What Does WAR Have Against Catchers?
Consider, for a minute, Sean Smith’s ranking of the top 500 position players by career WAR. The list, for the most part, makes sense to me. I think it computes fairly well with many fans intuitions about who the best players ever were. Next, break the list down by position. It looks a little like this, if you look at the top 5:
- 1B: Gehrig (13), Foxx, Bagwell, Pujols, F. Thomas (43)
- 2B: Hornsby (9), Collins, LaJoie, Morgan, Gehringer(34)
- 3B: Schmidt (16), A. Rodriguez, Mathews, Boggs, Brett (30)
- SS: Wagner (6), G. Davis, Ripken, Yount, Vaughan (44)
- C: Bench (52), Fisk, I. Rodriguez, Carter, Berra (97)
- OF: Ruth (1), Bonds, Cobb, Mays, Aaron, Speaker, Musial, Williams, Mantle, Henderson (14)
Which of these is not like the others? Judging by career WAR, the best catcher ever ranks behind the 5 best at every other position, and the 10 best outfielders. This does not seems to fit the way baseball is really played. Let me explore this for a bit to see what we can learn about the way WAR works and the careers of catchers.
First, let us look briefly at the components of WAR. Look at this chart for Johnny Bench, WAR’s best catcher ever. To get WAR, you add together Batting Runs, Baserunning Runs, Grounded Into Double Plays, Reaching on Error, Fielding Runs (measured by Total Zone), infield Double Plays, Outfield Arm, a generic adjustment for Catchers, a positional adjustment, and the replacement adjustment. That gives you Runs Above Replacement, which converts to WAR by dividing by a number close to 10. None of this seems to disadvantage catchers, until you look a bit deeper. Two components challenge catchers, Batting Runs because of its heavy plate appearances component and Total Zone because of the difficulties of measuring catcher defense. Now these points are fairly obvious. So how does Smith, in particular, try to correct for this bias?
Smith adds an adjustment for catcher defense, first of all. This is necessary when you look at Bench’s numbers. By TZ, Bench is a defensive liability, but the catching adjustment makes him above average. Nonetheless, the adjustment strikes me as questionable. Compare Bench to Gary Carter or Ivan Rodriguez. For their career, Carter is 107 fielding runs above replacement (TZ + Catcher), while Rodriguez is 154 runs above replacement. Bench lags well behind at 72, though a large portion of the deficit comes from late in his career when he played 3rd base poorly. Stripping those years out, Bench still sits at 88, a substantially weaker defensive catcher than Gary Carter. That strikes me as idiosyncratic at best, and nonsense at worst.
Second, Smith adds a positional adjustment. This adjustment attempts to account for the relative scarcity of catchers, shortstops, etc. when compared to left fielders and first basemen. For his career, the positional adjustment adds 98 runs to Bench, 135 to Rodriguez, and 118 to Carter. In contrast, it subtracts 81 runs from Babe Ruth and 129 runs from Barry Bonds. Look back at our top fives, and you can see the logic behind the positional adjustment. Great outfielders are on every street corner, while great shortstops are substantially rarer. Nevertheless, the fact that catchers trail every other position by a substantial margin suggests that the positional adjustment is not large enough for catchers. The catching adjustment needs to correct for the difficulty of the position, but it also needs to account for the plate appearances lost solely because a player catches.
Finally, the replacement level for catchers is probably set too high. It is very hard to find a great catcher. Given that fact, which Smith’s chart clearly supports, it should necessarily be harder to replace the production of a great catcher. Until WAR does a better job recognizing how difficult the production of a great catcher is to replace, the position will be undervalued.
So what should do when you look at WAR, especially for catchers? First, recognize how little we can quantify catching defense. Until that is more accurately measured, the fielding component of WAR is borderline useless when comparing other positions to catchers. Second, adjust for position in your head. Looking at the list, I think Smith does a good job adjusting for most positions. His list recognizes the added value of a great shortstop like Honus Wagner over a more replaceable great left fielder like Rickey Henderson. The best catcher ever at #52? That seems like a positional adjustment that needs some tweaking. Finally, replacement level needs to be carefully pegged to each position. In the case of scarce positions like catcher and shortstop, replacement level needs to be lower than it is for more common positions like left field and first base. WAR already does this, but the gap needs to be accentuated. I still like WAR, and I consider it one of the best comprehensive stats around, as you can see from my frequent use of it on this site. Nonetheless, catchers sit in WAR’s blindspot at the moment, and until that problem can be adequately corrected WAR will not be quite as comprehensive at its aims to be.