Finally, awards season proper, perhaps inaugurated by last night’s Emmy Awards (more credible than the Grammys, less credible than the Gold Gloves), has begun. Perhaps the annual baseball awards would be more exciting if it somehow incorporated hours of inane red carpet banter, but instead we have to settle for seemingly endless arguments.
If you are reading this, you probably are at least somewhat aware of an added element to the arguments about baseball awards voting the last couple of seasons: the increasing popularity of Wins Above Replacement as a measure of player value. Although I personally have not experienced single-season WAR being used as a “conversation stopper” in player comparison, it seems that some people feel that happens far too often. That is unfortunate, because while WAR is a very useful tool for a getting a picture of a player’s overall contribution relative to his peers, it isn’t something that should be used to end those debates, but to recast them in a different, and hopefully better, fashion. Rather than explain WAR from the ground up (the FanGraphs Library has a good primer), or even to say who should win, today my goal is simply to show how I would use WAR in relation to the 2011 American League MVP Award in a way that probably isn’t too different from most other FanGraphs authors.
The first thing I should say is that the “Official Position” in the title is a bit ironic. After all, one stereotype of saber-nerds is that we all think in the same (probably robotic, perhaps cybernetic) fashion. That isn’t the case, of course, but since the point of this post is to demonstrate general principles for using WAR rather than very specific distinctions, it’s probably okay.
Second, although I know that many people like to argue about the meaning the “V” in “MVP,” I won’t be engaging that issue at length here. For the sake of this post, “Most Valuable” means something like “if you could guarantee the performance the season ahead of time, which player’s 2011 would you pay the most for without knowing ahead of time what the rest of the team is like.” Perhaps that’s a bit Rawlsian, but I’m a few hundred words into this thing already…
Let’s (finally) get to it: who does WAR say is the American League’s Most Valuable Player in 2011? Easy, let’s just look at the value leaderboard for the American League. As of this writing, Jacoby Ellsbury is at 8.5 WAR, just edging out Jose Bautista who is at 8.2. Assuming the vote is today, Ellsbury should win, right?
Well, maybe Ellsbury should win, but it isn’t that simple. For some seasons, just looking at the WAR leaderboard and picking the overall leader would might work. For example, in 2005, Alex Rodriguez was more than two wins better than any other position player, and although I think pitchers should be able to win the MVP, not even 2005 Johan Santana came close enough for me to have seriously considered him over A-Rod.
However, a brief glance over the WAR leaderboards from the past shows that a situation like 2005’s is a rarity; most seasons are more like the current one. So how do we decide who should win?
The first thing I would do is decide who the contenders are. Although I could make the cutoff larger, for the sake of space I will make my initial and imprecise cutoff at about 1.5 wins from the current WAR (Ellsbury at 8.5). Although I think that players like Ian Kinsler, Alex Gordon, and a number of others should get at least some down-ballot support, for now we are left with leaves us with Ellsbury, Bautista, Dustin Pedroia, Curtis Granderson, and also two pitchers in CC Sabathia and Justin Verlander. Again, I think a pitcher should be able to win the MVP, but for this season, with Sabathia and Verlander just on the cusp, I don’t think they are quite in the same league as Ellsbury or Bautista, though the reality is that you can make a legitimate case for any of these guys and not be too far off the mark.
Pedroia and Granderson are a bit more difficult to knock off, as their listed values with respect to Ellsbury and Bautista are significantly weighted by their fielding values for the season: very positively in Pedroia’s case (plus 17 runs), somewhat negatively in Granderson’s (minus seven). If you think that one or both of those fielding ratings is too low, then conceivably one or both players might really be in the area of eight WAR, and thus close contenders. On the other hand, some might find Pedroia’s rating in particular too high, in which case he falls back to the pack. I’m not going to get into an extended discussion of fielding metrics here (there will be more later in the week). I am simply acknowledging that they are rightfully the most controversial component of FanGraphs and other total value metrics. Indeed, if you make enough “adjustments” to the defensive ratings of all the top players, you could get a very different ranking of players indeed! While my preference of UZR, which is what is used as the fielding component useful and is rightly used as the fielding component of WAR here at FanGraphs, the individual player pages also have Dewan’s Defensive Runs Saved, or you can check other fielding metrics out there. I would also recommend looking at the ongoing results of the 2011 Fans Scouting Report, which is very useful.
With those (qualified) “eliminations” aside, we are down to Ellsbury (8.5) and Bautista (8.2). Even if the fielding issue weren’t there, for practical purposes the 8.5 and 8.2 WAR are the same. WAR isn’t precise enough to make that distinction. Again, that doesn’t knock its overall usefulness (and it should be noted, the primary intended use of the WAR concept in general is not to decide annual awards), it is just the nature of the beast. With that in mind, how can we decide who has “really” been most valuable?
We know that the same offense from a center fielder like Ellsbury is more valuable that from a corner outfielder like Bautista, but a positional adjustment part of WAR (which also accounts for multi-position players like Ben Zobrist and, briefly this season, Bautista), accounts for those differences. The non-stolen base element of baserunning is accounted for by UBR, and the differences aren’t that great, either. Really, then, the decision between Ellsbury and Bautista comes down to fielding versus hitting.
As mentioned earlier, it isn’t just the fielding portion of WAR that is subject to uncertainty. However, it is true that the estimation of batting runs is on much firmer footing that fielding runs. Thus, it is tempting to say that given two players of roughly the same WAR, one should go with the player who has the better batting as the more “sure thing,” which would make Bautista the winner.
Before we let ourselves off of the hook that easily, let’s be a bit more careful. After all, the uncertainty around fielding statistics does not simply imply that Ellsbury’s rating should actually be lower than plus 14 runs and/or Bautista’s higher than minus 4.5 runs. The opposite might be the case: Ellsbury might have been better than fourteen runs and Bautista worse! Variance goes both ways, after all.
As I said at the beginning, I’m not posting this to use WAR and other metrics as a sledgehammer to make a case for one MVP candidate over another. I suppose if you asked me today, I would say Bautista, but I could easily be convinced that Ellsbury was the man, and certainly wouldn’t think any worse of someone for differing with me. The main thing I hope to have shown in my own long-winded way is that WAR itself can at be a useful (if imprecise) tool for assembling who should be on the ballot. Perhaps by looking at the components from different angles (different fielding metrics and scouting reports, WPA/LI), we can refine our thinking a bit. My way of looking at it might be summed up like this: WAR isn’t problematic because its imprecision prevents it from being a conversation-stopper. Rather, WAR succeeds because it gives us a framework to help get a better conversation started.
WAR gives us a pretty good idea of who the legitimate candidates for the award are this year. Splitting hairs among players who all had great seasons is not what it was designed to do, nor is it a wise use of the metric. In most years, WAR will not tell you who exactly should get the #1 spot on an MVP ballot, but it will probably give you a good idea of who should be in the top 5. From there, use your best judgment, and understand that there are usually multiple valid opinions, with none of them being obviously more correct than the others.