Back in February, Lance Hedrick showed the world that shaking your grounds before making an espresso could both increase the extraction and reduce the shot time of your espresso. The effect on extraction was striking, but it was the fact that it also decreased shot time that really fascinated us. That finding implied that shaking could let you grind finer without causing clogging, potentially allowing higher extractions without negative flavours creeping in.

Just a few months later, shaking appeared at the World Barista Championship, when Dominika Piotrowska, representing Austria, used shaking on stage for her routine. Domenika explained to us this week one very tangible benefit from shaking grinds: 

“One significant finding was the remarkable consistency I achieved in extraction time. The difference between shots was only 0.1 seconds each time (e.g., 20.0/20.1, 20.0/20.1). This level of consistency was unprecedented for me.”

Dominika Piotrowksa dosing into blind shakers at the 2024 World Barista Championship.

Our own testing, on the other hand, gave contradictory results. Shaking seemed to work in some situations and not others, and we never saw the combination of increased extraction and decreased shot time that made shaking so exciting. In some sense, this is what we expected: in our first post on the topic, we explored the possible reasons that shaking might work, and concluded that “the ‘best’ distribution method overall will depend on the grinder, the roast, even the ambient conditions.”

This topic has opened many new doors for espresso making, but weirdly — no one can quite work out why it works … and sometimes doesn’t work in the way you’d expect. Yes, we expected shaking to work better in some situations than others, but we wanted to know why it worked for Lance, and not so well for us. Does it depend on the coffee, or the grinder? Or something else entirely?

In our previous posts, we explored ‘densification’ — the idea that fines would combine with larger particles to create more spherical aggregates. While we still don’t have a complete answer, after some new testing this week, what we have seen suggests that this isn’t happening. In fact, we see the opposite: 

Shaking liberates fines!

The Shaking in the Basket Test

On our travels, we’ve noticed quite a few baristas starting to shake directly in the filter basket, whilst holding a dosing cup on the top like a lid. So we wanted to test the efficacy of that approach compared to using a purpose built shaker. We had a theory that possibly a lot of fines could be sieved through the holes in the bottom of the basket whilst shaking and we wanted to find out if that was a problem, or if it would affect extraction yields. But in the process of testing this, we seem to have hit upon something quite a lot more interesting. 

This week our Dean of Studies, Jem Challender teamed up with Lloyd Meadows from Tortoise Espresso to test shaking using a Mythos grinder and a Synesso and found that shaking made very little difference in extraction or shot time with one coffee, and slightly increased both extraction and shot time with another. None of these results showed a statistically significant effect, but they are at least consistent with the idea that which coffee you use could affect how well shaking works.

We then repeated the tests with the same coffee used in the second test, but using an EK43 instead. This time, shaking seemed to increase extraction, although the difference was not statistically significant (p ~ 0.1). It did however decrease the shot time by a couple of seconds, and that difference was significant (p < 0.05).

Extraction yields and shot time averages for shaking vs not shaking. Results are an average of 5 shots per test. In each test, we distributed the coffee using the AutoComb. The coffee we used was a Colombian, Reinaldo Parra ‘Las Pinas’ Washed Espresso.

These results suggest that the choice of grinder is also part of the reason that shaking works better in some tests than others. In our last post we observed that the grinders that Lance used for his tests (the EG-1 and DF64) both produce less fines than other single-dosing grinders, which could explain why his results were so different to ours.

If it is the behaviour of fines that determine how shaking works, then it’s no surprise that the effect is different with different coffees and grinders, and even in different locations. The amount of fines you get while grinding depends on the coffee and the grinder — as well as many other factors, including temperature, the moisture content of the beans, the condition and alignment of your burrs. This alone can explain why we get different results than Lance, and even different results from one test to the next.

Free Your Fines

But what is it that fines are doing when we shake the grounds? You remember early on in the shaking discussion, coffee researcher and roaster Dr Mark Al-Shemmeri, theorised that shaking could have an effect similar to ‘densification’; a process used in industrial roasteries when packing pre-ground coffee. A densifier is essentially a large mixer, and the mixing process smooths out the jagged edges of larger coffee particles, and forces fines to lodge inside their pores. This produces aggregates that act like denser, more spherical coffee particles and reduce the number of free fines in the grounds.

When we tried measuring the particle size distribution of grounds after shaking, however, we found that the opposite was happening. Using Professor Abbott’s Grind Size Analysis app, we saw an increase in the number of particles below 100µm after shaking.

But here’s where it gets quite bizarre: During our most recent tests, when pulling shots with shaken grounds, the portafilters seemed to get dirty more quickly. Take a look at these photos below. Each photo shows the residue left behind on the bottom of the portafilter after pulling just one shot. The portafilter was spotlessly clean prior to pulling each shot.

Shaking grounds releases fines from the coffee. The inside of a portafilter after making one espresso shows the difference that shaking makes. Without shaking (left), the portafilter still looks clean. After shaking in a dosing cup with the filter basket on top (middle) or with a blind basket (right) a lot more fines and residue are visible.

We tested this several times, and each time the result was the same. When we made a shot without shaking, as seen on the left, a very small amount of fines collected on the base of the portafilter as you would expect. Then we tested the effect after shaking for 5 seconds with a dosing cup inserted directly on top of the filter basket (middle) — cocktail bartender style. To check that this effect wasn’t simply the result of fines escaping through the holes in the basket, we also tried shaking in the same dosing cup but with a blind basket on top. Shaking appears to increase the amount of undissolved solids coming through the holes in the portafilter — and in the case of shaking in the filter basket the amount of residue is dramatic. With both methods the increase in the amount of solid on the floor of the portafilter is obvious, but we’re guessing this is only a fraction of what is being washed down into the cup below. 

If shaking releases more solids from the coffee bed, perhaps this also explains the flavour and tactile effects we saw in our previous post. Miruna Tudose and the team at the fabulous Origo Coffee in Bucharest reported a “somewhat powdery texture, and a lack of clarity” in their shots after shaking the grounds.

Please note, for the photos above, we used an HE% basket from Pesado, which has pretty large holes to allow high flow. With that in mind, we also checked how much coffee escapes through the holes in the basket if you really go for it with the shaking. After 30 seconds of forceful shaking, we were able to collect around 0.1 g of coffee — not a huge amount, but perhaps enough to have an effect on how the shot runs.

To Shake or Not to Shake?

At the end of all these tests, our conclusion remains the same: whether or not shaking increases extraction and decreases shot time seems to vary, depending on your coffee, your grinder, and many other factors. It may well be that shaking is more beneficial if you have a high-end grinder, like the EG-1.

If you’re going to shake, then it’s probably better — and definitely cleaner — not to shake directly in the filter basket. A dedicated tool like the Blind Shaker from Weber Workshops is in any case far more ergonomic than trying to shake a heavy portafilter.

But whether you shake the coffee in the portafilter or in a dedicated shaker, keep an eye on your spouts. You might find you need to rinse the portafilter more often than usual to keep the coffee tasting clean and sweet over the course of a long day’s espresso making.

Ultimately, the only way to know if shaking will work for you is to try it out for yourself for every coffee that you serve, checking the effect on flavour as well as extraction before making a decision. But if you have the right setup, and the right coffee, to benefit from shaking, it’s a marvel how effective it can be.

Methodology

The Kit We Used

• For these tests we used a Synesso Hydra. 
• The preinfusion pressure was running 3 bar and pump pressure was set to 8 bar. 
• We used Pesado HE% Baskets. 
• The EK43 has titanium coated burrs. 
• We used a Rhinowares stainless steel dosing cup.
• We tamped using a Puqpress and distributed using the AutoComb for all shots. 
• We did not tap the basket vertically. 

How We Shake

• For the ‘Shaking in Basket’ method we dosed the coffee into the basket using the dosing cup; left the dosing cup in place (upside down above the basket) and shook the portafilter and dosing cup like a cocktail shaker for 18 shakes (five seconds of shaking).
• For the ‘shaking out of basket’ method, we placed a blind basket on top of the dosing cup to stop the coffee falling out and shook that like a cocktail shaker, also for 18 shakes. (That’s not a bad shaker hack if you’re on a budget by the way.)
• For the ‘No Shake’ method we did everything the same as the aforementioned shaking methods, just no shake.
• Aside from the three pics we’ve already shown you (above) we pulled another three repetitions of each method which you can see below. 

No shake

Shaken out of basket

Shaken in basket

A Future Experiment

Seeing as there’s a difference in the amount of grinds sitting on the floor of the portafilter from these three approaches, there’s very likely to be a difference in the amount of solid in the cup. So how to test that? One idea we’ve had is to pour an espresso through a filter paper; wait for it to drain; then pour some more water through the paper to rinse out the dissolved solid. After that, we might be able to use a modified version of Prof Abbott’s grind analysis app to count the particles that are trapped in the filter paper and measure their diameters.