How does the 'no distribution distribution' method work?

[JakeLiefer]

Over the past couple of years there's been a move away from distribution methods. As new and improved grinders have come on the scene, the preferred method seems to be the 'no distribution distribution' method, with a mound in the center and a tamp to distribute the grinds in the basket. Fans say it results in faster, cleaner, and consistent dosing, with an added benefit of reduced channelling. I'm not discounting this, but I'm just confused how this happens. Could someone with a better understanding of fluid dynamics and physics explain this?

In my mind, part of the importance of distribution is to ensure that grinds are all the way up to the wall of the basket. With the no distribution method, it seems like there's a larger chance of the grinds not evenly making it to the edges. I'm probably completely off with this and I know many others are confused too, so any explanation would be helpful, especially when many of us were schooled in elaborate distribution methods and debates. Is there a 'golden ratio' of mound height to basket width? Does it even matter?

[Andy Schecter]

As new and improved grinders have come on the scene, the preferred method seems to be the 'no distribution distribution' method, with a mound in the center and a tamp to distribute the grinds in the basket. Fans say it results in faster, cleaner, and consistent dosing, with an added benefit of reduced channelling. I'm not discounting this, but I'm just confused how this happens. Could someone with a better understanding of fluid dynamics and physics explain this?

The deafening lack of response you've gotten so far suggests that:

(1) Nobody with a rudimentary understanding of fluid dynamics and/or physics would ever claim this, and
(2) The "no distribution method" doesn't reduce channeling; it's just that good grinders make channeling less of a problem anyway.

[Philip Search]

While my understanding of fluid dynamics is limited, there are a few things Ive been thinking about:

1. how much head-space you leave in the basket. The technique of no distribution works best with machines that produce a water flow to the center of the basket (la marzocco and synesso are good examples of this) and when using a basket that gives you some head-space for water to build up before it begins to soak into the coffee. Another thing to consider in this is how the water flows through the coffee bed once full pressure is reached (this happens several seconds into the shot, as the head space has to fill first, then pressure builds).

2. is it possible to re distribute coffee and gain an even density?

3. is an even density going to give you the best taste?

I have spent a long time thinking and experimenting with this, and here is what i know, a slightly denser center produces a more even extraction of the whole puck (based on testing what was left in a puck after extraction; when used with the before mentioned basket. Shots done this way seem to taste better (subjective). Lastly it is impossible for multiple baristi on a bar to conform to a re distribution technique that leaves a error margin of less than .5 grams +/- and even distribution. This is not the case with the other technique.

What tastes best to you?

[JakeLiefer]

While my understanding of fluid dynamics is limited, there are a few things Ive been thinking about:

1. how much head-space you leave in the basket. The technique of no distribution works best with machines that produce a water flow to the center of the basket (la marzocco and synesso are good examples of this) and when using a basket that gives you some head-space for water to build up before it begins to soak into the coffee. Another thing to consider in this is how the water flows through the coffee bed once full pressure is reached (this happens several seconds into the shot, as the head space has to fill first, then pressure builds).

2. is it possible to re distribute coffee and gain an even density?

3. is an even density going to give you the best taste?

I have spent a long time thinking and experimenting with this, and here is what i know, a slightly denser center produces a more even extraction of the whole puck (based on testing what was left in a puck after extraction; when used with the before mentioned basket. Shots done this way seem to taste better (subjective). Lastly it is impossible for multiple baristi on a bar to conform to a re distribution technique that leaves a error margin of less than .5 grams +/- and even distribution. This is not the case with the other technique.

What tastes best to you?

Hi Phillip,
You pose some interesting questions concerning how we look at puck extraction. How are you achieving a denser center puck? I'd like to replicate some of these experiments and see what happens. I imagine using different tamper bases affect the density or how much the center mound piles up? Currently, we're running a Mazzer Major with the Elvinator installed (http://www.espressoparts.com/ELVINATOR_V1), so while it produces a much better throw, I still feel the need to distribute. I'm looking to upgrade to a timed, doserless conical grinder to improve consistency and cleanliness. As you mention, not having to distribute increases the consistency, which can in turn increase the taste. I just want to make sure that the 'no distribution method' (we need a name for this!) still ensures a proper extraction.

[Philip Search]

Jake, Ive been super busy at work and haven't had a chance to reply... sorry. I only have a minute here now, so expect a more detailed response when I have more time.

A good starting place would be to do this experiment with your grinder you have by setting the grind with your normal method and weighing out where you dose is at. Now, at that grind, weigh your dose out, put it back in the chamber and do your normal distribution (tricky i know with out spilling grinds). Next do the same thing but don't re-distribute simply put it all into the center of the basket and tamp lightly to settle, then pull off, then go down with pressure as you normally would (don't tap if you do that). This will create a denser center due to the grinds falling in the middle of the basket. Watch and taste both extractions. They should be a bit different. NOW do the reverse, set the grind and dose by doing the dose to the center technique and then reverse the experiment. Lastly, as a comparison, at both grind settings, don't redistribute, but simply move the portafilter in such a way as to dose in a circle with coffee falling evenly around the basket (easiest with a naked).

Please post what you thought and found when you did this, noting some controlling factors like grinder temp (how many shots in a row etc.), ambient humidity, your espresso machine etc.

Hope that give you a good starting place.

p.s. it is sometime neat to cut the puck apart with a razor into cross sections (i do it into about 5 or so) and see if there are any weird spots or what not. Im not saying you can tell much from this (unless you happen to have a chem lab, and a dryer handy) but its fun.

Heres to good coffee.

Philip

[Philip Search]

Just one more quick note, people often think that the density of the puck is established at tamping... this isn't the case, it is reinforced there, and can be changed somewhat, but the density of the puck is mainly established at the initial dose (exception would be some elaborate tamping techniques).

[nick]

Ok.

First off, some folks have noted over the years that the word "distribution" is probably not the best one to use in this context. "Re-distribution" is more accurate. "Leveling" sort of captures the idea of the final "swipe" or "level-off." Anyhoo...

It's helpful to think about why people were "fingerstrike leveling" (another term) in the first place:
1) mass control : either adding mass to the resulting puck by a controlled "smoothing" downward, and/or by removing part of the mass or pile with your level-off, you could, with training and practice, be left with a fairly consistent mass of coffee grounds to tamp and brew
2) lateral density control : because the dosing mechanisms weren't delivering coffee into the baskets in a way that created a fairly even lateral density distribution, baristas would redistribute the coffee to sort of fill-in lower density areas.

Over recent years, there are two espresso grinder features that have become more popular: timers and different dosing mechanisms.

You can, with relative precision, correlate mass of the dose with a duration of grinding time. This can help, if not completely resolve the mass-control benefit of fingerstrike leveling.

As far as lateral density control, some of the newer grinders are delivering coffee more evenly. The prolific Mazzer dosers notoriously throw "back-left," while the Compak, Anfim, modified doser (Elvinator, etc.), and the better grind-on-demand models are delivering coffee more centrally.

Some folks are looking at their resulting pile, untouched, and deciding that it looks good enough to tamp without messing with it manually. With some tapping and/or settling, you can get a fairly uniform (in lateral density) basket-full of coffee that indeed doesn't really need to be messed with. Frankly, messing with it would probably introduce more problems than benefits.

Anyway, that's my two cents on the subject.

The precision of the timers on some of these grinders is in doubt. Frankly, it's the speed of the grinders that I see as the biggest problem: the faster the grinder, the smaller the margin of error to deliver a consistent dose.

All of this, as with most things espresso-related, doesn't exist in a vacuum. I mean that how much you're dosing will determine whether or not you're just mashing your dose up against the shower screen, and the shape of your tamper will affect how the pressure exerted by the tamper piston directs the coffee and the force in its various directions. For the record, I'm a mild-curve-tamper proponent. I think flat-bottomed tampers suck.

[nick]

P.S. to answer your initial questions, Jake, I've never heard that it's necessarily cleaner, consistent, or that it reduces channeling.

Right back at you though: why do you say that "part of the importance of distribution is to ensure that grinds are all the way up to the wall of the basket?" I don't understand what you mean.

For the record: I don't have any clue what Philip means by "The technique of no distribution works best with machines that produce a water flow to the center of the basket." How does a La Marzocco or Synesso's shower screen have anything to do with this? Because of the central dispersion screw design? But the water is directed outward, not downward from that point.

"Ambient humidity?"

[Philip Search]

Hey Nick, I totally agree that most forget the main purpose of redistributing coffee had a lot more to do with barista adapting to grinding per shot as a method of volume control. I work on a couple of different machines currently, and after a couple of shots, the initial drops of water off of machines with dispersion screws and screens in good working order tend to fall more to the center of the basket as opposed to other machines with dispertion blocks that have a rear or from flow pattern. Example would be a simonelli etc. Yes, ambient humidity, because it drastically effects certain aspects of grind quality if your experimenting over a period of time to cool your grinder. There have been some interesting studies commisioned by con agra (sp?) in relation to grain milling and the calebration of grain mills that correlate this. I have seen many examples of shops with multiple baristi who would back up claims of reduced channeling and faster bar flow with less waste when using a no level approach, while I readily

[Philip Search]

Oops, phone posting mistake... To finish, I readily acknowledge some may not see these benifits as they are endowed with god like barista skills. As for tamps, I can't seem to detect a difference between the two I use most on the bar(one flat, one convex), but I think the variables are fairly interdependent. Shrug.

[nick]

I think that reduced channeling (with no-redistribution techniques) may be from the increased attention on dosing properly since there's no redistribution to (errantly) rely on.

When watching baristas work (that is, when watching 95th-percentile baristas who 'should' know how to pull a good shot), I often watch their eyes to see where their focus is. Often they pay little if any real attention to their dosing, a moderate amount of attention to their redistribution technique, and hyper-focused attention on their tamp. I think it'd be easy to argue that they're completely backwards in their attentiveness.

As for ambient humidity, sorry Philip, but I still don't get it. Does ambient humidity make a difference? Sure it does. Is it a relevant metric in this context, especially in the list that you wrote? Sorry, but I really don't think so.

As for flat vs. convex, I think it's one of those things: when your lateral density distribution is sound, flat vs. convex really doesn't matter, but when you're a little off (and who isn't at least some times?), convex helps apply a little lateral pressure in conjunction with the downward pressure, rather than a flat tamper piston which provides lateral pressure only after a certain amount of downward pressure is achieved. That lateral pressure helps reinforce the coffee's adhesion with the sides of the basket.

Sort of like having a nice big head on a golf driver... though Trish hates my analogies and hates golf analogies most of all!

[Philip Search]

Hmm, a couple more thoughts; humidity is important to me as a relevant factor because I think that it contributes quite a bit to channeling as to clumping and where your grind and dose need to be set at. I feel that along with age of espresso it is one of the variables that is most overlooked by baristi. As for tamping, I totally get what you are saying, in some cases, except that with a normal tamp procedure (is there such a thing?) you are only tamping the top 1/3 of the coffee bed. That said, there is a good chance that your correct for the simple fact that the tamp is only effective up until the head space is full and the water pressure hits the coffee putting much more force than the initial tamp into play. My query is, since the puck is generally saturated at this point, would channels have already formed? I have a few other thoughts rolling around like small marbles on a hill, so Ill have to play around a bunch more today and see what happens switching between tampers.

[JakeLiefer]

A good starting place would be to do this experiment with your grinder you have by setting the grind with your normal method and weighing out where you dose is at. Now, at that grind, weigh your dose out, put it back in the chamber and do your normal distribution (tricky i know with out spilling grinds). Next do the same thing but don't re-distribute simply put it all into the center of the basket and tamp lightly to settle, then pull off, then go down with pressure as you normally would (don't tap if you do that). This will create a denser center due to the grinds falling in the middle of the basket. Watch and taste both extractions. They should be a bit different. NOW do the reverse, set the grind and dose by doing the dose to the center technique and then reverse the experiment. Lastly, as a comparison, at both grind settings, don't redistribute, but simply move the portafilter in such a way as to dose in a circle with coffee falling evenly around the basket (easiest with a naked).

Unfortunately as I was getting ready for this experiment, my scale broke. As soon as I get another reliable gram scale, I'll post up the results.

Right back at you though: why do you say that "part of the importance of distribution is to ensure that grinds are all the way up to the wall of the basket?" I don't understand what you mean.

When I've seen channeling, it tends to be near the basket walls. This seems to be the 'weakest link' in the espresso puck, and I've assumed that a poor distribution often leads to this channeling. I've worked mainly with Mazzers with the left throw, so the right side was often the culprit of channeling. I think the theory was that extra attention needs to be paid to the walls, as this tends to get the least coffee but most likely to cause issues.