The Myth of Spring Training Records: Do They Predict MLB Regular Season Success?
Spring training is one of baseball’s most cherished traditions. Under the warm Arizona and Florida suns, players shake off the rust, rookies chase roster spots, and fans get their first glimpse of the upcoming season. Games in the Cactus and Grapefruit Leagues spark optimism—or concern—based on early standings. But does a strong (or weak) spring training record actually foreshadow regular season performance?
The short answer, backed by decades of data: No, there’s virtually no meaningful correlation.
Multiple statistical analyses over the years have examined the relationship between spring training winning percentages and regular season results. The findings are consistent: the predictive power is extremely weak.
One comprehensive study of team-seasons from 2003–2017 found an R² value of just 0.05—meaning only 5% of the variance in regular season win-loss records could be explained by spring training performance.
Another analysis pegged it at around 4.8%.
A broader look estimated that roughly 11% of a team’s talent level might be reflected in its spring record, but that’s still far too low to be reliably predictive.
For context, an R² below 0.20 is generally considered weak, and values this low approach statistical noise.
Here’s a visual representation of the data. These scatter plots show spring training winning percentages plotted against regular season results—note the lack of a clear trend line.
Even advanced metrics like spring run differential (which accounts for margin of victory) show similarly poor correlation with regular season success.
Why Spring Records Don’t Translate
Spring training isn’t designed to produce competitive balance or meaningful outcomes:
Rosters are fluid: Starters often play limited innings, while prospects and non-roster invitees get extended looks.
Experimentation reigns: Managers test lineups, pitchers work on new pitches, and veterans focus on building arm strength rather than winning.
Motivation differs: Games don’t count, so intensity varies. A team trailing in the 9th might empty the bench rather than mount a serious comeback.
Small sample size: Most teams play 25–35 games, prone to random variance.
Health emerging from camp often matters more than wins and losses—one analysis noted strong correlations between injury-free springs and regular season success.
Real-World Examples
History is full of teams that defied their spring results:
Strong springs followed by disappointment: Teams like the 2019 Mets or various Royals squads dominated exhibition play only to struggle once games counted.
Poor springs leading to success: Contenders like the Yankees and Braves have frequently posted sub-.500 spring records before thriving in the regular season.
In 2025, the Mariners posted the worst full spring record in franchise history (11-19-2), yet historical patterns suggest it won’t doom their season.
The opposite holds too—bad teams often perform decently in spring when rebuilding with young talent.
What Does Matter in Spring?
While team records are meaningless, certain individual metrics can offer clues:
Pitch velocity, spin rate, and movement tend to carry over reliably.
For hitters, max exit velocity shows some year-to-year stability.
But even these are better viewed as small adjustments to preseason projections, not definitive predictors.
The Bottom Line
Spring training is invaluable for player development and roster decisions, but its standings are entertainment, not prophecy. As one analyst put it: enjoy the games, but don’t bet the season on March results.
When Opening Day arrives, the real test begins—and past springs offer few clues about who will pass it.
