Here is a strategy that has worked well for me.
Take a result you want to understand and try to prove it on your own.
If you succeed, great! If not, peek at the beginning of the proof and close the proof the moment you reach an insight you weren't able to think up on your own. Try to prove this insight, and try to complete the proof with this insight at hand.
If you succeed, great! If not, repeat.
In my experience, trying and failing to prove a result gives you a good appreciation of the proofs that do work. Constructions which might otherwise seem artificial often make sense as overcoming some barrier or other that you stumbled upon.