Extraction yield or taste assessment: For coffee professionals, which came first?

Coffee preparation techniques besides espresso like pourover.
User avatar
TomC
Team HB
Posts: 10534
Joined: 13 years ago

#1: Post by TomC »

For those who might better understand the history of the SCAA/CQI charts for extraction, I have a general, random, and likely useless question. We all know of the charts, and the tests that were performed to determine what created the "ideal" beverage strength/yield. But what I'm curious to know is if they first brewed to the specific, precise extraction yields (objective data) then tested for flavor/taste assessments (subjective, collaborative data) or vice-versa? Chicken/egg, egg/chicken?

I have no idea if it even matters, but the thought crossed my mind and I didn't know the answer. I figured someone here would. It would seem that in the middle of the last century, oven drying would be carried out after a brew was made to determine the extraction after the brews were assessed, so it leads me to think the later not the former.

Has this same test been conducted with current roasting ideologies using the finest sourced coffee that our modern age has to offer, on a large scale? I hate seeing data about roasting/extraction and finding a lot of the subjective opinions and judgments made were based on lesser quality low grown or naturals from over a half century ago.
Join us and support Artisan Roasting Software=https://artisan-scope.org/donate/

User avatar
another_jim
Team HB
Posts: 13871
Joined: 19 years ago

#2: Post by another_jim »

TomC wrote: I'm curious to know is if they first brewed to the specific, precise extraction yields (objective data) then tested for flavor/taste assessments (subjective, collaborative data) or vice-versa? Chicken/egg, egg/chicken?
The tasting was done by taste panels of both regular and experienced coffee people. By contemporary standards, it would be a terrible research design to use known brews unless you were very confident of your double blind procedures. As a rule, you should do subjective assessments first, and objective measures later whenever possible, since this eliminates the possibility of blinding breaking down.

Sivetz reports that the CQI was financed by equipment vendors, who also ran the experiments. On the one hand, this is good; since product testing was already being done very well in the 50s. On the other hand, if the companies were simply confirming brewing targets they had established for their equipment; the research technically needs to be replicated.

My personal opinion is that this is probably unnecessary; and that the CQI data is good enough regardless of any flaws in the research design. For instance, you don't need to know the exact temperature to adjust a thermostat; you just need to know whether the room feels too hot or too cold. In the same way, you don't need to know the precise extraction parameters to fine tune a coffee; you just need to know whether it tastes over or under extracted. The CQI data is certainly sufficient to allow people to develop their palates to that point.
Jim Schulman

MWJB
Posts: 429
Joined: 11 years ago

#3: Post by MWJB »

Hi Tom,

It's hard to say for sure on a lot of these points because much of the detail on the research is hard to find, CBI/CBC did print books on the subject, but they didn't do the sensory research, as far as I can tell (David Walsh also tried to track a lot of this stuff down and seemingly met with a few dead ends).

Initially the National Coffee Associaton (a trade body) conducted research on strength vs yield preference. Their result was wider spread on %TDS (1.04 to 1.39) and a little lower on yield (17.5 to 21.2%). A later study by the Midwest Research institute gave the 1.15-1.35/18-22% box we are familiar with. This chart gave a range of ratios at which these targets could be met (gall/lb). There is no reason to assume that the respondent samples for these consumer studies were made up of anything other than a representative sample of the population (e.g. not aimed specifically at connoisseurs, indeed doing so would probably skew results). There is no evidence that I have seen/heard that specialty/single origin/lot/high grown coffees were the target of the research.

Later, the CBI (subsequently renamed the CBC) under EE Lockhart established the 'golden ratio' that would give you the best odds of landing somewhere in this box, as well as protocols for time & grind. They determined this to be ~56g/l, if you look at the SCAA chart, this line (106g/half gall) intersects the box at the bottom left and exits the box around 21% at the top right. The CBI/CBC research seemed to be more about the line, than the box. The box (identified by the MRI) was the range, but the line (+/- a couple of g/l) the mechanism to give you the most time/leeway in the box.

The Nordic Coffee Association has apparently run its own research and ended up with a box higher in %TDS, but within the 18-22% yield range. The SCAE declare that their chart pertains specifically to drip brewed coffee, using a bleached white paper filter at 50-65g/l. It too targets 18-22% yield, but at strengths that overlap the SCAA & NCA boxes. The 18-22% range is common to all three bodies. This is what confuses me with suggestions that Nordic coffee is inherently sour...is it inherently sour, or are lighter roasts harder to extract & often on the low side of nominal yield in the cup?

The chart adopted by the SCAA, from the CBI/CBC, does not specifically apply to drip coffee...well, it may be the SCAA's intention, but it was not the intention of the CBI/CBC. It works for drip coffee, but was aimed at large batch brewing & home manual brewing equally, using contemporary methods (percolator, drip, French press, vac pot).

Sorry for the long preamble...
TomC wrote:Has this same test been conducted with current roasting ideologies using the finest sourced coffee that our modern age has to offer, on a large scale? I hate seeing data about roasting/extraction and finding a lot of the subjective opinions and judgments made were based on lesser quality low grown or naturals from over a half century ago.
The SCAA did reinforce the MRI/CBI/CBC extraction range some years ago. The Nordic CA & SCAE are both specialty based organisations and any research they have conducted will be much more recent than the 50's/60's.

The roast/coffee varietal is perhaps less pertinent, unless the roast &/or less soluble bean varietal prevents you from getting 'in the box' with your brew method. Even if it does, coffee is not suddenly toxic, nor vile within a wider range than the ideal box suggests. It starts to actually taste like coffee between 12-14%, but it's a generic taste, certain coffees may taste more to one's palate at the lower end, just off the chart, or at the higher end. Reducing fines seems to be able to push the box up...this perhaps requires it's own 'box', but this box wouldn't really be relevant to the masses, or even the majority of home specialty brewers.

You would be better placed to know than me, but I don't really see much said (in US & Euro specialty circles) that suggests coffee is repeatably/consistently preferable outside the 18-22% extraction yield range (whether brewed or espresso). (There is also the concept of immersion yield for steeped brews, but this has only been identified after the aforementioned bodies devised their charts).

Another thing to consider is that there are many new brewers (without specialty body involvement/analysis during development) hitting the market all the time, if they, or your method, skew your average yields, you may be drawn towards coffees that taste better in those yield regions? How many times do you hear that certain coffees are best brewed in a French press - tea like, floral, notes of jasmine...?

Overextraction and under extraction apply to commodity grade & specialty coffee equally...it's just one might be easier to overextract & the other easier to underextract.

Generally, I think a lot of assumptions are made around the 50's research, based on what we consider normal in our relatively niche sphere today, that may not necessarily conflict with that research, but equally may not share the same motivation? (ball-park 'good' cups for the masses vs exploiting specific characteristics of certain farms/lots).

One thing I do wonder though, is whether significant regional preferences exist, many Asian recipes (80-100g/l & over) seem to suggest very strong, or under extracted (technically, not objectionable on the palate) coffee?

User avatar
endlesscycles
Posts: 921
Joined: 14 years ago

#4: Post by endlesscycles »

I've been running into something recently.

Strength is somewhat personal or regional, but extraction is supposed to be more universally agreed on. 20% is supposed to be "right", right?

Lately, I've found some of my roasts tasting bready at the uneven extractions of my home grinder (a well used Encore) and Chemex make. Kinda boring and baked. These are standard 60g/L brews, 1.3% TDS, ~19% yield. I take home the VST, but not the EK.

My roasts are all sub 10min rarely dropping below 10F/min, so not classic baked territory. I spend about 4min between 320-392, which might be overdeveloping the sugars. I don't know. I'm tightening up there to see if more sweetness shows up in less than EK quality extractions.

Targeting 15% extraction at home on the Baratza creates syrupy nectar with great fruit acidity and an amazing depth of sweetness; more like the 22% yield brews at work with the EK. I've long suspected light but well developed roasts over-extract in a more dulling way than adding astringent bitters. Maybe the over-extracted fines from the worn grinder are blurring over the under-extracted boulders. Maybe.

Separately, I'm having a coffee and a grinder that together aren't getting above 17% regardless of grind (Fetco Brew) That coffee (light roast) and other grinders are doing fine. That grinder and other coffees (dark and decaf) are doing fine. This stumps me. The brew is ultimately flat and boring. The dark roast of the very same coffee is hitting 20% and tastes full and lively, and the only difference is an extra 90sec getting to the doormat of 2nd crack. Again, these burrs may also need to be replaced; they've ground about 3,000lbs of coffee.

These experiences echo Rao's "little hump", as posted by James Hoffman. http://www.jimseven.com/2010/11/08/the-double-hump/ Rao says somewhere in his latest (great) book that if you can't get into the 19-20% box, targeting 15% can be a useful strategy.

I don't really know whats going on other than I haven't been enjoying my coffee at home as much as I used to, and exploring this area of under-extraction has been new and actually quite nice. It's making me wonder if 80g/L is a worthwhile brew ratio, sometimes.
-Marshall Hance
Asheville, NC

MWJB
Posts: 429
Joined: 11 years ago

#5: Post by MWJB »

endlesscycles wrote:I've been running into something recently.

Strength is somewhat personal or regional, but extraction is supposed to be more universally agreed on. 20% is supposed to be "right", right?.
Whilst it often may be "right" & a good place to start, is there really absolute consensus on that? Do people assume that 20% is the "cross hair" just because it sits halfway between 18% & 22%? Did the MRI/CBC ever identify a precise target point, if so, why bother to include the less than ideal, "ideal %"?

A point between 18.5-20.5% is often mooted to be the target/sweetspot (varying between 19%, 19.5%, 20% depending on various sources...but being realistic, we can usually expect to get within a small range rather than to the decimal place), but method, unevenness of extraction & grind quality can probably cause it to drift wider than that, or at least (as you note) make it logical to stop short, making the most of what you can get, rather than banging your head against a brick wall. But then there is the possibility that just because that's as far as you can get in a specific scenario, it doesn't necessarily mean that a higher yield won't show improvement? (Theoretically at least).

Looking back at my percolation brews, best cups are 18% to over 21%. Immersion yields 20% to 26% depending on brewer (but this is a slightly different calculation, those same brews translated directly to "extraction yield" in the CBC/SCAA sense would read lower, so it's not as telling as Imm. yield & probably not wise to make direct comparisons between the two, but they still fall 18-22% as "ext. yield"...with the odd, juicy, fruity 17%).

Both your instances do make references to "well used burrs".

Just a thought here, not trying to be contentious, if the ideal box was defined by sensory testing by a representative sample of the population (non coffee geeks, nor folks used to doing a lot of cupping), would "great fruit acidity" be high on their list of positive attributes? Not saying that it isn't an attribute, just may there be a bias amongst super tasters (dislike bitter) & folk who typically look for & prize acidity in coffee, compared to the man in the street who expects "coffee/roast flavours", after all there are plenty on this forum who are cringing right now just because we said, "great fruit acidity". OK, OK...I'll stop saying it now.... :wink: