In programming, I think of myself as a slow thinker, but quite fast programmer. I very often find myself writing some code, re-reading it and throwing it away, because I realized it doesn't really work the way it should work. I do the same things with emails. Whenever I write a long email, the first version is most likely going to be discarded, but it still helps me organize my thoughts.
When solving a problem, always make sure you understand the root cause of it. Don't be satisfied with fixing the symptoms. If things don't work as you expect, make sure you completely understand why it is so.
Eventually, it becomes a habit to think twice before committing to something. You might still experiment, draw things on a whiteboard, do whatever helps you think, but you should not consider the side-products of that process to be the result.