I don't know if anybody knows the answer to this question, but here it goes.  When you are in a league where you are allowed to draft your AAA players, is there a greater likelihood that those players will not perform as well as they usually do if you were had drafted them as a veteran player?  Or is the probability that they are a dud about the same as if you had drafted them as a regular pick on your 25 man roster?  For instance, I have drafted Jeff Reardon in the past for my 25 man roster and he performed pretty well.  However when I chose him for a AAA player, he was pretty bad.  It would make sense to me that a AAA player is more apt to turn out to be a dud. 

No.  Whether a player is drafted in the "regular draft" or as a AAA, makes no difference.  They are the same player with the same underlying stats.  The only difference between AAA versions of players are the "real player" is when you have random AAA.  Random AAA are not normalized to their real year-- they are either normalized to the last year (e.g. 2010) or not normalized at all.

Typically you will see more variation in AAA players, but that is simply a result of smaller sample sizes due to (typically) fewer PAs.

Well I know for a fact that 81 Lonnie Smith does pretty well when drafted on a 25 man roster, but is horrible when selected as a AAA.

That, sir, is a silly notion.

Some people will never accept the fact that this game is run entirely based on statistics and that differences they see from one season to another have everything to do with random variation and league makeup and nothing to do with the sim manipulating their players.

I understand the game is run on statistics, I just didn't know if there was something built into the software when a player is drafted as AAA if they "lose" some of their statistics because they are considered a rookie coming up from the minor leagues.  Thanks Zubinsom for answering my question. 

Sorry dahsdebater. Although I truly believe that game is supposed to based on "randomness", and more than likely, it probabIy is.  However, I just haven't been turned yet to the "randmoness" theory, especially when my team wins 9 in a row and then promptly lose 8 in a row. I wonder exactly how many times that's ever happened in real baseball in the major leagues.

regular AAA are normalized to the current year.  U-pick AAA are real players and may be normalized to their own year. if that's true then there will be some statistical difference.