We're so used to seeing the gifted Starlin Castro so natural and comfortable at the plate that seeing him struggle makes us incredibly uncomfortable. We want to find answers.
There have been a lot of attempts to analyze Castro and most of them seem to conclude that he is getting worse. Perhaps no piece exemplifies this than the one Dan Bernstein wrote a few days ago.
Bernstein attempts to make the piece look analytical by quoting new metrics, most of which focus on current value, not predictive numbers. So all it really does is prove one thing: Starlin Castro is playing poorly. It doesn't mean he is worse. The unfortunate truth is that you get more listeners when you focus on the negative on talk radio. The article struck me as one where the conclusion was decided on before the research. That often leads to poor analysis because you are looking for numbers that prove what you've already decided to believe. There's a saying I like to quote from the Scottish poet Andrew Lang:
An unsophisticated forecaster uses statistics as a drunken man uses lamp posts -- for support rather than illumination.
With all due respect, I think that applies here. This is not to call Bernstein "unsophisticated", but I do think the analysis itself can be described that way.
To be fair, Bernstein backtracks and does say this is a small sample size and there is plenty of time to re-write this. That is probably the most important thing he wrote, yet it gets lost in the mumbo-jumbo of what I call snapshot analysis -- which is to pick a particular moment in a player's career, quote results-oriented numbers, and throw out everything else the player has done in his career to that point. We're talking about two months worth of numbers for a player 4 years from his peak years.
But this is more than just small sample size.
You'll often here me talk about process and how they often precede results. There are always growing pains, even for a 3 year veteran like Castro -- especially when they're just 23 years old. Especially when that player is asked to reverse lifelong habits and learn how to succeed with a new approach and adapt to a new organizational culture. One important step in this process is an increase in the pitches Castro is seeing at the plate, from 3.46 to a career high 3.85 this season. That is just one step, however, as it is clear that what Castro needs to learn is to take pitches with a specific purpose and develop a plan to set up pitchers. Right now it's clearly the other way around, but it's early in the process and according to his hitting instructor James Rowson, Castro has the intelligence to eventually figure this out.
We talked about process when we analyzed Jeff Samardzija as a starter as he made his transition to a starter a couple of seasons ago. More to the point, we talked about it when we said Jorge Soler's good pitch recognition skills would eventually translate to walks -- which is what we've also been saying about Albert Almora this year. We've also written about Javier Baez's transition and how, despite the low walk total, there were signs that the process was improving as he was becoming more adept (if still inconsistent) at working counts and waiting for a pitch to drive. It's an idea that has since been backed up by his manager Dave Keller and Baseball America's Ben Badler.
The problem is that measuring process is much more difficult than measuring results. It's abstract. It requires a certain level of comfort with uncertainty. Results are easily defined. Process is not. There is no easy answer and you can sometimes be hopelessly wrong, which is the nature of the game with process-oriented analysis. I believe that a team that can figure out how to do this well will have a competitive advantage over those that strictly measure results.
So, back to Starlin Castro. Because there is no easy answer, I've asked a few of my baseball industry contacts -- all of them with a strong combination of statistical and scouting knowledge. I talked to those in other organizations to get a more unbiased view. The conclusion was nearly unanimous.
Starlin Castro is in a slump.
That's it. Not one of them was worried about an actual decline in his physical skills. There was acknowledgement that he was struggling with his approach but that the struggle was almost certainly temporary.
Jed Hoyer also talked about this when he talked with Buster Olney about players improving their approach and that if they didn't, they'll have to find players who will. Of course, that immediately brought up the subject of Castro. This is how Hoyer replied,
"He's just in a slump right now. He's not a guy that we really worry about. He's such an unbelievably gifted hitter. I'm pretty confident that he's gonna have a .350-.400 batting average month and he'll even out his stats a bit. He is 23 years old. Sometimes we forget how young he is because he was up in the big leagues at 20. I think he does need to improve his plate discipline over time. I think he will. I think he's going to grow into a lot more power. I think Starlin...he's a big part of our future and he'll figure out the on-base thing. I think right now that he's frustrated because a lot of hits haven't fallen in but with his hands and his gifts he'll have a good month and get right back on track."
That pretty much echoes what the people I talked to said about Castro. Once he has that good month and gets back on track, there will be new numbers to analyze -- ones that will likely lead to very different conclusions.