This is best explanation I've seen as to why many of us can't taste the advertised flavors for coffees from some roasters.

Sample roasts tend to be fast and short, and the immersion brewing method used for cupping is designed to maximize extraction from what may be an under-developed roast. The idea is to be able to taste everything the green coffee has to offer -- and any flaws it may have -- without introducting any roast flavors. That sort of roast probably won't work well for typical brewing or espresso extraction methods. It's quite an art to translate a sample roast into a production roast that brings out all the flavors detected in cupping.

What baffles me is why some roasters put tasting notes from the cupping on the bag instead of tasting notes from the production roast.

A friend of mine has an ikawa sample roaster and uses a 1kg perforated drum roaster for his production roaster. (very small company) Those two Roasters will never produce the same results and his sample roaster is way better than his production roaster. So I'm sure he gets wonderful results for his ikawa but that doesn't get to the customer.

Sample roast cupping is to determine any presence of green defects. If you look at a cupping score sheet from the SCA it's about finding green defects and is more pointed in that direction rather than the customers direction in terms of roasted coffee. Roasters are probably cupping all of yesterday's coffee the next day for things like quality control but if it has some roast defect or some other characteristic comes out they aren't going to change the label on the bag just because one batch didn't line up. what makes it in print is going to be a desired roast profile or what they got from a sample roaster over the actual roast which may not line up at all.

I've never thought about asking the roaster their cupping process for what makes it to the label but it's probably worth asking before you buy. Also it's probably worth asking if they do any tasting from other brew methods like espresso or pour over rather than just the immersion method for cupping. A roaster than does that is cupping for the customer.

Green defects are just that, green defects.

Roasting defects are entirely different are they not?

green defects are different than roasting defects but aside from visual characteristics cupping is the method to determine the defects for both. to score coffee you don't discriminate and just grade what you taste even if you can pull out what's a roasting defect or not but with that said there are guidelines for roasting when cupping that may disqualify a roast but that's more for determining its market value before it hits the roaster.

At the roaster level, they will cup coffee using sample roasters to evaluate the coffee themselves before production roasting and to check quality. After production roasts, they will cup the coffee again for internal QC to determine things like their own roast quality and check for consistency with their original cupping notes. This is all for internal checks of course and what actually gets printed to the label has a different motivation which may not be as transparent as you would like.