I learn the hardest things first, not the easiest thing (ie I flip right to the Black-Scholes formula when learning options, gradient descent in neural networks, etc etc).
I find that by going after the hardest things first, you learn the most information in the quickest way as all supporting things must be researched on the way.
If you only go after easy basic things and try to work your way up, the pressure (and rate of growth) will never compare.