AI Baseball: Oakland Ballers Hand Control to Artificial Intelligence

The Allure and Apprehension of AI in Baseball Management

A memorable scene from “The Simpsons” features Mr. Burns, a calculating businessman, assembling a softball team comprised of professional Major League Baseball players to secure a wager. However, during a crucial moment, he substitutes an eight-time National League all-star, Darryl Strawberry, with the unexpected choice of Homer Simpson.

Mr. Burns justifies his decision, stating, “You’re a left-hander, and so is the pitcher. Utilizing a right-handed batter in this situation is known as playing the percentages,” adding, “It’s a strategy employed by astute managers to achieve victory.”

Modern, competitive baseball relies heavily on mathematical analysis, with teams employing numerous data engineers to scrutinize detailed statistics that inform managerial choices. Nevertheless, as illustrated in the “Simpsons” episode, there's a risk of overcomplicating baseball analytics to an excessive degree.

The Oakland Ballers' Experiment with Artificial Intelligence

The Oakland Ballers, an independent Pioneer League baseball team, explored the concept of “playing the percentages” by allowing an AI to manage the team for a single game.

Founded by tech entrepreneur Paul Freedman, the Ballers emerged as a response to the relocation of the Oakland A’s, a Major League Baseball team whose departure was widely criticized as a detrimental managerial decision. Despite not being a Major League team, the Oakland Ballers—often referred to as the Oakland B’s—have cultivated a dedicated fan base that rallied in protest against the A’s move. Remarkably, after just two seasons, the Ballers secured Oakland’s first baseball championship since 1989.

“The Oakland Ballers possess a unique position, functioning like a major league team within a minor league market,” Freedman explained to TechCrunch. “This allows us a level of creative freedom, enabling experimentation with concepts that MLB or NBA leagues wouldn’t consider at this time.”

Minor league teams frequently serve as testing grounds for new technologies before their implementation in the majors, such as instant-replay reviews for challenged calls or automated ball-strike systems. The Ballers embraced this role, particularly given Freedman’s tech background, while also incorporating elements of playful innovation, piloting ideas unlikely to be adopted by Major League Baseball.

Previously, the team partnered with Fan Controlled Sports, allowing fans to make managerial decisions during a late-season game. This resulted in a loss, as fans prioritized amusing choices over strategic ones—at one point, requesting a pitcher to bat.

Developing the AI Manager

Following their postseason qualification, the Ballers collaborated with the AI company Distillery to develop AI software capable of managing a baseball game in real-time.

“Baseball provides an ideal environment for this initial experiment due to its data-intensive nature and analytically driven decision-making,” Freedman stated to TechCrunch. “The game’s pace allows for adjustments to be made after each pitch.”

Distillery utilized OpenAI’s ChatGPT, training it on over a century of baseball data and analytics, including Ballers game information, to simulate the decision-making process of Ballers manager Aaron Miles.

“The AI essentially determined what our human coach would have done—the strategic insight and conceptual framework originated with [Miles], while the AI’s strength lay in data utilization and pattern recognition,” Freedman clarified. “Therefore, I believe the role of human ingenuity remains secure, with AI serving as a tool to enhance, rather than replace, decision-making.”

The AI's Performance and Fan Reaction

The AI-managed game proceeded seamlessly, with the AI replicating all of Miles’ anticipated choices regarding pitching changes, lineup construction, and pinch hitters. The sole instance of Miles overriding the AI was to substitute the starting catcher due to illness.

Miles accepted the AI’s temporary management role with good humor—acknowledging that his position wasn’t genuinely threatened. A video posted on the Ballers’ Instagram shows Miles approaching home plate before the game to greet the opposing manager, extending the tablet running the AI for a handshake instead of his own hand.

However, the AI’s involvement sparked discontent among Oakland fans, who expressed concerns about companies like OpenAI—the provider of Distillery’s baseball AI—prioritizing “winning” the AI race over ensuring the safety and thorough testing of their products. For many, the experiment felt like a betrayal, mirroring the corporate actions that led to three professional sports franchises leaving Oakland within five years.

“The Ballers are now pandering to Bay Area tech enthusiasts instead of baseball fans,” one commenter wrote. “This is the end for Oakland.”

This negative response was unexpected, and Freedman does not plan to repeat the AI experiment. However, the fan reaction highlights a broader cultural debate surrounding technology in baseball and beyond.

“It’s never pleasant to have your fans express their disapproval,” Freedman admitted. “But it’s beneficial that this technology’s advantages and disadvantages are being discussed now, rather than waiting a decade when it might be too late.”