The Error of the Reached on Error

A hitter slaps a ground ball between second and third and it rolls just past the diving shortstop’s (let’s call him “Jerek Deter”) glove into left field. IT’S A HIT! In the official statistics, the hitter gets an at-bat (and thus also a plate appearance) and a hit. The hitter’s batting average and on-base percentage have increased.

Now imagine the exact same action by the hitter, the swing, the path, speed, spin on the ball, and speed to first base. This time, however, the defender just gets to the ball, but fumbles around and the hitter takes first base. In the official statistics, the hitter gets an at-bat (and thus also a plate appearance)… but no hit. The hitter’s batting average and on-base percentage have just decreased. Makes sense, yes? Uh, no.

People have complained about silliness of the how errors are treated, particularly with respect to how it has resulted in an incredibly messy pitching stat in ERA that has come to prevail over a much more logical runs-allowed (RA) scale for pitching metrics. The subjective nature of the scorekeeper’s decision on whether certain balls are errors is perhaps best highlighted by the fact that players have called the press box and had plays scored as errors overturned and called hits. But the problem I’m concerned with isn’t scorekeepers, it’s the rules for scorekeeping and how official baseball statistics represent events. While most writing on this topic focuses (understandably) on the effect of errors on the earned run rule and on the distortions intrinsic to ERA, my concern here is how reached on error is a fly in the ointment of offensive statistics as well.

Part of the issue is the idea that ruling a play a “reached on error” puts all of the onus on the fielder… yet it somehow also counts (in batting average and on-base percentage) against the hitter. Why should the hitter be “punished,” especially since on most errors he has at least made some contribution (e.g., putting the ball into play, running hard to beat the throw, etc.) to it not being an out? I do think the way reached on error is recorded is unfair in that way. There are differences within the sabermetric community about how much skill is involved in reached on errors and whether it should be included in metrics like wOBA (as in this discussion). That is yet another discussion. For now, I’m concerned with the accuracy problem.

The problem is that, because of the way “official” MLB statistics are recorded, we have plate appearances (in hitters’ records, at least) that are “blank spaces” that actually distort the record of what has happened. This is most obvious by looking at on-base percentage, which might be better thought of as “not-out percentage.” Hits, walks, and hit-by-pitches count as instances of getting on base. Reached on errors count in the denominator as at-bats (and thus plate appearances), but not in the numerator, even though no out occurred on the play (another argument lurks in the denominator about sacrifice hits [bunts] as opposed to sacrifice flies).

My personal perference would be take the most simple path and call what is now a “reached on error” a hit. But if that doesn’t seem appropriate it could also be treated as a separate category like hit by pitch. That is another event in which the hitter seemingly (in a non-technical, non-sabermetric sense) plays an even “more passive” role and is still included in the official statistics. Even if reached on error doesn’t have the hitter skill component some would like for sabermetric stats and projections, that doesn’t mean it shouldn’t be given a different treatment that it currently is in official statistics. Intentional walks are somewhat analogous: while they are somewhat related to player skill (they are generally given to hitters who are thought to be too likely to do damage in a particularly situation), the player plays an apparently passive role; furthermore, like a reached on error, someone else’s subjective judgment of an objective situation is involved (in this case, the manager rather than the scorekeeper).

Intentional walks also provide a historical precedent against “we haven’t done it this way before, what about history?” sorts of arguments: prior to the mid-1950s, they weren’t recorded separately from walks, yet they have been added to official statistics. Hit by pitch wasn’t recorded earlier, and neither were caught stealings. If it makes our records of what a player has done more logical and accurate, what is the objection? Reached on errors don’t occur in the overall numbers and with the regularity of hit by pitches and caught stealings. I realize that I’m picking nits here. But this is a blog about baseball, what did you expect?