Sorry, Bryce: Matt Williams is Right

Yesterday, Bryce Harper returned from the disabled list. This is good news for the Nationals, since Bryce Harper is good at baseball. Having more good baseball players is not a bad thing for a team trying to win, so a returning Harper is a net positive for the organization. However, Harper’s return is not entirely without controversy.

As Wendy Thurm noted after her conversation with Ryan Zimmerman a few weeks ago, Zimmerman enjoyed playing the outfield more than he enjoyed playing third base. His shoulder issues, and the mental pressure that came with making the throw across the diamond, were not a factor in the outfield, allowing him to enjoy the game in a way that he wasn’t at third base. However, Harper’s return means that there is not an outfield spot for Zimmerman any longer, and on Monday night, he went back to third base.

Before the game, Bryce Harper publicly disagreed with the decision.

“I think (Zimmerman) should be playing left,” Harper said. “Rendon’s a good third baseman. He should be playing third. We’ve got one of the best second basemen in the league in Danny Espinosa. Of course, we want the best-hitting lineup in there. I think Rendon playing third and Zim playing left is something that would be good for this team. I think that should be what’s happening.”

Essentially, Harper’s return forced Matt Williams to make a choice between two players: Denard Span and Danny Espinosa. If Span plays, Zimmerman has to play third, pushing Rendon back to second base, which puts Espinosa back on the bench. If Espinosa plays, then Rendon can play third, Harper shifts to center field, and an outfield spot opens up for Zimmerman. While Harper and Zimmerman are the two most notable players in the story, this is really an evaluation of Span and Espinosa.

And when you evaluate those two players, this shouldn’t even be a particularly hard decision. Here is what the pair have done over the past three calendar years.

Name	PA	AVG	OBP	SLG	wOBA	wRC+	BsR	Off	Def	WAR
Denard Span	1621	0.272	0.324	0.382	0.312	96	10.0	3.2	22.7	8.1
Danny Espinosa	1421	0.228	0.296	0.367	0.292	81	0.9	-29.5	24.5	4.1

And here is what they have done just this year, for those who still think a half season’s worth of performance should drive roster decisions.

Name	PA	AVG	OBP	SLG	wOBA	wRC+	BsR	Off	Def	WAR
Denard Span	335	0.265	0.312	0.388	0.309	95	3.8	2.1	-0.7	1.2
Danny Espinosa	268	0.217	0.284	0.348	0.278	74	0.2	-7.7	2.6	0.3

As plain as it can be: Denard Span is a better player than Danny Espinosa, and it isn’t particularly close. Span has a consistent track record as a roughly league average hitter, and his baserunning value pushes him a little above that mark in total offense; Espinosa isn’t anywhere close to that. Over the course of a full season, the difference in offensive performance between the two is worth about a win and a half.

And defensively, they’re roughly equivalent in the field. Over 5,600 innings in center field, Span’s career UZR/150 is +5. In 3,600 innings at second base, Espinosa’s UZR/150 is +5. Given that Espinosa is 27 and Span is 30, you might give a slight edge to Espinosa’s glove going forward just due to aging curves, but the gap is not going to be particularly large. At most, you might estimate that Espinosa is a few runs better per season than Span is in the outfield.

But, of course, the question isn’t just how well they stack up individually, but rather, how well the team’s entire defense stacks up when one or the other is in the line-up. The argument for Espinosa essentially revolves around Rendon and Zimmerman. If Zimmerman is more comfortable in the outfield than at third base, and third base is Rendon’s natural position, maybe they make up the offensive gap that exists between Span and Espinosa?

We don’t have the tools necessary to make definitive declarations about Zimmerman’s defensive value as a left fielder, but we can make some assumptions based on what we do know about his defense at third base and defense in general. While Zimmerman certainly is no longer the quality defender at third base that he was earlier in his career, there isn’t much in the way of evidence that suggests that he’s going to be a disaster there.

His shoulder problems began back in 2011 and the throwing issues followed soon after, significantly decreasing his defensive value as an infielder, but he’s still played nearly 3,500 innings at third base since the start of 2011 and has been decent enough to avoid disaster status. His UZR/150 over that stretch is -6, while DRS actually has him even closer to league average. He hasn’t been good there, but it’s not like he’s spent several years showing why he needs to be moved off the position immediately.

Last year was his worst year at third base by UZR, and he is getting older, so perhaps you want to be aggressive with the weighting and project him as a -10 third baseman going forward. That’s within the realm of reason, especially given that Zimmerman has a stated preference to not play there going forward. But how much better does we think he’s going to be in the outfield, realistically? He’s not fast and he doesn’t have much experience in the outfield, so while the early returns have him being good enough to play out there, he’s not going to be an above average defensive outfielder.

Speed score isn’t a perfect measure by any stretch of the imagination, but it does a decent job of bucketing players as “fast” or “slow” based on their stolen bases and distribution of extra base hits. A league average speed score is about 4.5; Zimmerman’s career average is 3.6. This year, he’s at 1.2. It’s probably fair to say he’s a below average runner, and only getting slower. For reference, here is the 2011-2014 UZR/150 for left fielders with a speed score below 3.5.

Name	Inn	Speed	UZR/150
Raul Ibanez	2,785	3.5	-15.5
Matt Holliday	4,107	3.1	-3.8
Josh Willingham	2,735	2.5	-10.5
Delmon Young	1,755	1.8	-12.5
Michael Morse	2,092	1.6	-21.2

Yeah. Zimmerman is almost certainly more mobile than Morse, Ibanez, or Young, but being better than the worst defensive players in baseball doesn’t make you good. It would probably be charitable to project Zimmerman as anything close to an average defensive left fielder, given his size and lack of experience, and he might very well be just as poor in left field as he would be at third base. If there is a significant defensive upgrade to be gained, it’s probably no more than five runs over a full season, or a third of the offensive difference between Span and Espinosa.

And it’s very unlikely that Rendon make sup the other two-thirds of that difference by shifting from second base to third base. Since he has fewer than 1,000 inning in his career at both positions, defensive metrics don’t tell us much, but neither UZR nor DRS suggest that Rendon has been significantly better at third base than he has been at second base. Given his depth of experience at third base, it’s probably fair to expect him to be better at that position going forward, but the magnitude of the gap is just not going to be that large.

And we haven’t even touched on Harper himself, who would have to slide over to center field without Span in the lineup. Even if we take very aggressive stances on the defensive value of Zimmerman as an LF and Rendon as a 3B, we’d still have to mitigate those gains by pushing Harper into center field, where he’s likely to be less valuable.

To boil things down, we have a lot of evidence that suggests that there’s a big gap between Span and Espinosa at the plate. We have basically no evidence that suggests that the defensive improvement from putting Rendon back at third base and Zimmerman in the outfield would even cancel out that gap, much less make the Nationals better. To have a strong preference for Espinosa over Span, one would have to think that there could be something like a 25 run difference in defensive value to be gained by having Rendon play third, Zimmerman play left, and Harper play center. There’s just no reason to think that’s true.

And even the intangible argument about Zimmerman’s comfort level lacks much in the way of support. While I think a player’s preference for position should be included in the discussion of where he is going to play, the data does nothing for Zimmerman’s case that he feels less pressure in the outfield and enjoys the game more; he has a .496 wOBA in 40 plate appearances as a third baseman and a .267 wOBA in 108 plate appearances as a left fielder. The samples are tiny and the numbers are basically useless, but the do-it-for-Zimmerman’s-feelings case would be stronger if he weren’t crushing the ball while playing third and hitting like Danny Espinosa while enjoying the game in left field.

Not that his feelings and preferences don’t matter; they do. They just don’t matter enough right now, given the team’s current options for fielding a line-up. The fact that Zimmerman wants to play the outfield should be a goal for the Nationals to get him there at some point, but realistically, it’s going to take a trade or two to make it a good idea for the franchise.

If they could trade Denard Span for a better second baseman than Danny Espinosa, that might not be a terrible solution. But there are a lot of teams that need second baseman, and there aren’t a lot of them available. Getting Zimmerman off of third base after this season is a good idea, but playing Danny Espinosa over Denard Span right now just to placate his preference is not. And Zimmerman, for his part, has been willing to go back to third base to help the team win.

And that’s where Williams put him, because that’s the current alignment that gives the Nationals the best chance to win baseball games in 2014. That will likely change in the future, and the front office can take proactive steps to help Zimmerman get back to the outfield in the future, but as long as the choice is between Denard Span and Danny Espinosa, Ryan Zimmerman should play third base.