Sure, the big difference is that at a University they go through a bunch of things that come up and they talk about how to avoid them and why. So everyone learns how to do a bubble sort and then you learn why that is so slow, both from an analysis point of view and an algorithmic point of view, and then you move on to quicksort or insertion sorts.

Whereas someone just picking it up as they go along might independently come up with a bubble sort, it works, and then just keep using and reusing it because hey, it works doesn't it? And then when its the bottle neck and they need something better they don't even think to look at the 'working' code they look elsewhere. It is a sort of unintentional blindness.