Distributing for Espresso

March 22, 2015

Last week I sent out a survey with a single simple question:

How do you distribute for espresso?

Distribution is the act of evening out the coffee grinds in an espresso machine’s basket before tamping. This helps the water pass through all of the coffee grinds at the same speed and pressure, increasing the evenness of the extraction. More on that later.

I received nearly 3000 responses (from Barista Hustle subscribers only) and an incredible array of “other” answers. Baristas around the world are distributing with needles, toothpicks, credit cards, knives, cake spatulas (?), love taps (??), “new tools I made”, something called the “Chicago Drop”, Gwilym Davies, gift cards, Pallo cleaning tools, "my girlfriend” and many others. It was fun to read.

What’s more important is the majority and what they’re doing.

Here are the responses and their percentages:

Finger Swipe - 30%
Hand/Palm Grooming - 20.6%
Just tamp it as it comes - 14%
Stockfleth - 10.4%
Other - 7.9%
Dosing Tools - 5.9%

The vast majority of the "other" responses were either:

Vertical taps, “collapses" or “drops" on the grinder forks/bench - 2-3%
Tapping on the side of the portafilter with fingers or palm - 2-3%
Other stuff - 2-3%

Upon reflection I really should’ve included horizontal and vertical taps as an answer even though they’re only 3% or so. I believe some responders may have also chosen Hand/Palm Grooming option for tapping on the side. Thanks for taking the time to write your method in, and please excuse the omission.

I have rather strong opinions about distribution but - as with a lot of coffee techniques - I have never really tested it thoroughly. My method makes sense to me and feels consistent, yet still it looks like I might distribute differently to ~90% of you! This is both worrying and comforting.

So I set out to do some really intensive testing this week. I had designed an experiment and was half way through when I realised some inaccuracies were creeping in. I don’t want to publish bad data or mislead people when I know I can do better. I’m angling to have improved results ready within a week or two.

Instead, this week I want to explore the concept of distribution and create a framework for thinking about and evaluating it. I’ll also voice my opinions of the most popular methods. Hopefully I’m not too far off what the experiment will tell us!

There are a few metrics one needs to think about when finding an excellent distribution method. In order of importance, they are extraction, consistency, speed and cleanliness. Let’s go through them in that order.

Extraction

An excellent distribution method will enable an even extraction. The best thing about this is that there’s a ceiling. There is a maximum to the evenness provided by a distribution method (namely, perfectly even). So obviously, we should be chasing that maximum.

Note: distribution won’t improve the evenness of grind sizes, roasting, baskets etc. It will only improve itself: a noble goal regardless of the others. Everything helps!

The only way to really measure extraction effectiveness of a distribution method is to prepare identical shots with different distribution methods and A) taste the difference and/or B) measure the difference in extractions. When tasting, a superior distribution will produce the sweetest and strongest espresso. It may also present less extraction taints than the other. When measuring, a superior method will produce a higher strength and extraction.

Disclaimer: I’m not saying that stronger and highly extracted espressos are always better. I’m only saying that a better distribution will allow more of the coffee to be extracted, which produces an espresso with a higher strength and extraction.

Consistency

An excellent distribution method will create many extractions that are very similar. Wether they are better or not is up to the previous section. For consistency, all I care about is, well, consistency. The best way to measure this is by pulling a large number of shots and measuring their strength and extraction. The tighter the grouping, the better the consistency.

Another factor of consistency is how well the method can be transferred to another Barista. If you and your team mate are individually consistent but different to each other than the method has failed. An excellent method is easy to teach, doesn’t suffer from human error too badly and is transferrable across a large number of Baristas.

Speed

An excellent distribution method is fast. Really fast. If you’re making 500 coffees a day, the difference between spending 2 seconds and 5 seconds to distribute is 25 minutes!! It should also be part of a “flow” or progression from grinding to tamping. If you have to pick up extra tools, swap portafilter hands or make extraneous movements then the speed of the distribution will suffer.

That said, speed must sometimes be sacrificed to improve quality of extraction and consistency. As always, balance and intelligent compromises are essential.

Cleanliness

An excellent distribution method keeps the Barista and Bar clean. If your skin touches coffee it will dry up and get dirty. This damages your skin, creates mess down the line (eg. dirty cups), slows you down to regularly clean your hands and creates a minor contamination hazard (coffee brewing is a pasteurisation of sorts, but you can never be too careful). I am a strong advocate for never touching coffee with your hands or fingers: the advantages (extremely few) will never outweigh the disadvantages (many).

Tantamount to personal hygiene is that of the bar. Distribution shouldn’t create any mess. If it does you’re wasting coffee and decreasing accuracy of the amount of coffee in the basket. You’re also doing future you a disservice by making them clean up your mess.

Secondary Indicators

Here are some other less obvious ways to measure the effectiveness of a distribution method.

Where and when the espresso emerges from the basket with a naked portafilter can tell you if there’s areas of lower density in the coffee grinds. This is obviously much less accurate than measuring extraction. If the espresso is emerging first from the edges then into the middle, your method has a central tendency which reduces flow in the middle of the basket. If there are patches that emerge first, then there is significantly less coffee in that region.

The last one that I thought up while designing this experiment is to measure density of the tamped coffee. To do this, one needs a dosing tool or piece of plastic/metal that scrapes out coffee grinds above a certain level in the basket. This would leave ~95% of the coffee grinds in the basket taking up a consistent volume of space in the basket. After distributing and tamping each basket, one could scrape the coffee grounds away, leaving a very consistent volume of grinds in the basket each time. If this consistent volume of space held differing weights of coffee grinds each time then the method is inconsistent.

How do they stack up?

So we have our four considerations for distribution effectiveness. Now let’s apply them to the most popular methods!

Some of these methods utilise vertical taps or “collapses” to varying degrees. That’s all well and good for vertical density of grounds, but the main issues like horizontal distribution are still holding them back.

Just Tamp It

The laziest and fastest of all the methods. Just tamp whatever the grinder gives you. Sure, you can move the handle around beneath a doserless grinder to improve distribution around the basket, but you’re being lazy.

The grounds might be a mound, they might be totally random, or you might be able to get a half-decent distribution from the way your grinder doses it up. Regardless, I’m of the opinion that you need to distribute no matter the delivery of the grinder.

What I already know is that this method has a fast speed and moderate cleanliness. In the experiment, I’m predicting the most uneven extractions and lowest consistency of them all.

Stockfleth

The Stockfleth uses your finger to spread the coffee around the top layer of the basket, discarding anything above a prescribed layer. Some baristas claim this method allows them to control the dose more accurately. I’m calling their bluff. Humans are not accurate or sensitive enough to perform this task to the required levels of consistency. We can’t sense the density of coffee beneath our finger without pressing on the coffee, which would defeat the purpose of sensing the density in the first place.

Second to my density argument, the Stockfleth method fails to distribute coffee to the lower edges of the basket. Because it only manipulates the top layer of coffee, Stockflething neglects the lower layers leading to uneven distribution of grounds. It may look like the coffee is evenly distributed, but it’s just a façade for the grinds beneath.

Waste can be minimised by sweeping excess grounds into a container or grinder doser for later use. This is rarely the case, especially with the prevalence of doserless grinders.

What I know already: Cleanliness and skin health are terrible. Potential wastage is high. It also takes way too long.

What I predict: mildly inconsistent and uneven extractions.

Of course, the Stockfleth is much more effective if you're overfilling baskets and heavily under-extracting your espresso. The method relies on the basket being packed up to the brim with dense coffee grounds. Lower the dose to within the recommended range of a bigger basket (a la VST) and it falls over.

Finger Swipe

The Finger swipe is very similar to Stockflething. It’s inconsistent, unclean, and wasteful if you’re not reusing leftover grinds.
Also, Newsflash: the edge of your finger isn’t straight!

I’m predicting inconsistent and uneven extractions once again.

Hand/Palm Grooming

Some respondents may have chosen this answer for tapping the side of the portafilter with their hand. I’m evaluating this method later.

Grooming the coffee grinds with your hand or palm shares the problems of both the Stockfleth and Finger Swipe, with the added bonus of getting your palm dirty as well! It also - usually - moves more coffee to the centre of the basket rather than the edges.

Same predictions as above.

Dosing Tools

Depending on the tool, you can get mixed results. For best horizontal distribution you need dosing tools that operate below the rim of the basket (eg. the curved Scottie Callaghan tools) but they put less coffee in the middle of the basket and more at the edges, reducing evenness.

Tools can be fast and fairly consistent if you have developed a nice method. Keeping that consistent between members of staff is another matter.

Cleanliness and wastage is also minimal if you’re disciplined; though can be difficult when trying to push coffee right to the edges as some will inevitably fall out.

Other tools like knives and credit cards will give you a seemingly perfectly flat surface on top, but won’t help your distribution of grinds lower in the basket.

Same predictions again.

Tapping Vertically and Horizontally

I feel as though this is the best compromise between speed and quality. combined with a perfectionists cleanliness. The name says it all.

Horizontal taps will move the coffee right to the edges of the basket. That same motion will even out the distribution as denser areas will become stiff, allowing loose grinds to fill the rest. This motion affects the entire mass of coffee; not just the top layer a la Stockfleth.

Vertical taps on a solid surface will then remove any air pockets within the grinds. If there is any unevenness, you can easily see it because the grinds will sit at different heights across the basket. With that visual clue, you can revert to more horizontal taps to even it out.

This method is super clean, super fast and my personal favourite. I’m predicting (and hoping) that it'll will win the experiment. It’s what I use every day, and a method that I’ve prescribed to many, many espresso bars around the world. I’m also totally happy to be proven wrong. This project is as much about learning for you as it is for me!

So there’s lots of things to think about when evaluating distribution methods. The more even, consistent, faster and cleaner, the better! I’m really looking forward to seeing some hard data to sort this issue out (at least temporarily). If I can’t execute the experiment properly before next weekend, I’ll be posting a thorough guide to how I distribute with the tapping method.

Write your predictions in the comments! I'll be in there and active as per usual.

If you have found this useful and want to enjoy delicious coffee with the rest of the community – register for our monthly Superlatives coffee subscription. Or if you just want to keep up with every thing Barista Hustle - sign up to the Newsletter.