Espresso puck preparation

Why an even bed is the aim

Sourced

Espresso forces water through a shallow bed of fine coffee in well under a minute. If part of the bed lets water through more easily than the rest, more water goes that way, and the coffee along that path gives up more while the rest gives up less.

One published study gives this some weight. Its authors built a mathematical model of espresso extraction and compared it with shots pulled across a range of grind settings in a café. The model predicted that finer grinding would always extract more. The measurements did not agree: extraction rose as the grind got finer and then fell. The authors' reading, which they present as an inference and did not observe directly, is that at the finest settings parts of the bed clogged and the water flowed unevenly. It was one café, one coffee, one grinder and one machine.

That is the evidence that unevenness matters. It does not say which preparation routine prevents it.

The usual sequence

CoffeeHQ explanation

Dose, and why the basket has a say

CoffeeHQ explanation

A basket is made for a range of doses. Its maker usually states the range, and it is worth finding, because the same weight of coffee sits at a different depth in a different basket. Too much and the tamped bed touches the shower screen when the portafilter is locked in; too little and there is a gap above the bed for water to pool in.

This is why a recipe copied from somebody with a different basket often does not work. The dose in a recipe belongs to the basket it was written for. Changing the dose also changes the strength of the drink and how the shot flows, so it is one variable among several and not a free adjustment.

Distribution: stirring and levelling are different jobs

CoffeeHQ explanation

Coffee ground fine enough for espresso often leaves the grinder in clumps. One study of grinding compared ways of reducing the electric charge coffee picks up as it is ground, and found that adding a very small amount of water to the whole beans cut the charge by at least half and reduced clumping. Espresso from a dark roast ground that way ran longer and came out more concentrated; a light roast of the same coffee showed no significant change. The study used one grinder. Static is one reason grounds cling together and to the grinder. How much a grinder clumps is a property of that grinder and that coffee.

Stirring the dose with thin needles, the technique usually called WDT, is aimed at those clumps. A levelling or distribution tool that spins on top of the basket does something else: it flattens the surface and does not reach into the bed. The two are often sold as alternatives and are not.

No published study testing either tool against the taste or the extraction of the shot was found for this page. The case for stirring is the reasoning above and a great deal of baristas' experience. The page that asks whether a WDT tool is needed reaches a conditional answer for the same reason.

Tamping: level matters, and force is a convention

Accounts differ

Tamping presses the loose bed into a puck. A figure of thirty pounds of force is widely repeated. It is a teaching convention, and no source establishing it as a requirement was found.

What can be said from the definition of the job is narrower. Coffee compresses until the particles are in contact and then compresses very little more, so beyond that point extra force has little left to do. A tamp that is not level leaves the bed thicker on one side than the other, which is the unevenness the whole exercise is meant to avoid. So the practical advice is the dull one: press until the bed stops giving, keep the tamper flat, and do it the same way each time.

A tamper that is too small for the basket leaves a ring of loose coffee at the edge. That is a matter of fit, and the basket's maker states the diameter.

Reading the result

CoffeeHQ explanation

A portafilter with the bottom cut away shows the underside of the basket while the shot runs. Baristas read it for evenness: coffee appearing across most of the surface at about the same moment and drawing together into one stream, as against a jet from one spot or a region that stays dry. The reading is a craft, not a measurement, and a shot can look untidy and taste good.

The cup is the test. A shot that tastes both sharp and harsh at once, and that changing the grind does not improve, is the usual sign that the fault is in the bed and not the setting. The dial-in assistant treats that pattern as its own case for the same reason.

If the shots are uneven

CoffeeHQ explanation

Each row names one thing to change. These are common causes in practice, not the only ones.

A jet sprays from one spot under a bottomless portafilter

Usually because: Water has found an easier path through part of the bed, often a clump or a gap at the edge.

Change first: Stir the dose before tamping, and change nothing else.

Bottomless portafilter spraying

Shots vary from one to the next with nothing changed

Usually because: The dose is varying, or the routine is, more often than the grinder.

Change first: Weigh every dose for a day before touching the grind.

Inconsistent dose from the grinder

The puck is marked by the shower screen

Usually because: The dose is too large for the basket, so the bed touches the screen.

Change first: Reduce the dose to within the basket's stated range.

Basket overfilled or underfilled

Careful preparation makes no difference at all

Usually because: The basket may be a pressurised one, where a small exit hole and not the coffee sets the flow.

Change first: Check which basket is fitted before changing the routine.

WDT makes no difference

What puck preparation cannot fix

CoffeeHQ explanation

It cannot make up for a grinder that has no usable adjustment at espresso fineness, and it does nothing in a pressurised basket or a capsule machine, where the restriction is built into the basket or capsule. It cannot rescue stale coffee. And it is not the place to look first when a shot is simply too strong or too weak, which is a question of dose and yield.

What is practice here and not evidence

CoffeeHQ explanation

It gives no tamping force, no number of stirs and no dose, because no source read supports one. It does not say that any tool improves the taste of espresso: no study testing that was found. It reports one study's finding about uneven extraction with that study's limits, and one study's finding about static, and labels the rest as practice.

What to do next

• Espresso dial-in assistant — Separates a fault in the bed from a grind that is simply set wrong.
• Espresso pressure and flow — What the puck's resistance does to pressure and flow once the pump starts.
• WDT tool — A decision tree on whether stirring is worth adding to your routine.
• Diagnose with a bottomless portafilter — How to read the underside of the basket, and what it cannot tell you.
• Espresso foundations lesson — Dose, yield, ratio and puck preparation taught together, with an exercise.