The Next Big Thing in Defense?
On Tuesday, in Grapefruit League play, those in Clearwater, Fla., witnessed what might possibly represent the future of defensive alignment. They at least saw how aggressive and creative first-year Phillies manager Gabe Kapler is prepared to be.
As Matt Gelb reported for The Athletic, non-roster invite outfielder Collin Cowgill and Tommy Joseph (who is playing some outfield this spring) were told before the game that, if Tigers switch-hitting prospect Victor Reyes batted left-handed, they would swap corner-outfield positions. Reyes did bat left-handed, and when he did, Cowgill and Joseph swapped their positions mid-inning.
While I don’t have video evidence to offer as proof, there is a box score that recorded this curiosity for posterity.

Per Gelb, it seems as though the Phillies are considering this as a regular practice.
“This is not one of those situations that will catch the opposition by surprise,” Kapler, the rookie manager, said. “It’s actually not a brand-new concept. It’s just more straightforward. It’s putting a guy who has a really good defensive capability in the spot where you think the ball is going to go. Is it a little unorthodox relative to what is done every day? Sure. But, at the same time, we want to be prepared for it during the season.
“I played, what, 3 2/3 innings in left and a third of an inning in right?” Joseph said. “That was pretty neat. That was part of the ‘Be Bold’ slogan we’re working with here. We’re trying to be trailblazers when it comes to the shifting. That’s the hitter that it worked out for.”
Infield defensive shifts have become commonplace, as we know. While they declined last season, in terms of total volume, teams have simultaneously become more interested in outfield shifting. Perhaps the next generation of defensive alignment was on display Tuesday afternoon — one that includes just shifting in one direction or another actually switching positions based on a hitter’s batted-ball tendencies.
Perhaps the Phillies’ motivation is to mask the defense of Rhys Hoskins, who is converting to left field after the club signed Carlos Santana to a three-year, $60 million deal earlier this offseason.
But batted balls lifted in the air are distributed differently, more evenly, in the outfield than in infield, so even if what we saw Tuesday becomes a regular practice, it would still likely be less common — and perhaps have less effect — than infield defensive-alignment practices.
Last season, 29.1% of line drives and fly balls were pulled, 36.1% were hit to center, and 34.8% were hit to the opposite field. That’s in comparison to ground balls, 53.7% of which were pulled last season, 33.6% hit to center, and 12.7% to the opposite field.
Most ground balls are pulled, while fly balls are more evenly distributed — even, perhaps, with a slight tendency toward the opposite field. Only six batters last year (of the 317 to record 90-plus plate appearances) hit 50% or more of their batted balls to a particular field. All six favored the opposite field, let by Joe Mauer (63.1%).
So even if this is a new thing, it’s probably more limited in scope and potential impact. Yet even if it’s a small thing, it’s something — particularly if a club is fielding a particularly weak or questionable outfield option like Hoskins. Even little things are important at a time when all the low-hanging analytical fruits have been picked. The Phillies should want Hoskins to be exposed to as few batted balls as possible. When the Phillies face, say, Mauer or a similar left-handed batted-ball profile, positioning becomes more important, as such batters concentrate their batted balls to the opposite field.
What the Phillies did Tuesday is not so much a shift as it is a switch. We’ll have to see if it’s part of the next generation of defensive alignment. Maybe it’s nothing more than showing that Kapler is going to be an open-minded, progressive manager. If nothing else, spring is a time to explore and, to date, Tuesday’s experiment is one of the more interesting we’ve seen.
A Cleveland native, FanGraphs writer Travis Sawchik is the author of the New York Times bestselling book, Big Data Baseball. He also contributes to The Athletic Cleveland, and has written for the Pittsburgh Tribune-Review, among other outlets. Follow him on Twitter @Travis_Sawchik.
Shouldn’t this approach be more applicable to ballpark configurations than to individual batters? For example, the Red Sox and Yankees might do LF/RF switches depending on whether they were at Fenway (big RF, tiny LF) or Yankee Stadium (LF much bigger than RF). Don’t remember seeing this done with regulars.
Your last two (excellent) articles on the change of defense might fall on many deaf ears due to the nerdiness of them but clubs understand. Maybe part of the reason position players are losing value as free agents?
At Fangraphs the “nerdiness” of an article ensures it won’t fall on deaf ears.
This reminds me of the Mets switching d’Arnaud and Cabrera between 2b and 3b all game. Except on purpose, instead of out of roster-desperation.
How important are line drives for that? Aren’t most liners hits anyway and just tossed back after a bounce? I think that is interesting because while liners are pretty evenly distributed real fly balls are only pulled 23%.
But of course there are not just line drive singles and liners directly to outfielders that even pujols could handle but also liners to the gaps which are very tough. So not sure how that plays.
Btw im sceptical on hoskins in the outfield. He was not super terrible in short sample but he is slow. And with slow I don’t mean just for an OF but even for a 1b (in LF only kemp was slower last year and even at 1b he was like bottom third), eric also said 1b only.
In fact even santana was faster than him, maybe phillies end up switching rhys and Carlos:)
There was an outstanding analysis on this over at BP yesterday, analyzing through run totals if this switch is even productive. They discuss line drives.
Baseball prospectus just talked about this topic yesterday
https://www.baseballprospectus.com/news/article/38145/bp-wayback-machine-corner-outfield-inefficiency/
Right at the bottom of that link, there’s a comment from somebody who asks the perfectly reasonable question: why not just shift the CF sideways?
I’m a cricket fan, so I’m used to what in baseball would be considered extreme re-positionings in the field. Baseball, on the other hand, is just now getting used to the shift (as a general proposition), and there’s still a residual reluctance to buy into it wholesale.
But I’d need to see a whole lot more evidence before I bought into something weird like swapping corner outfielders on a per-batter basis, rather than just IF/OF/whatever shifts. It’s not like teams have quite worked out how to maximise that sort of shift in the first place.
Every time I think score keeping has got as complicated as possible, they add another layer.
This is hilarious… We do this on the little league team I coach. Different in a way (i.e. we have few 9 year olds who can really cover OF, but kids are starting to really hit the ball by then, so I shift my best outfielder around based on the batter. It works incredibly well. At that age, it is pretty easy to see who is going to pull the ball and hit line drives or fly balls.
Why wouldn’t this be a common approach? Seems inevitable.
I feel like a lot of opposite field fly balls are of the lazy, can of corn variety. I would think the pulled fly balls are the ones you’d want a good defender going after.
Yeah, what we need is the angle distribution of “interesting” chances, like in the 20 – 80% catch probability by Tango’s metric.
But those balls, if they’re hits, are fielded more quickly and are less likely to become doubles/triples, no?
The sample of line drives/fly balls includes home runs, which are mostly pulled and nearly all uncatchable. If we restricted the sample to balls in play, does the distribution change dramatically enough to make a difference? I tried to figure this out myself, but if FG’s splits offer a way to exclude home runs, I couldn’t find it.
Great point, according to pod 33% of pulled flies are homers vs 5% for oppo. Defense doesn’t have to worry about homers so the number of bip to pull field is only around 20%. Sure an occasional pulled homer could be robbed at the fence but that is really rare.
I was watching the game on MLB.tv and saw this happening live. I got your back if anyone says this didn’t happen.
The pure number-crunching efficiencies of these defensive shifts are one thing, but there could be a thing as too much, too soon.
People (professional athletes in particular) can be creatures of habit, and oftentimes it takes significant time and retraining for players to have the proper comfort level and regain reaction time to be able to play at the same high level when constantly faced with off-nominal conditions. Basically, in early training, teams will be asking players to be constantly “out of position” based on the conventions they may be used to.
Early spring training is absolutely the right time to introduce these experiments and get everybody in the same frame of mind, but a team that goes too far in this direction may end up preventing themselves from improving as much as they could through mid-season callups and trades. It will be necessary to play the long game, and commit to implementing these tactics at the lowest levels of minor league ball, and wait for the rest of the league to reach a certain level of critical mass before these types of moves are truly effective.
The direction of the batted ball isn’t that important. It’s the direction of a *well-hit* ball that matters more, I would think.
Great point. We would need to take it a step further and see if pulled fly balls/liners are hit harder than opposite field.