ORIGINAL: Godzilla Blitz2
I like the idea that the peaks are unknown. Makes it more challenging.
This becomes the issue.
You lose some of the 'coolness' of the aging model. If you can't figure out that a player is going to peak at 23 (I played with the XML to have 2% of players peak at that age), then you'll never be able to develop them correctly, because you won't put them at the right minor league levels and they won't ever reach their ratings peak.
So the aging model allows for some cool career curves that would have a player peak at 23, but not to start to slow decline until 37 (in my league it should happen once every 4,000 players or so). However because of the way the players develop, you won't ever actually see it happen because you won't have the player who peaks at age 23 in the minors by age 21 for them to reach their ratings potential - because you always are going to play the percentages and have a 21 year old player in A or AA - because if his peak is 27, that is where he needs to be at age 21.
This leads me to the conclusion that not being able to discern a players peak beforehand actually takes away one of the best things about the peak/decline/cliff model, most obviously the player who peaks at a young age will rarely develop correctly.

