I think there should be occasional 'freak' variability in progression/regression of player potential. Which is to say, in the current system, I can more or less gauge how a player will progress throughout their career based solely off their very first training camp. If a player has a massive gain, he will continue to gain for the next several seasons. If he has a massive loss, he will continue to go down. If he remains static, he will remain more or less static with only minor loss/gain thereafter. There should instead be a handful of rare 'late bloomers' (players who started off negative or static and have later dramatic swings) and 'early flatliners' (players who started off very positively but then quickly hit their ceiling). This isn't terrible frequent in the NFL, but it definitely does happen and should be reflected here.