Not long ago, I got a set of cores to practice on. And within an hour I was taking down the 6-pin pretty quickly. Of course this meant I got all confident and decided to randomly re-pin the lock to give myself a little more challenge... 3 or 4 hours worth of challenge, as it turned out.
Humbled by this, I did what seemed to be the only reasonable thing: make a spreadsheet with all the pin permutations I can create with the locks I have, and see if I can find a pattern. After close to a hundred different permutations, I looked over my notes and realized the obvious fact that guarding short pins between tall pins makes picking harder. For now, it's satisfying to chew through a 4-pin that I previously marked as "wicked hard" ... as a light warmup.
So it goes back to the adage that practice makes perfect. Tracking it in a spreadsheet helps measure progress.
My excel macro-fu is weak, so I can't generate the list of permutations inside the spreadsheet yet, otherwise I'd post my scoring spreadsheet. Maybe one day I'll come up with a way to compute an approximate difficulty score of a given arrangement of pins. For those wanting to make something similar, the FACT() and COUNTA() functions will be your friends. It could be argued that factorial is the wrong function in the presence of 2 pins of the same size - if you can tell your #5 pins (for example) apart, it's probably worth swapping them around to see if they make any noticeable difference.