Coffee Ratios by Method

"1:16" only means something once you know what the two numbers are counting. Here, the first number is coffee and the second is water, both by mass — one gram of coffee for every sixteen grams of water — and almost every ratio quoted anywhere in specialty coffee follows that same order. Get the order backwards once and a "light" 1:17 reads as a syrupy 17:1, which is a different drink entirely.

Why mass, not scoops

A scoop measures volume, and volume is a poor stand-in for the thing that actually matters, which is how much ground coffee is in contact with the water. A rounded scoop of a dense, dark-roasted bean can weigh noticeably more than the same scoop of a light roast with an oilier, more irregular surface, so "two scoops" quietly drifts by several grams depending on which bag it came from. A gram is a gram regardless of roast, bean density or how finely it's ground, which is the whole reason a cheap kitchen scale outperforms a scoop for anyone who wants the same cup twice. Water is close enough to one gram per millilitre that treating the second number as either a weight or a volume barely matters in practice — one less conversion standing between the ratio and the kettle.

Espresso counts a completely different thing

Every ratio for a gravity-fed method describes water poured onto the grounds. Espresso's ratio describes the opposite end of the process: what came out the bottom, weighed against the dry dose that went into the portafilter. An espresso portafilter locks under pressure, so a fair chunk of what went in never makes it to the cup at all — it stays behind, soaked into the used puck — which is exactly why quoting "water in" for espresso would overstate what you actually drink. Once a ratio is read as output against input rather than input against input, an espresso number that otherwise looks unusually low next to a pour-over or French press stops looking like an anomaly.

"A cup" is not a fixed amount

More ratio confusion traces back to the word "cup" than to the ratio itself. The little lines etched into a drip machine's carafe generally mark roughly 5 fluid ounces each, a recipe or a calculator that says "cup" almost always means the 8-ounce US customary cup, and the mug sitting on the counter typically holds somewhere between 10 and 12 ounces, sometimes more. Scale "brew four cups" by the wrong one of those three and the finished pot can be off by nearly double, which shows up tasting like a ratio mistake even though the ratio itself was never touched — only the size of cup feeding into it was wrong.

Where the numbers actually cluster

Filter methods — pour-over, the Chemex, the V60, Kalita Wave and drip machines — mostly sit in the high teens by grams of water per gram of coffee, because the water only passes through once. Immersion methods spread far wider: a French press sits closer to the middle of that same range, while cold brew and nitro cold brew run down into the single digits and low teens on purpose, built to be diluted rather than served straight from the steep. Pressure and traditional methods — espresso, the Moka pot, Turkish coffee, Vietnamese phin — cluster tight and low for a different reason again: small, concentrated servings by design, not a shrunken filter cup.

Finding your method's own page

Open the ratio page for whatever's actually in front of you for the exact numbers at three strengths, worked in real grams for that method's typical dose. If the cup already tastes wrong at the right ratio, a ratio adjustment usually isn't the fix — the fix pages name the more likely culprit before you touch the scale again.

Does a stronger ratio mean more caffeine?

Not directly. A ratio is a taste-strength dial, and caffeine doesn't track taste — it tracks the actual coffee dose and the extraction method itself. A big, deliberately weak pot from a drip machine can still carry more total caffeine than a single strong espresso shot, purely because there's so much more of it. The caffeine estimator is built around dose and method rather than around how strong a cup tastes for exactly that reason.

Why don't two sources agree on the same method's ratio?

Because a single "correct" number for something like pour-over would be false precision — real sources cluster in a range rather than a point, and where they genuinely disagree by more than rounding, the honest move is to widen the stated range rather than pick a side. The methodology page covers where each range comes from and how disagreements between sources get handled.