In 2016, as the leadoff hitter for the Chicago Cubs, Dexter Fowler contributed a 129 wRC+ for the eventual World Series champions. This mark denotes that the center fielder was 29% better than a league-average hitter, which was a career-best for the soon-to-be free agent, who signed with the St. Louis Cardinals that December.

Fowler’s disastrous 2018 campaign has led to something of a public misconception that he turned into a pumpkin the second he landed in St. Louis. In 2017, his offense was only moderately worse—Fowler was 21% above-average at the plate and belted a career-high 18 home runs in 118 games. But Fowler’s defense, long a shortcoming in his game but statistically better over the previous two seasons, reverted back to firmly below-average, and with the ascension of career center fielder Tommy Pham in 2017, the Cardinals moved Fowler to right field.

While most of the attention to Fowler’s 2018 has focused on his terrible offense—only six hitters with as many plate appearances as Fowler had a wRC+ worse than his 62—a big part of the collapse of his total value was his lack of defensive contribution. Fowler was a bad hitter, yes, but he wasn’t nearly as catastrophic at the plate as 2013 Pete Kozma (wRC+ of 49), and yet because Kozma provided positive value in the field (and to a lesser extent on the bases), he was at least above replacement level.

Fowler had subpar defensive metrics in 2018, though the degree to which he seemed to look worse than the previous year in center field, an ostensibly more difficult position to handle, suggests he was a victim of small sample sizes making him look worse than his true talent level. But even if Fowler turned out to be a great defensive right fielder (probably less likely than that he is as bad as he was last year), the potential value just isn’t as much as it was for Pete Kozma at shortstop. In 2016, Fowler’s Cubs teammate Jason Heyward was an elite defensive right fielder and a good base runner, but was firmly below-average as an overall player because of his lackluster offense which, by wRC+, was better (72) than what Fowler contributed last year.

Playing Dexter Fowler in right field limits his potential defensive downside, but it also severely limits his overall upside as a player. Fowler’s projected defensive value is slightly lower than his defensive value from 2017 (when he was in center field; positional adjustments mean that an average center fielder is projected for more defensive runs than an average right fielder), and in that season, Fowler was slightly above-average by FanGraphs Wins Above Replacement and slightly below-average by Baseball Reference Wins Above Replacement. And Fowler replicating his 2017 season is considered a bit of a long shot–ZiPS projects him for a 102 wRC+, slightly above average but below-average for a right fielder, and with poor defense to boot. In 430 projected plate appearances, Fowler projects for 0.9 WAR. This is a substantial improvement from his 2018, but it still makes him a below-average MLB player and probably not a player worthy of being a regular starter.

Sometimes teams cannot avoid playing subpar players–the aforementioned 2013 Cardinals’ alternative to Pete Kozma was Daniel Descalso, who one could argue was better but he certainly wasn’t ideal–but the 2019 Cardinals have multiple alternatives who project to be not only better than Fowler, but perhaps considerably better (and now that Bryce Harper is on the Philadelphia Phillies, there isn’t a free agent available who would improve the situation further). By either raw ZiPS WAR projection or by taking it on a per-plate appearance basis, both Tyler O’Neill and Jose Martinez project as superior players to Fowler.

• Tyler O’Neill: 2.0 zWAR, 2.4 zWAR/600 PA
• Jose Martinez: 1.8 zWAR, 2.0 zWAR/600 PA
• Dexter Fowler: 0.9 zWAR, 1.3 zWAR/600 PA

This three-horse race for right field playing time (when Marcell Ozuna is healthy and there isn’t a need to play one of them in left field, at least) is made all the more intriguing by the fact that the three represent unique molds. O’Neill is a highly regarded prospect, is easily the youngest of the trio at 23, and has the single most dynamic skill of the group (his power, which enabled him to be above-average at the plate in 2018 despite a K/BB rate of over 8/1). Martinez is the most established hitter, providing more value at the plate for the Cardinals last season than anybody but Matt Carpenter, even if his defense left something to be desired.

Fowler is the one with the strongest expectations, given his $16.5 million average annual salary, and making him the everyday right fielder is therefore the “safe” decision from a “justifying your past transactions” standpoint. O’Neill is the young kid who still needs to pay his dues and prove himself; Martinez is the late bloomer who is just happy to be there; Fowler is the marquee free agent signing. And despite that his 2018 was by far the worst season any of the three have ever experienced at the Major League level, Fowler is a safe pick to play in the sense that he will almost certainly be better than he was last year, and his high salary comes with an implicit level of credibility.

Spring Training statistics border on meaningless in their totality, much less when the team’s leaders in plate appearances 1. Have 27 of them apiece; 2. Are Rangel Ravelo and Max Schrock. But given the question marks of these players (well, O’Neill and Fowler mostly; Martinez seems like a relatively safe bet to be a good bat, poor glove guy), it’s understandable to at least glance at the numbers and see how they confirm or deny our preconceived notions. And (and I can’t stress this enough–these are comically small sample sizes) Fowler has just two hits, both singles, in 16 plate appearances, to go along with five strikeouts and no walks. Meanwhile, not only has Tyler O’Neill been the team’s best hitter, sporting a 1.242 OPS in 26 plate appearances but doing so via the thing we know he can do (he has four home runs) and via the thing we were less certain about (his walk rate is 19.2% and his strikeout rate is a far more manageable 26.9%).

The point of looking at these statistics isn’t to project anything–they mean almost nothing. But say O’Neill does emerge, showing similar plate discipline numbers to what he had in the minors (he was never exactly Joey Votto, but he wasn’t close to as wretched in this regard as he was in the big leagues), and say Fowler continues his downward slide, maybe not to last season’s depths but to a point where he is very clearly an inferior hitter to O’Neill or Martinez. What do the Cardinals do?

I am a firm believer that contracts should not factor into assigning playing time, but there is something to considering salaries when determining whether or not to cut players from the roster altogether. Particularly in the case of multi-year contracts (and Dexter Fowler has three years remaining on his contract), jumping the gun could cost the team future value. Say, hypothetically, that Fowler is now a one-win player. Do you want him starting? Probably not. Do you want to be paying him as much you’re paying him? Certainly not (I’m assuming that the money would be used on somebody better). But a one-win player does have value. He particularly has value in 2020 or 2021 if Marcell Ozuna doesn’t re-sign, if the Cardinals move Jose Martinez to first base after a potential Paul Goldschmidt free agency departure, if Harrison Bader’s promising rookie campaign turns out to be a flash in the pan. If the Cardinals cut Dexter Fowler today, they wouldn’t have him in 2020 or 2021 either, but they would still have the obligation to pay him.

The Cardinals should try to make things work with Fowler, but “make things work” doesn’t necessarily mean trying to force him into the #2 spot in the batting order, a spot in which, mathematically, teams should play their best hitter, while Peak Dexter Fowler would be no better than the third-best hitter in this lineup. It doesn’t even necessarily mean starting him. But even if he doesn’t start every or even most games, there is still a potential role for Dexter Fowler in gradually re-establishing himself as a viable big-league player. He can be a pinch-hitter. He can play some center field (even if he isn’t great at it). He can pinch-run and he can mentor young players. This kind of role player may not merit $16.5 million, but his salary is a foregone conclusion. And if Fowler can be a role player, not only is this better than nothing, but it could help buy time in case he is able to become something close to what he has been in the past.