“Dogma is a principle or set of principles laid down by an authority as incontrovertibly true. It serves as part of the primary basis of an ideology or belief system, and it cannot be changed or discarded without affecting the very system’s paradigm, or the ideology itself.”
I am often asked what recipes to use for espresso, pour over and cafetière. My answers a described through numbers. I roast to profiles. These profiles are refined to those suitable for ‘filter’ and ‘espresso’. I use Cropster, a VST refractometer, scales, water with a certain chemistry and a stop watch. I log all my roasts. I track roastery conditions before, during and after each batch. I log quite a lot of my in-house brews – what was the TDS, time, grind size, brew method. I produce and store a whole bunch of data.
What am I getting at? There is an awful lot of discussion about brewing technique on the internet. A lot of these techniques have begun to introduce numbers. These numbers have been derived from ‘gadgets’ that can be purchased and used by anyone who has the required budget. The result of which, is that I find a number of people in my peer group brewing to numbers. The same can be said of roasting for that matter. I too fall foul of this when approaching a new coffee. I lean towards the numbers to give me the result that I’m after and tend to trust that those numbers, or boundaries, will yield something tasty. Invariably this is correct. The numbers will get us in to the ball park in which we’re seeking. It will guide us towards producing something tasty.
So what’s the point? Taste. You must taste the result. Brew or roast by numbers, but taste. Taste the output. 20 – 25% development time may be a prescribed recipe, but 15.5% may taste better. 20% extraction yield may be seen as the median and therefore the most balanced brew by number. However, 18.2% may taste better.
When experimenting with profiling of any description. First taste, then measure the results. Brew/roast forward based on which taste better, irrespective of the number.