sully, I do think there is a formula, but it takes into account some %'s leading to probablility, based on the psych eval (i.e. it is not a sure thing that a guy will leave if he is self centered, just raises the odds.
I also think the formula is by position, everyonce in a while you will see a pretty bad (in terms of guard ratings) PG leave, while dozens of better SG's stay. this is also why sometimes one position on one team gets raped, if you recall, I lost 3 junior PF's from my UConn team a few years back, two of whom averaged less than 5 pts and shot less than 40% from the floor.
I once retained a 1050 SF junior, from a team that won a PIT, in the tourny, I was not even thinking he would stay, so I gave him 100% distro in several of the games, just to see what he would do - then he stayed, so I would not think that advancing far in the PIT has an extreme effect on going early.
I do think making the final 4 NT does, but so does high ratings vs others at one's position.
we all used to not play guys to try and keep them, it almost seems like that has backfired as much as worked as of late - not sure what others think of that comment, or any of the other ones for that matter -
all highly speculative at this point, although a coach did some analysis work on this at one point, can't remember who he was, he might read this and chime in