Why Don’t the Rockies Use Four Outfielders?

© Kevin Jairaj-USA TODAY Sports

The Colorado Rockies are projected to be gobsmackingly bad in 2022. Look no further than the summary of this season’s positional power rankings: They have three positions that rank 30th and six below 20th, which works out to a cumulative last-place finish. Most of it traces back to a lack of certifiable talent on Colorado’s roster. But some of it, inevitably, is a function of Coors Field. Today, I will mainly focus on the fact that at home, the Rockies allow lots and lots of runs. The common thinking is that this is because Coors is an environment conducive to home runs. While true, there’s another factor that arguably matters more. Check out this graph:

When a fly ball or line drive is hit at Coors, the resulting .459 BABIP has led all of major league baseball by a laughably wide margin for the past few years. The gap between the Rockies in first place and the second-place Red Sox (.421) is equal to that between second place and the 15th-place Orioles (.393). If you’re wondering why, the outfield at Coors is absolutely enormous, so much so that it’s hard to believe just three men patrol it. The thin air helps the ball travel, but crucially, there’s also a lot of space for it to land. It’s a two-part mechanism that captures why offense can get out of control in the Rockies’ home park.

As such, it’d be in the team’s best interest to limit the number of balls that land for hits. We can at least tell from Baseball Savant that the Rockies are shifting their outfielders to the left against right-handed hitters and vice versa, which research from Rob Arthur has shown to save roughly 20 runs per season. Perhaps related to that, they surrendered the lowest BABIP on fly balls and line drives (.423) on record last season. (Even after accounting for a league-wide drop in BABIP, it’s still extremely significant.)

But maybe the Rockies can do better. Think – what’s better than an alignment with three fielders? That’s right, an alignment with four. Now, even in the age of analytics, four-man outfields are seldom seen, reserved for the most adventurous of clubs against the most grounder-averse of sluggers. We’re dealing with special circumstances, however, and as they say, modern problems require modern solutions. Taking this a step further, what if the Rockies used a four-man outfield in every possible situation? What happens on the ground doesn’t seem that dangerous, after all. It’s the type of stupid that fascinates me to no end. It’s also the type of stupid that just might work.

It’s time for some, but not all, of the math. First, let’s establish when to employ a four-man outfield. We’re going to limit ourselves to instances with no runners on base, since it’s nearly impossible to control them with just three infielders. As for which batters to align against, I’m going to ignore them in the interest of time and my well-being. Besides, the focus here is the big picture and not so much the nitty-gritty details. I also altered which balls count as fly balls and line drives to exclude ones that outfielders couldn’t plausibly reach.

Using that custom definition, the Rockies have allowed 2,266 fly balls and line drives from 2018–21, for a .385 BABIP. If a four-man outfield was used in each instance, how many runs would the Rockies have saved? It’s hard to say, and even an intricate model might have trouble spitting out a precise number. But let’s assume such an extensive effort would have helped the team keep pace with the second-place Red Sox, who allowed a .343 BABIP during that same span. Maybe a 42-point decrease is an optimistic figure, but then again, there’s quite literally a lot of room for improvement.

To work out the Rockies’ gain, I first took the 78 extra outs created from a reduction in BABIP and distributed them in proportion to the actual number of singles, doubles, and triples generated by those fly balls and line drives. It’s way simpler than it sounds:

Theoretical Post-Shift Outcomes
PA Result Before After Delta
Single 266 237 -29
Double 392 349 -43
Triple 58 52 -6
Home Run 407 407 0
Outs/Errors 1,143 1,221 78

Because doubles constitute the majority of in-play base hits, they receive the greatest share of outs, followed by singles and then triples. Home runs remain unchanged, as they aren’t factored into BABIP and are untouchable by any possible outfield configuration. Now, here’s the fun part. The run value associated with a given result shifts with each season, but for convenience, I used the linear weights from 2021 only. Since a single was worth 0.73 runs relative to an out, calculating how many theoretical runs the Rockies surrendered on 266 of them is a simple matter of multiplication.

Repeat the process for doubles and triples, compare the total runs allowed before and after vigorously implementing a four-man outfield, and bam! We arrive at 73.2 runs saved over the course of three-and-a-half seasons. So far, so good for our Denver denizens. But every maneuver comes at a cost. That’s especially true in baseball: You can’t just yank away an infielder, place him far from his original home, and count on everything else to remain constant.

What’s one of the biggest reasons why teams are reluctant to use four-man outfields? That lack of a fourth infielder is huge. It sometimes isn’t enough even with one, which is why infield shifts that maximize available resources have exploded in popularity. But by relying on three infielders, teams are essentially leaving a third or more of the infield completely bare. And while it’s true that players are aiming to lift the ball, a lot of the action still occurs near or on the ground.

Based on images I’ve seen of four-man outfields, I’m going to assume the Rockies organize into a standard infield alignment against left-handed hitters, minus the third baseman. It’s a little harder to imagine a procedure against right-handed hitters because there isn’t an obvious one, but I decided to take a page out of the Rays’ playbook:

You can see two fielders positioned at the pull side, with the first baseman covering the opposite field but not so much as to completely neglect his base. It’s a valiant effort with limited manpower, and yet there’s a gaping hole in center – a ball knocked up the middle with enough force is certainly a single. With creative positioning, you win some, you lose some.

The hitter featured there is none other than Marcus Semien. He didn’t hit a ball towards center, so we won’t know what would have actually happened. But hey, we can make a series of educated guesses. It’s time to once again crunch the numbers, finding out how many runs the Rockies would stand to lose from their reckless pursuit of outfield security. As with fly balls and line drives, I adjusted the definition of a groundball to exclude those not impacted by the shape of the infield. From 2018–21, Rockies pitchers allowed 1,261 specific groundballs to lefties and 1,631 to righties. That gives us a good sample to start our investigation.

Unlike with fly balls and line drives, however, I arrived at a different methodology. This is chiefly for two reasons. First, because there are a limited number of outcomes on a groundball, converting lost outs into whole singles would represent an overcorrection. Second, while the overall sample size is robust, specific situations are lacking in quantity. For instance, there weren’t many instances where the Rockies shifted against a righty hitter, which are necessary for our calculus.

Here’s what I did instead. With left-handed hitters, I split those 1,261 groundballs by direction – pulled, straightaway, or opposite. Next, I looked at the league-wide numbers from 2018–21 and queried lefties’ performances on each groundball type against standard or shifted alignments. When a lefty hit a groundball the opposite way against a shifted infield as opposed to a standard one, he picked up 130 points of wOBA on average. Since three-man infields also create an opposite-field weakness, I applied that gain in wOBA to the actual opposite-field groundballs allowed by Rockies pitchers.

Furthermore, based on the Rockies’ shift rates against lefties during that time span, I also made a proportion of pulled grounders more valuable. Sticking to a three-man infield doesn’t just permanentize the downside of a shift; it also removes its upside. Naturally, the only groundballs I didn’t touch were the up-the-middle variety. Below is a summary in table form:

Change in Performance vs. Lefties
Direction Before After Delta
Pulled .220 .267 .047
Straightaway .304 .304 .000
Opposite .494 .624 .130

As is usually the case, right-handed hitters proved to be more of a hassle. The problem is that the standard righty shift doesn’t leave the region around second base so vulnerable – the second baseman is still there, ready to snatch up anything to his left. When righties have hit a straightaway grounder into the shift since 2018, they’ve posted a .363 wOBA, as opposed to a modest .299 mark against a regular, symmetrical infield. In the end, I doubled the 64-point difference in wOBA and set that as the expected outcome with a three-man infield. I know that isn’t exactly scientific, but it’s reasonable nonetheless. In sum:

Change in Performance vs. Righties
Direction Before After Delta
Pulled .268 .277 .009
Straightaway .299 .427 .128
Opposite .507 .507 .000

With all the groundwork finally laid out, only the final calculations remain. Unfortunately, the results are heartbreaking to this baseball fool. From earlier, we know the “four outfielders or death” approach saves 73.2 runs. But, as a consequence, it also allows approximately 117.3 runs. The net loss, therefore, is 44.1 runs, or roughly four wins over the course of three-and-a-half seasons. Cue the sad trombone sound effect.

As you can tell, I came into this project expecting better results. The BABIP on fly balls and line drives in Coors is unfathomably high, so it seemed like trying to stop the bleeding at all costs would benefit the Rockies no matter what. But doing so ultimately left too many holes in the infield, and again, the reality is that there are many more grounders than there are air balls, even though the latter are far more damaging. They could be ignored entirely with a humongous reduction in air BABIP, but it’s to an extent that doesn’t seem feasible even with a cast of defensive superstars.

This doesn’t mean four-man outfields are a bad idea for the Rockies, however. If anything, I still endorse them! So how could they make the most of them? For one, pumping the brakes against right-handed hitters alone would drastically improve results. If we remove righties entirely from our earlier calculations, the net loss in runs relative to an out shrinks to a mere 1.4 runs. Also, the Rockies wouldn’t actually be deploying four outfielders against so many batters. In real life, groundball rates and other batted ball tendencies of individual hitters should determine whether or not to fortify the outfield. Against a slap hitter, it’s a no go. But against an air-heavy guy, that’s a green light. In theory, this sort of efficiency would save additional runs. What’s better, the Rockies’ circumstances will likely let them be more liberal with four-man outfields, though I’ve yet to work an exact threshold.

Optimizing when to use a four-man outfield won’t gift the Rockies enough wins to become contenders. Moreover, if the new CBA is any indication, it’ll probably become obsolete following this season. It is, however, a step in the right direction. As it is, the Rockies find themselves in an unfavorable situation due to effects of their home park. So far, they’ve been behind the curve in terms of how to address those effects. But it isn’t enough to follow the curve – no, for future success, they need to be ahead of it. That’s no easy task, but it’s a necessary one. Anything could help, including, say, an extra defender who can track down the copious balls hit towards that gargantuan turf. What’s there to lose?





Justin is a contributor at FanGraphs. His previous work can be found at Prospects365 and Dodgers Digest. His less serious work can be found on Twitter @justinochoi.

53 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Kenny Ockermember
7 months ago

Nice try, Eric Hosmer.