Hey guys! This blog post is specifically applicable to my settlement game concept, but it might be useful for other designers.
The important paragraph is the following:
The randomness of cards is extremely limited. Thus, they should not be used to represent random success odds against a consistent target, but consistent success rates against an unknown target. An adventurer in DND always has a 2-in-6 chance to break open a door. However, a Club-n card can beat any non-Club card at combat, or any Club card less than n. This is a completely different paradigm to approaching random scenarios and player agency.
I should clarify that cards are not limited in the random situations they generate, but in application. Their strength is their randomness of opposition so to speak, but they cannot be used to compute random odds of success like dice can.