Robot MLB Umpires are Coming! Whether you like it or not
Has Tim Anderson and his bat flipping had you upset all year? If so, I’d advise clicking out of this article now. As times have changed, so has baseball. Like all things, it has to adapt and evolve to stay relevant. Hell, even Elton John was recently seen leaving the studio with Future. To the MLB’s credit, they have made some moves to modernize the game like implementing instant replay and trying to cut down on the length of games. Whether or not instant replay adds to the game is a different debate entirely, but there’s no doubt it was the MLB heeding to something that was in high demand for some time. Last season, umpires missed 34,294 ball/strike calls according to data collected by Boston University Master Lecturer Mark T. Williams and his graduate students.
34,294. That’s the number. In a way, it’s sort of like trying to envision what exactly a billion of anything looks like, as it is very hard to comprehend. To give some reference, Williams’ study, conducted over 11 seasons of Major League Baseball, points out that one in every five calls made by an umpire in a given game is incorrect in some way. On top of this, umpires demonstrate a definitive bias when making calls with two strikes. It’s not their fault, of course, as humans experience biases and lapses in judgment all the time. There isn’t a single person on the planet that could step behind the plate and deliver 100% of calls correctly. The point isn’t to pick out certain umpires and demand they practice so as they make more right calls. Rather, there shouldn’t be a human behind the plate relying solely on his own intuition to make ball and strike calls. Instead, we need robots.
Robot Umpires
It sounds kind of dystopian, doesn’t it? Yes, I’ve seen iRobot. In fact, generally speaking, I’m not a fan of robots being used in everyday life and slowly being implemented into the workforce. I feel this gives credibility to my claim, though, as even though I have this bias about robots I’m still able to see that they have a place in baseball. I’ll even say that when I first started hearing murmurs that robotic umpires would eventually make their way into baseball, I was terrified. I was staunchly opposed to the idea. That is, until I saw this video.
It’s one thing to be acutely aware of the terrible ball/strike calls that go on each and every day in the MLB, but to see a slew of absolutely atrocious calls like that in a row really puts things into perspective. One of the calls that was hardest to watch in that video happened during a Rangers vs. Rays game on April 9, 2013. Here it is in all its glory in a separate video.
Yes, it is what it looked like. The Rangers’ Joe Nathan threw a ball outside and well below Zobrist’s knees, but home plate umpire Scott Barry decided to call it a strike and send the Rays home packing. If you continue watching, you can even see Nathan mouthing “WOW” as he walks off the mound, in disbelief. Sure, this is an extreme example with the outcome of the game being directly and immediately affected, but I’d argue that every missed ball/strike call is just as detrimental to the players on the field.
HBO’s Real Sports with Bryant Gumbel did a segment on the idea of robot umpires and raised a few unique points from different perspectives. In this first clip, former pro (and current MLB analyst) Eric Byrnes is being interviewed about the matter. As a hitter, he feels that when a catcher takes a pitch that is in and of itself a ball and frames it to be called a strike, that’s essentially “stealing” a ball that he as a batter earned. Watch the one-minute clip below to see the true passion and raw emotion he feels as he raises his point.
There’s not really a number that you can put on it, but there’s no doubt in my mind that enough of these bad calls amassed against a single batter has made or broken his career in the sense that he may be sent down to the minors before he gets a chance to flourish, etc. Of course the counter argument would be that hitters need to learn to adapt on the fly, whatever the situation is. The question I would raise, though, is: are hitters supposed to adapt to opposing pitchers or umpires? Of course each pitcher is going to bring something new and challenging to the table, but when you get a pitch that’s called a strike two ball lengths off the plate, you as a hitter now have to have in the back of your mind that the strike zone is exponentially bigger than what you’ve been taught your whole life.
In another clip from Real Sports, Yale professor Toby Moskowitz, who had analyzed all the pitch data from the last few seasons preceding the episode’s air date in 2016, is interviewed.
He touches on just how much the umpire’s ball/strike calls change based on the count. When you’re looking at the same pitch, he says, with a 3-0 count and an 0-2 count, the umpire will call it a strike correctly 89% of the time with the count at 3-0, but only 59% of the time with an 0-2 count. Based on the presumption that we’re fine with the fact that umpires only get the pitch call right 89% of the time on a 3-0 count (which I’d argue shouldn’t be acceptable), there’s a 30% drop-off rate in accuracy if the count is 0-2. To reiterate, this is on the exact same pitch. Moskowitz attributes it to the “omission bias,” which is to say that an umpire doesn’t want to determine the fate of the hitter, so he’d rather see the count go to 3-1 or 1-2. This all happens on a subconscious level, though, so there really isn’t any training that can be done to prevent it.
In a game so dominated by statistics and saber-metrics, it’s a wonder that the MLB isn’t already tapping into this technology that is available for the taking. If you can improve the accuracy of a game, and if the MLB is increasingly concerned with getting game times down, why not implement technology that will ensure those two things?
