Coffee Data Science
Basket Size by Dose for Espresso
Data Breakdown
Recently, Lance Hedrick was looking at the differences between 49 mm baskets and 58 mm baskets. He was curious if you could dial in a smaller basket to achieve the same result as a larger basket. He looked at taste and Extraction Yield (EY). I explored the EY data to see what I could see. He originally used box plots, but for small data sets, I prefer scatter plots.
Total Dissolved Solids (TDS) is measured using a refractometer, and this number combined with the output weight of the shot and the input weight of the coffee is used to determine the percentage of coffee extracted into the cup, called Extraction Yield (EY).
Design of Experiment
Lance collected 48 shots in total across these variables:
- Two Machines:
- Decent (49 mm Sworks basket modification)
- La Micra Marzocco (58 mm basket)
He did many tests before the video to show that the two machines functioned essentially the same for the profiles.
2. Two Coffees:
- Ethiopia Yirgecheffe. Chelchele. Washed Heirloom
- Colombia El Jordan. Tolima. Washed Caturra Colombia Castillo
3. Two Doses adjusted to be equivalent volume in each basket.
- 17.5g and 28.1g on the 58 mm
- 12.5g and 20g on the 49 mm
4. Then 6 shots for these 8 total variants.
Data Analysis
I first looked at the data unsorted, but it was a mess. It doesn’t show any pattern.
For these types of data sets, I prefer scatter plots over box plots, and I also prefer sorted paired datasets. So I compared the best to the best and the worst to the worst for each variable.
In the sorted case, the difference between the 49 mm and 58 mm basket is not statistically significant (p<0.05) for a two tailed paired t-test.
What about breaking this graph by coffee or dose? No trend was found.
What about dose? Does dosing less improve EY? Statistically, dosing less decreases EY.
We can split this by coffee and machine. For coffee, the Colombian coffee seems more affected by dose. For machine, the data is very intermixed.
This data showed basket width does not matter so much, with all else being equal. The data also shows a larger dose leads to a higher EY, which is another way to say that less headspace (from a higher dose) increases EY.
If you like, follow me on Twitter, YouTube, and Instagram where I post videos of espresso shots on different machines and espresso related stuff. You can also find me on LinkedIn. You can also follow me on Medium and Subscribe.
Further readings of mine:
My Second Book: Advanced Espresso
My First Book: Engineering Better Espresso
