Kind of an abstract question, but how can you doublecheck whether your mechanics are too obscure or your tutorials too vague without/before betatesting?
I don't mean it in the FOLLOW THE YELLOW PAINT YOU SHEEP school of problemsolving, but finding that unknown unknown. There are ideas and assumptions that (you) always hold in your head, but the players most likely don't. While I got THE FEEL that I wanted, I don't want to simply hope that players will get what I'm going for.