Spring training is now in full swing and in just a few days, the real (fake) games begin! As the Cactus League starts up, a familiar mantra will begin to reemerge: “Spring training stats don’t matter.” The fact that Stefen Romero has a career .361 batting average during the spring has no bearing on his ability to maintain that average during the regular season. Players just don’t accumulate enough plate appearances or innings pitched during spring training for most stats to become predictive—for the most part.

Most of you are familiar with the pitfalls of using small sample sizes to analyze performance. In brief, statistics require a large enough sample size to become predictive enough to begin analyzing a player’s performance. Spring training is the ultimate small sample size trap. Batters see 60 to 70 plate appearances at most and pitchers just 15 to 20 innings pitched. But there are a few stats that stabilize quickly enough that we can glean a few things from a player’s spring performance.

Beyond just watching the Cactus League games with a scout’s eye, here are a few stats that you can pay attention to this spring.

Pitch velocity and movement

The velocity and shape of individual pitches stabilize fairly quickly. Since pitchers are throwing so many pitches per at-bat, the sample grows far quicker than accumulating plate appearances. Unfortunately, there are just a few stadiums in Arizona that are equipped with PITCHf/x technology (the Mariners are one of the lucky ones). Furthermore, the Cactus League PITCHf/x data we do receive is notoriously miscalibrated. The World Baseball Classic will give us some additional data to play with but it’ll be limited to games played in major league stadiums.

For pitchers working on a new pitch, this PITCHf/x data can give us some idea of what that new pitch will look like. We can also monitor pitchers for any extreme jump or drop in velocity—minor differences will either be calibration errors or corrected during the regular season. For the Mariners, pay attention to Felix Hernandez’s fastball velocity during his spring starts. His trainer claims that his offseason workout routine will be able to add a few ticks to Felix’s velocity. This is where he can prove it.

Strikeout and walk rate

By the end of spring training, pitchers who have seen the most innings will have faced enough batters that their strikeout and walk rates will be close to their stabilization points. For these two stats, we have to also take into account the quality of competition during the spring. Many teams are evaluating their youngsters during the spring which doesn’t necessarily mirror the competitive environment during the regular season. Pitchers may also still be stretching out their arms and pitching at less than maximum effort. Batters will also be close to reaching the stabilization point for their strikeout and walk rates—and the same caveats apply for them too.

Again, we should be looking for any major swings in either stat. The pitcher I’m most interested in watching is Yovani Gallardo. He’s coming off a disaster of a year, and if he’s able to regain his velocity and command, he could be a sneaky bounce back candidate. Among the position players, I’m going to be watching Dan Vogelbach’s walk rate and Mitch Haniger’s strikeout rate.

Exit Velocity

A batter’s exit velocity usually stabilizes somewhere around 40 balls in play—the same is true for pitchers (though their data is a little more wonky). Like PITCHf/x, only a few Cactus League stadiums are equipped with TrackMan technology. We may get some additional Statcast data during the World Baseball Classic but it won’t be nearly enough to draw any conclusions from. Pay attention to the beat writers and MLB.com to see if there are any exit velocity stats that are shared during Spring Training. These might be our only glimpses at major changes in exit velocity for batters or pitchers.