Imagining the All-World Defense Team

The Cubs have had baseball’s best defense this season. They rank first in Defensive Runs Saved, with 58 runs above average, giving them a 10-run lead over the second-place Houston Astros. They rank first in Ultimate Zone Rating, with 50 runs saved, giving them a 15-run lead over the second-place Toronto Blue Jays. They’re turning a historically high number of balls in play into outs, and while a number of factors influence a team’s BABIP, Chicago’s elite defense is among the most important.

We’ve seen what a truly elite defense can do for a mediocre pitching staff over the past couple years, with the Kansas City Royals. We’re seeing what a truly elite defense can do with a great pitching staff right now, with the Chicago Cubs. But, while the 2013 Royals had the best defense on record during the current 14-year era of advanced defensive metrics, it’s not like that team reached the ceiling of what a defense can be. That team wasn’t built by a mad-scientist general manager whose only goal for the season was to experiment with the upper bounds of defensive performance. It was just a really, really good defense. I’d like to play the role of that mad-scientist general manager for a second, based off this chat question I received on Tuesday:

Apologies to Joe for making you wait. Hopefully you understand that your question was a bit complex for an instantaneous chat-room response. I’d like to think that your excellent question being rewarded with a full article makes up for the delay.

So, first thing’s first. Our roster. We want the best defensive player at each position. The best way to achieve this would probably be to take the three best center fielders and disperse them among the outfield, and divide the four best shortstops up among the infield. But that’s a bit too hypothetical even for this ridiculous hypothetical. So instead what I’ve done is simply look toward the rest-of-season ZiPS projections to identify the top-rated fielder at each position.

Our defensive alignment looks like this:

Hell of a defense! Hell of a team, probably. Pitchers don’t receive defensive projections, hence the absence of a pitcher on the image. It’s also worth noting that ZiPS’ defensive projections don’t include framing, hence Salvador Perez’s choice as our top defensive catcher. With framing included, Perez’s value plummets, and we likely end up with someone like Buster Posey or Yasmani Grandal instead. I don’t think this team will mind.

The numbers. That’s what this is all about. What kind of numbers could this team put up, defensively? For the first method of answering that question, I simply prorated each player’s ZiPS runs-saved projection to 700 plate appearances. No subs on this team.

Runs saved: 103! That’s a ton of runs. That’s about 10 extra wins, relative to a league-average defense. For context, 103 runs saved would be a record by either defensive metric, dating back to 2003. The DRS record belongs to the 2005 Phillies, at 97 runs. The UZR record belongs to the 2009 Mariners, at 86 runs saved. Using a 50/50 split of the two, the component runs-saved record belongs to the 2013 Royals, at 88. Which… is shockingly close to our all-world defense team above. Puts into perspective how impressive that Royals defense was.

But, of course, we’re working with projections here, and projections are inherently conservative due to the nature of regression. And this is a hypothetical, so let’s get buckwild. We want the absolute ceiling, and the absolute ceiling ought to be higher than 103 runs. So instead of using a projection, let’s use the best individual defensive season that each of our eight fielders has already posted:

Now we’re talking. If Perez’s 2013, Rizzo’s 2013, Pedroia’s 2011, Machado’s 2013, Simmons’ 2013, Gordon’s 2014, Kiermaier’s 2015, and Heyward’s 2014 all happened in the same season, for the same team, that team would’ve saved damn near 200 runs on defense. Now, I know what you’re thinking, about how putting all these guys on the same field would likely yield diminishing returns defensively and that the runs-saved total wouldn’t actually reach 195 the way it would in a vacuum. And to that well-reasoned response, I simply stick my tongue out, blow a raspberry, and move on.

So now, to finally get to Joe’s actual question. What would the ERA-FIP spread look like with this team? What kind of ERA might a league-average staff put up? And what might that team’s BABIP be?

In an attempt to answer this question, I gathered every team’s runs-saved total since 2003 — using a 50/50 split of DRS and UZR — as well as every team’s ERA, FIP, and BABIP. Then I let Excel do some work.

As expected, defense and BABIP allowed correlate fairly strongly, with our runs-saved measure accounting for about 33% of the variance in a team’s BABIP. Our 103-run defense with a league-average pitching staff turns what would be a .299 BABIP into something closer to a .270 BABIP. Our otherworldly 195-run defense would be expected to produce something like a .260 BABIP, which would be among the lowest in MLB history. Also, higher than what the Cubs are currently allowing this year. Just a reminder that what the Cubs’ pitchers and defense are doing to balls in play this year is insane.

And now, what all that means for our team’s run prevention:

The 103-run defense is good for a little more than a half-run swing in team ERA, turning a league-average 4.19 ERA into about a 3.60 ERA. For context, this year’s Cubs are outperforming their FIP by 0.69 runs per nine, and that’s the largest difference since at least 2003. Our 195-run defense would yield a team ERA about a full run lower than its FIP. A league-average staff with our otherworldly defense would suddenly became baseball’s best run-prevention unit.

This concludes an exploration in pushing the boundaries of hypothetical roster construction.