the deep end
Here’s an analogy I’ve been contemplating for explaining the unknown unknowns concept. In particular, the assumption that one’s own experience is complete. They’re like a kid who thinks he’s a deep sea diver because his ears popped when he swam to the bottom of the deep end. The analogy is particularly suited to How bad can it be? types of conversations.
Now I may not know everything about the ocean or how deep it is, but I know it exists and that it contains things you won’t find in a pool. The frustration dealing with the kid is that his experience is so shaped by the pool environment, he can’t imagine the ocean.
First, there’s a failure to appreciate the range of problems encountered. The ocean is really deep (bad, difficult, ...). You can try to extrapolate from the pool (ok, my ears will hurt a little more), but the reality is that if you dive deep enough in the ocean you will experience problems like the bends that cannot be predicted. The problems don’t just get more difficult, they get different.
Second and relatedly, there are the unworkable solutions. Problem: can’t breathe/work/play tennis in the pool. Solution: drain the pool. OK, that works. But you can’t drain the ocean. Revised solution: use a bigger bucket. No, sorry, that still won’t work. At a certain point, removing the problem is no longer a viable solution and we need a solution that looks more like a workaround. This is almost the opposite of the five whys approach. Instead of continuing to dig deeper, just accept the circumstances and work with what you’ve got.
Imagine you have only experienced diving in a pool when you see someone getting into a giant copper diving helmet. Such a ridiculously limiting piece of equipment. Why bother with all that when you can just dive down, hold your breath, and pop back up whenever you need air?
The unspecified conversations which make me think of this analogy are generally related to the field of software development, where it’s very easy (and typical) for random commenters to critique a situation given only the meagerest of a problem description, but I’d say it crops up everywhere problems and solutions are complex enough that context free tidbits can leak out and make the whole operation look ridiculous.