>>14228268(2/x)
By learning the fundamentals of your field with the historical context in mind, all the proofs will be motivated not by "le magic of rationalism," but by common, fully graspable human imperfection. While you see University or Textbook (not all textbooks are guilty) folk building their space tower, you will remain grounded as they ascend. But someday, when your fundamentals are strong, you will have a launchpad which allows you to travel anywhere in the metaphorical solar system with relative ease.
There are still concrete strategies. Understanding the structure of per-requisites is useful for guiding yourself. Having strong multi-variable calculus chops seems to open the most doors. Set Theory requires no pre-reqs and can construct all higher level math, though getting from Set Theory to Differential Equations is a long journey, and there is plenty of room to stretch your mind in any small sector of math.
You could try and plan a roadmap ahead of time, but you could also simply start with something that seems interesting and keep answering questions you have, being brutally honest with things you do not understand, not ever tolerating hand waives ("that's just the way it is"), and digging deeper and deeper answering the basic questions. I recently did this with Elliptic Curve Cryptography and ended up learning and grasping fundamental theorems of Modern Algebra in such a way that I now have a string of proofs I can construct that lead from an intuitive understanding of groups all the way up to a working toy-implementation of ECC.
I recommend this approach, though it is time consuming and sometimes frustrating how few resources respect a man's need to understand rather than being told it is so. You would think a field like math would be all about that but its not always such a popular way to teach math these days.
Don't grind for arbitrary marks, scores or grades. In math, understanding trumps knowing, as most truths were constructed, not given.
