This is why Ozuna could bounce back this year

The Cardinals have added Paul Goldschmidt to their lineup and Andrew Miller to their bullpen so far this winter, and they'll likely welcome back the injured Alex Reyes to their pitching staff in 2019 as well. Those are three potentially large upgrades if all goes well, but the more interesting question may not be about the new faces. It might be about the returning big-ticket names who weren't quite themselves in '18. What can you expect from Adam Wainwright or Dexter Fowler? What version of Marcell Ozuna are you going to see?
Let's tackle that last question here, because it's a complicated one. On the surface, Ozuna's first season as a Cardinal was a disappointing one. By just about every measure you can think of, he took a major step back from the stellar 2017 season with Miami that had made him such an appealing trade target.
2017
.312/.376/.548 • .924 OPS • 37 homers • 144 wRC+
2018
.280/.325/.433 • .758 OPS • 23 homers • 106 wRC+ 
That's bad, obviously. That's worse. There were 277 players with 400 plate appearances in both seasons, and his 115-point slugging drop was one of the 10 largest. His 166-point drop in OPS was the seventh-largest. 
You might think of a few reasons why his numbers fell, like more strikeouts, more grounders, a drop in hard-hit rate. Any or all of those could have helped cause such a decline, especially since it was known that he was playing through a shoulder injury.
The thing is, none of that happened. Looking at the plate discipline and Statcast™ metrics, Ozuna looked like he was essentially the same hitter in both years.
2017
21% strikeout • 9% walk • 45% hard-hit • 47% ground-ball • .367 Expected wOBA
2018
18% strikeout • 6% walk • 45% hard-hit • 47% ground-ball • .353 Expected wOBA
Ozuna walked slightly less, but he struck out less, too. His hard-hit, ground-ball and launch-angle metrics were unchanged. His Expected wOBA, a Statcast™ metric that accounts for both quality of contact and amount of contact, was in the 84th percentile in 2017 ... and fell only to the 80th percentile in 2018.

Let's make it even more complicated. We know Ozuna's shoulder injury was severe enough that it required October surgery, but the differences in his first half and second halves were massive -- with a much better performance after the break.
2018, first half
.268/.309/.385 • .693 OPS • 10 homers • 89 wRC+
2018, second half 
.299/.351/.506 • .857 OPS • 13 homers • 131 wRC+ 
In the first half, he hit like Jordy Mercer or Tucker Barnhart. In the second half, he hit like Javier Báez or Matt Carpenter, despite having a shoulder problem that was clearly not getting better. 
So what happened here? How could his season step back by that much, without the underlying metrics changing, and then still be that strong in the second half anyway -- and what does it mean about expectations for 2019? Let's find out.
Some of it was unfortunate outcomes ...
When you hit the ball that hard and don't have success, you have to figure that at least part of it was the ball just not finding grass, and in this case, that's true. (Somewhat. This is part of the story, not all of it.)
One way we can track that is by looking at "barrels," a Statcast™ metric that describes a batted ball that has the perfect combination of exit velocity and launch angle. You can read about how it works here, but the short version is that a barrel has a minimum expectation of a .500 average and a 1.500 slugging percentage, and the shorter version is that as a hitter, it's just about the best possible thing you can do. Khris Davis (70) and J.D. Martinez (69) led the Majors in barrels last year, while light hitters like Dee Gordon had just one.
Ozuna had 46 barrels in 2018, slightly more than the 44 he had in '17. The difference is that in '17, only four of them turned into outs; he hit .909 with a 3.227 slugging percentage. (As we said, these are high-value batted balls.) In '18, a full 16 of them turned into outs, the second-most in baseball. He hit "only" .682 on them.
So what did outs those look like? Perhaps less impressive than you'd think. Maybe you were hoping for lots of fantastic plays. Instead, it's a lot of "at-em" balls. Some were plays where the fielder made it look a lot harder than they needed to, like this one in Miami.

Others were balls that were hit hard and looked good off the bat, but didn't quite make it all the way out. 

Whatever it was, he had four times as many high-value batted balls find gloves in 2018 as he did in '17. What was that about? 
...but more of it may have been approach and health
Here's the thing about his identical-looking hard-hit rates: They weren't necessarily evenly distributed.
It's true that Ozuna had a 45 percent hard-hit rate in both his good 2017 and poor '18, and that he had a 47 percent ground-ball rate in both years as well.
But when you figure out how those were arranged, you start to see the issue. In 2017, just seven percent of his hard-hit balls -- the ones hit with 95 mph of exit velocity or more -- came on the ground. In '18, that jumped to 13 percent. He didn't hit more grounders overall, but he wasted more of his hard-hit balls on the ground.
In addition, you can really see the difference when you look just at fly balls and line drives and look at his average distance (in feet) between the two seasons. Interestingly enough, Ozuna's balls hit in the air had some added distances to center and right field. But to his pull side, he lost a shocking 42 feet of distance, more than enough to turn home runs into base hits or outs.

Maybe that's the true effect of the shoulder issue, that his power didn't disappear so much as it changed. You can see why this is a problem when you look at Ozuna's career home run spray chart; while he goes deep to center or right every so often, it's clear that his preferred long-ball territory is to left.

So that all makes sense, except for this part: If the shoulder was still injured late in the year, badly enough that it required the offseason procedure, how did he manage to hit so well in the second half? 
Here's one theory: Maybe the shoulder was already feeling a little better. Ozuna was placed on the disabled list on Aug. 22, the day he received a cortisone shot, and he returned on Sept. 2. We broke it up into "first half" and "second half" above, but let's go even deeper here. From the All-Star break through the disabled list trip, he hit .293/.347/.474 (121 wRC+), which, to be fair, was already a big step up from the first half.
After the rest and the shot, he returned to hit a stellar .306/.355/.551 (144 wRC+). While it's always smart to somewhat discount September numbers, since the expanded rosters can lead to opposition that wasn't good enough to be in the big leagues in earlier months, it's also fair to point out that three of his home runs came off of All-Stars Josh Hader and Ross Stripling as well as Tigers closer Shane Greene.
Again, we can't say for sure how much and when the shoulder was affecting him. What we do know is that we're looking at a player who was a star in 2017 and for the second half of '18, playing with an obvious injury and with underlying metrics that suggest the skill hasn't actually gone anywhere. 
"I'm going to get right," Ozuna said to MLB.com's Jen Langosch in October. "Big year. You watch. Big year for me next year."
Those are big words, the kind you'll hear plenty of players offer after a disappointing year. But from Ozuna, we've seen him have that year in the not-distant past. We've seen flickers of it in 2018. It's not at all hard to think that he could be just a big a part of the '19 success of St. Louis as Goldschmidt or Carpenter.