Spring Training stats should not determine starters

In a little less than two days, pitchers and catchers for your Florida Marlins will report. Stretching and conditioning will be done, and even though it has been fairly cold for South Florida standards, baseball will be played. And with Spring Training and baseball just around the corner, I am reminded by Dave Cameron of FanGraphs of one very important thing.

None of it means anything.

Sometimes us stat geeks espouse on how little a certain sample size is in terms of determining how good a player is. You certainly wouldn’t judge how good a player is by looking at his stats in June. Yet everywhere around the league, players will be fighting for positions on the roster based almost entirely on their batting average or ERA during Spring Training. And this, my friends, is an atrocity.

A little closer to home

There are two primary “battles” that will be going on this Spring Training for the Marlins. There is an open call for five starters to try and fill in three spots, though barring anything catastrophic, two of those spots are already guaranteed basically to Chris Volstad and Anibal Sanchez. That leaves Sean West, Rick VandenHurk, and Andrew Miller to fight for the fifth spot in the rotation. The other battle is at first base, where Gaby Sanchez and Logan Morrison will duel to see who gets to bat eighth for the Marlins this year.

I guarantee that none of these players will significantly separate themselves from their respective packs during Spring Training. I doubt any of them will bring about enough separation for them gain a starting job convincingly. Yet one of the two first basemen could easily bat .350 while the other bats .270 in their 100 respective PA, and the guy that bats .350 will have been “impressive” enough to win the job.

What’s the difference between a .350 hitter and a .270 hitter in that time scale? Assuming a walk rate of 10% for either gentleman, the difference is about seven hits, the difference between 24 and 31 hits. Significant? Not particularly, given the small sample and even smaller gap. And beyond that, someone will credit this seven-hit gap to the excelling player’s “new workout regimen” or his “offseason diet that has them in the best shape of his career,” or some other BS like that. Trust nothing of the sort! Most of these stories come out after a strong performance, which lends credence to the idea that only players who have good stats in Spring Training have “renewed work ethic” or “a new training schedule involving sleeping in an isobaric chamber” or something.

The need for scouting

Spring Training is mostly for established players to stretch out the kinks before the season starts and for young players to try and get on the major league roster. Performance in terms of batting average, wOBA, ERA, or FIP don’t matter. What does matter is what the coaches and scouts see when these players perform. This is where the need for scouting and proper coaching becomes most evident, especially when dealing with younger players who may actually have flaws in their game. You probably don’t have to key in too much to Hanley Ramirez or Dan Uggla no matter what their stats look like in Spring Training, but you probably do have to watch how Cameron Maybin’s swing is looking or out of what arm slot Sean West is throwing.

The importance of this critical scouting function can be diminished when you have statistics in front of you. I prefer my scouts stat-free, because (if I was a general manager or something) I’m paying my scouts to watch for problems or improvements, not to do statistical analysis. I think a lot of what managers do wrong is put good faith into small samples while not doing what they and their coaches are really supposed to do, which is use a critical eye and their experience to watch for problems in their players. If a coach tells me a player has a kink in his swing, I won’t argue too much because I can’t tell. If his reasoning is that he’s batted .200 the last two weeks, I’m going to dismiss his notion. Scouts and coaches need to use their eyes in Spring Training the most, because at this point everyone is just trying out stuff to see how it works. Leave the statistical analysis to people who may better understand it.

Why does it matter?

It doesn’t, right? Statistically, no. I think ST still serves a purpose, no matter how meaningless the games are. Whether us fans watch it or not is irrelevant. What I want to see and hear out of ST is that the coaches are watching and evaluating players based on their skillsets and not their stats. It’s a good thing ST isn’t televised locally, because I think I would tear my hair out hearing about jobs being won or lost against inferior competition and based on small samples.