In tepid defense of umpires

Umpires are remarkably good at their jobs. Trust me on this one.

If you hear about one really bad blown call in a week, it’s been a bad week, and there are thousands of plays, tens of thousands of pitches every seven days. But we only really notice the umpire when a mistake is made, and when we notice a mistake we get a sense of injustice. We want to fix things, and MLB doesn’t fix things.

As technology advances, it’s clear that umpires are increasingly obsolete. Sure, pitchf/x, the hardware and software system that tracks pitches from the point they leave the pitcher’s hand through the strikezone to the catcher’s mitt, isn’t 100% accurate yet, but it’s darn close. And instant replay gives us the chance to make field calls with close to 100% accuracy.

 Jeff Sullivan wrote an interesting op-ed piece for Fangraps (the best site on the Internet) about whether bad calls are bad for baseball. His controversial position: Well, they might not be so bad. After all, they create memorable controversies, they make people watch and care, and they reinforce the game’s human element. The case study in egregiously bad calls creating great press is the infamous Armando Galarraga / Jim Joyce affair. A perfect game was ruined, but the game is all the more memorable for that. Phillip Humber threw a perfect game, and nobody knows who he is (or, they wouldn’t, if the Atros weren’t so bad they needed him). Galarraga isn’t a Big League-caliber pitcher, either, and we would have long ago forgotten that game as a fluke if not for the controversy. So there’s something to the idea that blown calls can make the game somehow more interesting to us.

At the same time, that’s a cheap analysis. Recent tragic events have turned ordinary school days and trips to the movies into something more memorable, but for the wrong reasons. Baseball is a much less serious instance of the same effect. Perhaps the most famous World Series of all time happened in 1919—and it isn’t because of great baseball that we remember.

I’m an old-fashioned guy. I like tradition, in baseball and elsewhere. I kind of like umpires, but I’m souring on the whole idea. I would embrace an automated system at this point.

However, it’s interesting that as baseball wisdom moves toward quantitative analysis it has grown less forgiving of umpires. In fact, randomness is more apparently part of the game today than ever before, and that’s exactly what umpires introduce.

Voros McCracken (read the article, it’s a terrific piece) revolutionized the baseball world when he proposed the idea that developed into what we call today ‘batting average on balls in play,’ or BABIP. Basically, it seems to us today that when a batter hits a baseball, his coordination and skill might help him make good contact, his strength might help him hit the ball hard, and he has some limited control over which part of the field he’ll target, but on the whole his ability to direct the baseball to a certain point is virtually non-existent. In other words, all a batter can really do is make good contact and pray that the ball doesn’t go right to the shortstop, or left fielder, or whatever. Contemplating this proposal, a batter who hits a ball well, but right to the right fielder, might have had just as good an-bat as someone who hits a double to the gap. He just got unlucky, as it were.

This seems like it can’t be right, but there’s some proven truth to it. Think about a baseball like a coin. We’re all agreed that the result of a coin flip is random—50/50. Of course, that’s not true. A coin is a physical object subject to the rules of physics. We could calculate how a coin will fall, based on the force of the flip, air resistance, and any number of other factors. We could build a robot that flips a coin the same way every time, getting the same result. But all of those factors are so complicated and so difficult to control that when we flip a coin, the result is, for our intents and purposes, random. The same is true with baseball. The quality of the batter’s contact, the humidity, the speed of the pitch, the type of pitch, the playing surface, the center fielder’s mood, and any number of other things can affect whether a ball turns into a hit or an out. It’s easiest just to reduce all of those incomprehensible factors to a single number. For a coin, it’s 50%. For BABIP, it’s usually somewhere around 35%.

BABIP measures how often the balls a batter hits between the lines (that aren’t home runs) actually result in base hits. Being a fast runner helps. Making good contact helps. But in the end, people casually and somewhat accurately consider BABIP equivalent to “luck.” In the fantasy universe, a player with a career BABIP of .350 whose BABIP currently sits at .100 is probably getting unlucky. Conversely, every season some unheralded player will break out of the gate with great numbers. Even if his OBP is .400 after a month, if his BABIP is .600, you can bet he’ll be regressing. Pitchers’ BABIP-against is also widely consulted.

This has been a long tangent. Weren’t we talking about umpiring!? Well, I propose that much of what umpires get wrong is just an extension of the randomness that happens every day, anyway.

Take, for example, the blown call at first that was the case study for Sullivan’s article (above). It was a close play, although the call was clearly wrong. The wisdom behind BABIP tells us that whether the ball was fielded in playable range or not is somewhat arbitrary. Longoria hit that ball almost right up the middle. Had the ball gone maybe 6 inches closer to the first base line, Longoria would have reached base safely. That the ball instead rolled through the shortstop’s range was totally arbitrary. In order for a batter to have a chance to be the victim of a missed call in the field, he has to be the beneficiary of some luck in the first place.

So really, what umpires do is just add a little randomness. They supplement the cosmic randomness that governs all baseball. It’s problematic to us because of the preventability of it all, of course. We could fix it! And we probably should. But really, we overestimate the impact umpires have. Armando Galarraga was the beneficiary of lots of good luck before he had some really bad luck on the last out of that game. We don’t consider perfect games destroyed when a line drive happens to fly right at the second basemen.

My final analogy returns to the coin flip. Suppose I challenged you to flip a coin ten times and get heads each time. Obviously, the odds of you doing that are pretty long—a lot like throwing a perfect game. Now suppose you succeed and flip ten consecutive heads, but on the tenth I mistakenly call it a tails, and determine that you’ve lost the challenge.

It’d be unfair, sure. It’d be outrageous because it’s so correctible and demonstrably wrong. But really, the odds of that tenth flipped coin actually coming up tails were pretty good anyway, and you had to get outrageously lucky just to get to that point. If there’s a miniscule chance that I mess up a call, that element of randomness will be significantly less influential on the number of heads and tails that you flip than the coin flips themselves.

So maybe we should relax a bit about umpires. Maybe the human management they bring to the game is worthwhile for its little nuances—the satisfying punch out gestures, the hilarious ejections, the friendly rapport with the catcher, and the occasional bickering with players. Personally, I think I’ll take it.