Coffee extraction yield: what the 18 to 22 percent window means

Brew a filter coffee at 20 percent extraction and four fifths of the ground coffee never leaves the filter. That is the arithmetic behind two figures professional brewers argue about and almost nobody measures at home. One of them is brew strength. The other is how much of the dry coffee actually dissolved. Together they place a cup somewhere on the brewing control chart, a grid that has told the trade what good coffee looks like since the 1950s.

That chart is now being rebuilt by the researchers who tested it. What they found does not throw out the old numbers so much as show how much room sits around them.

Two numbers describe every brew

The first is brew strength, measured as total dissolved solids: the share of the finished drink that is coffee rather than water. The second is percent extraction, the share of the dry grounds that ended up dissolved. A refractometer reads the first in seconds. The second is calculated from it, once the dose and the finished weight are known.

The Specialty Coffee Association draws a box around part of that space and calls it the golden cup standard. It spans 1.15 to 1.35 percent dissolved solids and an extraction of 18 to 22 percent. Inside that box, in theory, sits the coffee everyone is trying to make.

What the extraction figure does to taste

Pull too little out and the cup is thin and sharp, sour in a way that has nothing to do with the bright acidity of a good Mexican lot. Pull too much and the drink turns dry and hollow, with a bitterness that settles at the back of the tongue and stays there. Caffeine gets blamed for that far more often than it deserves, as the chemistry of a bitter cup makes clear.

Grind size is the lever. Finer grounds give the water more surface to work on and the figure climbs, coarser grounds drop it. The brew ratio moves strength almost independently, which is why two cups can read the same on a refractometer and taste nothing alike.

Where the brewing control chart came from

The chart was developed in the 1950s and its shape has barely changed. Nine regions on a grid. Vertical lines separate weak from strong along the strength axis, horizontal lines separate underdeveloped from bitter along the extraction axis. The middle square is the target and the other eight are diagnoses.

It has outlived most of the equipment it was drawn for, and for decades nobody went back to test it. When a group at the University of California, Davis did, they began by brewing to fixed points on the grid and handing the cups to a trained panel of 12.

The 118 drinkers who did not read the chart

In 2021 the same lab published what happened when 118 self reported black coffee drinkers rated 27 brewing profiles on a nine point scale. Liking did not pile up inside the box. It split into two groups with different preferences, and both spread well past the edges of that rectangle.

The paper's own conclusion is that the chart's idea of an ideal cup should be reconsidered. That is not the same as saying the box is wrong. It says the box is narrower than the people drinking from it. If a cup that lands outside it tastes right, the measurement is not the mistaken part, and the ratio you settled on is doing more work than any grid.

What the new coffee extraction yield chart shows

In 2023 the Davis group published replacement charts in the Journal of Food Science: a sensory version mapping which attributes turn up where, a consumer version showing both preference groups, and one that combines them. They are a proposal from a research team, not an adopted standard, and the association's box has not moved.

The most useful finding for a home kitchen came out of the temperature work. Drip brew coffee made at 87, 90 and 93 degrees Celsius showed no appreciable difference in sensory profile once strength and extraction were held fixed. The kettle dial is the thing most people reach for first and it is the one that changes least. Weigh the dose, weigh the water, then move the grinder a notch and taste what that does.