Written by 23:58 Feed Views: 3

The Pulled Fly Ball Revolution Was Always Underway

Troy Taormina-USA TODAY Sports

I’ll lead with this: I’m not certain the Launch Angle RevolutionTM was ever really a thing — or at least, it wasn’t a thing in the way we thought it was. In 2019, we were faced with an onslaught of home runs that needed an explanation, a genesis. It made sense to turn to launch angles: all else equal, if you hit balls higher, they tend to travel farther. As we’d later learn, juiced balls were much more a culprit than anything else. I wish I could find the sound byte for it – my squishy memory may have manufactured it – but I swear I recall Christian Yelich, perhaps the juiced ball’s most prominent (though, to be clear, not necessarily its biggest) beneficiary, scoffing at the concept of a “launch angle swing.” (Edit: It’s here! Thanks, Mike Petriello!) Although Yelich’s fly ball rate jumped 13.4 percentage points in 2019, he (arguably rightly) denounced the very idea of what everyone assumed had fueled his success.

There is, however, unquestionably another revolution afoot: the Pulled Fly Ball RevolutionTM. Inherently, it’s its own kind of launch angle revolution. But it’s also a spray angle revolution, and a pitch selection revolution, and a swing decision revolution. It is multifaceted and sprawling, and it is much more clearly defined than its predecessor. Here’s the percentage of batted ball events (BBE) that were pulled fly balls (PFBs, for short) by year:

The Pulled Fly Ball RevolutionTM

2018 7,293 126,283 5.8%
2019 7,609 125,751 6.1%
2020 2,817 43,972 6.4%
2021 8,113 121,702 6.7%
2022 8,432 124,265 6.8%
2023 8,767 124,232 7.1%

SOURCE: Statcast

There it is: an indisputable (unless you want to relitigate how fly balls are defined, or dispute Statcast’s stringer labels for fly balls), steady increase in pulled fly balls on a rate basis.

At its most fundamental level, this is just good hitting, good coaching, good whatever-you-want-to-call-it. Ben Clemens wrote just this morning about the merits of pulling the ball with authority. Indeed, all else equal, a pulled fly ball is by far the most productive batted ball:

wOBA on Contact by BBE Type (2021-23)

Batted Ball Type Pull Center Oppo
Popup .005 .010 .019
Fly ball .937 .333 .244
Line drive .755 .598 .624
Groundball .189 .250 .387

SOURCE: Statcast

Why? Hitters generate more power to their pull side, and — if you perhaps haven’t watched a game of baseball before — the shortest distance to the outfield walls is down the lines, toward the foul poles, rather than to straightaway center field. It’s a twofold recipe for success.

Truthfully, not everyone ought to hit fly balls. A hitter’s strength — not just his observed power vis-à-vis home runs or isolated power, but his sheer Aaron Judgian strength, his bat speed — should play a substantial role in dictating the attack angle of his swing. If one did desire to hit fly balls, though, one ought to do so to their pull side.

Hitters are becoming savvier batsmen (i.e., wielders of baseball bats). However, the data suggest it is actually quite difficult to pull a fly ball. Among the launch-spray combinations made up of [groundball, line drive, fly ball, popup] and [pull, center, oppo], a fly ball to the pull side is less likely than its fly ball siblings up the middle or the other way…

BBE Frequency, 2021-23

BBE Type Pull Center Oppo
Popup 1.5% 1.6% 3.9%
Fly ball 6.8% 9.5% 9.5%
Line drive 9.0% 8.4% 6.4%
Groundball 24.1% 13.0% 6.4%

SOURCE: Statcast

…not to mention significantly less frequent than several other launch-spray combinations, too.

There are prescriptions available beyond simply instructing a hitter to pull his fly balls (which I imagine is not a particularly effective coaching strategy). Launch angle correlates strongly with vertical pitch location (pitch height), and, likewise, spray angle correlates strongly with horizontal pitch location. Without meaningfully changing his swing mechanics, a hitter can make better swing decisions (“better” here meaning optimized for pulled fly balls) by offering more frequently at high and/or inside pitches.

Indeed, the league has done exactly this, not only swinging at more pitches up and in…

Swing% by Pitch Location

Year Location Inside Middle Outside
2021 Up 7.1% 11.0% 7.8%
2021 Middle 11.1% 17.3% 14.0%
2021 Down 7.7% 12.8% 11.2%
2023 Up 7.6% 11.4% 7.8%
2023 Middle 11.1% 17.3% 13.3%
2023 Down 7.9% 12.9% 10.7%
2021-23 Total 26.4% 41.2% 32.4%

SOURCE: Statcast

Frequencies calculated as a percent of the total within each year; that is, percentages for each 3-by-3 grid sum to 100%. These swing rates do not account for the number of pitches thrown in each location.

…but also making ever-so-slightly more contact on them, too:

Contact% by Pitch Location

Year Location Inside Middle Outside
2021 Up 83.4% 79.0% 74.8%
2021 Middle 92.0% 88.3% 84.2%
2021 Down 82.0% 83.6% 78.7%
2023 Up 85.2% 80.4% 75.9%
2023 Middle 92.3% 89.3% 84.5%
2023 Down 82.8% 84.2% 78.3%
2021-23 Total 87.2% 84.9% 80.4%

SOURCE: Statcast

Frequencies calculated within each cell, e.g., according to the unique swing rate for pitches low and away in 2022.

It’s not just the nominal, observed value of these types of batted balls that matter; it’s their perceived value, too, one that contributes to the underrating of hitters who routinely leverage a pull-and-lift approach to maximize the value of their batted balls. Again with the relitigating, but I’m not here to relitigate what differentiates (what Tom Tango calls) the play from the player. In fact, through my own research and in trusting those smarter than me, I have found that including spray angle makes xwOBA less predictive, not more.

Nevertheless, it’s clear that pulled fly balls are significantly underrated on the basis of expected wOBA on contact (xwOBAcon) for descriptive (backward-looking) purposes. Here’s the difference between actual and expected wOBAcon, where a positive value represents xwOBAcon underrating actual production:

wOBAcon – xwOBAcon (2021-23)

Batted Ball Type Pull Center Oppo
Popup -.017 -.044 -.006
Fly ball +.263 -.172 -.002
Line drive +.045 -.058 +.027
Groundball -.038 -.001 +.136

SOURCE: Statcast

(In Tango’s defense, you can see that the margin for error between the most-optimized and least-optimized batted ball types in terms of xwOBA — pulled fly balls and fly balls to center field, respectively — is pretty thin. A fraction of a second and/or a fraction of an inch separates a productive fly ball from an unproductive one. And that’s just one defense. But I digress.)

On a rate basis, each percentage point of pulled fly balls in excess of the league-average rate corresponded with about an 0.005 bump in actual minus expected wOBAcon (r2 = 0.29). Someone like Isaac Paredes, rightly lauded for his ability to pull and lift, would be expected to exceed his xwOBAcon by roughly 35 points in aggregate the last two years. His wOBAcon minus xwOBAcon the last two years? +0.056.

Sky Kalkman asked me which pitch types produced more or fewer pulled fly balls. Here you go, Sky!

PFB% by Pitch Type

Pitch Type 2021 2022 2023
Sweeper (SW) N/A 10.4% 12.1%
Slider (SL) 8.7% 8.9% 9.1%
Curve (CU) 7.9% 8.5% 8.5%
Changeup (CH) 6.8% 6.5% 7.4%
Cutter (FC) 6.9% 7.1% 7.4%
League Average 6.7% 6.8% 7.1%
Splitter (FS) 5.7% 5.5% 6.7%
Four-seamer (FF) 6.8% 6.8% 6.7%
Sinker (SI) 3.4% 3.5% 3.3%

SOURCE: Statcast

Sorted descending on 2023.

As you might expect, pitches with greater emphasis on lateral movement — and thus a greater likelihood of being located on the inside edge of the strike zone — have produced higher frequencies of pulled fly balls. Sinkers emphasize horizontal movement, too, but they are called sinkers for a reason. Cutters have largely escaped this fate, possibly benefiting from higher velocities relative to sliders or sweepers, which prevent hitters from getting around on them as easily. In fact, Ben, prolific and sharp as ever, also wrote about the steep platoon splits for sweepers, with opposite-handed hitters teeing off on them.

Need I ask rhetorically where sweepers end up against opposite-handed hitters? Or what opposite-handed hitters might do with those suboptimally located pitches?

PFB% by Pitch Type and Platoon Split

Pitcher Handedness LHP RHP
Sweeper (SW) 9.0% 14.6% 12.6% 11.2%
Slider (SL) 6.7% 10.4% 11.1% 8.1%
Curve (CU) 7.2% 8.8% 7.9% 8.7%
Cutter (FC) 5.8% 8.5% 8.6% 5.3%
Four-seamer (FF) 4.9% 7.6% 7.5% 5.9%
Changeup (CH) 9.5% 6.3% 6.1% 10.2%
Sinker (SI) 2.8% 3.0% 4.2% 3.3%
Splitter (FS) 9.1% 6.5% 5.5% 7.1%

SOURCE: Statcast

Sorted descending on LHP vs. RHH.

And this is just the tip of the iceberg. Do certain teams pull their fly balls more than others? They sure do, friend:

Yearly Team PFB% Trends

Team 2021 2022 2023
LAD 7.6% 7.8% 8.5%
SEA 7.8% 8.4% 8.4%
TEX 5.5% 7.8% 8.1%
MIN 7.5% 7.1% 8.0%
SFG 8.1% 7.9% 7.8%
LAA 7.3% 6.7% 7.7%
ATL 7.1% 7.4% 7.7%
NYM 6.9% 6.1% 7.6%
HOU 7.0% 7.5% 7.4%
DET 6.1% 5.9% 7.4%
BAL 7.0% 8.2% 7.3%
KCR 6.5% 6.9% 7.3%
NYY 6.0% 8.3% 7.3%
OAK 7.6% 5.9% 7.3%
STL 8.0% 8.8% 7.2%
TBR 7.4% 6.1% 7.2%
TOR 7.6% 5.9% 7.1%
SDP 5.8% 6.8% 7.1%
CIN 6.5% 6.6% 6.9%
PHI 5.8% 6.1% 6.9%
BOS 6.6% 5.8% 6.6%
PIT 5.4% 6.5% 6.4%
COL 5.8% 5.8% 6.4%
CHC 7.2% 6.3% 6.4%
WSN 4.7% 5.6% 6.3%
ARI 6.2% 6.9% 6.2%
MIL 6.3% 7.4% 6.2%
CHW 5.0% 5.2% 5.8%
CLE 6.6% 5.6% 5.3%
MIA 6.0% 5.6% 4.9%

SOURCE: Statcast

Sorted descending by 2023. Click headers to sort.

Team-level trends are affected by the static nature of roster composition: The majority of any team sticks around for the next year, thus creating a sort of inertia in a team’s batted ball metrics from year to year. Even if a hitter (or several) makes wholesale adjustments to their swing, their discipline, or their approach, it remains unlikely to move the needle dramatically. Furthermore, big year-over-year changes can possibly be chalked up to noise, or fluke, or what have you, like the Tampa Bay Rays’ team PFB% plunging 1.3 percentage points in 2022 only to rise back up to its previous levels last year. Still, caveats aside, teams appear to have an “approach” — and while I’ve claimed that some year-over-year changes can’t be trusted, certainly some should be. We just won’t know without hindsight. But 2023’s biggest movers — the Dodgers, the Mets, the Tigers on the upswing; the Yankees, the Cardinals, the Marlins on the downswing — may have implications on 2024 performance. Who knows! We’ll see.

Perhaps more interesting (and more pertinent) are the player-level trends, since I suspect that’s what you’re most curious about. Who hits pulled fly balls the most often? The least often? Who uses pulled fly balls to offset a lack of true plus(-plus) power, to create the illusion of that power? Is it everyone you’d expect? (Answer: Probably.) Here’s a table of the top- and bottom-10 PFB hitters last year (min. 200 BBE). The full table is prohibitively long, so if you’d like to investigate specific hitter-seasons not featured here, you can use the Pitch Leaderboard or inquire in the comments:

2023 PFB% Leaders/Laggards

SOURCE: Statcast

One last question interested me most: Who hits the most pulled fly balls against velocity? This could suggest many things — an approach that favors inside pitches, or an aptitude for making swing decisions early — but I imagine it suggests bat speed above all. Here are the leaders and laggards who, with at least 50 BBE against pitches 95 mph or faster, have the highest rates of PFB%. The league-average PFB% against such pitches was just 4.4% in 2023 (whereas it was 7.5% for all pitches below 95 mph). It’s Anthony Santander’s world, we’re just living in it:

95+ mph PFB% Leaders/Laggards

SOURCE: Statcast

Is it bad to see visible decline (or at least profound fluctuation) in PFB% against high velocity? Frankly, I didn’t research it. You can! I implore you to. But I will say this: Ronald Acuña Jr. had one of the steepest drop-offs of any player last year (down from 11.1% in 2022 to 2.5%), and he compiled one of the most remarkable seasons of all time. So anecdotally and haphazardly (but I hope correctly), I will hypothesize that, like most things, it depends on the hitter. Yordan Alvarez and Austin Riley? Two of baseball’s most feared and lethal sluggers? Zero-point-zero-zero PFB% against heat. In and of itself it’s not a death knell, and even steep drop-offs require additional context (as pretty much all good analysis does). But surely there’s a signal in there somewhere. It’s definitely not just bat speed, though.

What can we expect this year? Typically it’s unwise to extrapolate linearly, but it certainly seems like we’ll see continued growth in PFB rates league-wide. Like, how wrong will I be if I guess 7.3% or 7.4%? Until there’s some kind of sea change, some kind of inflection point, we shouldn’t expect otherwise, especially as hitters dial in on this particular highly optimized subset of batted balls.

What might that inflection point look like? I anticipate that in the ever-evolving game of chess between hitters and pitchers, game theory will push pitchers to stratify how they select, locate, and shape their pitches according to the hitter’s handedness. This already happens, to be clear, but it may become increasingly black-and-white, especially for offerings with strongly lateral movement profiles. The rub: It’s easier said than done.

One thing that stands out to me, though: pitchers tend to locate away more than in; hitters swing more often inside; and hitters make less contact per swing outside. There’s merit to inducing fewer swings; a called strike is a strike, after all, and inciting more swings (despite lower contact rates) invites the natural chaos imbued in every batted ball event. But batted balls the other way are also dramatically less damaging than pull-side batted balls such that allowing a few extra of those would be worth avoiding making mistakes on the inner half. And, again, based on the swing and contact rates, pitchers probably wouldn’t be allowing any extra batted balls. There’d be fewer! And they’d be less damaging! It feels so simple! It can’t be this simple.

At any rate, I think any movement toward the aforementioned sea change will be a slow burn. Until then, I’ll have an eye on 2024’s early trends, and now you will too.


(Visited 3 times, 1 visits today)