Predicting Baseball’s Future Just Got Even More Complicated—Here’s Why
Baseball forecasting is a game of constant evolution, and PECOTA 2026 is no exception. This year’s updates may seem modest at first glance, but they reveal deeper challenges that keep analysts like us on our toes. But here’s where it gets controversial... While last year’s system leapfrogged in accuracy for pitcher predictions by hundreds of runs, our latest tweak—switching from StuffPro to ArsenalPro metrics—didn’t move the needle. Why? ArsenalPro’s context-heavy approach struggles with the unpredictability of a pitcher’s season-long performance. So, we’re sticking with what works—for now.
Batter projections saw a slight boost in home run forecasting, but the real story lies in the broader challenges ahead. And this is the part most people miss... Two major hurdles stand out as we look to the future.
First, deciphering minor-league performances at the Double-A and High-A levels remains a puzzle. These leagues are crucial for identifying major-league talent, yet the publicly available data is shockingly limited. While we’re grateful for play-by-play and pitch result data, the gap in quality compared to Triple-A and Low-A is staggering. Triple-A and Low-A provide not just pitch data but also Statcast measurements of batted balls—a luxury absent for Double-A and High-A. This disparity puts public analysts at a massive disadvantage compared to teams and their vendors, who often trade this information behind closed doors. Shouldn’t this data be standardized and made public with MLB’s stamp of approval?
MLB’s recent decision to standardize technology across minor-league affiliates is a step in the right direction, though not without controversy. Hopefully, it will pave the way for Double-A and High-A data to finally see the light of day. Fingers crossed.
The second challenge? It’s unsolvable—at least for now. It stems from the economic divide in MLB. Teams like the Phillies, with older, expensive players locked into long-term contracts, are easier to forecast. But teams like the Brewers, who avoid long-term commitments and rely on players with shaky histories, throw a wrench into the works. Is this a fair playing field for analysts?
Established players are predictable; fledgling ones are not. Large-market teams, with their hefty free-agent signings, often stick with players even if they underperform—a sunk-cost fallacy that makes forecasting easier. Small-market teams, however, have no such loyalty. They bench underperforming players in favor of promising minor-leaguers, armed with insider knowledge that analysts can only dream of. Does this give small-market teams an unfair advantage in unpredictability?
At Baseball Prospectus, we often say that projection accuracy hinges on playing time, not just player measurement. When teams become unpredictable in their lineup decisions, our job doubles in difficulty. We’re left guessing not just about player performance, but also about who will even take the field.
Don’t get us wrong—we love the surprises as much as any fan. But as minor-league data improves, we’re eager to better predict the unpredictable. Even if lineup cards still throw us curveballs, we’ll be ready.
What do you think? Is the lack of standardized minor-league data holding baseball analysis back? Or is the unpredictability part of the game’s charm? Let us know in the comments.
This article is free, but if you’ve enjoyed diving into the complexities of baseball forecasting, consider subscribing to Baseball Prospectus. Your support helps us continue public research and analysis in an increasingly proprietary world. Subscribe now.
