My theory is something like this - teams completely discount the phenomenon of players occasionally peaking in their 30s, and that leads to a much steeper aging curve.
If I'm signing a guy long-term, does it really make sense to assume he'll age like Justin Turner, or Jed Lowrie, or Daniel Murphy, and randomly become a much better hitter in their 30s than they were in their 20s? So why even include that type of player when you're modeling a player's aging curve and deciding how much to offer in free agency?
Looking at average WAR by age by definition inflates your averages, because you end up excluding players who flamed out early, and including players who improved later in their careers. It doesn't really give you a valid expectation that you need to take a random 30-year old and project what he's going to do going forward. Doesn't it make more sense to look at guys who were already good and look at how they aged?
I pulled up 26 guys who put up 3 or more WAR as 30 year olds from 2010 to 2013, and then looked at their next 5 years. I excluded guys who "don't really count" because they were one-year wonders and it happened to be in their age-30 season (Ryan Roberts?!?!). That leaves the following cohort - Pujols, Holliday, Teixeira, Uggla, Dunn, Bautista, Granderson, Zobrist, Victorino, Phillips, Hamilton, Swisher, Hill, Pagan, Ethier, Gonzalez, Cabrera, Choo, Cano, Molina, Mauer, Pence, Martin, Encarnacion, Hardy, Escobar.
I think that's a pretty good group. Well here's their combined aging curve.
Age: Average WAR
30: 5.0
31: 3.1
32: 2.6
33: 1.9
34: 1.5
35: 0.9
So the -0.5 WAR per season aging curve looks pretty good...if you assume they've already fallen off a cliff in the first place.
But look at those numbers and you get a pretty good idea of why teams aren't jumping to sign Harper and Machado to 10-year deals. If guys drop that quickly to just "above average regular" level, then the upside of their early years really isn't worth the premium of paying them on the back end.
And the downside isn't just for 30-year olds. Here is the aging curve for the 10 best 25-year olds in 2012 and 2013 (Goldy, Upton, Belt, Seager, Posey, McCutchen, Jackson, Reddick, Kipnis, Jennings).
Age: Average WAR
25: 4.8
26: 4.0
27: 3.8
28: 3.7
29: 2.7
30: 2.6
So to convince yourself to sign Harper or Machado to a 10-year deal, you have to first convince yourself they won't age like an average 25-year old, and THEN convince yourself they won't age like an average 30-year old.
Poll
