Puck preparation: distribution, tamping, and why they're not the same problem
Short answer: distribution makes the coffee bed uniform in density across its area. Tamping compresses that bed to a consistent height. They are different problems, and a tamper cannot solve the first one — it presses down on the whole surface at once, so it compacts whatever arrangement it finds rather than rearranging it. Most puck-preparation problems are distribution problems wearing a tamping costume.
Training tends to collapse the two into a single motion: dose, level, tamp, lock in. That works well enough to produce coffee, which is why the conflation survives. It also means that when shots go wrong, the diagnosis often lands on the wrong step.
Two different problems, stated precisely
Distribution is a horizontal problem. After grinding, coffee lands in the basket unevenly — mounded under the grinder chute, sparse at the edges, with clumps of stuck-together particles scattered through it. What you want is the same mass of coffee per unit area everywhere, with no clumps and no voids. That is a question of where the coffee is, across the circle of the basket.
Tamping is a vertical problem. Once the coffee is arranged, it needs to be compressed to a consistent, level height so that the bed has uniform thickness and enough integrity to resist water without collapsing. That is a question of how far down the coffee goes, and whether the surface it presents is flat and parallel to the shower screen.
The reason tamping cannot substitute for distribution is mechanical. A flat tamper base contacts the entire surface simultaneously and applies force downward. Coffee under a dense region is already dense; coffee under a sparse region is already sparse. Pressing them both down compresses both, in proportion, and the density difference persists — now in a shorter bed. The gradient survives the tamp because nothing in the tamp moves coffee laterally.
You can level a heap with a tamper. You cannot equalise it.
What each failure does to the shot
Distribution failure produces channelling. Water entering the bed takes the path of least resistance. Where the coffee is less dense, or where a void sits between clumps, flow concentrates. That region receives far more water than its share and over-extracts to bitterness; the dense regions receive too little and stay under-extracted and sour. The cup contains both faults at once, which is why channelled shots taste simultaneously harsh and thin and resist correction by grind adjustment — grinding finer raises resistance everywhere, including in the channel, without removing the difference in resistance that created it. This is covered in detail in the piece on channelling.
The visible signs: a shot that starts fast, sprays or splits into erratic streams, runs pale early, and finishes short of the expected yield in less than the expected time. Afterwards, holes or craters in the spent puck, sometimes with visible tunnels.
Tamping failure produces a wedge. If the tamper goes in tilted, one side of the bed is thicker and denser than the other. Water passes preferentially through the thin side. The failure looks similar in the cup — uneven extraction — but the puck tells a different story: a sloped surface, and erosion concentrated along one edge rather than in isolated craters.
Both together produce shots that behave differently every time, which is the condition most often blamed on the machine.
| Distribution failure | Tamping failure | |
|---|---|---|
| The defect | Density varies across the bed's area | Bed height or angle varies |
| Cause | Clumping, mounded dose, static, no intervention | Tilted tamp, inconsistent seating |
| Puck evidence | Craters, holes, tunnels, wet patches | Sloped surface, one-sided erosion |
| Shot evidence | Erratic or split streams, fast pale start | Consistent bias to one side of the basket |
| Fixed by | Redistributing before tamping | Attention to level, not to force |
| Not fixed by | Tamping harder | Tamping harder |
The upstream cause: what the grinder hands you
Distribution work exists because grinders do not deliver coffee in the state you want it.
Clumping. Ground coffee sticks to itself. Fines in particular agglomerate, and burr geometry, retention in the grinder's exit path, and how quickly the grounds leave all affect how much of that reaches the basket. A clump is a locally dense object surrounded by lower density — a channelling site delivered ready-made.
Static. Grinding generates charge. Charged particles repel each other and cling to surfaces, which scatters fines, makes the dose spray unevenly, and leaves grounds on the rim and outside the basket. Static rises when the coffee and the air are dry, which is why the same grinder behaves differently in different weather.
Chute geometry and dose shape. Coffee falls in a stream and forms a cone. A cone is, by definition, a bed of varying depth.
Two upstream measures reduce how much distribution work is needed. Dosing into the basket in a way that spreads the stream — rotating the basket under the chute, or dosing into a cup and pouring — starts from a flatter bed. And a small amount of water added to the beans before grinding reduces static substantially, which reduces both scatter and clumping. Neither eliminates the need to distribute, but both reduce the size of the problem being solved by hand.
Distribution techniques and what they trade
Levelling tools. A disc with a shaped underside, rotated on the rim of the basket, sweeps the top of the dose flat and lightly compresses it. Fast, repeatable, easy to train, and it fixes the surface reliably. Its limit is depth: it acts on the top layer, so it flattens a cone but does not break up clumps sitting lower in the bed. Good at the visible part of the problem, partial on the hidden part.
Needle-based stirring. A set of fine needles moved through the full depth of the dose breaks clumps and mixes the bed throughout — the approach usually referred to as WDT. This is the most thorough method available, because it addresses density through the whole bed rather than at the surface. Its costs are time (several seconds per shot, which is real during a rush), a tool to keep clean, and technique dependence: done carelessly it can compact one region or leave the surface rough, and an aggressive stir close to the basket wall can loosen the edge.
Drop and settle. Tapping the basket on the bench, or dropping the loaded portafilter a short distance, settles the grounds and collapses voids. Cheap and fast. It also tends to migrate fines downward and can compact the base unevenly, so it works better as a supplement than as the whole method.
Combinations. Common practice in specialty bars is a needle stir followed by a level, which addresses depth and then surface. Common practice in high-volume bars is a level alone, because it costs almost nothing per drink. Both are defensible; they are answers to different constraints.
Where none of it is worth it. If your grinder delivers a loose, clump-free dose and your shots are already even, elaborate distribution routines add seconds per drink and buy nothing measurable. The test is the shot and the puck, not the ritual. A busy bar should work out what a few extra seconds per shot costs across a morning, and spend that time only where it changes the cup. Improving the grinder — or how the dose leaves it — usually returns more than adding a step at the bench.
Tamping: level matters, force mostly does not
Coffee compresses until it stops compressing. Beyond the point at which the bed has reached its packed density, additional force does not meaningfully reduce its volume further; it just loads your wrist and the basket. The consequence is that the difference between a moderate tamp and a very heavy one is small in the cup, provided both are firm enough to consolidate the bed — and the difference between a level tamp and a tilted one is large.
This makes force the wrong thing to standardise on. Calibrated pressure is measurable and therefore attractive as a training target, but it is optimising a variable with a shallow response curve while leaving a steep one — angle — unmeasured.
What to attend to instead:
Level. The tamper base should be parallel to the basket. Seating the tamper on the rim, or bracing the forearm so the wrist does not introduce a tilt, does more for consistency than any pressure target.
Consistency, not magnitude. Whatever force you use, use the same one. The bed's resistance should be a constant you are not changing shot to shot.
Tamper fit. A base substantially smaller than the basket leaves an unconsolidated ring at the wall, which is a channelling path along the edge. Match the tamper to the basket.
No tapping after the tamp. Striking the side of the portafilter after tamping to knock down loose grounds cracks the consolidated bed and separates it from the basket wall. It creates the fault it is meant to tidy away.
One tamp. Re-tamping a bed that has already been compressed and then disturbed does not restore it.
Basket fill and headspace
How much coffee goes in interacts with both problems.
Under-filling leaves a large gap between the puck surface and the shower screen. When water arrives it pools in that space and lands on the bed as a body of water rather than being pressed evenly through it, which disturbs the surface and can lift the bed. Under-filled baskets channel readily.
Over-filling leaves no clearance. The screen presses into the bed on locking in, imprinting the screen pattern and compressing the centre more than the edges. It also means grounds get caught on the group gasket, which stops it sealing.
The right fill leaves a small clearance — enough that the screen does not contact the puck, little enough that water does not accumulate above it. The practical check is to lock in, unlock immediately, and look: a clean bed with a light impression is correct, a deeply imprinted one means too much coffee, and no mark at all with a visibly deep gap means too little. Note that a basket's stated capacity is a guide, and different coffees at the same mass occupy different volumes depending on roast level and grind.
Where the machine wets the bed gently before applying full pressure, the under-filling problem is reduced but not removed — pre-infusion makes the arrival of water less violent, which is one of the reasons it improves consistency on imperfect pucks.
Diagnosing which problem you have
Read the shot, then read the puck.
During the shot. Streams that split, spray, or come out at an angle indicate uneven flow. Note where on the basket it happens. A consistent bias to one side across many shots points to tamping angle or a mechanical issue with how the basket seats. Randomly located faults that move shot to shot point to distribution.
After the shot. Knock the puck out intact and look at it, then break it open.
- Craters, pits or visible tunnels: channelling from uneven density. Distribution.
- A sloped or wedge-shaped puck: tamp angle.
- Erosion around the whole circumference: an edge-sealing problem — tamper fit, or a bed that was disturbed after tamping.
- A puck that is wet and mushy throughout with a sound surface: usually a dose or grind issue rather than preparation.
- A screen pattern pressed deeply into the surface: over-filled basket.
The controlled test. Take one coffee, one grind setting and one dose. Pull three shots with your normal routine, three with a thorough needle stir added, and three with deliberate attention to tamp level and nothing else changed. Time all nine and taste them. Whichever intervention tightens the spread of shot times is telling you which problem you actually have. This is worth twenty minutes once, and it saves a great deal of guessing afterwards.
The complete guide to extraction variables sets puck preparation in the context of grind, dose, ratio, pressure and temperature — useful once you have established that preparation is not the limiting factor.
Frequently asked questions
Can tamping harder fix channelling? No. Tamping compresses the bed vertically and cannot move coffee sideways, so a density difference across the bed survives the tamp. More force compresses both the dense and the sparse regions and leaves the difference between them intact. Channelling is addressed before the tamp, not during it.
Is WDT worth the time on a busy bar? It depends on what your grinder delivers and what your service pace allows. If shots are erratic and the puck shows craters, needle stirring is the most effective single intervention available. If shots are already even, it costs several seconds per drink for no measurable gain, and reducing clumping at the grinder is a better use of the effort.
How much tamping pressure should I use? Enough to consolidate the bed, applied the same way every time. The bed stops compressing meaningfully above a moderate force, so exact pressure matters much less than keeping the tamper level and consistent. Standardise the angle before standardising the force.
Why do my shots change when the weather does? Static. Grinding generates charge, and low humidity increases it, which scatters fines and worsens clumping. A small amount of water added to the beans before grinding reduces it and is the usual remedy.
Should I tap the portafilter after tamping to knock down loose grounds? No. It cracks the consolidated bed and can break the seal at the basket wall, creating an edge channel. Deal with loose grounds by dosing more cleanly or by wiping the rim.
This article is part of the Vea Group knowledge base. Vea Group S.p.A. is an Italian designer and manufacturer of professional and premium coffee machines, with heritage dating to 1919 and production in Chignolo d'Isola, Italy, and Suzhou, China.