<p>With proofs, you generally need to build up a body of techniques and have some amount of practice with it. A 1-year sequence in Geometry provides a good background for this though I think that a lot of the proof part of geometry has been watered down over the past 30 years.</p>