9 things you need to know about the shift

The shift has essentially taken over baseball during the past decade, going from an oddity when it began to crop up 10 years ago into such a regular occurrence these days that it's almost weird when certain hitters aren't being shifted against. More than 17 percent of plate appearances end on a shift these days, up from just over 13 percent only two years ago. (It is not a new idea; Ted Williams saw versions of it 70 years ago.)
But what does a shift -- generally considered to be three infielders to one side of second base -- actually do? Maybe not what you think.
Many smart researchers have attempted to figure this out, but they were restricted by the limitations of the available data. Today, at Baseball Savant, in both leaderboard and searchable form, we've made Statcast™ shift and positioning data available. Now you can do things like find out which batters have had to face four outfielders (Joey Gallo, Logan Morrison, Kyle Seager and Justin Smoak) or which center fielder plays the deepest (Alex Gordon) or which first baseman plays the furthest from the bag (Adrián González).
As importantly, for the first time, you can see what the shift does when the ball is not in play, and while that may not seem like a big deal, it really is. Sure, the shift swallows up grounders, but it also changes the way hitters and pitches approach their jobs. What if the true effect of the shift is to increase walks, not prevent infield hits? What if the shift doesn't actually work at all?
Let's find out. Here are nine things you need to know about the shift, right now.
1. No team uses infield shifts more than the Astros.
This probably shouldn't surprise you, because Houston is the team that does things like this against Gallo.

So far in 2018, the Astros have shifted on 43.5 percent of plate appearances, the most in baseball. In '17, they shifted 34 percent of the time -- the most in baseball. In '16, it was also 34 percent, which was -- wait for it -- the most in the Major Leagues. 
Shifting doesn't by itself make you successful, of course; the Orioles have the third-most shifts this year, and the Cubs have the fewest. But it's clearly a strategy the defending World Series champions use well. They're responsible for nearly three-quarters of all four-man outfields in 2018, too.
2. No team has added more shifts than the Royals.
Around the Majors, teams are shifting about five percent more than they did last year, and most clubs have held steady. Twenty-five teams are within 10 percentage points -- one way or the other -- from last year. Four are shifting 10-15 percent more often ... and then there is Kansas City, which is shifting an astonishing 34 percent more than it did in 2017.
After years of being reluctant to use the shift -- they were 28th in 2016 and 25th in '17 -- the Royals were open about attempting to use it more in '18.
"Let's give it a wholehearted try and see," manager Ned Yost said this spring. "Let's get out of our comfort level a little bit, or at least me out of my comfort level."
The numbers back that up. Kansas City is now second behind only Houston in percentage of infield shifts, and no team shifts against right-handed batters more often than the Royals (36 percent of the time). The biggest decline? That'd be in Milwaukee, where last year's third-highest-shifting team has dropped down to league average.
3. No righty hitter is getting shifted against more than Kris Bryant (54.8 percent).
You'd think maybe this would be the slow-footed Albert Pujols or the notoriously pull-heavy Brian Dozier, but the only righty hitter to see a shift more than half the time is Bryant. When you look at where he's put his ground balls since the start of 2017, it's not hard to see why. In '18 alone, Bryant has pulled 84 percent of his grounders.

If anything, the question here may be why it took so long, after being shifted 39 percent of the time in 2016 and 33 percent in '17. It's been nearly two years since we wrote about Bryant's near-total inability to hit with power to the opposite field, and over the past three seasons, very few regular hitters have performed worse to their opposite site. He grounds everything to the left, and he can't hit with power to the right. Why wouldn't you shift him?
4. The Yankees and Cardinals make minor adjustments the most.
We define shifts as "three infielders to one side of second base," but that's not the only way a team can change their positioning. You might have a second baseman playing deep, while everyone else stays home. Maybe the shortstop plays behind the pitcher, but not quite to the right of the second-base bag. For now, we group those other non-traditional, non-three infielder plays as "strategic" positioning.
No team is doing it on more of their pitches than the Yanks, who are doing it 18.1 percent of the time. No team has had it in place to end more plate appearances than the Cards, who have had it on 17.3 percent of the time. Strategic positioning changes don't look like overshifts, but they have impacts, too.
5. The shift does a good job of lowering batting averages. (But there's a lot more to it.)
The entire point of the shift, as it's generally understood, is to take away singles by positioning an extra infielder where one wouldn't have been before. While it gives some singles back by moving fielders out of their usual positions, or by allowing some easy bunt hits for those who dare to take them, it does this part of its job well.

You might be tempted to simply look at league production vs the shift (.276 batting average on balls in play since the start of 2017) against the same without a shift on (.306) and say that it's working, but that's not really a fair way to do it. Not everyone is shifted equally, so you're then comparing two different groups of players with different skills and talent levels. Chris Davis is almost always shifted, for example, while Jose Altuve never is.
Instead, from this point forward, we compared the 201 hitters who saw at least 100 pitches both against the shift and not against the shift since the start of 2017. Then we compared just how those hitters performed against the shift and not against it. We're talking stars like Aaron Judge, Charlie Blackmon, Mike Trout and Giancarlo Stanton, as well as lesser-known names like Steve Pearce, Daniel Nava and Austin Hedges.
When making contact, those hitters batted .299 without the shift and .281 against the shift. It hurt righties (.302 without the shift, .277 with the shift) even more than lefties (.298 without, .283 with), but it lowered averages on both sides.
The shift does its job. Right? Well ...

6. The shift also looks like it increases walks.
There's more to life than just singles, obviously. Now that we have pitch-level data on positioning, we can look at what happens when the ball is not put into play. Here, it's pretty clear. Our group of 201 players walked 9.8 percent of the time against the shift and nine percent of the time without the shift. Eight-tenths of a percent doesn't sound like much, does it? As it happens, that's almost exactly the jump we've seen overall in the Majors from 2015-17. On that scale, that was almost 1,800 more walks. It's hard to say why, for certain, but it's happening.
7. The shift may be slightly increasing home runs, too.
There's a few reasons for all the home runs in baseball these days, and the shift may be one of them. Justin Turnersaid as much when we asked last year, when he said "you don't beat the shift by hitting around it or through it, you beat the shift by hitting over it."
Since the start of 2017, 4.6 percent of plate appearances against the shift have ended with homers vs. just three percent without the shift. That's a bit of a flawed look for reasons we stated above, so again looking just at our 201 players who saw both the shift and not the shift, we saw home runs on 6.7 percent of all contacted batted balls against the shift and 6.1 percent without the shift.
Again, small increases, but across many thousands of batted balls, small increases add up to dozens or hundreds more occurrences.
8. The shift may decrease the number of fastballs hitters see.
Our sample of 201 hitters -- the ones who have seen at least 100 pitches against the shift and also not against the shift -- saw 52.9 percent fastballs (four-seam, two-seam, sinker) when the shift was on. When they were facing a non-shift alignment, that jumped to 54.7 percent, which is to say, a few thousand more fastballs. 
9. The net effect of the shift may be far less than you think.
So the shift takes away some singles, more than it gives back. It may add back some homers, it probably increases walks and ... well, what does it actually end up with?
If we look at our 201 players again, we can see that their wOBA (a number just like OBP, except it gives more credit for extra-base hits) against the shift in 2017-18 has been .336. Without the shift, it was ... .335. That's close enough as to be essentially identical. For lefty batters, it hurt them a little -- .326 vs .334 without the shift. (The righties, interestingly, performed better against the shift.) 
Of course, even this is incomplete. This is measuring what happened, not what didn't happen -- that is, the value of getting a Gallo or a Davis to change his primary swing in the first place. That's an entirely different conversation, however. The more we learn about the shift, the more we learn there's much more to learn.