More machine than man: Managing in Major League Baseball

In Star Wars Episode VI: Return of the Jedi, Obi-Wan Kenobi imparts this advice to Luke Skywalker regarding nemesis Darth Vader:

“He’s more machine now than man”

Right now in Major League Baseball, that statement could also be used to describe the state of most managerial situations. Once upon a time, the manager brought his own brand of baseball to a club, and then the general manager supported him in that task. But now, with advanced analytics embedded into every aspect of decision-making, the relationship seems to be reversed: the GM sets the team philosophy and the field general is a conduit for it.

Twins fans saw that play out before their eyes in Game 2 of the recent Wild Card series. After pitching five innings of one-run ball, Jose Berrios was lifted from the contest having thrown just 75 pitches. Considering La Makina could be seen pleading with Rocco Baldelli to let him continue, the only rationale for the quick hook seemed to be the oft-cited “third time through the order” bugaboo, in which the quality of pitching returns diminishes each plate appearance from the opposing hitter.

Quick sidebar here: I cannot pin the WC series loss on this single event. Score two runs in two games, and a team will most likely lose both of them. However, the decision to pull Berrios is potentially troublesome for its lack of “reading the room” capability.

Upon getting George Springer to ground out and end the top of the 5th, Berrios exhorted mightily in a pump-up gesture towards his teammates. In the bottom of the inning? The Twins rally and tie the game. It seemed like an emotional turning point—until the air leaked out of the balloon when Jose wasn’t allowed to continue.

What do the numbers say? I don’t have a subscription anywhere to compile large statistics easily, but here are some tables (courtesy of Baseball-Reference.com) that tell the story of times-through-the-order stats by starting pitchers the last few years:

The categories for starting pitchers do tick in the wrong direction the more times a batter is faced. However, I would have expected a more dramatic turn for the worse rather than the gradual uptick seen here. Interestingly, SP numbers actually get better the fourth time through an order. Small sample size for sure, as pitchers simply don’t reach that rare territory anymore, but it would seem to indicate that when they do, they succeed.

More specifically, here are the career numbers for Berrios...

Baseball-Reference.com

Jose’s career pretty much mirrors that of MLB as a whole the last two years: great the first go-round, worse the second, and about the same to a little worse the third time through. Nothing in that data seems to indicate a “fall off a cliff” between the 2nd and 3rd time through the order.

To be honest, this entire conversation is one that troubles me. Though analytics strive to provide the best possible outcome in any scenario (as well as shield the decision-makers from criticism), is managing strictly by the numbers always the best—or most entertaining, though that is a different conversation—approach all the time?

To quote another ‘80s movie:

“He’s not a machine—he’s a man!”

Perhaps Rocco Baldelli agreed with the decision to yank Berrios. Perhaps not. Only he can answer that query. However, if he is strictly managing on what the numbers tell him, it begs the question of why he needs to be in the dugout in the first place. Yes, that statement is a bit facetious—someone needs to direct on-field decisions—but how much are current skippers merely figureheads for the number-crunchers as opposed to actual decision-makers?

Numbers can tell us what has happened, as well as the likelihood of what will happen, in any given scenario. But they are not perfect predictors of future success, especially in small sample sizes (like, say, a three-game playoff series). A manager’s gut instinct is no better, of course, and over the long haul is likely to be considerably worse in terms of decision-making. But right now, as with so many other aspects of life, there seems to be no room for middle ground. Over a 162-game grind, I’m much more comfortable playing probabilities. But in a short playoff series? There might be a need to play a hunch or read a situation in the moment. Right now, by and large, MLB managers are not given the chance to do that.