Do spring stats matter? Perhaps more than you think

ISO, baserunning aggressiveness can offer hints at regular-season performance

Do spring stats matter? Perhaps more than you think

From the Grapefruit and Cactus Leagues to a computer screen near you, baseball players and pundits have long repeated the same refrain: Spring Training stats are meaningless.

Regardless as to whether you agree with the decree, calls to ignore preseason data often seem dogmatic, even harsh.

And according to at least one baseball researcher, demonstrably wrong.

Dan Rosenheck, a sports editor at The Economist, defended the predictive value of certain Spring Training stats at the Sloan Sports Analytics Conference in Boston on Saturday. The stats in question -- often referred to as peripherals, which include strikeout and walk rates, isolated power (ISO, a simple way to calculate a hitter's raw power by taking a batter's slugging percentage and subtracting his average) and baserunning aggressiveness -- tend to stabilize faster than their surface-number counterparts. As such, one could argue that peripherals begin to hold predictive weight in relatively quick order.

"It is true that the correlations between Spring Training and regular-season numbers in most fantasy baseball categories, such as batting average or ERA, are very weak. However, those statistics also show very weak relationships from one season to the next; pitchers frequently go from having an elite ERA one year to an awful one the next," Rosenheck said via email.

"We know that you get better year-to-year predictions by focusing on less noisy stats that stabilize faster, such as strikeout rate and walk rate. So why don't we apply those principles to Spring Training as well?"

In cases when a player lacks big league experience -- think the green and the oft-injured -- Rosenheck ascribes an even greater importance to spring stats.

"When we have very little other information about a player -- such as rookies or Cuban imports -- Spring Training is the best data we have and should be given a fair amount of weight," Rosenheck said via email.

To be clear: Rosenheck does not believe that spring stats are incontrovertibly prescient; rather, they can be used to guide a spring projection toward the regular-season truth.

For example: Harken back to 2010, the year Jose Bautista "came out of nowhere" to belt a big league-best 54 home runs.

Bautista mashed well enough that spring -- in fact, his performance was more than two standard deviations above the mean -- to warrant a sizable regular-season projection adjustment (for comprehensive baseball projections throughout Spring Training, visit In 57 Grapefruit League at-bats, he posted an outrageous .420 ISO with just one strikeout.

Another example can be found in data from last year, when the Indians' Michael Brantley finished third in the American League MVP Award race.

Although Brantley entered 2014 with a reputation as a solid player, few viewed him as an offensive star. Spring stats, however, seemed to suggest that the Tribe outfielder was set to move up in class. Across 18 Cactus League contests, Brantley tallied a .240 ISO, with five walks to just two K's.

But as most know, outlier performances don't always forecast future success. Sometimes, they serve as red flags.

Just ask Anthony Rizzo, whose strong first half-season in the Majors during 2012 led many to anoint him as the face of Windy City baseball. However, three doubles and no homers in 48 at-bats during Spring Training 2013 resulted in a .063 Cactus League ISO.

The aforementioned anecdotes aside, though telling, represent a sample of extreme outlier performance. Spring stats that fall closer to the mean do less to move the projection needle.

So what's the takeway? Should hitting and pitching spring stats be taken seriously or with a grain of salt?

Pay attention, Rosenheck said, but focus more on a third element of the game: basestealing aggressiveness, arguably the most predictive spring stat of them all.

This makes sense. Steals, unlike most stats, have about as much to do with organizational philosophy as a player's actual ability level. So an aggressive spring basestealer would seem to have both the talent to run and his manager's confidence to do so.

With games having recently begun, the baseball community is still weeks away from parsing March data and predicting the risers and fallers for April and beyond.

But by tracking spring stats and making player-performance adjustments through the proper lens, you could prove to be the most prescient among your friends.

Zachary Finkelstein is a fantasy editor for This story was not subject to the approval of Major League Baseball or its clubs.