In my head this breaks down like:
Split: generate alternative hypotheses. “I believe X. What not-X might I come to believe, and why?”
Commit: flesh out those hypotheses with contingent commitments for action. This is how you know you actually generated a real hypothesis rather than just “maybe the house elves did it, wow, this rationalist stuff is easy.”
Practice. It might be a great rationality workout to take very low-probability hypotheses seriously and build an action plan. “If I come to believe house elves did it, I will submit myself for psychological evaluation. If I pass that evaluation, I will dedicate my life to documenting their existence.”
Your last point about flinching from hypotheses because they’re undesirable is key. That flinch is the enemy of equanimity, the acid that dissolves you to make room for Moloch.