5 the extraction is fine but strength has slipped below the comfortable band; at 1:1.5 strength is generous while extraction sits under 18 %. The default survives because it is the only column where both metrics are simultaneously inside their windows.

When to leave 1:2 — and where to go

Condition → ratio direction → target
Condition in the cup or log Direction New target Trade-off to watch
Dark roast destined for milk Shorten 1:1.5 – 1:1.8 Extraction drops — raise temperature 1 °C to compensate
Light roast, closed and sour at 1:2 Extend 1:2.2 – 1:2.6 Strength falls — verify TDS stays ≥ 8.5 %
Bitter, drying finish at 1:2, time ≤ 32 s Shorten 1:1.8 – 1:2.0 If bitterness persists, it is temperature or water, not ratio
Thin, weak cup with good time Shorten 1:1.8 Do this before touching the grind — fewer side effects
Shot time outside 25–32 s Keep ratio 1:2.0 Time errors are grind errors — fix grind first, ratio second
Dose 0.5 g off basket spec Keep ratio 1:2.0 Correct dose first; ratio tables assume dose within ±0.5 g

Which lever changes what

Lever comparison — grind, ratio and dose
Lever Strength (TDS) Extraction yield Shot time Use when
Grind finer ↑ slightly ↑ ↑ Time is under 25 s or EY is low with fast flow
Grind coarser ↓ slightly ↓ ↓ Time exceeds 35 s or EY is over 23 %
Extend ratio ↓ ↑ ↑ (longer pour) Time is fine but the cup is sour or under-extracted
Shorten ratio ↑ ↓ ↓ (shorter pour) Cup is weak, or bitter with correct time
Raise temperature ≈ ↑ ≈ Sourness survives an in-window EY — solubility limited

Three-shot verification

Lock the ratio and prove it before trusting it. A ratio is only “your default” after three consecutive shots land inside every window.

Verification protocol
Step Action Hold constant Pass criterion
1 Weigh dose 18.0 g ± 0.2, program yield 36.0 g ± 0.5 Grind, temperature, puck prep Ratio reads 1:1.98 – 1:2.02
2 Adjust grind only until time lands 25–32 s Dose, yield, temperature Three shots inside the time window
3 Measure TDS in duplicate on the last shot Everything 8.5 – 11.5 % TDS, duplicates within 0.1 %
4 Compute EY = (36 × TDS) ÷ 18 Everything 18 – 22 % extraction yield
Run your own numbers: use the ratio and formula tables to get the ratio class, extraction yield and the next grind correction at a glance.