Last November, much ado was made, I mean a heck-of-a-lot-of-ado was made, about Matt Kemp’s WAR. So much ado that Jonah Keri was still making much ado about it last week. According to the old bWAR, Kemp had posted the best single-season from a hitter since Barry Bonds “preposterous ’04 campaign.” Kemp’s 10.0 bWAR was so abundantly better than Ryan Braun’s 7.7, Sport’s Illustrated’s Cliff Corcoran could go so far as to claim “MVP voting continues to defy logic or reason” because the BBWAA had the audacity to choose Braun over Kemp (by the slimmest of margins). (To be fair, Corcoran was hardly alone in his outrage, he merely mainstreamed a widely-held position.)
While it’s hardly true that everybody made their MVP arguments based upon WAR alone, nor was Kemp by any means an illogical choice, this type of reductive brow-beating has been an unfortunate side effect of WAR’s popularity. The inconclusiveness of arguments based solely on single-season discrepancies in WAR is reinforced by the Baseball-Reference update. As a result of the reformulation, Matt Kemp’s 2011 bWAR dropped to 7.5, only 0.1 better than Ryan Braun’s, a margin which is, to put it lightly, inconsequential, as bWAR’s margin of error is, at minimum, about 10%.
In general, the 2011 awards debates were embroiled in discussions of WAR to a previously unprecedented degree.* It’s interesting to look back and see what’s different about the leaderboards in the wake of Baseball-Reference’s update.** I don’t have enough time to analyze in great detail the table that follows, but a couple things jumped out at me as ways into thinking about the changes in B-R’s formula. For starters, Clayton Kershaw and Roy Halladay, who were neck-and-neck according to the old bWAR (0.2/2.7% advantage to Kershaw) and fWAR (1.4/17.1% advantage to Halladay), are now significantly separated (3.8/37.6% advantage to Doc). The dominance of Doc and Cliff Lee under the new formula likely stems from the new park factor adjustments (for road starts/interleague play) and the adjustment of run to wins calculations based upon “the length of outings.” Anyway, comment if their any other aspects of this new date which you find interesting and make sure to check that Sean Foreman explanation.
*Probably the most irresponsible article from the ’11 MVP debates was one written by Sean Hartnett of CBS who supported the headline “WAR Statistic Proves Verlander’s Worthiness” by citing the 0.1 difference in fWAR between Verlander and Jose Bautista. Ugh.
**I had to piece together the “Old bWAR” leaderboard from my own notes, as well as ballots, blogs, etc. posted in October and November of last season, so, as such, I cannot absolutely promise that they are accurate, although, in most cases I verified the totals in multiple places.