Wednesday, March 30, 2011
A reminder about spring stats
Posted by Brad JohnsonWe all know what everyone says about spring stats: They don't count. Despite that, spring stats still can convey information about a player and they can even reflect a change in a player's skill set. Someone who pays attention to such information can learn something prior to the season despite sample size issues, a weaker slate of opponents, and stats that are not counted for any number of reasons. It simply requires a "user beware" warning.
Leading sabermetricians don't ignore spring stats. Why would they? It's free information and we analytic types love information. According to BaseballPress.com's Nate Springfield, John Dewan of Baseball Information Solutions has successfully predicted breakout campaigns at a 60 percent rate using spring training slugging percentage. You can learn more about his technique here.
With that said, the purpose of this brief article is to remind everyone that small sample sizes are particularly prone to luck. And we can see that luck via BABIP. Below are three surprising performances of the spring that have been talked about as indicative of a breakout campaign.
Asdrubal Cabrera has put together a .364/.426/.600 triple slash with three home runs. He also has a .426 BABIP.
Mark Trumbo has helped alleviate concerns about the injured Kendrys Morales by posting a .297/.316/.662 slash along with six home runs. His BABIP? .407. Two walks against 20 strikeouts is also a little worrisome.
Alex Gordon has people back on his side after throwing together a .343/.459/.729 spring with six home runs. His BABIP was .436.
This is not to say any of these players can't or won't break out this year. Gordon in particular still has some spiffy stats even after you normalize his line thanks to 14 extra base hits out of 24 and 12 walks against 15 strikeouts. But before we get too excited about a breakout spring potentially carrying into the regular season, we need to remember to do our due diligence. Be sure to adjust those spring numbers.
Follow Brad on Twitter @baseballAteam. Email him at pitchin432 AT Yahoo.com
I’m wondering where the hell this 60 percent rate comes from, because I look at that list of 18 guys from 2010, and I see three that had a significant improvement over their career rates—Bautista, Rasmus, and Cruz. Other guys had good years like Tulowitzki), but perfectly in line with what they had already done.
Frankly, I think this system is a load, and I’d like to see the proof that it posts a 60% rate of predicting breakout from previously established levels of performance.