MARGINAL GAINS - LIVING WITH THE PERGTAMP

This is the first in a series of posts looking at the Marginal Gains we can make in improving what we do everyday. For those who are not familiar with the term it refers to those small incremental improvements that can, over time, aggregate into a larger step forward.

The early posts will be focussed on extraction, but later posts will branch out into service, efficiency etc....

Lets start with an easy one, The Pergtamp.

If your'e not familiar with The Pergtamp, have a quick read of this from one of its creators, Matt Perger. http://mattperger.com/Pergtamp

The concept and the science are solid and well thought out and there are some bold ( but qualified) claims for how it can improve your extraction, but what is it like to use and do those claims hold true?

For clarities sake, lets separate the anecdotal evidence from the empirical data, and look at that first.

You instantly notice that the build quality is, as you would expect, excellent.  A well turned, comfortable wood handle paired with the stainless base that is a darker tone to most other grades of steel used in tampers. The edge is as sharp as Matt describes it and it's this, combined with the increased diameter,  that makes the difference when you use it. You can feel it fit better in the basket, and this forces you to take more care over your tamping routine, which can only make your more consistent.

The increased diameter is interesting as I feel it's value can be felt in many was. We have seen what seems to be a marked decrease in channels from the edges of the basket, although we have never tracked this. Lets consider why.

A coffee puck is a bed of irregularly shaped granules, water flow through this bed is dependent on a number of factors, but for now lets consider the puck density and voidage. Voidage is that volume not filled by solid materials and will provide the lowest resistance to fluid flow. So If our goal is to achieve an even extraction, then it is logical that a even puck density is important, this has been known for some time and is the basis of distribution techniques. Channelling occurs when there are areas of low density within the bed, again nothing new. But, when the bed is contained there are often significant increases in voidage at the interface of the container and the bed, or in our case the puck and the basket. What the Pergtamp does is mitigate these voidage effects by allowing us to create a more even density gradient over the entire bed. Logically when used in conjunction with an even distribution technique, this should allow us to extract more evenly across the bed.

Whether this is the case or not is difficult to measure, after all it does't necessarily follow that an increase in the evenness of extraction leads to an increase in the total extraction. 

But what about the total extraction? Does the Pergtamp deliver in the "significant and noticeable increase" that Matt expects?  Here's where we get to the data.

For some context, we are a multi-roster shop, running only 8kg of each single-origin espresso. We measure the extraction yield after dial-in when we have set our recipe. 

We use a La Marzocco Strada EP with Nuova Simonelli Mythos 1 Grinders and VST 18g baskets. Our water specifications, if you are interested, can be found here.

What the data allows us to look at is the average extraction over a period of time, as this would cover espressos from a range of roasters, regions and varieties. The Pre-Pergtamp era was a Reg Barber 58.35mm flat base.

The data can be viewed here.

But here are the highlights.

Pre-Pergtamp

Mean Average - 20.89%

Standard Deviation from Mean - 0.71

Post-Pergtamp

Mean Average - 20.84%

Standard Deviation from Mean - 0.61

What's immediately apparent is how similar the results are, so much so that it is not far fetched to say that there has been no significant change in extraction since using the Pergtamp. The 0.1 difference in Std Deviation could be argued to be significant, showing that the Pergtamp facilitates a closer grouping of extractions across a range of espresso, and so perhaps more even extraction. But in reality 0.1 is not a significant enough a difference to manifest any effect in taste. This become more apparent when we take a closer look at the data, what's apparent is that the Pre-Perg set contains an obvious outlier @ 22.44%. If we remove the upper and lower outliers for both sets, the differences shrink even further.

Pre-Pergtamp

Mean Average - 20.83%

Standard Deviation from Mean - 0.55

Average Absolute Deviation - 0.46

Post-Pergtamp

Mean Average - 20.87%

Standard Deviation from Mean - 0.53

Average Absolute Deviation - 0.44

Now the results are even more similar, with both the mean average and both standard and absolute deviations showing small enough differences as to be insignificant.

These are, of course, isolated tests over a very small sample set. But I believe that the conditions of the testing encompassing  many different espressos, show a realistic expectation of the results for further comparable investigations.

So where does this leave us? Does the Pergtamp have an empirical effect on extraction? For us it seems not, but that's not to say that for different systems and set ups it wont have an effect. I'd postulate that we may already achieving the maximum extraction capable by our combination of equipment, so much so that switching to the Pergtamp simply couldn't give us any added extraction. That said, we have benefited from a saving in efficiency due to reduced channelling as well as an implied reduction in wastage. If there is an increase in evenness of extraction, it is too hard to tell at this time.

We still use the Pergtamp and will continue to for the foreseeable future.