When you apply tension, only one of the pins will be binding against the inside of its chamber because of tolerance defects. The pin chambers are different diameters, they're not perfectly circular, and they're misaligned. The pins are all different as well, not identical in size or shape like you might think. The differences can't be seen with the naked eye unless the lock is very poorly made, but these defects are all present in even the highest quality locks.
To simplify things, imagine you had a cylinder where all the pin chambers were exactly the same, but each of the pins were different in size. The biggest pin would bind first, it's impossible for them all to bind at the same time or for any of the others to bind before the biggest pin.
The binding order is completely random because there isn't supposed to be one. In a perfectly engineered lock, everything would bind similtaeneously and thus need to be set simultaeneously in order for the plug to turn. But it can't be done, and that's why picking is possible in the first place. Very old and well worn locks will sometimes pick straight from back to front because of the way things wear down but this is still pretty uncommon. Anyway, if the binding order was predictable then it wouldn't be much of a puzzle would it?

Raking exploits the same mechanical weaknesses as single pin picking, binding order still applies... you're just not paying attention to it, because you're attacking multiple pins at the same time and in a random fashion.