What's the typical advice here? Something about writing down 100 ideas, diving into detail on your favorite 20, actually build a working prototype for 10 of th

What's the typical advice here? Something about writing down 100 ideas, diving into detail on your favorite 20, actually build a working prototype for 10 of them, and 2 or 3 will stand out, so pick one of those? I read that somewhere, possibly here. Maybe someone can link the source, if they have it.