Did you guys know that Spring Training starts this week? It does! Finally, after more than 20 weeks without the Mariners, baseball is back. However, because it's been so long and people are so excited for BASEBALL, once Spring Training starts in earnest and new statistics begin to pile up, quite a few folks tend to put a bit too much stock in these Spring Training numbers. Most people who frequent Lookout Landing seem to be at least somewhat aware of small sample size-related pitfalls, but it can still be easy to fall into this trap. You should absolutely rejoice in the return of baseball, but please make sure to do so responsibly. More often than not, Spring Training numbers don't mean a lot.
For example, take a look at the following following table. This compares some offensive numbers that have been put up by the Mariners during Spring Training and during the regular season over the past three seasons.
The offensive numbers generated by most teams down in/around Phoenix tend to be fairly inflated. This is due to a lot of factors, including smaller, hitter-friendly ballparks; effects from the warm, sunny weather; and the use of absurdly ineffective minor leaguers to finish up some Spring Training games.
If Dustin Ackley happens to have a monster ST this year, with an OPS above 1.000, it's pretty clear that we wouldn't necessarily expect him to perform similarly in the regular season. Right? That much is obvious. But maybe there's still some sort correlation between how well a player does in Spring Training versus how well they do in the regular season? Maybe? To find out, I put together the following chart, which shows the Spring Training OPS vs. the regular season OPS for every Mariners player over the last three years who had at least 40 AB during Spring Training and then went on to have an appreciable amount of playing time in the following regular season. (I'm only going to look at offensive stats during this post. Maybe I'll come back and do an analogous pitching-related post later.)
Hey! Look! A slight positive correlation! Now if only that R-squared value were hundreds of times larger this might mean something... but it isn't. So it doesn't. Although a player's performance during Spring Training might be reflective of how he'll perform in the regular season (e.g., Logan Morrison had a 2014 ST OPS of 0.735 and a 2014 regular season OPS of 0.735 - wow!), this chart suggests that it's pretty futile to try and predict a player's regular season performance based upon their Spring Training numbers. A few key examples:
- In 2013, Michael Morse and absolutely destroyed the opposition during Spring Training, putting up OPS's of 1.332 and 1.149, respectively. Unfortunately, their regular season numbers were ~half as good: Morse had an OPS of just 0.651 and Montero had an OPS of 0.590 in ~100 PA before being demoted.
- In 2014, Kyle Seager struggled in Spring Training, managing an OPS of just 0.518 (the lowest of anyone on the team not named Willie)... and then he went on to have a breakout season, putting up a career-high OPS of 0.788, and earning a selection to the All-Star team, a gold glove, and a $100M contract. Woo!
Looking at these numbers, I'm reminded of the fact that Morse had a killer start to the season in 2013 (before hurting his hand) whereas Kyle had a very slow start in 2014. So maybe Spring Training numbers, while not necessarily indicative of a player's whole season, can at least provide insight into a player's performance during the beginning of the season?
*Includes contribution from Brendan Ryan in 2013, but his 0.386 April OPS messed with the x-axis range, so his pitiful data point is not shown.
Oh. I guess not... This plot provides us with an even weaker correlation (and a negative one, at that). It's important to remember that these are small sample sizes, but that's kinda the whole point of this post... Spring Training is lousy with small sample sizes. Try not to forget this fact.
Of course, it's almost certainly better for a player to perform well during Spring Training, as opposed to playing poorly. With that in mind, it's good to get excited when Stefen Romero hits a ST home run. And it's okay to be disappointed if Seager has two strikeouts and a pop up in first ST start. Just do your best to avoid letting those feelings influence your regular season expectations too much. And don't let them cause you undue anxiety or worry. And whatever you do, don't camp out in the Lookout Landing game thread during the second week of April, after Austin Jackson maybe gets off to a slow start, complaining about the fact that Lloyd should be starting James Jones in center field because Jonesy had a 0.390 batting average in Spring Training. Please don't do that. If you do that, no one will like you. They'll probably leave you some (deservedly) snarky comments. Thank you in advance.
2015 baseball is almost here. I'm excited.