WDT and Self-Leveling-Tampers: The Physics of Even Resistance
WDT and Self-Leveling-Tampers: The Physics Behind Even Resistance
An espresso puck is a packed bed with water flowing through it. Whatever is uneven inside it, the water finds without mercy. We have gathered what has actually been measured on WDT, tamping pressure and self-leveling tampers – including the places where the evidence is thinner than the price tag suggests.
Between roughly 5 and 20 kilograms, tamping pressure demonstrably makes no difference. Three independent measurement series, one of them with 150 extractions from Basel, find no significant difference in time, TDS or yield.
WDT demonstrably reduces the spread in flow resistance. That it raises the average extraction yield is not documented.
Self-Leveling-Tampers deliver reproducibility of a quantity whose variation has no consequences. Their real benefit lies in ergonomics and process standardisation – not in a proven advantage in the cup. That is our reading of the data; we are still happy to sell these tampers, just with a different reasoning.
The physics of flow resistance
Darcy's law describes the puck as a packed bed with flow through it: the volumetric flow depends on permeability, area and pressure difference, divided by viscosity and bed height. The decisive quantity is permeability – and it varies dramatically.
Corrochano et al. (2015) measured values between 2.59 × 10−14 and 3.36 × 10−13 square metres, a factor of 17. A bulk density 30 percent higher lowered permeability by a factor of three to four.
And that is the core of the problem: if the same pressure sits across the entire puck but the local permeability differs, the water has to flow at locally different speeds. And it reinforces itself – where more flows, fines get washed out and the resistance drops further.
That is why a shot with channeling tastes sour and bitter at the same time. It is not a contradiction, it is two differently extracted zones mixed into one cup.
Peer-reviewedCameron and Hendon (2020) showed this elegantly: assuming homogeneous flow, their model predicts that yield rises monotonically with a finer grind. The experiment shows a maximum instead – yield drops off again at very fine settings. The difference between model and measurement is the signature of channeling.
A more recent 2026 paper in Royal Society Open Science modelled permeability using X-ray micro-tomography across eleven grind steps. An interesting side finding: coffee particles do have internal pores, but these are mostly isolated. The water flows almost exclusively through the spaces between the particles. The authors rank grind size and dose as dominant, and tamping as important but secondary.
Where clumps come from
Grinding creates triboelectric charge – in orders of magnitude you would not expect. Méndez Harper et al. (2024) measured up to 120 nanocoulombs per gram on dark roasts in Matter and compare that with particles in volcanic plumes and thunderstorm cells.
The decisive part for puck prep: fine particles charge opposite to coarse ones. They attract each other and form aggregates. Those are exactly the clumps that WDT breaks apart. The polarity flips at a water content of around 2 percent by mass, and a drop of water before grinding pushes the charge to practically zero – the physical basis of RDT.
Tamping pressure: three series, one result
Here the evidence is unusually clear, and it contradicts the 15-kilogram rule of thumb quite plainly.
The same finding shows up in two further sources: a series of 40 shots across 5, 10, 15 and 20 kilograms found p-values from 0.30 to over 0.60. And Cameron and Hendon write in their paper that they tried different tamping pressures and observed no noteworthy variation in shot time or yield.
An important limitation we do not want to gloss over: below roughly 5 kilograms there is no solid data. The lower threshold has not been cleanly determined. Anyone claiming that everything is irrelevant from 2 kilograms upwards is making the number up.
What counts instead, according to the widely held view, is levelness – tamping straight and level rather than hard. That follows well from Darcy and the permeability measurements. But we will stay honest here too: we found no controlled study that quantifies tamper tilt as an isolated variable. The mechanism is plausible, the direct measurement is missing.
What WDT actually does
The Weiss Distribution Technique goes back to John Weiss, who described it in a coffee forum in December 2005. Weiss holds a doctorate in biochemistry and spent 35 years as a computer science professor – an enthusiast, not a manufacturer. The name, incidentally, was coined by someone else. Weiss originally used a single dissecting needle; today's tools work with several fine needles.
Community measurementThe best available series covers 24 shots with live measurement of pressure, flow and weight. The result: with mere raking of the surface, the bed's peak resistance varied by around 40 percent. With WDT through the full bed height, the peaks sat much closer together, and runaway fast shots practically disappeared.
But here too the honest half: even with deep WDT, 15 to 25 percent of shots remained outliers on the high side. And all 24 spent pucks showed a slight dip in the middle – a zone of lower density that comes from the circular grinding motion and that WDT does not visibly fix.
Another series compared WDT with direct dosing and found the difference “only slightly better” and not statistically significant. With very fresh coffee high in CO2, WDT even came off worse there. The author still keeps using WDT – which we can understand very well.
There is no peer-reviewed study on WDT. What has been measured is a proxy: the variability of puck resistance. There the effect is clear and points in the expected direction. That is not a documented rise in average extraction yield. The most robust single finding: deep beats superficial. Needles thinner than 1 millimetre, through the full bed height.
Self-Leveling-Tampers, honestly
Spring-loaded tampers deliver a defined force. The usual spring ratings sit exactly in the range for which the three series above consistently show no effect:
| Tamper | Spring force | What the maker claims |
|---|---|---|
| Weber Workshops Really Nice Tamper | 9 kg preinstalled, 12 kg included | no claims about extraction or taste |
| Normcore V4 | approx. 6.8 / 11.3 / 13.6 kg | repeatability, no cup effect claimed |
| The Force Tamper | no kg figure on the maker's site | claims “perfectly even extraction” – we found no evidence for it |
We find it remarkable how differently honest the communication is here. Weber sells on mechanics and build quality and claims nothing at all about the cup. Normcore stays with repeatability. With other suppliers this turns into a claim of effect for which no independent tests exist. We have written up the direct model comparison under Normcore vs. Weber Workshops tampers.
And one finding that gives pause: in a controlled series across three grinders, a well-known distribution tool led to significantly lower TDS and lower yield – and the spread was no smaller than without it. A levelling tool is not the same thing as a Self-Leveling-Tamper, but it is the only controlled series on this category of device that we found, and it comes out negative.
Our stance on it, and it really is only ours: a Self-Leveling-Tamper is a good tool for ergonomics, wrist relief and process standardisation. In a café everyone tamps the same way with it, regardless of physical strength or how the day is going. That is a perfectly legitimate reason to buy one. “Better extraction” is not. If you sort your budget by evidence, you invest first in the grinder, then in dosing and recipe, then in a WDT needle for a few francs – and only after that in the tamper. In the Puck-Prep & channeling simulator you can play through how distribution, tamping and grind size interact.
Frequently asked questions
How hard should I tamp?
Between roughly 5 and 20 kilograms the difference is demonstrably negligible. Three independent series – one of them with 150 extractions – found no significant differences in shot time, TDS or yield. More important than the force is tamping straight and level. Below 5 kg there is no solid data.
What does WDT actually achieve?
What is documented is a much smaller spread in flow resistance: with mere raking of the surface, peak resistance varied by around 40 percent, with deep WDT considerably less. That is not a documented rise in average extraction yield. So far no peer-reviewed study on WDT exists.
How thick should WDT needles be?
Thin. In a comparative series, 0.4-millimetre needles through the full bed height gave the best consistency. The rule of thumb in the literature is that needles over 1 millimetre in diameter are counterproductive, because they churn the bed rather than de-clump it.
Is a Self-Leveling-Tamper worth it?
For ergonomics, wrist relief and consistent work across a team: yes. For better extraction: there is no evidence, because the quantity these tampers make reproducible – the tamping force – is demonstrably without consequence in the relevant range. That is our assessment, not the whole industry's.
Why does an espresso taste sour and bitter at the same time?
That is the signature of channeling. At the channel it over-extracts locally, which tastes bitter, in the rest of the puck it under-extracts, which tastes sour. Both land in the same cup. Cameron et al. (2020) measured that at fine grind settings up to 13.1 percent of the bed surface is no longer reached.
Roastery tip
You hear channeling before you taste it: when the shot suddenly speeds up and sprays lighter, a channel has torn open. For practising, a forgiving coffee works well – Cozy Chocolate lets a lot slide. Wild Peach, by contrast, shows every distribution mistake immediately, because acidity and bitterness sit far apart there. If you want to learn to work cleanly, practise with the difficult coffee. Our recommendation, one of many.
