Stats Prove MLB Umpires Call 99.5 Percent of Plays Correctly

Gil Imber@RefereeOrganistAnalyst IIOctober 26, 2011

ARLINGTON, TX - OCTOBER 22:  Mike Napoli #25 of the Texas Rangers reacts after first base umpire Ron Kulpa calls Matt Holliday #7 of the St. Louis Cardinals safe after a tag in the fourth inning during Game Three of the MLB World Series at Rangers Ballpark in Arlington on October 22, 2011 in Arlington, Texas.  (Photo by Ezra Shaw/Getty Images)
Ezra Shaw/Getty Images

There is a saying among sports officials in baseball, basketball, and any other sport with two teams: "Every time we make a close call, we earn a temporary friend and a permanent enemy."

Before the advent of television replay, umpires were questioned but never objectively criticized. Before super slow-mo, a close call was always right if it benefited one's own team and wrong if it didn't.

Before technology came into the picture, second guessing an umpire was purely subjective.

These days, every close play is reviewed multiple times, in multiple speeds, across multiple angles. In the World Series, non-umpires by the names of Buck, McCarver, and even La Russa criticize the state of umpiring on national television.

During Game 1, Fox turned to an infrared camera slowed to display each and every one of its 100 frames per second to demonstrate a missed call.

In Game 3, Fox identified the second missed call of the series using traditional replay technology.

As a result, umpire Ron Kulpa was very harshly criticized for blowing an "obvious" call—a pool reporter even asked if his being from St. Louis had anything to do with it.

Coincidentally, both missed calls did benefit the Cardinals. Indeed, the Texas Rangers lost both Games 1 and 3, though not at all because of the missed calls.

Meet instant replay, a process that takes significantly longer than the original play to complete.

Had MLB employed expanded instant replay during this World Series, no more than two calls might have been overturned as a result.

Considering 27 outs per game per team, there have been at least 270 calls made in this year's World Series. That is a minimum of 270 calls of which the umpires missed two—a missed call every 135 calls.

That equates to an accuracy of 99.3 percent, or one missed call every two-and-a-half games. We'll round that to three games due to our low-balling of calls made per game.

Officials have another saying when it comes to statistics: "100 percent of the time, we upset at least 50 percent of the crowd."

To further demonstrate the issue of umpire accuracy, we turn to a curiously phrased study in MLB umpiring.

ESPN previously touted that MLB umpires miss an egregious 20 percent of close calls.

But ESPN didn't especially define what missing 20 percent means.

Considering our previous model of 54 outs per game equating to a minimum of 54 calls per game, we look more closely at ESPN's data.

ESPN found 47 incorrect calls over the course of 184 games, or one missed call every four games.

With 184 games examined, ESPN analyzed 9,936 calls (54 minimum calls per game times 184 games).

If, out of 9,936 calls, 47 were deemed incorrect, that means umpires missed approximately 0.47 percent of all calls thrown their way.

That comes out to approximately 99.5 percent of calls that were not missed.

Had expanded replay been in place, umpire Jim Joyce and pitcher Armando Gallaraga may have never had the chance to share their tremendous sportsmanship with the world.
Had expanded replay been in place, umpire Jim Joyce and pitcher Armando Gallaraga may have never had the chance to share their tremendous sportsmanship with the world.Kevin Winter/Getty Images

Welcome to the wonderful world of statistics, where ESPN can take a study showing MLB umpires miss less than half of one percent of calls and turn it into a headline accusing umps of missing 20 percent of calls that presumably shouldn't be missed.

When Rangers catcher Mike Napoli was asked about the controversial call in Game 3, he was adamant he had tagged his runner before the runner arrived at first base.

Napoli, of course, was correct.

However, when asked about instant replay during a follow-up question, Napoli stated he was unequivocally opposed to expanding instant replay to determine tag plays.

Proponents for keeping replay out of the more routine plays of baseball often cite a desire to "preserve the human element."

The "human element" refers to umpires' calls which sparingly might be incorrect, but which always must be final.

So how does one go about quantifying the human element?

Simple. We've already done that. If the human element refers to an incorrect call, ESPN has already determined this occurs once every four games.

We're looking at a 25 percent chance of a missed call occurring in any given game, and an even lesser chance that the missed call will be a game-changer.

For those keeping score, the last missed call that was a game changer occurred on July 26, 2011. Umpire Jerry Meals incorrectly ruled Braves runner Julio Lugo safe at home plate to break a tie score in the bottom of the 19th inning of a Pirates-Braves game.

By that point, the Pirates-Braves game had persevered through a minimum of 111 correct calls over the course of six hours, 38 minutes.

Meals's miss in the sixth hour, 39th minute cemented his crew's accuracy at no less than 99.1 percent for the entire contest.

How would expanded instant replay impact the classic umpire-manager argument?
How would expanded instant replay impact the classic umpire-manager argument?Bob Levey/Getty Images

In the end, the "human element" purists and replay advocates are arguing over less than one half of one percent of all calls made.

Reading this article, the baseball purists will say umpires are so close to perfect, we need that tiniest of imperfections to remind us that umpires are human.

Meanwhile, the replay advocates will say disruptions to the game will be minimal, since only 0.5 percent of all calls would be overturned.

ESPN's study showed that only 1.3 calls per game were deemed close enough to warrant instant replay review. For replay advocates, this proves games will be minimally affected if only one or two calls per game will be reviewed.

Meanwhile, the purists will counter with the fact that only 20 percent of these reviewed calls would be overturned: why stop play when an overturn will only occur in one out of every four games?

It is a philosophical debate that has subjectively made its way into player union meetings, owners' meetings and into the office of Commissioner Bud Selig.

At least now, thanks to statistics, we can insert some objectivity into the debate.

Whether or not MLB institutes expanded instant replay to raise 99.5 percent to 100 percent, the highest of standards in umpiring is extremely real.

As Hall of Fame umpire Nestor Chylak once said, "They expect an umpire to be perfect on Opening Day and to improve as the season goes on."

By "they," Chylak referred to fans, players, coaches, the media...everyone.

July's 19-inning Pirates-Braves contest may never had ended without umpire Jerry Meal's missed call at the plate
July's 19-inning Pirates-Braves contest may never had ended without umpire Jerry Meal's missed call at the plateOtto Greule Jr/Getty Images

With expanded instant replay, many are hoping Chylak's quote will become the new reality while others are hoping to leave the call-making to the men in blue.

As a footnote to this piece, you may be wondering why I left out ball/strike calls, in which umpires' 99.5 percent accuracy dips to the 95 percent level.

Plate umpires' ball/strike calls are graded by the computerized Pitch f/x tracking system.

The reason is simple—many people simply cannot bring themselves to accept Pitch f/x's judgement.

Throughout these playoffs, the television commentators who proudly replayed close pitches using the famous FoxTrax box could not bring themselves to agree with the computer's determination.

From Terry Francona to Buck, McCarver and everybody in between, the prevailing attitude in regards to Pitch f/x was best summed up by Francona during Game 1 of the ALCS.

Francona disagreed with a plate umpire's call of "ball" on a 2-2 fastball located near the outer edge of home plate. As Francona bemoaned the "missed call," FoxTrax flashed on the screen, conclusively showing the pitch was located several inches off the plate and had correctly been ruled ball three.

Instead of eating crow, Francona brazenly remarked, "I don't care what FoxTrax says. That was a good pitch."

Even if replay does expand to include safe/out and fair/foul calls, computers cannot be used to call a strike zone.

No one seems to agree with the computer's judgement.

Calling balls and strikes via computer would require fans and broadcasters to accept the computer as an accurate arbiter
Calling balls and strikes via computer would require fans and broadcasters to accept the computer as an accurate arbiterDoug Pensinger/Getty Images

After the Cardinals' loss in Game 4 of the 2011 World Series, St. Louis fans took to the blogosphere to complain about plate umpire Kulpa's allegedly unfair strike zone.

The claim?

Kulpa was atoning for his Game 3 mistake by calling an unfair pro-Texas zone. However, contrary to St. Louis fans' claims, Kulpa called a Game 3 zone with 95.6 percent accuracy.

He missed all of seven pitches, including two against St. Louis and five against Texas.

If anything, Kulpa's Game 4 zone was pro-Cardinals, not pro-Rangers, proving once and for all that sour grapes from losing fans carry very little logic and objective weight.

In a recent article for Bleacher Report, I asked who should call balls and strikes in Major League Baseball—a computer or a human.

With 127 votes in, the answer is clear. The umpire wins, 93 percent to seven.


The latest in the sports world, emailed daily.