Pedro Martinez was never ejected from a major league game for arguing balls and strikes, but the Boston Red Sox’s Hall of Fame pitcher admits he should have been once. In a start against the New York Yankees at Fenway Park on July 25, 2003, Martinez began his tirade by wagging his finger at home plate umpire Dana DeMuth.
He and Yankees starter David Wells were annually among the sport’s stingiest at issuing walks. Wells had walked only six in 19 prior outings that season, yet walked five that afternoon. Martinez issued four free passes — more than he had in his four previous starts combined.
A base on balls to Derek Jeter in the seventh inning sparked Martinez’s eruption, which escalated beyond the initial disapproving finger. He framed the strike zone box with his hands, asking, “Where is your strike zone? Is it up? Is it down? Is it in? Or is it out?”
DeMuth replied, “I’m trying to do my job.”
Martinez’s retorted, “So am I. Why don’t you try and do mine? And I will try and do yours.”
As he said that, Martinez removed his glove, extended his arm and offered the mitt to DeMuth.
The real culprit of his infamous outburst, Martinez contended, was Umpiring’s Big Brother, an evaluation system that included a pair of cameras literally lurking over each shoulder.
“They brought out the QuesTec,” Martinez said, “so he was hesitant to call anything around the plate a strike.”
QuesTec, a small Long Island-based company, made its national debut 20 years ago last month as a camera- and software-based broadcast system for rendering the strike zone during NBC’s coverage of the 1997 National League Championship Series and World Series.
The image-processing technology, which first appeared under SuperVision branding, had roots in military research funding through its engineering partner, Atlantic Aerospace Electronics Corporation, which refined the system after coming on board in 1998. By 2001, a more precise product from QuesTec was adopted by Major League Baseball as an evaluation tool for umpires in an effort to standardize the strike zone.
Tom Glavine and Greg Maddux, Hall of Fame teammates for the Atlanta Braves, were among the biggest beneficiaries of a widening strike zone that granted calls several inches off each corner of the plate. Star pitchers were thought to have earned the benefit of the doubt on borderlines pitches, receiving more strikes than rookies. The height of the strike zone appeared to shrink, with few strikes called above the belt or even at the knees.
Worst of all: each umpire seemed to have his own interpretation of exactly what was a ball or strike. The league, led by former executive vice president of operations, Sandy Alderson, sought uniformity in the strike zone, seeking to restore the rulebook definition.
“I felt strongly that the strike zone needed to be sort of the cornerstone of umpiring and that it couldn’t be individualized,” Alderson, now the New York Mets’ general manager, told SportTechie. “It had to be more standard, and we had to figure out a way to promulgate that and then enforce it. That was the technology that was available at the time.”
Depending on whom you asked, QuesTec became either a technological marvel or maligned, misappropriated machinery — also rife with corporate mismanagement, as many later reported. What is indelible two decades later is its legacy in reshaping the sport: a more universal and predictable strike zone, a decline in offense, record numbers of ground balls and maybe even a trace cause of the modern home run surge.
Most of all, though many are loath to admit as much, QuesTec succeeded.
“QuesTec had its growing pains,” former umpire Clint Fagan said, “but those were necessary growing pains to get to where we are today.”
“An absolute joke”
The game that many observers point to as the turning point for change took place during QuesTec’s very first playoff series in operation. In Game 5 of the 1997 NLCS, Florida Marlins starter Livan Hernandez threw a complete-game victory with 15 strikeouts, a performance marred by an exaggeratedly wide strike zone called by home plate umpire Eric Gregg.
As former QuesTec CEO Ed Plumacher told the Palm Beach Post in 2003, “There were a lot of pitches that weren’t shown that day because it was too controversial. I mean, the ones we did show were bad enough.”
Former Cy Young winner David Cone said a friend recently sent him a video from the 1996 World Series, in which he pitched Game 3 against Glavine. He was shocked by the zone from that night.
“There were balls — for both of us — six inches or more off the plate that were routinely called strikes,” Cone said.
QuesTec’s umpiring system would became a polarizing topic around baseball, with numerous umpires and players voicing resounding scorn and derision. Curt Schilling angrily smashed a $5,000 camera and reportedly asked to have the cameras removed from his home park. Glavine called himself “the poster child” after struggling in QuesTec-equipped ballparks. Barry Bonds deemed the sport a computer game, pegging QuesTec “the worst thing that can happen in baseball.” Umpires gave scorching reviews: one called it “an absolute joke” and another likened the job to “umpiring a video game, not a baseball game.” Counsel for their union, the World Umpire Association, called the system an “embarrassment to Major League Baseball” and filed a grievance to an arbitrator in 2003 (later resolved in 2004).
“This has been an overwhelming battle for us,” Plumacher told Newsday in 2003. “We feel like a pawn in a major dispute. Whenever something happens, blame it on QuesTec.”
For many pitchers and umpires, the name QuesTec became a profanity and, though the system was replaced in 2009, its moniker lingers around the sport as a label for successor editions of the evaluation system, a branded eponym like Xerox and Kleenex morphing into a generic term. (Plumacher could not be reached for comment by SportTechie.) Much of the umpires’ acrimony was also derived from the fallout of a labor dispute that prompted mass resignations from 54 umpires in 1999, of which 22 were accepted (and only some later reinstated). Supervision of the umpires was consolidated from a department under each league office, American and National, into one at MLB headquarters.
The league initially installed QuesTec’s evaluation product without umpire assent and in only a handful of ballparks in 2001. Formally known as the Umpire Information System, it later peaked at 11 ballparks by 2008, before the technology was supplanted by SportVision’s Pitch F/X in all 30 parks for 2009 (and, subsequently, this past year by Statcast’s radar-based program). QuesTec relied on four cameras installed in the park: two up high and two in the photo wells near the dugouts. The cameras automatically detected the outer boundaries of the strike zone, but an on-site employee, often making $50 or $100 per game, would set the upper or lower boundaries by drawing lines on still photos of each pitch, given the varied heights of the batters. This, too, was a point of contention from the umpires.
Mike Port, who was MLB’s VP of umpiring from 2005 to 2011, called QuesTec a helpful first step to improving accountability.
“QuesTec was the black-and-white TV, if you will,” he said. “The thing had to start somewhere.”
Port extended his analogy by likening Pitch F/X to a standard color television and Statcast to high-definition. Sure, there were glitches. Two that Port relayed included a borderline full-count pitch to Yankees catcher Jorge Posada in a postseason game that, upon review, the QuesTec operator had set the lower boundary of his strike zone at his ankles; and one pitch thrown by a former Cleveland Indians southpaw in which the ball appeared to move vertically, prompting Port to joke with general managers about the pitcher who empirically had the best stuff.
Overall, however, Port said, “I think the system has brought a lot of standardization for the strike zone.”
Go north-south, young man
The impact of the umpire monitoring system on game play was immediate. The 2000 season saw record numbers of home runs and total pitches (both broken in 2017), as well as the majors’ highest scoring rate in 70 years and the most walks in 50 years. In the first two seasons of QuesTec’s implementation, however, scoring decreased 10 percent and walk rates declined even more.
Florida sports economics professor Brian Mills, whose research interests including training and monitoring in the workplace, has studied this topic extensively and found that, soon after QuesTec was introduced, “There was a large upward shift in the rate that umpires called strikes.” Mills said the response was predictable behavior from employees being watched. After a couple of years, strike rates were generally consistent in all stadiums, whether they actually had the QuesTec system installed or not.
A 2003 ESPN study by Nate Silver (subsequently of FiveThirtyEight fame) and Keith Woolner (currently the Cleveland Indians’ principal data scientist) found the overall difference between QuesTec-equipped parks and the others to be “negligible.” A similar analysis compiled by the Elias Sports Bureau indicated a minuscule difference between ballparks with and without QuesTec in a variety of categories, suggesting that discrepancies were, at best, exaggerated if not completely unfounded. (Port once told the New York Times, “A lot of people are of the impression that a particular umpire not in a QuesTec park would run wild and become an outlaw. But there is virtually no difference.”)
The rule book strike zone had been last modified in 1996 to expand the height of the zone, emphasizing its northern and southern extremities over extensions of its eastward and westward frontiers.
“We always thought — those of us that are old enough — the strike zone used to be more vertical,” Cone said, “and it turned more horizontal and that QuesTec would get back to calling the high strike. Everything was always about calling the high strike again and less off the plate as much. Gradually, I think that’s kind of happened since QuesTec started 20 years ago.”
Conversations at the time primarily centered around ensuring the high strike just above the belt was called, but few knee-high strikes were being called either. That prompted the Playing Rules Committee to redefine the lower boundary as well to coax umpires into stretching their zone lower.
The formal definition now reads, “The strike zone is that area over home plate the upper limit of which is a horizontal line at the midpoint between the top of the shoulders and the top of the uniform pants, and the lower level is a line at the hollow beneath the kneecap. The strike zone shall be determined from the batter’s stance as the batter is prepared to swing at a pitched ball.”
QuesTec evaluations had a more profound and largely unexpected effect on the bottom of the strike zone.
“There’s no question that umpires now are calling the strike zone more consistently, to the point where they’re calling pitches at the bottom of the strike zone that, to me, were never even intended to be called strikes,” Alderson said, adding: “With the advent of QuesTec and Pitch F/X and TrackMan, now we’ve been able to identify the bottom of the strike zone and grade out umpires against the totality of the strike zone, including that lower end, so they’re just calling it more down there.”
Research on publicly available Pitch F/X data — which began governing the umpire evaluations in 2009 — has shown that the strike zone lowered by the full diameter of a baseball. This was all the inducement pitchers needed. With the St. Louis Cardinals and Pittsburgh Pirates leading the way, pitchers bombarded the bottom of the strike zone, and by 2014 and 2015, batters were hitting record numbers of ground balls, plunging offense further.
“It’s very easy to see that the strike zone bottom lowered by about three inches, that pitchers noticed this and threw there more often, that batters noticed that and they swung there more often, and they got worse outcomes when they did,” Mills said. “That absolutely happened.”
Hitters later adjusted. Many began swinging with a slight uppercut to lift the ball. Beginning in late 2015, home run totals surged; there were more homers hit in 2017 than in any season in major league history. The terms exit velocity and launch angle — generated by the Statcast system — have permeated the daily discourse. (There are also some studies indicating that the composition of the baseball itself changed somewhat as well, albeit still residing within the rulebook parameters.)
“I can’t prove it and I don’t know of anybody who could, but I do think there might be some second or third correlation to ‘Hey, the strike zone moved down, guys had to start swinging at that pitch, they changed their swing,’” an American League scout said. “It’s not like you’re seeing these awesome flat-planed swings that are built to hit .330. You’re seeing swings that are built to elevate the ball and hit it out of the park.”
Alderson deemed the connection between QuesTec and home runs to be a rung or two more tenuous, calling that link “a three- or four-step process.” His theory on the uptick in homers is more directly a result of infield shifts. Low strikes were inducing ground balls; more infield shifts meant more of those grounders were being converted into outs; and newly available batted-ball data, such as exit velocity and launch angle, helped hitters realized that hard-hit balls in the air are more likely to be hits even when struck into the shift than attempts to hit the ball the other way.
“Ultimately, I think the fact that the zone is being called more consistently and more frequently at the bottom of the zone, as it should be under the current rules, hitters had to figure out a way to adjust to that and counterattack in the low part of the strike zone,” Alderson said.
Bruce Weber, who wrote a book about umpires in 2009 called “As They See ‘Em,” told Gelf Magazine: “QuesTec is kind of a Pavlovian tool: You call a narrower strike zone, or you get a bad grade. As it turns out, it hasn’t been an awful thing for the game, which is why I don’t want you to think that my view of QuesTec is entirely dim.
“By the end of the 1990s, the strike zone started to resemble something seriously other than what it was supposed to be. There were really egregious examples of terrible, terrible strike zones, where either the hitters or the pitchers were at a terrible disadvantage. One of the points I make in the book is that the strike zone is like the fulcrum of a seesaw — it really is the thing you adjust in order to make sure the offense and the defense are on a level playing field. During the 1990s, the seesaw was out of whack, and QuesTec helped bring it back, that’s for sure.”
An umpiring ‘master’s degree’
The umpires contended that QuesTec was “inaccurate and unreliable,” as union attorney Larry Gibson said in an arbitration hearing. Concerns about its technical accuracy, however, were never proven, and MLB permitted a two-inch buffer zone on either corner of the plate. The idea was to overcompensate for the margin of error and not score umpires as having made the wrong call on anything close.
“There are problems, but it’s a pretty good system,” Yale professor Robert Adair, who independently reviewed the system’s accuracy, told ESPN.
As former commissioner Bud Selig told Sports Illustrated’s Tom Verducci in 2003, “I think QuesTec can get more sophisticated and more refined and yes, ultimately we need to get QuesTec in every park. You need a device to evaluate people, or help them. And if it isn’t QuesTec, it’s going to be something else. But the people we have in charge of this really believe in QuesTec. And until I have evidence otherwise it’s going to remain.”
Paul Baim, who at the time was director of software engineering at Atlantic Aerospace Electronics Corporation (now a division of L-3 Communications), told the PBS program Scientific American Frontiers in a March 19, 2002, episode dedicated to baseball that QuesTec was accurate within 1 1/4 inches for the commercial TV product and within 1/2 inch for the umpire product. Moreover, tweaking calls on borderline pitches was not the intention of the system anyway.
“The issue here is not helping umpires get those pitches ‘right’ that they missed by half an inch,” Baim told interviewer Alan Alda. “The point is not to split hairs.
“The point is to give them a view of pitches that might be within the upper part of the strike zone that they’re not necessarily used to calling a strike that in fact was a strike. Or pitches that curve a lot and fall through the bottom of the strike zone and get caught by the catcher down near the ground and to everyone watching looked like, well, that obviously must have been low when in fact it passed through the strike zone because it was curving so much. Those are the pitches that we’re trying to help the umpires get another view of.”
Fagan made his major league debut in 2011, so he never worked a game with the original QuesTec system, but said the scores from its successor, ZE, were seen as less important as the areas to improve. (Even with the new technology, that two-inch buffer remains.) He said that working the Arizona Fall League was like earning “a master’s degree” for a minor league umpire because that was the the first assignment in which their strike zones were graded.
“It was more of a teaching tool to indicate what type of trend an umpire had and how to correct those tendencies,” said Fagan, who is now working as an insurance agent and taking law school classes.
Relatedly, Port remembers veteran umpire Joe West often repeating the refrain, “Don’t tell me that I missed a call. Tell me why I missed a call.” Port said that, through his last full season at MLB in 2010, the staff average was between 95 and 96 percent accurate. In other words, home plate umps would miss one of every 20 or 25 pitches. The Boston Globe reported in 2015 that nearly every umpire was between 92 and 97 percent, with that 92 percent figure representing an elite grade back in QuesTec’s inaugural 2001 season. Fagan said that Phil Cuzzi, the home plate ump for 2017’s World Series Game 1, annually rates atop the ZE’s ball-strike leaderboard.
Umpires, of course, still have individual interpretations of the zone and well-known tendencies, but those variances have been mitigated. Postseason assignments became, in part, dependent on their strike zone scoring. Commissioner Rob Manfred, who oversaw labor relations during the 2000s, said Alderson’s most important achievement with the umpires was that he “instilled a culture of accountability,” as he told Baseball America, adding: “The really uniform strike zone that we have today is part of Sandy’s legacy.”
Glavine, who told Weber that he needed two years to adapt to the QuesTec-affected strike zone, later came around on the basic technology, endorsing the company’s later performance-training tool, PitchSight. Indeed, some the umpires that previously spurned the QuesTec system eventually came to embrace the results upon conforming to its standards and rating highly.
“When I was around them in the dressing room,” Port said, “it became even — aside from their earlier apprehensions — like a scholastic score. ‘I got a 96 on my last game, what did you get?’”
Perfection or excellence?
SportTechie spoke with three former QuesTec operators: one who worked on only the broadcast tool, one who worked only on the umpire system and one who worked on both. All said their jobs were straightforward and had little oversight. One came to the realization that even “the average umpire was still remarkably accurate.” Jeff Thurs, who operated the TV system in 2001 and 2002 at Milwaukee’s Miller Park, said his role was to make sure the dugout cameras were properly aligned before each game and then number every pitch so he could call up a previous at bat if requested by the producer.
For part of the 2001 season, he sat next to his friend Jeff Blume, who worked a few games on each system. They sat at a table in a networking room, watching the game on a TV monitor. His task was to set the strike zone’s height lines and check if the image of the pitch crossing the plate was clear and unobstructed, then burn the results to a CD-ROM.
A former project manager for the UIS, Ivan Santucci, spoke to Baseball Prospectus about his role back in 2004, dispelling some concerns about the subjectivity of the QuesTec system. He noted that operators don’t need to discern the midpoint of a hitter’s torso but rather only the top of his belt buckle at which point the system automatically elevated the top line 2 1/2 ball widths. Santucci said, “There is no judgment being made here.”
QuesTec’s TV tool was less precise than the umpire evaluation product, not the many baseball fans would have tempered their reactions any.
“They’re going to throw that box up there, and they’re going to be as critical as any critic in the world — the movie critic, food critic, whatever they might be,” Fagan said. “I guarantee you baseball fanatics are going to be the biggest critics when it comes to enforcement of the strike zone.”
Historically, the center field shot of pitcher and batter was never directly centered; the camera was a little off to the side to avoid obstruction, causing optical illusions of the strike zone. QuesTec and other such boxes were meant to provide clarity.
“I think we were always concerned whether what the viewer at home was seeing was reality, if you will,” NBC coordinating producer Ricky Diamond said. “It was deceptive. You thought you saw the strike zone, but you didn’t.”
The boxes used by the varying networks appear under different branding, such as K-Zone, FoxTrax and PitchTrax. ESPN recently experimented with a 3-D rendered version. Many observers have begun calling for fully automated strike zones — i.e. robot umps.
“I think in baseball there’s a current movement toward perfection when, instead, perhaps the overall movement should be towards excellence,” Port said.
Manfred has said that the technology to call balls and strikes better than the umpires will be available “some day,” but he believes the people currently have a smaller margin of error.
“These guys are the best of the best, and they’re very well trained,” Mills said. “I don’t think you’re going to find humans that are any better.”