FanPost

Why Does WAR Despise Lefties?

Earlier this week, Bill James created some controversy on social media with a cryptic message about the statistical use of Wins Above Replacement (WAR).

The tweet provided no further clarification, and immediately generated confusion from baseball gurus like Tom Tango. But the war against WAR is nothing new, perhaps most famously being labeled a "semi-junk stat" by Tom Verducci.

I myself reference WAR frequently, and find it a useful way to show the impact of players in a given season. But I wouldn't be the first to point out its flaws as a value generator.

The most common critiques of WAR include:

  1. There's no standard definition: Baseball-Reference (bWAR), Baseball Prospectus (WARP), and FanGraphs (fWAR) all have their own formulas that can create wildly different numbers. In perhaps my favorite example, 19th century pitcher Jim McCormick tallied 76 bWAR (27th all time) and just 40 fWAR (163rd all time). According to the former, he is a borderline Hall of Famer that provided more "value" than 300 game winners Tom Glavine and Don Sutton, and the latter claims he was less valuable than A.J. Burnett and Dan Haren.
  2. The formula is always subject to change: For instance, last year FanGraphs added catcher framing to its formula, and multiple players saw double digit changes to their fWAR. So Buster Posey gained 12.4 wins overnight without collecting a single hit or throwing out a base runner. Say what you will about traditional stats, but outside of Hollywood, no one is adding/subtracting hits, home runs, or wins for retired players.
  3. It doesn't weigh plays on importance: We've gotten to the point that every play can be broken down to attribute runs scored or lost to every player involved, including the pitcher, hitter, base runner, and fielders. But the value is the same in the fourth inning of a blowout as the ninth inning of a tie game.

But allow me to throw an entirely new argument into the "WAR is flawed" discussion: it severely punishes southpaws.

The Position Conundrum

In baseball, there's a defensive spectrum that ranks each position based on its difficulty to play. Players routinely move down the spectrum as they age or skills decline (think Buster Posey moving from catcher to first base), and while the game is evolving, they typically don't move up the spectrum.

WAR accounts for this with the positional adjustment (Rpos for bWAR), one of the six components that go into bWAR and fWAR. Here's the current Rpos breakdown for bWAR:

  • C: +9 runs
  • SS: +7 runs
  • 2B: +3 runs
  • CF: +2.5 runs
  • 3B: +2 runs
  • RF: -7 runs
  • LF: -7 runs
  • 1B: -9.5 runs
  • DH: -15 runs

To calculate a player's Rpos, you multiply his number by every defensive inning he played, then divide by the possible number of innings in a season (1,350). Posey caught 846.1 innings in 2019 and played first base in 24, so his RPos would be ((846.1 x 9) + (24 x -9.5))/1,350, or 5.5 (rounded to 6).

Even great defenders at lower spectrum positions will have their defensive value eaten away by Rpos.

The Penalty for Lefties

Everything from the previous section seems fair, until you realize that baseball was built with right-handed throwers in mind. Because players run the bases counter-clockwise, left-handed throwers are all but disqualified from playing catcher, second base, shortstop, or third base due to awkward throws.

Take another look at the spectrum and you'll notice those four positions represent 80 percent of the chances to receive a positive positional adjustment. In other words, if I throw left-handed, I can either play center field, pitch, or my career will begin with negative WAR. That sucks.

How Badly Are Lefties Punished?

Let's compare four of the statistically greatest fielders of all time, two that throw left-handed and two that throw right-handed (chosen largely based on number of Gold Gloves won):

Rfield dWAR Gold Gloves oWAR
Name Thr Pos Tot Rank Tot Rank Tot Pos. Rank Tot Rank HOF?
Barry Bonds L LF 195 6 7.6 384 8 8 143.6 3 No
Keith Hernandez L 1B 117 37 1.3 >1,000 11 1 46.3 214 No
Brooks Robinson R 3B 293 1 39.1 3 16 1 47.7 200 Yes
Ozzie Smith R SS 239 4 44.2 1 13 1 48.8 188 Yes

Rfield is a Baseball-Reference statistic that counts the number of runs a player saved in the field above replacement level, while dWAR and oWAR are the number of wins attributed to his offense and defense.

Robinson and Smith were elite defenders, and saved two to three times more runs in the field than Hernandez. But their defensive value is 40 times more, which is explained primarily by the fact that Hernandez lost 103 of his 117 defensive runs saved by playing all but 38 of his 17,279 innings at 1B. Meanwhile, the three players have nearly identical value provided on offense.

I'm not trying to argue that Hernandez belongs in Cooperstown, but bWAR is surely doing him no favors. It doesn't get any better if we look at fWAR, which gives him a defensive value of .3 wins and the 3,394th all-time rating despite being the most decorated defensive first baseman of all-time.

Think I'm cherry-picking the data by highlighting Hernandez? Of the 55 lefties currently on the top 400 of the career Rfield leaderboard, only 16 rank in the top 1,000 for career dWAR (of those, only Bonds, Darin Erstad, Brett Gardner, Ricky Henderson, and Jason Heyward weren't predominantly CF). Only one out of 55 (Henderson) is in the Hall of Fame, compared to 31 righties.

It may seem odd to include Bonds in this list, because he was much better with the bat and his lower number of Gold Gloves and lack of Hall of Fame plaque are both likely explained by performance enhancing drugs. But it's still noteworthy that he's top 10 all-time in Rfield, but because only 1,387 of his 24,296 innings were played at a prime defensive position (CF), his defensive value closely resembles Paul Lo Duca and Martin Prado (zero combined Gold Gloves).

Why Does This Matter?

For many left-handed throwers, their defensive position was likely dictated by skill instead of their throwing hand. Left-handed pitchers have also had the benefit (until 2020) of guaranteed jobs in the bullpen based less on skill than handedness.

But I couldn't help noticing that Baseball-Reference lists Bonds, Hernandez, and Robinson all at 6' or 6'1" tall and 180 or 185 pounds (Bonds' weight likely from his earlier years). I'm no scout, but if the former two were right-handed, would their defensive skill have allowed them to handle the hot corner?

Well, let's compare two current Giants infielders across several bWAR factors:

Name Thr Pos RField RPos dWAR oWAR bWAR
Brandon Belt L 1B 48 -56 -1.3 18 23
Pablo Sandoval R* 3B -22 14 -1.1 22.4 20.1

The asterisk for Sandoval is because he is naturally left-handed, but taught himself to throw right-handed at age nine because he wanted to play shortstop. These two players have had remarkably similar values, but Belt is a significantly better fielder.

If we instead put all 9,645 (and counting) of Sandoval's defensive innings at 1B, his RPos becomes -68 and it's unlikely his RField can make up the difference. It would be difficult to predict his defensive ratings exclusively playing 1B, but he does have a similar body type as Prince Fielder.

Both players are listed as 5'11", while Fielder is 275 pounds compared to Sandoval's weight of 268. In 11,591.2 innings (20 percent more than Sandoval), Fielder put up -100 Rfield, -98 Rpos, and -20.5 dWAR, combined with 34.1 oWAR.

Let's say Sandoval is a better fielder than the ironically named Fielder, and only give him -15 dWAR at 1B. That still divides his bWAR by three (20.1 to 6.2), just because we changed him from a righty to a lefty.

So the next time you see someone use WAR as a way of evaluating a player's career, just remember that this statistic is by no means the black and white number it attempts to represent.

This FanPost is reader-generated, and it does not necessarily reflect the views of McCovey Chronicles. If the author uses filler to achieve the minimum word requirement, a moderator may edit the FanPost for his or her own amusement.

Recent FanPosts

In This FanPost

Teams