The Royals Try a New Shift

After a decade of hand-wringing and tedious arguments on both sides, MLB restricted defensive shifts this past offseason. Much has already been written about the pros and cons of this decision, and I’m not going to take the time to recapitulate all of those arguments here. One debate in particular really caught my eye, though: Would teams still play an overshift-esque alignment by moving an outfielder to the shallow right field position occupied by shifted second basemen in pre-restriction shifts?

I expected it to be a rare tactic, but still one that came up from time to time. Five-man infields already existed; in fact, I ran the math on when they might make sense in 2019 when the Dodgers tried one. The exact conclusion of that piece isn’t important; the point is that teams sometimes thought a five-man infield was the best defensive alignment when any defense was allowed, so they would surely prefer it with restrictions on other alignments in place.

One thing I absolutely didn’t expect: a five-man infield against Joey Gallo. The Royals had other plans, however. They pinched in their right fielder and rotated the rest of the outfield to compensate for it against him in his first at-bat of the season. The result was amusing, particularly if you’re a fan of the best laid plans going awry:

The Royals weren’t giving up so easily, though. They continued running a five-ish man infield against Gallo, and they’ve even expanded their use of the tactic against other left-handed hitters. Per MLB data architect Tom Tango, they’ve already used a roughly similar alignment 19 times this year:

This is the 2-OF configuration that the Royals are using https://t.co/lS97gpEt4T pic.twitter.com/cYISIa9Wnn

— Tangotiger ? (@tangotiger) April 9, 2023

A general analysis of this defense would be interesting, but it’s beyond the scope of this article. You’d need exact starting positions for each fielder on each play, as well as detailed batted ball data for the hitters involved. I don’t have access to that. What I do have is some spare time and a little bit of creativity, as well as some granular descriptions of Gallo’s batted ball tendencies. Can it truly make sense to remove outfielders against one of the air-happiest hitters in baseball? Let’s do some approximating and find out.

The question we’re trying to answer here is straightforward: Does moving a fielder from the outfield to a glorified infield position take production away from Gallo? To answer that, we’ll need two things. First, how many base hits does the extra fielder take away? Second, how many extra hits are the Royals giving back by putting fewer players in the outfield? If we can come up with estimates for those two numbers, the rest is easy. We just compare the two and pick the bigger one.

First things first: we need to know how frequently Gallo hits grounders, line drives, and fly balls. Easy enough! Here are his career numbers pulled from his player page:

Now we have an idea of how many batted balls will be affected by the displaced fielder. Grounders will be easier to field, MJ Melendez’s gaffe notwithstanding. Fly balls will be tougher to field with the new alignment. Line drives are trickier, so we’ll treat those separately.

One complication with working out what happens with grounders is that there isn’t a lot of data on Gallo against five-man infields. To handle that, I sliced his batted ball data up into two buckets: opposite field and pull plus straightaway. The five-man infield looks like an unshifted defense against opposite-field grounders and like an overshift against pulled and straightaway grounders. How has Gallo performed on those types of hits against those types of defenses? Here:

There’s just one problem with that lovely chart: Gallo has hit exactly five career opposite-field grounders against a standard defense. He rarely hits the ball on the ground, rarely hits it the other way when he does, and has rarely faced unshifted defenses (until this year, of course).

Still, I have good news. We can completely ignore that side of the table, because nothing the Royals choose to do can change anything about that. They can shuffle their outfielders all they want; there will still be two infielders on the left side. All we care about is the difference between Gallo’s production to the pull side against standard defenses and how much worse he does when the defense shifts.

For every 100 batted balls, Gallo hits 29.3 grounders, as we covered up above. Out of those 29.3, he pulls a ton of them, with 90% of his batted balls either up the middle or pulled. That makes the math straightforward. For every 100 batted balls, there will be 26.4 pulled grounders. Adding an extra fielder in shallow right is worth 171 points of wOBA, in expectation, on those grounders. After doing a bit of conversion to turn that into runs, the Royals are saving 3.8 runs in expectation per 100 batted balls, specifically on grounders, by bringing their right fielder in close.

How many runs are they giving back with a two-man outfield? That’s tremendously hard to measure, but I’m going to take a stab at it. The biggest problem with a two-man outfield is guarding the lines. If you look at the diagram up above, you can see that both foul lines are wide open compared to a normal defensive alignment.

I got extremely approximate here, using hit coordinate data from the Statcast feed to approximate Gallo’s batted ball landing spots. I found that 15% of his fly balls were down one line or the other and hit far enough that infielders had no play, which passes the eye test as well. We care only about a subset of those line-hugging fly balls, though: the ones that would have been outs against a regular defense. Roughly 44% of those fly balls turned into outs. Turning half of those into doubles and half into singles increased Gallo’s wOBA on these field-boundary fly balls from .960 to 1.426. Those numbers are huge; Gallo rips a lot of pulled home runs.

More important for our purposes than the raw production numbers are the changes. For every 100 batted balls, Gallo hits 45.2 fly balls, as we covered up above. If 15% of those are down the line, that works out to 6.8 batted balls per 100. That means that turning some of his outs into hits thanks to the two-outfielder defense costs the Royals 2.6 runs in expected production per 100 batted balls.

Lastly, we need to cover what happens on line drives, and that’s far from easy. I decided to apply the same criteria that I did for fly balls, but I’m a lot less confident that this method works here. For example, a line drive to dead center could split the two-outfielder configuration, but would be an easy out for a regular outfield. On the other hand, a line drive into the gap might be directly at a fielder who would otherwise have no chance. I treated those two possibilities as a wash, but again, I’m not sure that’s the right decision.

Regardless of whether or not that’s the correct decision, it’s the decision I made. Therefore, I mostly performed the same calculation as above, with one slight adjustment: I widened the definition of “line-hugging” somewhat because the fielders will have less time to get to harder-hit balls. I didn’t make any adjustments for line drives that the short right fielder can now catch, mostly because I didn’t know how to. Let’s treat that as a tiebreaker; if the math works out close to even, the short right fielder’s influence on line drives right at them, whether it’s turning hits to outs or doubles to singles, is a good enough reason to shift.

With that preamble out of the way, here are the numbers: 26% of Gallo’s line drives have hugged the line, almost exclusively to the pull side. They all already mostly turn into hits anyway, so there isn’t quite as much to be gained here, but if the outs again turn into half singles and half doubles, that would raise his wOBA on that subset of line drives from 1.004 to 1.184. Gallo hits 19.4 line drives per 100 batted balls. After plugging all these numbers in, the Royals are costing themselves 0.8 runs per 100 batted balls.

Now, we can put everything together. On grounders, the Royals’ unconventional shift saves them 3.8 runs per 100 batted balls. On balls in the air, it costs them a combined 3.4 runs per 100 batted balls. That’s close enough that we should apply the tiebreaker, which is in favor of shifting. In other words, bringing in their right fielder against Gallo saved the Royals a tiny part of a run, at least per my calculations. We’re doing math in per-100-batted ball increments, but if you want to think in terms of plate appearances, my math says that the Royals save roughly 0.2 runs per 100 PAs. That’s the equivalent of roughly two points of wOBA, not exactly a huge savings.

In the end, perhaps the Royals’ shift is much ado about nothing. Or perhaps my analysis is wrong; I had to make a lot of assumptions to come to a conclusion, and some of them are closer to guesswork than scientific knowledge. I’m most surprised by how narrow the gap between the two options appears to be; it’s close to being a pure toss-up. That’s only the case because Gallo is so extreme of a hitter, though.

Against a more normal hitter, this strategy would make far less sense. Tons of Gallo’s fly ball contact leaves the yard – nearly 30% in his career, and over 50% when he pulls the ball. When he puts the ball on the ground, he almost never hits it the other way, and shifting against him lowers his production on grounders by a ton, because he’s generally hitting the ball quite hard, which means infielders have less time to react. If you take a hitter who’s leaving the yard less frequently or who doesn’t hit his grounders quite as hard, I don’t think the math behind this alignment would hold up.

The Royals are still using it all over the place – they used it against multiple Giants lefties over the weekend. It’s far too soon to say whether it’s working. We might never know if it’s working; there were arguments over how much difference infield shifts made even when everyone was using them for years and years. Let’s leave it at this, then: The Royals brought in an extra infielder against Joey Gallo, and the math says that it was a reasonable decision. Baseball with extra constraints is fun!