Reframe problem in manuscript (i.e. goal is to detect trend and make good decisions, under what scenarios do you have good enough uncertainty to do that reliably)

[aebratt]