Automated Ball-Strike Systems: How Robot Umpires Have Been Tested
How automated ball-strike systems work, where they were tested from the Atlantic League in 2019 to the minors and spring training, and what the full and challenge versions each change.
An automated ball-strike system, usually shortened to ABS and nicknamed the robot umpire, is a tracking system that measures where each pitch crosses home plate and compares that location with a strike zone drawn for the batter at the plate. The result can be sent to the home-plate umpire, who announces it, or it can be used only when a player disputes the umpire's call. No version tested so far has removed the human umpire from the field.
The idea sounds simple and has taken years of testing, because every part of it contains a decision: how the zone is defined, how accurate the tracking must be, who gets to question a call, and what the game loses when a human judgment becomes a measurement. This article describes what was tested and where, explains how the two main versions differ, and works through the geometry of the zone. Everything here is about testing and design. Whether and how an automated system is used in a given major-league game changes with the rules, so check MLB's official site for the current status before assuming anything.
What an automated ball-strike system is made of
Three components make up any ABS, and the differences between tests come from how each is built.
- Tracking hardware. Radar or cameras follow the ball in flight and record its position when it reaches home plate. The 2019 Atlantic League test used a Doppler radar system; later systems in the minor leagues and in major-league parks have used multi-camera optical tracking. Either kind reports a location with a margin of error, which matters for borderline pitches.
- A zone. Each batter needs a top and a bottom for the zone, and the width comes from home plate. In the tests the top and bottom were set from the batter's height, measured in advance, rather than from his stance in the box.
- A way to communicate the result. In the full version the call reaches the home-plate umpire through an earpiece or a device, and he announces it as if he had made it. In the challenge version the system sits idle until a player asks for a review.
Notice what is missing from that list: nothing in it calls a check swing, a hit by pitch, a foul tip or a catcher's interference. The human umpire still does all of that, and still decides when the ball is live.
The zone the machine calls, and the zone in the rulebook
The strike zone as written in the official rules is a three-dimensional region. In the form it took through the 2025 season, it is the area over home plate between a level at the midpoint between the top of the batter's shoulders and the top of the uniform pants and a level at the hollow beneath the kneecap, and it is measured as the batter takes his natural stance. A pitch is a strike if any part of the ball passes through any part of that region, and the plate itself is 17 inches wide while a baseball is just under 3 inches across. The strike zone article explains the rule and its history in detail.
An automated system cannot judge a stance, so it uses a simpler zone: a flat plane over the plate, with a top and bottom computed as fixed shares of the batter's measured height. The shares reported for the tests are about 53.5 percent of height for the top and about 27 percent for the bottom. Treat those figures as approximate; they are not in the rule book, and the exact values are part of what the testing adjusted.
A worked example shows what that does. For a batter standing 72 inches tall, the reported shares put the top of the zone at about 38.5 inches and the bottom at about 19.4 inches, a zone about 19 inches high. For a batter 68 inches tall, the same shares give a top near 36.4 inches and a bottom near 18.4 inches, a zone about 18 inches high. A pitch that crosses the plane at 19 inches on the first batter is a ball by a hair on the first and a strike by a hair on the second. That kind of fixed rule is the point of the system, and it is also why the way the zone is defined is the main argument in the whole project.
The 2019 Atlantic League test
The first widely noticed test took place in the Atlantic League of Professional Baseball, an independent league whose clubs are not affiliated with the majors, under an agreement between that league and Major League Baseball announced earlier in 2019. The system was used at the league's All-Star Game in July 2019 and then in some regular-season games, with a radar tracking each pitch and the home-plate umpire wearing an earpiece to hear the call.
The Atlantic League is a useful place to experiment because it carries no standings pressure for major-league clubs and has been used for several rules trials. It does not, however, resemble the majors in its talent or its pitch quality, and reports at the time said that some of the first complaints concerned pitches the system called differently from a human, especially breaking balls near the lower edge. The early disagreement was about how the zone should be drawn. Specific numbers from that test are best taken from the league's and MLB's own reports.
The minor leagues and the Fall League
After the independent league, the tests moved into affiliated baseball. The Arizona Fall League, where each organization sends prospects after the regular season, used the system in some games in 2019. In 2021 the full system was used in the Low-A Southeast league, which gave the first season-long sample in affiliated baseball, with pitches tracked and called by the machine in the parks that carried it.
In 2022 the full system moved up to a handful of Triple-A parks, and in 2023 the testing widened to every Triple-A park. There the full system was used in some series and the challenge system in others, so that MLB could compare the two in similar conditions with the same players and umpires. By 2024 the challenge version was widely used in Triple-A. The design lesson of those seasons is that the same technology can be used in very different ways, and the debate moved from "does it work" to "what should it replace".
Full ABS versus the challenge system
The two versions answer different questions.
| Feature | Full ABS | Challenge system |
|---|---|---|
| Who calls each pitch | The system, relayed by the umpire | The umpire |
| When the system is used | Every pitch | Only when a player challenges |
| Who may challenge | Not applicable | The batter, the pitcher or the catcher |
| Effect on game flow | Small, one call per pitch | Pauses play for each challenge |
| Effect on catchers | Little reward for pitch framing | Framing still matters until the pitch is challenged |
| Typical complaint | Breaking balls and borderline pitches feel wrong | The limit on challenges, and the time each takes |
The full version is simplest to understand and the hardest to accept. Every ball and strike is a measurement, and the human umpire becomes a signal relay. The challenge version keeps the umpire as the primary judge and adds an appeal, in the same spirit as the replay review for plays on the field. The instant replay article describes how that older system works and why leagues limit appeals.
How a challenge works in the tests
The version used in Triple-A and later in spring training followed a standard pattern, and the details are worth knowing because they decide how much the system changes a game.
- Only three players can challenge: the batter, the pitcher or the catcher. Managers and coaches cannot, and players may not look to the dugout for advice.
- The challenge must be immediate. The player taps his helmet or cap right after the call, before the next pitch and before he does anything else, such as stepping out or turning toward a coach.
- The system answers within seconds. The result is shown on the stadium video board and announced, and the umpire's call is either upheld or overturned.
- Challenges are limited. In the versions tested, each team had a small number of challenges per game (two was the common number), and a team that wins a challenge keeps it, while a team that loses one uses it up. Extra innings added provisions, and the exact numbers are set in the rules for each league, so confirm current details at MLB.
The limit is what keeps the game from turning into a sequence of appeals. A team that spends its challenges early on routine pitches has none left for a ninth-inning walk-off at-bat, so players choose when to use them. That choice is a new strategic layer: the batter, the pitcher and the catcher each have a view of the pitch, and a team has to decide whose judgment to trust.
Spring training and the path to the majors
Spring training offered a different test: the same players who would use the system in the regular season, in front of the same umpires, with no standings at stake. In 2025, MLB used the challenge system in a number of spring-training ballparks in Arizona and Florida, so that hitters, pitchers and catchers could try it before any regular-season decision was made. The spring games were a trial, and the specific plans for regular-season use are set and changed by the league and the players' union, so this article does not describe them. Check MLB for the status.
What the spring trial adds is data on how often a challenge is used, how often it is successful, and how long it delays play. Those three numbers tell the league whether the system is a gimmick or a working rule, and they are tracked by the league, not by the historical record, which does not carry pitch-level data.
What the testing revealed
Several findings recur across the tests and reports.
- Borderline pitches are the whole story. Most pitches are clearly balls or clearly strikes, and a human umpire and a machine agree on those. The disagreements cluster at the edges of the zone, which is why the tracking margin of error matters.
- The zone shape is the real argument. A flat plane treats a curveball that drops through the front of the zone the same as a fastball that passes at the same height, while the three-dimensional rulebook zone gives credit for pitches that cross the back corner of the plate. The disagreement is partly about physics, which the physics of a curveball article explains.
- Height measurement is a decision. Measuring a batter's height in advance is consistent, but a batter who crouches or stands tall can have a zone that does not match what the rulebook says about his stance.
- Behavior is expected to change. A fixed, consistent zone rewards pitchers who throw strikes the machine will call and gives hitters a stable target. Whether it moves walk and strikeout rates is a question for season-long data, not for one game.
- Time matters. The challenge system adds a pause each time it is used, and any league adopting it has to weigh that cost against the arguments it removes.
None of these points is a verdict. Different leagues weighted them differently, and the choices between a full system and a challenge system reflect values, not just accuracy.
What changes for catchers, pitchers and hitters
Catchers feel the effect first. Pitch framing, the skill of receiving a borderline pitch so that it looks like a strike, is worth runs when a human umpire makes every call. In a full system it is worth almost nothing; in a challenge system it still matters on the pitches no one challenges, and the catcher also becomes a decision-maker about whether to ask for a review.
Pitchers who work the edges of the zone are affected next. A pitcher whose game depends on getting a call an inch or two off the plate loses that margin. A pitcher who throws strikes in the zone as drawn gains a clear target.
Hitters gain consistency, since the zone is the same from one umpire to the next and one game to the next. They also lose the ability to learn a particular umpire's tendencies, which has long been part of preparation. The box score article shows how walks and strikeouts are recorded, and those are the statistics a change in the zone would move first.
Where this fits among recent rule changes
Automated calls are one of several changes designed to reduce human variation or speed up play. The mound visit and pickoff limits, the intentional walk change and the pitch clock, which MLB adopted for the 2023 season, all move the game toward fixed rules. ABS differs from the others because it replaces judgment instead of limiting an action, and for that reason it has drawn the most discussion about what the game should preserve. Calls on obstruction and interference, covered in the obstruction versus interference article, show another place where the rules depend on an umpire's reading of intent and would be harder to automate.
Common questions
Do robot umpires call every pitch in the majors? That depends on the current rules, which this article cannot settle. In every test described here a human umpire remained behind the plate, and the system was either relaying calls or answering challenges. Check MLB for the current situation.
Why not just use the rulebook's three-dimensional zone? Because it depends on the batter's stance, which a tracking system cannot read well, and because a simple plane is easy to explain and to check. The cost is that the rule and the machine differ at the back of the plate.
Is the system more accurate than a human umpire? On measured location, a well-calibrated tracking system is consistent, though it has a margin of error. Whether that makes it better depends on whether you value matching the rulebook, matching the tradition of a human call, or avoiding visible mistakes.
Who can challenge? In the versions tested, the batter, the pitcher or the catcher, and only immediately after the call.
Will the strike zone change? The written zone has been revised before, and the tracked zone is a separate design choice. Both can be adjusted.
Where to go next
For the rule itself, read the strike zone: its official definition, its history and why it is hard to call. For another review system that limits appeals, read instant replay and the challenge system. Rules coverage is collected on the site's rules and umpiring topic page, and the minor-league section lists the leagues where many of these tests were held.