The Boston Red Sox’s decision-making came under scrutiny after manager Chad Tracy removed pitcher Sonny Gray from a crucial game against the New York Yankees. Despite all data pointing towards keeping Gray on the mound, Tracy opted for a change that led to a pivotal Yankees win.
The decision was questioned widely on September 30th, following the Yankees sweeping the Red Sox in the American League wild-card series. Tracy took Gray out in the fifth inning of Game 2, with the Red Sox trailing by just one run. Two runners were on base when left-handed reliever Erik Miller faced Cody Bellinger, who drove in a run, extending the Yankees’ lead.
Analytics play a significant role in baseball decisions, applying quantitative data to guide managers. Despite its convention, the data didn’t favor removing Gray. Bellinger approached Gray for the third time in the game, while the Yankees had just turned over their lineup. Gray, a right-handed pitcher, usually fares better the third time through the order, with a third-time-through xwOBA of .231.
Quality of Pitch data showed Gray’s curveball and sinker were effective throughout the game. Against Bellinger, Gray previously achieved three swinging strikes and a foul tip using those pitches. Thus, Tracy seemingly defied the analytics by pulling Gray from the game.
In his post-game comments, Tracy stated, “We’re trailing. We’re struggling to score runs. And when you get in spots where you’re struggling to score runs and you’re already trailing, you feel like you’ve got to be aggressive and go.” Tracy’s decision was rooted in instinct rather than data, stirring criticism over alleged excessive reliance on analytics.
Some media reports highlighted the perceived influence of Driveline, a consulting firm for the Red Sox’s R&D efforts, casting doubt on whether analytics played a clarifying role. Overall, the event spotlighted the dilemma of when to trust data versus intuition in high-pressure game situations.