Matthew Roberson: WAR is fine, paying players by it would be hell

NY Daily News· Julio Aguilar/Getty Images North America/TNS

NEW YORK — In the never-ending war between man and machine, the machines may have picked up another point.

According to a report from the general managers meetings in Carlsbad, Calif., Major League Baseball has proposed a new system for determining player salaries before free agency. Similar to their proposal from August, the new one suggests eliminating arbitration and granting players free agency once they turn 29-and-a-half.

That August proposal offered a predetermined dollar amount pooled among eligible players; the newest tweak would pay the players based on performance. The parameters are not your father’s home runs, RBIs, or pitcher wins, though. Instead, the league has pitched the idea of using Wins Above Replacement, the game’s statistic du jour that calculates the most all-encompassing totality of a player’s on-field value.

Specifically, the league proposed FanGraphs’ version of WAR, colloquially known as fWAR. The website Baseball-Reference has their own version of the statistic as well (bWAR), using a slightly different formula based on different underlying statistics. A major difference between fWAR and bWAR is the amount of weight they put on different defensive positions.

Catchers, for instance, receive a +12.5 positional adjustment in FanGraphs’ formula. First basemen, for context, receive an adjustment of -12.5 due to the perceived ease of their position, relatively speaking. With the added catcher bump, Buster Posey’s 2021 fWAR was 4.9, making him one of the 20 best position players in the league per FanGraphs. Baseball-Reference, meanwhile, had him at 3.5.

You’re starting to see the problem.

One agent told The Athletic that Albert Pujols will lead the league in stolen bases before the union agrees to the proposal. But if it somehow comes to be, there will be justified gripes about consistency and fairness. With two different versions out there, both used and cited regularly by the industry, choosing one that so highly fluctuates the value of certain positions is sure to ruffle the feathers of those nesting in a corner spot. Never mind designated hitters, a position of unquestioned importance, who would see their value tanked by fWAR. Giancarlo Stanton, who made 108 of his 134 starts last season at DH, would represent the anti-Posey. Stanton’s value was determined as half a win greater by bWAR (3.1) than it was by fWAR (2.6).

With literal millions of dollars potentially on the line, those decimal differences would be highly significant. Even after narrowing the focus to just fWAR, the list of probable issues with the system goes on and on. This is not to disparage or discount the statistic, but rather to highlight the absurdity of using it as gospel to determine how much a worker should be paid for their labor.

Consider the plight of the part-time player.

Under this new proposition, backups, platoon men and utility players would lose the ability to argue on behalf of their value. The current arbitration system lets anyone with at least three years of big-league experience present their case to an independent panel explaining why they should make more money. If that human element is removed, and fWAR becomes the be all, end all decider, those part timers are punished for the simple act of playing their role.

Take Chris Taylor for example. The Dodgers have used Taylor for over 700 innings at second base, shortstop, left field and center field. Before recently becoming more of a regular, Taylor filled in as needed whenever one of his teammates went down or simply needed a breather. It was a role the team specifically wanted Taylor for and he did it admirably, becoming the best utility man in the game. But with fewer starts than his contemporaries, Taylor’s fWAR was not as bloated.

In 2019, Taylor appeared in 124 of the Dodgers’ 162 games, hitting .262/.333/.462 in just over 400 plate appearances. Most everyday starters get to 600 plate appearances over the course of a full season. But with Taylor occupying a different role and doing it better than anyone — something a human would understand better than a computer — his fWAR came out at 1.7. That’s a solid season, but not one that would get him a two-year, $13.4 million contract, which is what he received.

On top of that, the $13.4 million was a way for the Dodgers to specifically avoid arbitration, where Taylor theoretically could have made even more money. Relief pitchers fall into the same boat, as they play a vital role but still wouldn’t have the innings and subsequent fWAR to warrant a big salary.

With FanGraphs prioritizing the sabermetric Fielding Independent Pitching (FIP) in their WAR formula, and Baseball-Reference sticking with good ol’ fashioned runs, a divide would also emerge between new school and old school, as well as pitchers with porous defenses and ones who play with Gold Glovers. We’ll go to the example board again and call up Nathan Eovaldi.

The Red Sox pitcher’s 2.79 FIP was MLB’s third-best last season. According to FIP, his actual ERA of 3.75 would have been a shinier 2.79 if he experienced league average results on balls in play. By using FIP in their WAR algorithm, FanGraphs rated Eovaldi’s season (5.6) higher than Baseball-Reference did (4.6). A pitcher like Robbie Ray, whose FIP was higher than his ERA, goes the other way (3.9 fWAR, 6.7 bWAR).

All of this is to say, it’s a terrible idea. The new proposal would complicate things more than it would correct them. Under the new scheme, which would reportedly be weighted for recency, imagine how a player like Michael Conforto would suffer from one outlier of a bad year. Not to play the world’s smallest violin for mega rich sports agents, but their whole job would also be greatly diminished if an algorithm determined their client’s next contract.

In submitting this idea to the world, MLB is trying to give the impression of progress and forward thinking. Instead, they came up with the contract negotiation equivalent of the despised BCS, college football’s postseason computer rating that was replaced by human voters eight years ago. College football is an apt comparison for another reason: MLB seems hellbent on stopping the players from getting the money they deserve.

Advertisement