Roast degree level based on color and lost weight

I am trying to determine my roast level based on several factors, but I get totally different numbers.

From the SweetMarias web: https://library.sweetmarias.com/sweet-marias-roasted-coffee-color-card/ I got this chart:

  • My latest roast weight loss was about 17.4%. In reality it was a bit lower because I lost some beans when dropping, but according to the above table I should be in the “burnt” zone.
  • Based on that same table and my IBTS drop temperature of 208C, and my BT at the drop was 185C (assume it’s lower due to 350g batch only) the roast is supposed to be “before FC”.
  • Visually it looks like Full City roast.
  • If I was to adjust (offset) the table based on my FC (which was at 201.5 C, 394.7 F), diff 17.3F, I’d be around the City roast level.

Any idea what would be the “best bet”?

I’ve been thinking about something similar recently.

Just to add to the confusion, I have a bean that is also roasted to 208C with a 15.5% weight loss. Using my eyeballs it looks like it should be a Full City or Full City+ roast and it should be well developed. What gives?

But I also have a Tonino color meter, and ground very fine it measures in the light to very light colour range. This is not what I expected based on the end temp and weight loss. When I grind it to about the coarseness of a v60, it reads as a medium roast. The Tonino manual says that only fine grinding will give accurate results so I am not sure that the coarser grind is even all that correct to begin with. But I have noticed that my espresso grinds look lighter than my pourover grinds and I don’t know what to make of that. Maybe that’s for another topic though.

I tried roasting another batch of the same bean to the same temp, but with a 1m shorter total time, and it was slightly lighter, as expected.

Next I am going to try roasting the same bean much more than I usually do. I want to try a really extended development time (like 40%) but keep the same end temp. I also want to try roasting really dark to see how the measurements turn out.

I have a feeling that having a kind of objective roast level measure for everyone is impossible given all the different variables and that I should try to come up with my own levels. That approach might work well for you too.