PECOTA 2026: Navigating the Future with PECOTA
Image credit: © Matt Marton-Imagn Images
This year's PECOTA system updates were subtle, building on last year's significant accuracy improvement in pitcher predictions (https://www.baseballprospectus.com/news/article/96300/pecota-week-pecota-2025-an-introduction/). We explored the impact of transitioning from StuffPro to ArsenalPro metrics (https://www.baseballprospectus.com/news/article/96026/introducing-new-arsenal-metrics/), but found no substantial gains. This outcome is expected, as Arsenal metrics consider context, which is challenging to predict comprehensively. Consequently, pitcher projections will maintain last year's methodology.
Batter projections saw a modest enhancement in home run forecasting, but otherwise remained unchanged.
Looking ahead, two significant challenges emerge.
The first challenge revolves around interpreting minor-league performances at the Double-A and High-A levels. These levels are crucial for identifying major-league talent, yet publicly available data is limited. While play-by-play and pitch result data is accessible, the quality gap between these levels and Triple-A and Low-A levels is significant. The latter levels provide detailed pitch data and Statcast measurements of batted balls, which are not publicly available for Double-A or High-A players. This disparity disadvantages public analysts compared to teams and their data providers. Standardizing and publicly releasing this information with MLB's quality assurance is essential.
MLB's recent decision to harmonize and limit technology used by major-league clubs across their minor-league affiliates is a controversial move. However, it offers hope for making Statcast and pitch data publicly available for Double-A and High-A levels. This development is eagerly anticipated.
The second challenge stems from fundamental differences in MLB economics. Teams with older, expensive players signed to long-term contracts are more predictable than those with short-term commitments and players with limited major-league experience. Forecasting established players is more straightforward than predicting the outcomes of new talent, especially given data limitations. Teams in large markets that invest heavily in free agents often have longer player histories and are subject to the sunk-cost fallacy, making them more stable.
Smaller markets, on the other hand, present unique challenges. Their reluctance to sign long-term contracts means they make fewer regrettable signings and are not bound by player obligations. Failing players can be benched for promising minor-league talents, which small-market teams have more information about. However, small markets with strong farm systems have numerous options and may experiment with many players throughout the season.
At Baseball Prospectus, we emphasize that projection accuracy is primarily driven by playing time, not player measurement. When teams' playing time becomes unpredictable, forecasting becomes more challenging. This complexity arises from limited player track records and uncertain playing time.
Despite these challenges, we at Baseball Prospectus remain committed to improving our minor-league data sources. As we gain better insights into unexpected players, we'll strive to anticipate their performance, even if surprises persist in team lineups. Thank you for reading.
Consider subscribing to Baseball Prospectus to support ongoing public baseball research and analysis in a proprietary environment. Subscribe now (https://www.baseballprospectus.com/subscriptions/?utmsource=site&utmmedium=banner&utmcampaign=freearticle).
