clock menu more-arrow no yes mobile

Filed under:

I am SO READY for Robo-Umps

Baseball has the technology to call balls and strikes correctly. Let’s use it!

MLB: World Series-Houston Astros at Washington Nationals Geoff Burke-USA TODAY Sports

Imagine a sport like, oh, the long jump. This is the track and field event where participants simply try to jump as far as possible. Then officials measure the distance from the starting point to the landing point with some sort of tape measure. Whoever has jumped the farthest, based on that measurement, wins.

Now imagine the same sport, but without the measuring tape. Officials watch the jump, see where the person landed, and then use the eye test to see who jumped the farthest. Is this jumper’s mark a centimeter farther than that jumper’s mark? Yeah, I guess, looks that way. OK, based on the official’s eye test we’ll give this jumper the gold medal.

It seems crazy that a sport would eschew available technology that could clearly measure important outcomes, doesn’t it?

So now let’s turn to baseball.

Baseball should embrace technology that makes the boundary conditions of the game more consistent

Using technology to determine balls and strikes is, of course, not as simple or obvious as using a tape measure for distance. But apparently baseball has developed computer applications that can do this pretty well. This “TrackMan” system is already in use behind the scenes at the major league level to evaluate umpires, and it enables baseball to collect the kind of data we see when we look at exit velocities and route efficiency.

If MLB is already using some form of “robo ump” technology to evaluate umpire strike- calling performance, why not use it to call balls and strikes?

That may indeed be the what’s in store, as MLB is currently testing TrackMan technology in the independent Atlantic League as well as in the Arizona Fall League. But you don’t have to look far to find resistance, with media commentators as well as players voicing concern.

Critics of an automated zone draw on emotional (and hence not very effective) arguments

Interestingly, few of the complaints focus on the accuracy or efficacy of the technology. There are some concerns that weather or altitude can throw off calibrations, but these are correctable problems, and the current trials outside of MLB and MiLB games are intended to work out those kinks. There was some note taken of an AFL player complaining about a robo-ump-called strike (see for example Evan Longoria’s reaction), but a review of the pitch suggests that the computer got it right.

Mostly there are fuzzy ruminations about the importance of “the human element” and a lot of nostalgia about great strike-zone inspired fights with umpires. A Washington Post column, for example, seems to imply that automated strike zones would eliminate interpersonal exchanges of all sorts, as though eliminating ball and strike disputes would mean there was nothing left to talk about in baseball:

“It’s another thing that takes away from the beauty of the game because it takes away from the conversation of the game,” Falsetti said.

A San Diego Union-Tribune article concludes that we don’t need technology, we just need Manfred to tell umpires to do better. It also includes this plea from Clayton Kershaw:

A lot of people have jobs and get paid as catchers because of how they present the baseball, and you’re basically taking that art form away. The catching position changes entirely at this point. Pitch framing is a huge deal. That’s taking jobs away.”

But automated systems would not cost a single job.

Home plate umpires would still be needed to call hit by pitches, foul tips, interference, and outs at home. And while those catchers who live and die with their framing might lose their roster spots, there are likely some offense-first catchers, or guys with amazing arms, who become more valuable.

Presumably MLB, when they do decide to move to this system, will phase it in gradually across the minor leagues so that a cohort of players will develop with that technology and the skillset adjustment won’t be sudden. Hitters, pitchers and catchers will surely need to make some adjustments, but players are doing this all the time.

Here’s a sensible way to look at this issue

Because many in baseball media have been so critical of this development, I was glad to see that Al Yellon took up this topic on Bleed Cubbie Blue, our sister site out of Chicago. Yellon provides a review of the Arizona Fall League experiment, summarizes some of the common arguments around this topic, and concludes with this spot-on observation:

To those who want “the human element” regarding ball-and-strike calls, I would argue that the “human element” ought to be what the players actually do, not what one guy standing behind home plate thinks the players did. Let’s get the calls right, or at least as many as it’s... humanly possible to get right.

Yellon has a proposal of his own: he suggests allowing each team one or two challenges each game, where umpires would consult the automated tracking system on a disputed call. Imagine the system in place for Grand Slam tennis matches when players can challenge a line call and the line judges consult the computer images to verify the call — it takes maybe 30 seconds from start to finish. If baseball finds fully automating the strike zone to be too great a change, Yellon’s suggestion could at least provide the means to eliminate those one or two truly egregious miscalls that can change an entire game.

Homeplate umpires do an amazing job. But that’s not good enough.

Umpires aren’t terrible at what they do. MLB estimates that they get it right 97% of the time. That means that a typical major league game, which may see 200 to 250 pitches, will have 6 - 8 that are clearly miscalled. While that’s not a lot statistically, if any one of those six pitches ends a rally or forces in a run, that small rate of error could have big consequences.

Independent researchers, moreover, think that umpires are missing a lot more than 3% of the time. This in-depth Boston University study, well worth a read, concludes that umpires are missing more like 13-14 calls per game; that they have sometimes predictable blind spots; and that the league does too little to weed out bad performance. They think a further embrace of technology is the only reasonable correction.

A few decades ago we accepted the imperfections of umpire-controlled strike zones because there was no other option. But now, as experiments using automated strike zones appear to be going smoothly, there is an alternative, and I hope it is adopted.

If perfection is possible, why not pursue it?