A common thing that happens to me is: Stuck on a moderately complex problem, usually in a field that I'm not intimately familiar with (ie. A complex mail server

A common thing that happens to me is: Stuck on a moderately complex problem, usually in a field that I'm not intimately familiar with (ie. A complex mail server config). Work on it, feeling frustrated late into evening. Go to sleep and sort of "dream about the problem" in a loosely chaotic way. Come morning, its not that I suddenly have a solution, but more often than not I've developed a good strategy for how to test the problem. By using this new method I'm able then to trial-and-error the problem in a systemic way, leading to a solution. TLDR: Sleeping often gives me successful 'approaches' to problems