Don’t expect baseball’s defensive shifts to keep working

A defensive shift by the Boston Red Sox. Barry Chin / Globe Staff/Globe Staff

One of the first things kids learn in Little League is where they’re supposed to stand. If you’re the first baseman, you go here. If you’re playing centerfield, you’re out there. The basic arrangement has been consistent for over a century. Now, though, things are changing.

It’s called the defensive shift. More and more, you find the second baseman on the wrong side of second base, or the third baseman back in the rightfield grass. Red Sox fans may remember the shift that was used against Ted Williams in the latter half of his career, or the variations tried against other players through the years. Recently, though, defensive shifts have become a much more common part of the game. According to data from Baseball Info Solutions, there were three times as many shifts in 2013 as in 2010, and so far in 2014 we’re on pace for another huge increase.

Clearly, Major League teams think there’s value in defensive shifts, but one theory suggests it may not work in the long run. It’s called the Lucas critique, and it reminds us that any change in tactics creates new, sometimes unpredictable risks. Defensive shifts may help teams plug holes that hitters have exploited in the past, but these same shifts open up huge spaces that hitters may learn to target in the future.

What’s the Lucas critique?

The Lucas critique is based on a simple but powerful insight formalized by the economist Robert Lucas. It says that you have to be really, really careful whenever you rely on historical data as a rationale for changing your policies or your tactics. Why? Because all of that historical data comes from a time when your new policy wasn’t in effect.

Let’s say, for instance, that you have reams of data about all of the robberies in the United States for the last hundred years. In all that time, no one has ever robbed Fort Knox. Yet, it would be a huge mistake if you decided to remove all the guards from Fort Knox and transfer them to more vulnerable places. You’d be handing thieves a new opportunity, one they’ve never had before. The whole reason Fort Knox has never been robbed is because it has been heavily guarded. The moment you change that, all of your historical data become meaningless. It’s based on a set of conditions that no longer apply.

What does this have to do with baseball?

Imagine you have detailed scouting reports showing that the most dangerous hitter you’re facing tends to pull the ball. When he comes up you might decide to shift all your players and overload one side of the field. That would certainly make it harder for him to pull the ball for a hit, but it also creates a new opportunity. The opposite side of the field is now totally unguarded, the baseball equivalent of Fort Knox. And you really don’t know if your opponent can take advantage of this big open space, because all of your scouting reports are about how he performed against the old, standard defense.

Have defensive shifts been working?

The fact that defensive shifts are being used more and more suggests that teams at least think they’re working.

We also know that as shifting has become more common, league-wide batting averages have been falling, though there may be other reasons for that, including rising strikeout rates and increased steroid testing.

Perhaps the best evidence for the value of defensive shifts is the impact they’re having on particular players. David Ortiz, for example, has come up against a lot of shifts in the past few years, and it looks like they’ve made a difference. Since 2013, when Ortiz has hit ground balls and short line drives -- the kinds of situations likely to be affected by a shift -- he has averaged .230 against the shift and .350 against a traditional defense. Mike Napoli, for his part, has hit .217 with a shift and .275 without.

Why wouldn’t they continue to work?

Defensive shifts don’t just cover vulnerable areas of the field, they create new ones. Some players, like Sox catcher A.J. Pierzynski, seem to have quickly figured out how to take advantage of shifts. Since 2013, Pierzynski has actually hit better against the shift (.366) than against a traditional defense (.198). Other players, like Ortiz and Napoli, might make similar adjustments.

Then again, they may not be able to adjust. Hitting is a very time-constrained, reflex-based skill, which doesn’t leave much room for conscious changes.

But even if some players can’t adjust, their teams can. The Red Sox could change their scouting and selection priorities to emphasize spray hitting and offensive unpredictability. There are players out there who rely on offensive versatility, and those players might well become more important and more valuable over time, while the pure pull hitters who can’t adjust become a kind of historical relic, like great spitballers.

What would Lucas say?

Whenever you’re trying to gauge the likely benefits of a new approach, in sports or in politics, remember Fort Knox. It’s not enough to think about the problem your new approach will solve. You also have to consider the risks you might inadvertently create. Defensive shifts may help teams plug holes that hitters have exploited in the past, but they also may open up huge spaces for hitters to target in the future. Sooner or later, hitters are going to find those spaces.