A Word on WAR

By: Andrew Muschel

Growing up a baseball fan in the 1990’s, I never heard the words “moneyball” or “sabermetrics”. We rooted for “traditional” statistics, hoping Todd Hundley would eclipse the Met and catcher records for single-season homeruns and Anthony Young would lose enough consecutive games to ensure his place among all-time goats. Fast forward to the present and one cannot help but notice how many intelligent members of the sports media not only include, but also focus on, new statistics, such as WAR (Wins Above Replacement), FIP (Fielding Independent Pitching), and UZR (Ultimate Zone Rating). The definition and scope of each of these variables, along with the many others, are well beyond the scope of this essay; however, the primary goal of each is to quantify that which had always been subjective: defense, speed, luck, and a player’s overall value. Specifically WAR, whose method of calculation has varying opinions (e.g. Fan Graph’s fWAR and Baseball Reference’s bWAR), attempts to understand “how much better a player is than what a team would typically have to replace that player,” measured in wins.

The debate over the use of these statistics, especially as replacements for the traditional ones, such as stolen bases, earned run average (ERA), and runs batted in (RBI), has waged on. While some from the “old-school” maintain that WAR, as the song indicates, is good for “absolutely nothing,” many mainstream media outlets have turned their attention to these decreasingly controversial statistics. For example, in the winter of 2012, the two frontrunners in the race for the American League’s Most Valuable Player, Miguel Cabrera and Mike Trout, embodied the two distinct schools of thought. At the time, the New York Times featured an article by Nate Silver explaining why Mike Trout, based on his overall value measured in WAR, should win. Conversely, Miguel Cabrera who had achieved the historic feat of winning the Triple Crown (leading the league in three traditional categories: homeruns, RBI’s, and batting average), was overrated based on superficial statistics.

The use and precision of these new statistics are remarkable. A casual fan of the Mets in 2013 may have noticed that after injuries, the team replaced its power-hitting Lucas Duda (career OPS, on-base percentage plus slugging percentage, .766) with the weaker hitting Eric Young Jr. (OPS .663) and Juan Lagares (OPS .633). Yet, during those months, that same fan may have observed the team’s win-percentage increase significantly. The reasoning behind the improvement was presumably what Young Jr. and Lagares contributed with their baserunning and defense, unlike Duda who often seemed incompetent at performing either. WAR tells the full story, with Duda’s 2013 WAR an embarrassing -.2 (-.9 lifetime) as opposed to Young Jr.’s .9 WAR and Lagares’ 3.7 WAR.

That said, it seems that even the “all-inclusive” WAR must be examined more closely before crowning it the lord of statistics. Both mainstream sites that produce a WAR calculation include mentions of limitations, such as “WAR is necessarily an approximation and will never be as precise or accurate as one would like”, and “You should always use more than one metric at a time when evaluating players.” However, the online media still seems to put great emphasis, perhaps overemphasis, on a number which ultimately includes at least some measure of subjectivity. Though WAR is often advantageous, helping sports fans understand a player’s value and assisting general managers in building better franchises, a little humility should be portrayed while using it. Bloggers seem to assume that WAR is the current gold standard to assess everything baseball, and some cynical blogs will attack the primitive folks (generally broadcasters) who still talk about homeruns, ERA, and batting average, as if those numbers have no meaning in our advanced society. These attacks are obnoxious at best, factually presumptuous at worst. After all, WAR is a number and baseball is played by real people. In the words of long time Met broadcaster Gary Cohen, “People who write from a remove have a different perspective. Either it’s a lyrical perspective, which is nice, or a statistical perspective, which is also useful. But when you’re actually up close and understand the human beings who play the game, it gives you a better perspective.” In fact, using the very concept of sabermetrics, an attempt to quantify every previously unnoticed detail, certain limitations on WAR’s power become apparent.

For example, when the Mets recently signed outfielder Curtis Granderson to a four year $60 million contract, a plethora of essays were penned critiquing the signing, primarily based on projections of his WAR.[1] Met fans have defended the signings, invoking the Mets’ morale since the collapses of 2007 and 2008, the Madoff fiasco, and five straight losing seasons, and how desperately the team needed a lift. Those immersed in pure numbers and statistics refute that morale is not a statistic and that the game is played on the field, not in the stands or in the hearts of fans. However, that argument is not so simple.

In their 2011 book Scorecasting, a book seeking to understand the hidden influences in sports, Moskowitz and Wertheim prove the existence of home field advantage and explain the source of the advantage (spoiler alert: referee bias). They cite evidence that the size of the crowd actually affects the magnitude of the advantage.[2] Do those using WAR to critique the Granderson signing consider how many fans he may bring to the stadium, thereby increasing the home field advantage? On a simpler level, do they account for the WAR of future free agents who may sign with the Mets because, with the first 8-figure annual salary awarded in the Alderson era, Granderson has showed that the Mets are looking to compete? Presumably not. And aside from future acquisitions, it has widely been reported that David Wright, who has averaged over 4.5 bWAR per season over his first ten, signed an extension with the Mets last year on the premise that they sign some legitimate free agents. Wright appeared excited about the Granderson signing, indicating that it may have played a role in keeping his WAR in New York.[3]

Take as another example the 2012 sensation that was R.A. Dickey. Dickey won the National League Cy Young award comfortably, despite some demanding that Kershaw be crowned based primarily on his superior WAR. But again, WAR does not tell the whole story. Returning first to attendance, in late September 2012, the Mets concluded a four-game series at home with the Pirates which had, in the first three games, drawn an average paid crowd of 23,416. For the final game of the series, they started their then 19-win ace, Dickey, and not surprisingly, their attendance inflated to 31,506. Perhaps that oversized crowd (by Citi Field standards) was of greater assistance to Dickey, who struck out 13 and won his 20th game, than the weaker previous crowds could have been to the previous pitchers. But that attendance number, or at least its significance, would not appear in the box score or be incorporated into any sabermetrics. As another example, a Met fan watching that game may have said that David Wright’s go-ahead three-run homerun was motivated by the opportunity to provide his beloved teammate the chance to win his 20th game. These emotions seem to have no place on the field or in the world of statistics, but consider this: In the 33 games started by Dickey, the Mets scored an average of 4.6 runs; in the 129 games he did not start, they scored a paltry 3.86 runs per game. Perhaps their feelings of camaraderie juiced their bats or maybe the opposing pitcher had to take more risks to compete with Dickey and failed. If the former sounds too magical, how else can we explain the fact that of Dickey’s seven no-decisions in 2012, he left the game in line for a loss in six but was redeemed by the generally ordinary offense’s heroic efforts.

Two other fascinating numbers quantifying Dickey’s otherwise uncalculated value were brought to light by Michael Salfino in the Wall Street Journal. Salfino pointed out that the unusual style of the knuckleball is not only confusing as its thrown to hitters, but also serves as a confusing contrast to the fastballs thrown by most pitchers, including Met relievers finishing for Dickey and starting pitchers throwing the day after Dickey against the same team. In his three years with the Mets, the relievers were much more successful following his performance than when following one of his teammates (3.44 to 4.26 ERA overall, 3.41 to 4.78 in 2012). Similarly, over the course of 2012, in the 132 plus innings thrown by Met starters the game after Dickey against the same opponent, they combined for an ERA of 2.38, significantly lower than their otherwise ERA of 4.56. If WAR’s goal is to “summarize a player’s total contributions to their (sic) team” (Fan Graphs), allow us to conclude that though it may incorporate many contributions, it remains impossible to quantify fully a unique human being and his contributions to a baseball team.

The game of baseball continues to evolve as does its statistics. Sports fans continue to benefit from precise statistics measuring almost any component of the game. My goal is not to elegitimize the validity and value of statistics such as WAR; merely to request from all of its users to engage in more humility, appreciating that despite its impressive thoroughness, there is more to a player than his WAR.  What is war good for? Something, just not everything.

*          *          *

Andrew Muschel, MS, is a psychology doctoral student and Met fan living in Queens, NY. He enjoys writing about his two topics of interest, psychology and sports, at his new blog, Psychology and Sports.