This whole discussion of BJ Upton, and more particularly what goes into free agent contract values is one of the more fascinating subjects (and most interesting applications of WAR) out there.
There is an interesting theory in the sabermetric community that you can basically settle in on an average market value of $$ per win that clubs will pay (last offseason, it was between $4M and $5M, closer to $5M) and that you can predict farily accurately what clubs will pay by using a simple model that projects age-related decline of about 0.5 WAR per year, and about 5% inflation on the $$/win rate each year. So, for instance,
in this interesting article from Fangraphs about a year ago (which dealt with whether each additional win provided by a premium player increases the $$/win rate he should get exponentially, or simply linearly), you see that this simple model (using a $5M/win starting rate)
exactly predicted CC Sabathia's contract a year ago.
As I alluded to in my earlier post, with the increased revenues from mega TV deals and Regional Sports Networks beginning to hit MLB, most expect the rate of $$/win paid this year to free agents to be closer to $5.5M/win this year (closer to a 10% increase over last year). If this is the case, then Upton's contract looks
almost exactly like market value, as JH suggests. If you assume Upton is a 3.5 WAR player next year (he was at 3.3 WAR last year, and has average 3.8 WAR over his last five seasons), that model predicts a total value of $74.5M over the next five years.
However, there are a lot of assumptions here. One is that $5.5M/win will be accurate, when this Hot Stove League season ends. We're not there yet; though the early indicators are that this value should be pretty accurate, we won't really know until ALL the Fat Ladies sing.
Another, as fenwaysouth has alreadysaid, is just how murky this "projection" business can be. As the article linked to explains, tweak any number of assumptions just a little, and all the suddent Upton looks pretty overpaid or underpaid, depending on how you think he'll do going forward.
Over the five seasons from 2008-2012, BJ Upton has averaged a little better than 148 games per season, and 3.8 WAR. However, his actual WAR has ranged from a low of 2.4 (in 144 games in 2009) to 5.0 (over 145 games in 2008). As Three Bagger notes, there are some red flags with respect to Upton's approach and some reason to believe his regression is a sign of things to come, but he (and JH) also note that his body type ages well and some of his five-tools are less appreciated than others (at least outside of those who actually have to open their pocketbooks to pay guys like BJ). Where will he come in next year? As I said, if he can repeat the year he had last year, he is getting paid pretty close to market value. If he slips to a baseline of, say, 2.8 WAR? The model suggests he's being overpaid by about $4M a year. If the true baseline for a player just hitting his peak (28) is closer to the average of his last five years (3.8 WAR), then it looks like the Braves paid about $10M less than they should have.
My own view is similar to Three Bagger's: I have some concerns about Upton's approach, and I think going forward I'd rather pay assuming he is a player closer to the lower end of the range he's displayed over the last five years than the upper end. I'd probably be comfortable assuming a baseline of 3.0 WAR rather than 3.5, BUT I also think Upton will age a bit more gracefully than the average player over the next five years (I'd be comfortable assuming a 0.3 WAR per year decline rather than 0.5). Based on those assumptions, IF the going rate is $5.5M/win, then Upton is being
slightly overpaid - that would project a contract value of $72M over five years, rather than the $75M he got.