WAR. The word itself is often followed by the phrase 'What is it good for?' because humor stopped evolving at some point in the 1970's. Lucky for us, though, baseball analytics have been evolving like mad since the 1970's, and WAR is pretty much the apex hunter in that evolutionary field.
What is WAR?
WAR stands for Wins Above Replacement. You can get it in Baseball Reference
flavor, with minor variations for the connoisseur.
There are three words in that sentence, and each needs its own explanation. First- Wins. It doesn't necessarily mean wins, per se. The metric is actually Runs, but using runs gets a little unwieldy. Baseball Reference pegs Mike Trout as 96.6 Runs Above Replacement last year, which sounds astounding (it actually is astounding). And if you want to use RAR, that's fine, but people will probably just say 'RAWWRR!' back at you and you'll feel sad. In order to avoid those negative emotions, we convert Runs Above Replacement to Wins Above Replacement (One Win equals tennish runs).So, say Nick Punto goes 0-4 through eight innings, but hits a game winning single in the ninth. That's not a win. It's a single. Say Yu Darvish pitches 7 innings, allowing one run, and the Rangers find a way to not score less than one run in a yu Darvish start (that would be a new trick for the 2014 Rangers that the 2013 version had trouble with). That's a Pitcher's win (which is kinda a junk stat), but it's not a Win in terms of WAR.
Replacement is, in a philosophical sense, the most important aspect to the stat. The idea is, once you get past a certain level, there's just not much differentiation between players. In terms of contribution, there's just not a huge difference, cumulatively, between what Adam Rosales gives you and what Alberto Gonzalez gives you. Establishing Replacement Level is important, because it gives us a good, quality baseline.
Above means above. It... it doesn't really merit its own paragraph.
One of the the great things about WAR is it normalizes for a lot of things. Want to compare players across a few generations? WAR is the best way we have to do it. Another of the great things about WAR is it adjusts for position (you already knew that, though, because you read the two links in the first paragraph, right?). As well, WAR includes data for defense.
An argument that we've faced late in the last few seasons is whether Miguel Cabrera or Mike Trout was the Most Valuable Player in the AL. WAR has tended towards Trout (10.4 for Trout in 2013 compared to Cabrera's 7.6, 10.0 to 6.8 in 2012). All this is really saying is Mike Trout was nearly (but not quite) as good as Cabrera offensively, while playing a more demanding position well defensively, while being a very good baserunner. Defense, position, and baserunning are all important, and WAR is the best way we have to communicate that.
What does WAR tell us about the 2013 Texas Rangers?
The Rangers face 2014 missing five key regulars from the 2013 team. If I had to quiz strangers on the street on who was had the highest fWAR (that's the Fangraphs
version of WAR) of the departures, I would guess that Nelson Cruz or Joe Nathan would be the most popular bet. However, that honor belongs to none other than Craig Gentry, whose 3.4 Wins Above Repalcement last year was second on the team among position players. Remember when I said position, defense, and baserunning was important? Fangraphs saw him as a good baserunner (5.8 Runs worth) and a phenomenal defender (14.3 Runs). Combined with playing a demanding position in Centerfield, that combined to earn him that 3.4 WAR mark, despite being only 2.6 Runs Above Replacement as a batter. They'll just have to make do with Shin-Shoo Choo as his replacement, and the 5.2 Wins worth of value he provided the Reds last year.
One thing we have to keep in mind is that no stat is or ever will be perfect. While my opinion is that WAR is baseball's answer to the Higgs-Boson, it does have flaws. Most notable is the defensive component. Quantifying defense is really, really hard, and we still haven't perfected it. Ergo, the defensive component of WAR varies, a lot, year-to-year. Theoretically, defensive skill doesn't fluctuate among players year-to-year, so it's a good guess that the metric measuring it shouldn't show year-to-year fluctuations. It's a flaw, but not a fatal one- just one to keep in mind.
As well, think of WAR as a Bowie knife instead of a scalpel. In other words, if Player A is worth 3.4 WAR, and player B is worth 4.2, in all reality they're about the same player, because we're not so confident in in the metric to say it's sharpened down that far. Luckily, we have scientists in the lab every single day to bring you better WAR.
... That last sentence sounds ominous.
As well, the different iterations of WAR make for some confusions. This is also a strength, because you could theoretically make up your own version of WAR according to your values. Think closers have ninth inning magic not available to ordinary relievers? Make a version of WAR that weighs Leverage (pLI) and saves!
The above kind of sounds silly, but it's a pretty well known fact that virtually every front office uses WAR in some form or fashion. It's just too good of a way to weigh disparate players, and try to find weaknesses on your own team and strengths available in the market.
At any route, my advice is to pick a version of WAR and stick with it. I prefer Fangraphs, mostly because they use weighted/regressed metrics in determining Runs Above Average, so their version (I find) has a bit more predictive value. You can then use WAR to weigh out how much a player should be paid... but that's an article for another day (foreshadowing!).
It's common for old-school/establishment types to look down on advanced metrics, for one reason or another. Yet, what methodology does the old school use to quantify defense and baserunning? Errors and steals are very, very coarse metrics. WAR is simply the best measure we have to combine all aspects of baseball, which every old-school baseball man in history would have to like. Plus, it's simple- player A is a 6, player B is a 2... Player A is lot better. You can get under the hood and things get complicated, sure... but the general essence of the stat is one everyone should appreciate and use.
Joseph Ursery has said ''RAWWRRR' at people, but 'people' is his very young daughter, so it's acceptable. He doesn't use it much in baseball context. You can follow him on Twitter at @thejoeursery.
Read or Share this story: http://www.wfaa.com/story/news/local/2014/08/20/14152332/