keyboard_arrow_uptop

It’s officially the dog days of spring training. The pitcher and catchers finally reported. They started playing games. The Will Ferrell thing happened. Everyone is wearing green today, even though that doesn’t make any sense for some teams whose name starts with “Red”. But a funny thing will happen in these next few weeks. Gone are the days when guys play every other day for four innings. Oh, they’re not up to every-single-day mode yet, but this is the part of spring training where you have to get up to full speed. Opening Day will be here before you know it!

Lately, I’ve been writing a lot about managers. I’ve found that some managers have the ability to help their hitters fight against getting worn down over the course of a season and that this ability can be worth a couple of wins over the course of a season. But on one of those recent articles, someone made an interesting comment. Is it possible that while some managers might see their players fall off over the course of a season, they are simply managers who are good at getting the team ready in spring training, and over time the team naturally regresses a bit? Are there managers who are good at “winning” spring training? If that’s the case, then perhaps some managers are getting a bad rap. They are good at getting guys up to speed quickly and make up their value there.

Once again, I’m using whether or not an individual pitch ends up as a strike as my main outcome. Using 2010-2014 data, I looked at all hitters who saw more than 500 pitches in that season. I focused exclusively on April (and the occasional late March) games. I calculated the chances that a given pitch would end up as a strike given the pitcher on the mound and the league-year average. But for hitters, I used last year’s rate on this metric (percentage of pitches that end up as strikes) to generate my estimate. The reason I did that was that I wanted to see whether managers had any special talent in getting their batters up to mid-season form in the spring.

I used the log-odds ratio method and then entered the manager into a regression equation (logit). I entered in each manager-season as its own separate entity. This gives us an idea for how much we should adjust each individual case upward or downward based on who the manager is. Also, since we are controlling for the expectations that we have for the pitcher and hitter, we can set that to any value we want to give each manager a “standard” set of circumstances. I chose to set that equal to a 50 percent chance that this particular pitch would end up as a strike. Then using the regression, I figured out what the regression believes is the “extra” contribution of a manager. If Bud Black was the manager, would we expect this pitch to be a strike 51 percent of the time? 49? 48.3?

Now, of course, there was variance among the managers, but the true test of whether something is a skill or is just luck is whether performance is repeatable. So, for managers who had managed in at least four of the past five years, I looked at what the regression picked out for them in each year and used an AR(1) intra-class correlation. This technique is sort of like a year-to-year correlation, just that it can take more than two time points. The result? There was no correlation from year to year (.06, for the curious.) Managers who seemed to have hitters who had a hot April one year were no more (or less) likely to have hitters who had a hot April next year.

But maybe it works for pitchers. I used much the same method, again looking at April and checked to see if the pitcher’s manager helped him to get more strikes while he was pitching early in the season, relative to what he had done last year. The answer was an intra-class correlation of .08. Nothing. It seems that there is no special talent that managers have for getting their pitchers ready for the season. There are of course teams who break out of the gate slow and some who break out fast, but it’s more likely to be a product of random variation than anything that the manager did. So while manager seem to have a good talent for helping players manage the grind of a long season, they don’t have any special powers when it comes to Spring Training.

The Season Isn’t Won in March
There’s a saying that you can’t win a pennant in April. Now it seems like you can’t win it in March either. While there is evidence that Spring Training stats actually do have some predictive validity, they mostly are just guys getting back to what they were 5 months earlier. It’s probably the case that Spring Training does exactly what it says on the label. It gets guys ready for the season. Nothing more, nothing less. I suppose that it’s possible for a manager to completely mess it up. He could tell everyone to read Chaucer, rather than run drills, and that would probably set everyone back, but no one would actually do that.

But soon, you’ll start hearing about teams who are having amazing camps and how the players look really good and how they are primed to have a big breakout April. It’s the “Best Shape of His Life” line for teams. And really, what else are you gonna say at this point in camp? There may actually be individual players who are ready to launch into the season, but from what we’ve seen here, it’s not likely that this effect will carry over on the team level. So this spring training season, try not to overthink it. In April, if all goes well, that’s a wonderful thing. And if it goes sour, it’s not fun. But don’t go crediting or blaming the manager. It probably isn’t his fault.

This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.

6/29
0
6/29
1
6/29
9

Related Articles

2/20
0
2/19
0
• Short Relief: The Perfect Number
2/11
0
You need to be logged in to comment. Login or Subscribe
lewist
3/17
Where did Ron Roenecke land? Seems like his Brewers have been ultra hot in April four times in a row
wonkothesane1
3/17
Is Chaucer analytically proven to be the worst spring reading though?
LlarryA
3/18
I'd be likely to give Tolstoy a nod. War and Peace wasn't bad until the end when Leo just stops the story and goes off on his political rants. And my wife HATED Anna Karenina for the utter humourlessness.

Finnegan's Wake (James Joyce) would be an interesting study, it ought to either win big or fail big. Kind of like Ozzie Guillen.