Calling Balls and Strikes: Why Computers Will Never Replace MLB Umpires

Use your ← → (arrow) keys to browse more stories
Calling Balls and Strikes: Why Computers Will Never Replace MLB Umpires
Doug Pensinger/Getty Images

After what appeared to be a decisive Cardinals victory during Game 7 of the 2011 World Series, Texas fans took to the Internet by storm to complain about the most elementary of issues—that an umpire's bad call lost their team a game.

Not the epic Game 6 collapse. Not poor pitching. Not an inability to hit. Not ill-advised roster moves. It's always easier to blame someone else.

On the other hand, I unilaterally attempted to absolve the umpires of sole responsibility for the Rangers loss—even while conversely pointing out that 14 of the umpire's 17 missed ball/strike calls had, in fact, benefited the Cardinals.

Yes, one of those 14 missed calls in St. Louis was that 3-2 ball call on an outside corner cutter to Yadier Molina which extended a one run Cardinals lead and prolonged an inning that should have ended.

To support the 14-of-17 statistic, I referenced Sportvision's Pitch f/x system, MLB-supported technology which records pitch location with an accuracy of within .5 inches.

Pitch f/x is not to be confused with MLB's Zone Evaluation system, which introduces post-game processing algorithms to evaluate umpire performances.

Instead, the Pitch f/x formula I referenced is used by the Umpire Ejection Fantasy League and employs Pitch f/x's maximum margin of error (one inch) to create a two-inch wide, 100% confidence interval for every pitch's horizontal location.

A similar 100% interval generating technique is employed for the variable vertical plot, while data distributed directly by MLB is used to calculate all variables and intervals.

How should balls and strikes be called in MLB?

Submit Vote vote to see results


The fantasy league judges pitches as correct, incorrect, or borderline based on these confidence intervals. If an interval falls completely within the strike zone, the pitch must be a strike and if it falls completely outside of the strike zone, the pitch must be a ball. If the interval captures only part, but not all of the strike zone, the pitch is considered borderline.

All borderline pitches result in the umpire's call being adjudged as correct. In other words, the umpire gets the benefit of the doubt—big time. A pitch can theoretically be located one inch off the plate, yet the call of "strike" can still be considered borderline, which makes it correct.

Further calculations are made in regards to breaking pitches which may or may have not captured the back part of home plate, though most of these pitches also end up being considered borderline and therefore, correct.

Yet in spite of the premise that equates borderline with correct, 17 pitches during the final MLB contest of 2011 were considered incorrect, and 14 of these benefited St. Louis.

Molina's pitch was centered approximately 2.7 inches from the edge of the outside corner, well outside Pitch f/x's margin of error.

Baseball is indeed a game of inches.

Some Cardinals fans denied the 14-of-17 claim, even denying the verified Molina miss. They referenced the logical argument of consistency and attempted to use Pitch f/x graphics to support the their theory in a "visualize this" sort of way.

Harry How/Getty Images
I'd like to see a manager try to argue with a computer

Some Rangers fans rejected the notion of absolving the umpire of blame. They accepted 14-of-17 and picked apart their opposition's theory of consistency, attempting to use Pitch f/x graphics to prove that the umpire was not consistent and irrevocably damaged their team.

No one seemed to agree about anything that had to do with balls and strikes, even with data clearly corroborating the 14-of-17 claim. Texas fans used Pitch f/x to criticize the umpire, while Cardinals fans used the technology to commend him.

Such an exchange perfectly illustrates why MLB will never hire robotic home plate umpires to call balls and strikes.

Make no mistake, the technology is there and it is fairly foolproof. The problem is not with the machines. Hawk-eye review in tennis proves that.

The problem lies within the culture of the game, that no one seems to agree with the computer 100% of the time, even when raw data overwhelmingly proves that a pitch is unequivocally a strike.

That includes some umpires, players, and coaches. Imagine a player or coach arguing with a computerized umpire. Now imagine a player or coach being ejected by a robotic umpire.

But umps, players and coaches aren't the only computer nay-sayers. Also opposed to a technological overhaul of baseball are many fans and most readily apparent, broadcasters.

How many times have we heard broadcasters say "I don't care what the box says, that pitch was..." or "FoxTrax says ..., but I disagree"? A computer behind the plate takes out a necessary clash within the game.

Could this be the new way of umpiring? Image courtesy /

Computers can't judge how a catcher positions himself to catch a pitch, nor can it sense a rowdy pitcher who misses his spots so badly, it becomes a safety hazard.

It is a philosophical given that—with a catcher set up on the outside corner—a wild pitch that barely nicks the inside corner and almost hits the batter will be called a ball.

Umpires police the game in this way to maintain safety and minimize danger. It tells the pitcher, "once you get yourself under control, you'll get strike calls." It is a positive element of the game and something a computer could never do.

There are so many intangibles that go into officiating—erraticism, working the zone, judging an attempt to strike the ball—a computer cannot account for those.

Computers can only definitively state whether a pitch has entered the rules book strike zone. That's it.

Baseball purists oppose computer precision and counter with umpire consistency.

Progressive types point out that computers are also consistent—extremely consistent.

Baseball purists refer to "the human element." This sometimes refers to an umpire's judgment in ruling interference over obstruction or ejecting a pitcher for the intentional bean-ball. I'd like to see a computer eject a pitcher for throwing at a batter's head.

But more often, when a traditionalist refers to "the human element," it implies a distaste for technology. It insinuates all traits which humans possess that a computer can never learn.

Jamie Squire/Getty Images
MLB Commissioner Bud Selig has long been opposed to expanding Instant Replay

During a time in which every major sport employs instant replay in some capacity, none use the technology as a substitute for the officials' judgment in determining fouls or penalties.

Even when reviewing a catch/incompletion call in football, the officials use technology only to determine the greater issue of possession. Replays never generate pass interference penalties.

When reviewing a who-touched-it-last out of bounds call in basketball, replays only dictate the last player to have touched the ball. They never result in a pushing foul being called to supersede the official's call of "no foul."

It is the same in baseball, a sport with no true penalty calls to make.

Balls and strikes are baseball's equivalent of determining marginal contact in basketball. Sure a computer can conclusively produce statistics to indicate a call's quality of correctness, yet it shouldn't trump an umpire's judgment.

I once explained Pitch f/x, computers, and baseball science to an experienced umpire. He kindly translated my words into baseball vernacular—horse###!

Intangibility always trumps technology.

Load More Stories

Follow B/R on Facebook

Out of Bounds


Subscribe Now

We will never share your email address

Thanks for signing up.