I minimized risk by having three advisors, whom I ended up relying on for different purposes. This was unusual in my school, but I found it worked well for me. One had a lot of experience and was great at dealing with paperwork and people problems. One turned out to be absolutely brilliant at idea generation and horrible at everything else. One was great for brainstorming and working on proofs. Neither of the latter two were what you might call "transactional" when it came to meetings and goals.

At the end of the process, it was my fellow students who who actually read my thesis and provided edits and suggestions. All for chocolate and cappuccino.