Coffee Data Science
Since returning to the office, I have had some time to compare the Kim Express and Decent Espresso machines Again. The caveat is that Kim Express works. So I’ve been slowly collecting data over the last few months and have some results ready to share. I’m content to say I’ve been able to close the gap between Decent and Kim, but there’s still room.
First, I made a lot of modifications to my shot profile for Decent, especially the Pump & Dump profile. This profile had the biggest impact, while at the same time I was working on better water supply.
Initially I used paper star on top puck, and I’ve since changed the shower screen to limit the water to the middle 25% of the puck. I’ve applied those changes to Kim as well, so my shots have gotten better across the board.
I didn’t feel ready to do a real comparison of the pairs, since I’d need a grinder at work or a machine at home (not that I don’t have another Kim). This creates three problems:
- There is a delay (30 minutes to 3 hours) between grinding and shooting for Kim, and I have done some experiments to understand how the age of grinding affects the flavor of the shot. There is a difference, but it also allows for higher extraction.
- Kim Express needs a finer grind. With the shower head change I dropped the water temperature from 116°C to 105°C, but the flow is so rapid partly because it is so even.
- Two Kim Express machines were used. I had a head gasket failure so I swapped machines when I replaced it.
We lose a bit of fairness in comparison, but my goal is to take the best photo for each machine.
Coffee Maker: A Decent Coffee Maker and Kim Express
Coffee grinder: Zero niche and Year
Coffee: Home Roasted Coffeemedium (First crack + 1 minute)
Shot Preparation: Staccato Tamped
Preinfusion: long, ~25 seconds
Filter basket: 20g Wafo Soe Spirit
Other Equipment: Acaia Pyxis Scale, TDS DiFluid R2 Meter
To evaluate the differences between techniques I used two sets of metrics: Final Score and Coffee Extraction.
Final result is the average of a scorecard of 7 metrics (pointed, wealthy, syrupy, sweet, sour, bitter, and aftertaste). These ratings were of course subjective, but they were calibrated to my tastes and helped me refine my shots. There is some variation in the results. My goal was to be consistent across each metric, but sometimes the granularity was hard.
Total Dissolved Solids (TDS) is measured using a refractometer and this number, combined with the output weight and input weight of the coffee, is used to determine the percentage of coffee in the cup, called Extraction Efficiency (EY).
Intensity Radius (IR) is defined as the radius from the origin on the control chart for TDS vs EY, so IR = sqrt( TDS² + EY²). This metric helps normalize shot efficiency across the entire output efficiency or brewing ratio.
I had 131 shots on Decent and 35 on Kim in 16 roasts. To put it in perspective, we can look at the control chart and see that all the shots fall into a similar bubble.
We can add flavor there, but there is no separation between the two machines.
We can look at each roast and find average TDS, EY, IR, and Taste. They look similar in terms of performance.
