The Outside Cornerby David Gassko
February 01, 2007
A short history of umpiring in baseball: When the first professional baseball league, the National Association, formed in 1871, umpires were still volunteers, chosen by the home team from a list of five names submitted by the visiting club. Umpiring became a paid profession only in 1878, when the National League decided to pay them $5 per game, and a year later, NL president William Hulbert formed the first umpiring staff, which consisted of 20 men for six teams.
But Hulbert was outdone in 1882 by H.D. McKnight, the president of the newly formed American Association. While the National League was reeling from a game-fixing scandal involving one of its umpires, the American Association formed the fist truly professional group of umpires, who were paid $140 a month and an additional $3 for every day they were on the road. The next year, the National Association adopted that system, and the modern professional umpire was born.
But for these early umpires, the ground was shaky and their job often dangerous. Fans jeered and mocked them, hurled insults and threats after calls they did not agree with, and occasionally got involved in physical altercations with umpires, throwing both objects and punches. Owners saw potential profit in this WWE-like atmosphere and often encouraged such behavior, sometimes even engaging in it themselves.
As baseball became professionalized, so did the umpiring system. Paltry umpire salaries were increased, league backing for umpires increased and fans backed down, and a two-umpire system took place, all near the turn of the century. By the 1930s, three-man crews were the norm, and in 1952, the modern four-man system model took place.
In 1935, National League umpire George Barr started the first umpiring school; in 1939, Bill McGowan of the American League opened the second. By the 1960s, it was almost impossible to become an umpire without having attended an umpiring school first. Today, umpires must graduate from either The Jim Evans Umpire Academy or The Harry Wendelstedt Umpire School to work in professional baseball.
The changes resulted in more consistent and better informed umpiring decisions. But as umpires became more professional on the field, they wanted to cement their professional status off the field as well. After more than two decades of struggles against Major League Baseball, the umpires finally unionized as the Major League Umpires Association in 1968, though it would be two years before the MLB would recognize the union, after a one-game strike on the first day of the 1970 playoffs. Subsequent strikes and strike threats promoted a rapid increase in umpire salaries—today, major league umpires make $100,000 to $300,000 depending on their experience.
In short, over time, the quality and consistency of umpiring in the major leagues has gone up, as strike zones have standardized, hometown biases have been removed, and training and experience requirements have greatly increased. But just how professional are modern umpires? Are they perfect? Of course not.
A Ball or a Strike
The first question that comes to mind is, how much do umpires differ in their strike zones? We all know that there are some umpires that will give the pitcher anything that’s close to the outside corner of the plate, some that will squeeze them on anything above the belt, etc. But what kind of an effect do umpires truly have; just how inconsistent are they?
One method we can use to answer this question is to look at spread in the percentage of pitches each umpire has called a strike, and compare that to the spread in the data we would expect from random chance alone. If we look at the last three years, we can mitigate the impact of variables we’re not accounting for, such as the quality of the pitcher on the mound or the lineup he’s facing. With three years of data, we have an average of almost 6,000 plate appearances for each umpire, which should allow those variables to even out for the most part.
So what happens when we run the test? We find that there is a definite effect, but it’s not that great. Regressed to the mean, the most extreme umpires are within 2% of the league average in terms of the percentage of pitches they call a strike. I estimate that this spread equates to a difference of about .20 runs a game between the most and least strike-happy umpires in baseball. It’s something, but considering that we’re looking at the extremes, it really isn’t very much at all.
Speaking of extremes, which major league umpires have the most extreme strike zones? Over the past three years, Doug Eddings, Bill Miller, and Bill Welke have been the most pitcher-friendly umpires, while Tim McClelland, Greg Gibson, and Jerry Layne have had the stingiest strike zones. It’s interesting that I immediately recognized four of those names, though I don’t know the names of most of the 95 umpires who have worked between 2004 and 2006. Eddings especially is an outlier, and his performance in the 2005 playoffs provides further evidence that perhaps he isn’t cut out for major league work.
But overall, we see that the professionalizing of the field has worked, as umpires have adopted relatively similar strike zones, and shown that no matter who is behind the plate, a strike will be called a strike and ball will be called a ball … most of the time.
Before the 2002 season, Major League Baseball installed a four-camera system produced by QuesTec and known as the Umpire Information System (UIS) in ten major league parks. The cameras record the position of every pitch, and whether or not it was a strike. They then compare that result to the umpire’s call, effectively serving as a check on personal strike zones that do not dovetail with the rulebook. Umpires who record particularly bad results may be disciplined or fired.
So the easy question to ask is, does it work? Well, if we take all umpire data from 1999-2001, and compare the spread in it (corrected for the expected spread) to the spread (again, corrected) in the 2004-06 data, we can directly answer that question. If QuesTec has forced umpires to call a more uniform strike zone, then the differences between umpires will be smaller than before QuesTec was installed.
And in fact, there has been a greater than 25% reduction in the true spread between umpires since the MLB first installed QuesTec in selected major league parks, which indeed supports the usefulness of the technology. It seems that umpires have curtailed the liberality of their definition of the strike zone for fear of losing their job, and gone with the rulebook definition instead.
Indeed, Eddings and Miller, for example, had the two largest strike zones between 1999 and 2001, calling 2.7% and 2.4% more strikes than expected, respectively (note: all numbers are “true rates,” and thus regressed to the mean). In the past three years, they have still called a higher number of strikes than anyone, but only 1.9% and 1.0% more than the league average, respectively. They may still be a pitcher’s best friend, but the bond isn’t nearly as strong as it used to be.
Baseball’s history with umpires has moved in a very clear direction. Volunteers chosen from the home team’s fans were replaced by paid umpires; then, rogue umpires were weeded out to ensure more consistency in umpiring decisions; that was followed by schools for aspiring umpires, which was followed by a more stringent process for selecting major league umpires. These days, only the cream of the crop gets to the big show. But all those changes have been made to achieve one thing: A standard set of rules for the game that would apply in the same way to every pitcher, batter, and team.
It’s clear that the MLB has been able to compress those differences in a significant manner through various training and technological methods. But it’s also clear that umpiring in the major leagues still isn’t perfect. In terms of correctly calling balls and strikes, it seems that the only perfectly reliable method is to use a machine that can correctly triangulate the position of each pitch as it crosses the plate (though it is not completely clear that a system like QuesTec can do so yet), perhaps with some human input in certain cases. It may be difficult to imagine the game without visible umpires and controversial calls, but it certainly seems that we are headed to an age of computers making the calls on balls and strikes instead of humans.
I, for one, welcome our QuesTec overlords.
References and Resources
Spectacular page on the history of umpiring
Wikipedia entry on umpires
Baseball Prospectus umpire statistics
David Gassko is a former consultant to a major league team. He welcomes comments via e-mail.