The War Over DJ LeMahieu’s WAR

I’d say I don’t know what to make of DJ LeMahieu anymore, except I’ve never had any idea what to make of DJ LeMahieu.

At a stringy 6-foot-4, he looks more like an outfielder than a second baseman, but he’s been a fixture in the middle of the diamond for Colorado since 2013. And he’s been pretty good, too: He took home a Gold Glove in 2014 and made the All-Star team in 2015.

This year, LeMahieu is having the best offensive year of his career. Long a high-average hitter with little patience or power, the 27-year-old is hitting .324/.386/.486, which will play at any position, in any ballpark. But whether this year represents a quantum leap for LeMahieu‘s all-around game is far less clear. That depends on how you feel about his defense.

In 2014, LeMahieu was an elite defender according to Defensive Runs Saved (which goes into Baseball Reference’s WAR calculations). His 16 runs saved led all NL second basemen, and if you’re that good defensively at an up-the-middle position, you can be a replacement-level hitter and still be a valuable player.

UZR (which is the defensive component of FanGraphs’ WAR), rated him at 10.7 runs above average that year — still tops among NL second basemen, but by a much smaller margin — and FRAA (the defensive component of WARP at Baseball Prospectus) credited him with only 1.8 runs saved. As a result, the three most popular public WAR estimators differed on LeMahieu by more than a win in 2014. That really doesn’t matter when you’re arguing whether Mike Trout was worth nine wins or 10, but for LeMahieu, it was the difference between being league average and replacement level.

And in the two years since, those three sites have continued to play tug-of-war over LeMahieu’s value:

2014: 1.5 (BR), 0.7 (FG), 0.2 (BP)

2015: 2.3 (BR), 1.9 (FG), 1.7 (BP)

2016: 1.7 (BR), 1.0 (FG), 2.0 (BP)

The Colorado second baseman is a special case: He plays in an extreme offensive environment at Coors Field, and a lot of his value comes from his glove. How confident are we that park adjustment numbers are exactly right? And even if we know LeMahieu’s a good defender, is there any way to be certain how good? Depends on what data you decide to collect, and how you collect it. For as exhaustively as we analyze baseball nowadays, individual defense is still incredibly difficult to measure precisely.

While WAR is the single most influential and important statistic of the sabermetric revolution, its value is not in its precision. WAR cobbles together disparate data sources, measuring discrete skills, attempting to weight them equitably for pitchers and position players, outfielders and infielders, glove-first shortstops and designated hitters, all rounded to a tenth of a win.

In order to do that, you have to not only measure quantitative data, but you also have to make methodological choices on what to value, and how, and how much to weight it. It would be suspicious if three different approaches to this problem didn’t disagree significantly from time to time. When they do disagree, it’s up to fans and analysts to take those three different WAR numbers and weight them against each other: Do you trust FanGraphs more than BP? If so, by how much? And how much do your own observations — the eye test — factor into the analysis?

Therefore, the value in this magic number is not so much in the number itself, but in the process, in teaching us how to come up with our own holistic value metrics, even if they only exist in our own minds. WAR gives us ground rules on what we ought to value, but what to make of LeMahieu is still open to interpretation.