I was tempted to begin this column with the almost-obligatory lyrics from "War," a No. 1 hit song in 1970. But when Edwin Starr sang that war was good for "absolutely nothing'' (say it, say it again), he was recording a protest against actual warfare. My protest, on the other hand, is about something much more widespread, threatening and serious.
I'm talking about the proliferation of WAR: Wins Above Replacement.
Wait! I'm not just some old curmudgeon sportswriter reflexively ranting against the latest technical advance. Well, yes, I am an old curmudgeon sportswriter who routinely rants about new technology -- in fact, my friends joke I would have complained about the introduction of electric light bulbs ("For crying out loud, what's wrong with a simple candle!?!") -- but hear me out on this WAR thing.
First, let me stress there are things I like about WAR.
1. Unlike most stats revered by curmudgeons (who tend to view RBIs as the most reliable and masculine of statistics), WAR not only acknowledges a player's defense, it embraces his fielding as a vital part of his performance. Other stats focus on offense; WAR makes defensive statistics a key part of its calculations.
2. WAR also is a very valuable tool for drawing general comparisons of players across baseball eras, particularly with groups of players. Colleague Dave Schoenfield recently wrote a wonderful analysis of the value of lower first-round draft picks, and he would not have been able to do so nearly as well without using WAR.
3. WAR can be fun, such as when you add up the WAR marks for individual players on a team's spring training roster and gauge whether the Pirates have any chance at a winning season.
So yes, I like WAR as a statistical measure.
My issue is this: I don't like the increasing overuse of (and overreliance on) WAR as THE definitive evaluation of a player's worth.
This was particularly true during the Mike Trout-Miguel Cabrera MVP debate last fall. For instance, consider this headline on an ESPNLosAngeles.com story in late September: "Mike Trout Is Your MVP (WAR Says So)."
That was just one of many stories focusing on WAR in the MVP race, where the stat became a big factor in analysis of the players, as Bleacher Report noted.
Now, Cabrera wound up winning the MVP by a wide margin, so WAR wasn't a decisive factor in the award vote. (I voted for Trout, though I did not base my ballot on WAR.) But I just found it tiresome to keep reading all the references to it, as if WAR was the only stat that should be considered, and leading a league in batting average and home runs and RBIs -- as Cabrera did in becoming the game's first Triple Crown winner since 1967 -- was somehow a mere accounting trick.
We saw it again during the recent debates over the Hall of Fame, when there were more mentions of the candidates' WAR record than you'd find in a presidential campaign. How this might have affected the case for Jack Morris is one of the reasons the over-reliance on WAR drives me nuts.
What's wrong with WAR?
1. Almost no one knows how to calculate WAR. I'm not just talking about curmudgeons who have trouble figuring out a batting average with a calculator (oh, for the glorious days of the Ready Reckoner!). Smart fans who really know their stats and their baseball (as well as being clever enough to work those new-fangled "universal" TV remotes) can't calculate it. Just the baseball-reference.com WAR explanatory page is more than 1,500 words, including several paragraphs explaining how the formula has changed (but not including the list of the players who were helped and hurt the most by those changes).
If we can't figure a stat out on our own, then how do we verify whether it is accurate?
2. Actually, we know it isn't always accurate because depending on your source -- FanGraphs or Baseball-reference.com -- you can get wildly different WAR scores. (For the record, ESPN.com is using baseball-reference's WAR.) For example:
Does Morris, in fact, belong in the Hall of Fame? No, he doesn't, according to baseball-reference.com, which gives him a WAR score of 39.3, tied for 145th all time among pitchers. Maybe he does, according to FanGraphs, which gives him a 56.9 WAR, 75th all time.
Who got the better end of last week's Justin Upton-Martin Prado trade, Atlanta or Arizona? Again, your analysis depends on your source. Baseball-reference has Upton's career WAR at 13.1 while FanGraphs has it at 17.1. That's a significant difference, in WAR terms. Prado's WAR is 15.8 in both and as high as 5.4 last year, according to baseball-reference, which could be good news for Diamondbacks fans. (Upton's two WAR numbers for the 2012 season were 2.5 by FanGraphs and 2.1 by baseball-reference.) But aren't most fans and experts hailing Atlanta for fleecing Arizona in the deal?
Who would you rather have had as your shortstop last year -- Derek Jeter (the All-Star Game starter and future Hall of Famer) or the Mariners' Brendan Ryan (who hit .194 with a .555 OPS)? Yes, this is a trick question. Because while any sane fan will pick Jeter (even with his ankle in a cast) over Ryan, baseball-reference gave Ryan a WAR of 3.3 last season and Jeter a WAR of 2.1, probably because Ryan is the superior fielder. FanGraphs' spread is just the opposite: Jeter at 3.2 and Ryan at 1.7.
And just in case you're thinking FanGraphs must feature the more reliable WAR, bear in mind that Ricky Nolasco was 13-9 with a 5.06 ERA in 2009 while pitching just 185 innings, but FanGraphs had his WAR at 4.3 that season.
Yes, I know those are outliers, and that most WAR scores between the two sites are much closer (as in Prado's numbers). But the very existence of such anomalies calls into question WAR's overall accuracy. If a player's batting average varied from .245 to .307 from ranking to ranking, would you trust either statistic?
3. What accounts for such differences? The main factor is the way the two sites measure fielding (which takes me back to Complaint No. 1). The fielding metrics used by baseball-reference (Baseball Info Solutions Defensive Runs Saved) seem to lift or lower their WAR scores much more than the fielding metrics used by FanGraphs (Ultimate Zone Rating). And that presents another issue with WAR.
Batting average, OPS, ERA, etc., are the products of such simple and certain mathematical calculations that even I can figure out armed with only a pencil and paper. WAR, however, is partially a product of fielding metrics that are by no means certainties.
If a player has a .333 batting average, we know he gets a hit once every three at-bats. It is a statistical fact. But if we say a shortstop has an Ultimate Zone Rating of 12.4, well, it means he's likely a very good fielder but we also have to assume that the theory behind UZR is indeed an accurate measure of fielding ability. It might be (more so with infielders) but it certainly isn't precise and inarguable.
In other words, most baseball stats are based entirely on indisputable math calculations. WAR has an element of theory and assumption to it.
As I wrote at the beginning, WAR can be a valuable measure. The problem is when it is used as the DEFINITIVE measure. Such as, "Mike Trout had a 10.7 WAR and Miguel Cabrera had a 6.9 WAR, so anyone who thinks Cabrera deserved to be the American League MVP should be strip-searched, tied to an anthill and forced to rely on dial-up for his Internet connection for the remainder of his pathetic life."
Look, all stats have their limitations. If a player has a .329 batting average, that probably means he's pretty good. But for an accurate measure of the player, we need more information. How many extra-base hits does he have? How many times does he walk? How many stolen bases? How many runs, how many RBIs? How many double plays has he hit into? The same is true for a player with 40 home runs. Does he have a .300 average to go with them, or a .230 average? Does he strike out a lot? How often does he walk?
The same approach should apply to WAR. We need to look at many stats to assess players, and one of them should be WAR. But it shouldn't be the only stat we look at or cite.