Sunday, November 14, 2010
Last week I wrote that the Red Sox should not trade for Adrian Gonzalez, but the more debatable point was that they should sign Prince Fielder next winter instead of the Padres‘ first baseman. The basis of my argument was that the Milwaukee’s slugger is younger with better comparable hitters and would not drop off as fast as Gonzalez with the bat.
The data and my calculations back up my assertion. In 2012, Fielder is projected to put up a 139 OPS+ at age twenty eight and then 137, 134, 124, 113, 116. By comparison, Gonzalez’s forecast is 142 OPS+ as a thirty year-old, and then 132, 116, 128. They look like similar hitters in 2012-2015, but many of Gonzalez’s comparables - and for that matter Fielder’s - were essentially out of baseball after the age of thirty three. Even after applying the Brewer first baseman’s comps to account for Petco, A_Gone has a projection that is only slightly better and still has the beyond thirty three issue.
Since both are likely to get more than a four year contract, the data suggests that Gonzalez is a riskier acquisition than Fielder. The Brewer first baseman’s bad body and lesser defensive reputation than A_Gone’s would seem to put on even ground a year away from free agency. I'll still take the younger guy looking for a break out season in the first two years of the contract.
For the numbers geeks, the projection calculation:
1) The forecast is calculated by first finding the player’s baseline production. It is simply a 60/30/10 model.
(YrOPS+ * 60%) + (Yr-1OPS+ * 30%) + (Yr-2OPS+ * 10%)
Then doing the same for each comparable hitter at the same age of the player to be projected.
Yr = most recent year
2) For each comp, the rate of change in OPS+ from their baseline is calculated through the end of their career or the age to be forecast. Each year’s rate of change is weighted based on their comp score to calculate a weighted average rate of change for each age to be projected.
(OPS+ - BL) / BL * CS
3) The weighted average rate of change for each age is applied to the player’s baseline production to find his forecast.
WavgRC * BLP