← Latest papers
💻 computer science

Inside Baseball: The Automated Ball-Strike System as an Object Lesson in Technological Rule Enforcement

This paper uses Major League Baseball's seven-year struggle to implement the Automated Ball-Strike System (ABS) as a case study to demonstrate that even seemingly clear rules require complex sociotechnical translation due to contested "ground truth" and the need to balance diverse stakeholder values, thereby challenging conventional evaluation paradigms in favor of practice-oriented approaches.

Original authors: Andrea Wen-Yi Wang, Waki Kamino, David Mimno, Karen Levy, Malte F. Jung

Published 2026-05-18
📖 6 min read🧠 Deep dive

Original authors: Andrea Wen-Yi Wang, Waki Kamino, David Mimno, Karen Levy, Malte F. Jung

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Idea: Why "Robot Umpires" Took Seven Years to Figure Out

Imagine you have a recipe for a perfect cake. The instructions are very clear: "Mix 2 cups of flour, 1 cup of sugar, and bake at 350 degrees." You would think that giving this recipe to a robot chef would be easy. The robot just follows the steps, right?

This paper argues that it's not that simple.

The authors studied Major League Baseball's (MLB) attempt to install a "Robot Umpire" (called the Automated Ball-Strike System, or ABS). The goal was to use cameras and computers to automatically decide if a pitch was a "ball" or a "strike." The rule for the "strike zone" is written clearly in the official baseball rulebook. It sounds like a straightforward math problem: Is the ball inside this invisible box? Yes = Strike. No = Ball.

However, it took the MLB seven years of trial and error to get the robot to work. The paper explains that even when a rule seems perfectly clear, turning it into technology is messy, complicated, and requires a lot of negotiation.

The Three Umpires: Who is the "Real" Rule?

The paper starts with an old joke about three umpires to explain the confusion:

  1. Umpire A: "I call them as I see them." (I might make mistakes, but I'm trying to see the truth.)
  2. Umpire B: "I call them as they are." (I am perfect. My call is the truth.)
  3. Umpire C: "They ain't nothin' until I call 'em." (The truth doesn't exist until I say so.)

The authors say that for a long time, baseball lived with Umpire C. Even though the rulebook had a written definition of the strike zone, the real strike zone was whatever the human umpire standing behind the plate decided in that moment. If the umpire said it was a strike, it was a strike, even if it looked like a ball to everyone else.

When MLB tried to bring in the Robot, they ran into a problem: What is the "Ground Truth"?

  • Is the truth the Rulebook? (The perfect, written definition).
  • Is the truth the Human Umpire? (What has actually happened in games for 100 years).
  • Is the truth the Robot? (The new technology).

It turns out, the Robot couldn't just copy the Rulebook. It had to copy the Human Umpire's habits, even when those habits broke the written rules.

The Four Reasons the Robot Had to Change the Rules

The paper details four main reasons why the "Robot Strike Zone" ended up looking different from the "Rulebook Strike Zone." Think of these as four different people pulling on a rope, and the final design is the knot in the middle.

1. The Robot's Eyes Can't See Everything (Technical Feasibility)

The Rule: The strike zone changes based on how the batter is standing (their "stance"). If they crouch, the zone gets lower.
The Problem: The robot cameras couldn't reliably figure out exactly where a batter's knees were bent in real-time. It was too glitchy.
The Fix: The robot stopped looking at the batter's stance and just looked at their height. It's like a security camera that can't tell if you are wearing a hat, so it just measures your height from the ground up. The rule changed because the technology couldn't handle the original rule.

2. The League Wants More Excitement (Economic Interests)

The Goal: Baseball fans want to see more action (hitting the ball) and fewer boring walks or strikeouts.
The Problem: If the robot followed the rulebook too perfectly, pitchers would strike out way more batters, and the game would get boring.
The Fix: The MLB tweaked the robot's numbers (making the zone slightly smaller) specifically to encourage more hits. They didn't pick the numbers because they were "mathematically correct"; they picked them because they made the game more fun to watch and sell more tickets.

3. Keeping the Game "Feeling" Like Baseball (Integrity and Stability)

The Rule: The rulebook says the strike zone is a 3D box over home plate.
The Problem: If the robot used a 3D box, it would call certain "curveballs" (balls that drop sharply) as strikes. But human umpires have never called those pitches strikes; they call them balls. If the robot started calling them strikes, players would be furious, and the game would feel "wrong."
The Fix: The robot was programmed to use a flat, 2D plane in the middle of the plate instead of a 3D box. This made the robot agree with the human umpires' old habits. The technology had to bend to fit the "culture" of the game, not the written law.

4. The Art of the Game (Gamesmanship)

The Issue: Should the robot call every pitch?
The Problem: If the robot calls everything perfectly, a famous skill called "pitch framing" (where catchers subtly move their gloves to trick the umpire into calling a ball a strike) disappears. Many fans and players love this as a strategic "art form."
The Fix: MLB decided not to use the robot for every call. Instead, they use a "Challenge System." The human umpire calls the pitch, but teams can challenge it if they think the robot disagrees. This keeps the "art" of the game alive. The robot is only used to settle arguments, not to replace the human element entirely.

The Main Lesson: The Robot Is the Rule

The authors conclude with a powerful idea. Usually, we think of technology as a tool that tries to copy a rule perfectly. We say, "How close did the robot get to the real rule?"

The paper says: Stop thinking that way.

In the real world, the technology becomes the rule.

  • When you drive, the "speed limit" isn't just the number on the sign; it's the number the police actually ticket you for (which might be 5 mph over the limit).
  • When you use social media, the "rules" aren't just the written guidelines; they are what the algorithm actually deletes.

The "Robot Umpire" didn't just enforce the rule; it rewrote the rule to fit what was possible, what was profitable, and what the players would accept.

Summary

The paper teaches us that automating a rule is never just a technical job. It is a social job. Even when a rule seems simple (like "is the ball in the box?"), turning it into a computer program requires negotiating with:

  • What the computers can actually do.
  • What the business wants to sell.
  • What the players and fans are used to.
  • What keeps the game exciting.

The "Robot Umpire" is a perfect example of how technology doesn't just follow the rules; it helps create them.

Drowning in papers in your field?

Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.

Try Digest →