clock menu more-arrow no yes mobile

Filed under:

Sabermetrics: Tyler White's Spring Stats

What Do Tyler White's Spring Stats Mean? Some Spring Batting Stats Mean More Than Others.

If you buy something from an SB Nation link, Vox Media may earn a commission. See our ethics statement.

Stacy Revere/Getty Images

At sabermetric-oriented blogs like TCB, you will frequently read, "Spring training statistics are meaningless," or "Ignore spring stats."  As a generalization, these statements make sense.  But some spring training stats have meaning in the context of contributing useful information to gauge a player's future.  We're not talking about all spring stats--batting average is the first spring stat which many people cite, but it isn't meaningful.  However, some batting stats that begin to stabilize in smaller samples may provide a small but useful predictive contribution.  The explanatory power may be relatively minor, but not meaningless.

One of the amazing stories of the Astros' spring is Tyler White's strong spring training performance which earned him a spot as the Astros' first baseman.   Focusing on the spring training batting stats which have some predictive value, what should we think about Tyler White's spring training offense?  Does his strong spring foretell a possible breakout as a major leaguer?

To gauge this question, Dan Rosenheck's research for The Economist, and award-nominated presentation at the MIT Sport Analytics Conference, will be used to identify spring training statistics which have some correlation with regular season performance.  His research defines such correlations as useful if the addition of these spring stats increase the predictive capability of regular season projections developed by systems such as ZIPS.

The three spring batting stats which may contribute some predictive value are walk rates (BB%), strike out rate (K%), and ISO on Contact (ISO-con).  K% and BB% are well recognized as batting stats which tend to stabilize quickly.  ISO-con is based on isolated power (ISO)--a mainstay for measuring power--utilizing batted ball contact, rather than at bats, as the denominator.  ISO-con also requires a smaller than normal sample to stabilize.  Again, it's not surprising that substantial increases in power during the spring are considered meaningful.  John Dewan has long contended that very large increases in a player's slugging percent in the spring are predictive. This 2013 Beyond the Boxscore article, working off of Dewan's contention, suggested that Jason Castro could have a power breakout based on his 2013 spring training performance.  (Was this right?  Yes. Look at Jason Castro's 2013 regular season stats.)

The research is aimed at identifying whether the spring training statistics can contribute any useful information to player projections.  For that reason, we will compare Tyler White's spring stats to his 2016 projection.  The Fangraphs Depth Chart projection, a combination of Steamer and ZIPS projections, will be used for this test.

Rosenheck says that the spring training stats tend to be more useful for young players with little or no major league experience. (We can check that box for Tyler White.)  In those cases, the projections are based primarily on minor league statistics.  Thus, the selected spring training stats are more likely to contribute to the predictive power of the projections.

The table below compares White's spring BB%, K%, and ISO-contact to the comparable rates from the 2016 projection.  I calculated the ISO-contact rates from the projection data on White's Fangraphs player page and the spring training batting data at  White's spring training batting stats are based on 61 plate appearances, by the way.

Tyler White

K% BB% ISO-con
Spring Stats 16.4% 14.8% 0.293

2016 Projection 18.0% 10.3% 0.174

Spring Difference -8.9% 43.4% 68.1%

As you can see, above, Tyler White's spring batting stats are better than his projection, across the board.  The lower K rate may represent a minor difference, but lower is better.  The projection expects White to experience a substantial drop off from his minor league walk rate (15% - 18% in A+, AA, and AAA).  But his spring training walk rate is more in line with his minor league walk rate.  The big improvement over his 2016 projection is shown by the ISO-contact.  The magniitude of the increase in power on contact is significant enough to have some meaning.

The Astros' spring training is located in the Grapefruit League which has its own stadiums and weather conditions that vary from the Cactus League as well as the major leagues.  Since these factors may affect power numbers, I revised the comparison to reflect the Grapefruit League average ISO-contact and the major league average ISO-contact.  The Grapefruit League average ISO-contact is shown in Rosenheck's MIT presentation.   I calculated the 2015 ML average ISO-contact for non-pitchers from the Fangraphs leaderboard.

The statistic I developed is shown as ISO-con +, which is based on a league average of 100, similar to wRC+.

Tyler White ISO-contact Compared to League Average

ISO-con avg. ISO-con+

Grapefruit League 0.178

2016 Projections 0.197

Based on comparisons to league average ISO-contact, White's spring training performance shows an even stronger improvement over the 2016 projection.  The projection expects White to exhibit below average power at the major league level.  White's Grapefruit League performance showed much higher power than the average for that league.

At the risk of placing too much weight on the small-ish sample that is spring training, these comparisons provide a reason for optimism about White's ability to hit at the major league level.  His spring performance suggests that the projections understate Tyler White's likely batting line.  Tyler White has worked on increasing his power, with tangible results in the Dominican Winter League, and the spring performance provides some hope that his power will be better than expected in the majors.

The recent fangraphs' article "Let's Find Some ML Comps for Tyler White" underscores the reason that power improvement is so important for White.  The article concludes "For the most part, White’s comps aren’t super inspiring." As a general rule, first basemen need to have batting skills beyond controlling the strike zone in order to stick at the position.  As the article notes:

For White to have a successful career as an average-or-better everyday first baseman, it feels like he’s going to have to add some power, like Carpenter did last year. White stands at just 5-foot-11, which may not bode well for a future power spike, but at the very least, he’s cognizant of the fact that a bit more power could go a long way to his game.

What do you expect from Tyler White this season?