Release Angles and the Illusion of Waste
Release angles comprise the vertical and horizontal angles at which a pitcher releases a pitch. They are the natural counterpart to approach angles, except they capture the initial angle of a pitch’s trajectory rather than its final angle upon crossing home plate. Release angles can tell us a lot — namely, where a pitch is headed (or, at least, intended to go). However, we already have plenty of data to describe a pitch’s flight path. We have its short-form movement (i.e., total inches of break), as well as its acceleration and velocity vectors in all three dimensions, not to mention its final location coordinates. We can pretty much map the entire trajectory without release angles. Like the last unrevealed letter in Wheel of Fortune, you theoretically need it to solve the puzzle, but you can probably infer the word or phrase just fine without it. What are release angles, then, if not just a different way to describe a pitch’s movement in space? What information do release angles add? (Michael Rosen adeptly provided an answer to that question here.)
When a pitcher throws a pitch, the pitch reaches home plate in a fraction of a second. The opposing hitter, then, has a fraction of a fraction of a second to discern a great many things about the pitch: its velocity, its shape, its probable final location, all to then ascertain whether or not he should swing. Given the impossibly small window of time in which to make a swing decision, much of a hitter’s behavior is influenced by the untold thousands of pitches he’s seen before, like a mental library of pitch shapes. One of the very first visual cues a hitter receives, aside from the pitcher’s release point, is the angle at which a pitch leaves the pitcher’s hand. This particular visual cue ought to enable a hitter to determine out-of-hand a prohibitively bad pitch — one that, on most occasions, will not find the zone. He can potentially make a snap decision with a fairly high degree of confidence that the pitch will miss the zone.
Release angles tell us plenty about what we do see about the pitch’s flight path, but they also tell us what we don’t see: where a pitch appears to be going, before we know where it actually ends up. And those two things aren’t always the same. In the ever-evolving game of baseball chess, a skilled pitcher could command good pitches with “bad” release angles, finding the zone with pitches that appear to have no business doing so. Is there evidence of this possibly being true? If so, which pitches achieve this illusion best?
Using Statcast data from 2022 through April 27 of this year, I calculated the average zone rate for each combination of vertical and horizontal release angles, rounded to the nearest tenth of a degree. I then judiciously (but ultimately arbitrarily) focused only on all pairwise release angles with average zone rates of 20% or less — the typically “bad” release angles in blue, below:
This analysis is agnostic of pitch type (theoretically, at least — more on that in a bit). If the initial release angle immediately signals to hitters that the pitch will miss the zone no matter what, that it is uncompetitive, does it really matter what kind of pitch it is? The shape is immaterial. (Again, in theory.)
Who throws the highest frequency of pitches that a hitter ought to quickly identify as probably wasteful? Tim Hill, followed by Tyler Rogers, which for him is an occupational hazard. The fewest? Logan Webb, with one. One! In almost 7,000 pitches. Rafael Montero, too, had zero in 2,500-plus pitches.
I was shocked that Blake Snell, subject of many think pieces on his unique brand of “command,” didn’t lead this metric on a rate basis. He does lead in terms of raw count, at 374, but as a percentage, his 6.5% clip, while elevated, isn’t egregiously high. Moreover, the number of Immediately Obvious Waste Pitches (IOWPs, for short) that a pitcher throws is decidedly less detrimental if he can induce the hitter to swing (Snell does just that), or if he can find the zone irrespective of the hitter’s swing decision. Case in point: Kyle Gibson and Justin Verlander have thrown roughly the same number of IOWPs (260 in 6,200-plus pitches for Gibson, 255 in 5,300-plus pitches for Verlander), yet one of them has induced swings on 44% of those pitches — the other, just 5%. Can you guess which is which?
Verlander is quite a bit better than Gibson at both inducing swings with IOWPs and finding the zone with them. In fact, Verlander is essentially the best. By the measure of “effective” IOWPs, he stands head and shoulders above everyone else — except, of all people, Dylan Floro (min. 1,500 pitches since the start of 2022). Verlander’s IOWPs find the zone a whopping 74% of the time (with his 44% swing rate), light-years ahead of the third-best pitcher on this list, Max Scherzer (47% zone, 30% swing), who, incidentally, is still pretty good. And right behind Scherzer? Blake “actually, his high walk rate is a good thing” Snell (40% zone, 28% swing), followed by flamethrower Hunter Greene (39% zone, 23% swing). It’s probably no coincidence that Verlander, Snell, and Greene, alongside other high-swing/high-IOWP guys like Nick Pivetta and Carlos Rodón, are stuff model darlings. Do challenging release angles make a good fastball? It’s not causation, but at the very least it’s correlation, and that’s pretty compelling to me.
It’s worth acknowledging that IOWPs are not created equal. They can miss the zone most of the time but not always, or miss it every time, without a doubt, plus everywhere in between. Under this definition there’s relative waste and absolute waste, and I imagine hitters can differentiate between the two. While it would behoove me to do so, I don’t make that distinction here.
Also, I know I said earlier that this analysis is agnostic of pitch type, and technically it is, but IOWPs are comprised almost exclusively of fastballs. Marquee exceptions to the rule: Rogers’ slider (Rogers is an exception to everything), Snell’s slider, Rich Hill’s curve. But it’s otherwise fastballs for days. Snell, of course, is perpetually fascinating: His IOWP fastballs find the zone half the time, his IOWP sliders literally never (since the start of 2022, at least). And, unlike his IOWP fastballs, those IOWP sliders don’t fool hitters into swinging, either. That is true waste.
Anyway, on its face, it seems bad that a pitcher might throw five or 10 IOWPs a start. Outside of Floro and Verlander, most pitchers find the zone less than half the time with their IOWPs (and induce swings even less often than that), which suggests that, on average, they may be more detrimental than they are beneficial. But I hypothesize that this is less about the magnitude of waste (although magnitude does matter — there must be an inflection point) than it is about creating uncertainty. It’s a leap in logic, but I assume pitchers who throw lots of IOWPs have more challenging release angles in general, that every pitch of theirs to some extent defies the visual cues and muscle memory that define swing decisions and plate discipline. IOWPs, then, are for some pitchers a necessary (and intentional) evil, one that sows doubt in a hitter’s mind and disrupts his mental library.
Of course, too many IOWPs is probably a bad thing. Is Hill particularly good? Adam Cimber? Bryan Hoeing? Max Castillo? There are two things they all share in common: league-leading rates of IOWPs and league-lagging rates of strikeouts. Then again, just steps behind them you’ll find the José Alvarados and Félix Bautistas and Jeff Hoffmans and Pete Fairbankses of the baseball world. The efficacy of IOWPs, as with so many other things pitching-related, is massively dependent on capital-‘s’ Stuff and command.
At the other end of the IOWP spectrum, we have the Kings of Command — George Kirby, Zack Wheeler, and Aaron Nola — who all basically never throwing IOWPs. Webb, too, who has been for some time one of baseball’s most effective and efficient starters. It wouldn’t be entirely surprising to learn that never throwing IOWPs can also be a good thing: If every pitch appears competitive out-of-hand, a hitter faces similar uncertainty in trying to decipher which pitches won’t find the zone. No doubt I could devote untold time and attention to the flip side of this coin, not to mention the hitter side of this puzzle as well.
Can this skill, the IOWP-that’s-actually-good, be taught, or is it innate? You can train a pitcher to release a pitch a certain way, perhaps a different way than what he’s used to, but will the pitch still have the same quality of stuff? The same command? For any pitcher, there’s no single path to glory, but the chosen path may not be the correct one if he’s not ideally equipped to traverse it.
I don’t find raw measurements for release angles especially interesting; they, like so many other metrics, are heavily context-dependent. But isolating the “bad” release angles that pitches somehow make “good” — well, that’s one way to leverage release angle data. I’m inclined to believe all this is more a fun curiosity, a novelty, than anything else. I’m not convinced I’ve even proven any point in particular. Mostly I’ve implied the existence of evidence and let the results speak for themselves. At any rate, the results are intriguing to me, and sometimes that’s all that matters.
Here’s a who’s who of pitchers (min. 5,000 pitches, 2022-24) and their select IOWP metrics:
Player Name | IOWP | Pitches | IOWP% ▴ | Zone/lOWP | Swing/lOWP | IOWP Swing/Zone Ratio |
---|---|---|---|---|---|---|
Blake Snell | 374 | 6.5% | 5,744 | 39.6% | 28.1% | 0.71 |
Justin Verlander | 255 | 4.7% | 5,382 | 73.7% | 44.3% | 0.60 |
Dean Kremer | 237 | 4.5% | 5,325 | 12.2% | 4.2% | 0.34 |
Kevin Gausman | 262 | 4.2% | 6,250 | 22.9% | 12.6% | 0.55 |
Kyle Gibson | 260 | 4.2% | 6,237 | 5.8% | 5.0% | 0.87 |
Taijuan Walker | 220 | 4.1% | 5,337 | 21.4% | 7.3% | 0.34 |
Jesús Luzardo | 203 | 4.0% | 5,102 | 9.9% | 6.4% | 0.65 |
Jordan Montgomery | 226 | 3.9% | 5,800 | 11.5% | 13.3% | 1.15 |
Nick Pivetta | 217 | 3.8% | 5,639 | 32.7% | 12.4% | 0.38 |
Miles Mikolas | 248 | 3.6% | 6,840 | 26.6% | 11.3% | 0.42 |
Zac Gallen | 214 | 3.2% | 6,730 | 17.8% | 8.9% | 0.50 |
José Berríos | 186 | 3.0% | 6,196 | 29.0% | 10.8% | 0.37 |
Lance Lynn | 168 | 2.0% | 5,631 | 9.5% | 8.9% | 0.94 |
Patrick Corbin | 178 | 3.0% | 5,974 | 16.3% | 10.1% | 0.62 |
Spencer Strider | 161 | 2.9% | 5,555 | 9.3% | 10.6% | 1.13 |
Yu Darvish | 155 | 2.8% | 5,578 | 8.4% | 5.2% | 0.62 |
Gerrit Cole | 180 | 2.7% | 6,555 | 7.8% | 3.3% | 0.43 |
Lucas Giolito | 160 | 2.7% | 5,943 | 1.9% | 3.1% | 1.67 |
Dylan Cease | 185 | 2.7% | 6,960 | 3.2% | 7.6% | 2.33 |
Sandy Alcantara | 153 | 2.6% | 5,969 | 12.4% | 7.8% | 0.63 |
Mitch Keller | 158 | 2.5% | 6,334 | 1.9% | 1.9% | 1.00 |
Tyler Anderson | 136 | 2.5% | 5,474 | 0.0% | 7.0% | n/a |
Luis Castillo | 158 | 2.5% | 6,365 | 5.7% | 3.2% | 0.56 |
Charlie Morton | 139 | 2.2% | 6,202 | 2.2% | 3.6% | 1.67 |
Cal Quantrill | 113 | 2.2% | 5,199 | 2.7% | 9.0% | 0.33 |
Reid Detmers | 114 | 2.2% | 5,249 | 4.4% | 3.5% | 0.80 |
Dane Dunning | 116 | 2.0% | 5,700 | 0.9% | 1.7% | 2.00 |
Patrick Sandoval | 100 | 1.8% | 5,605 | 4.0% | 2.0% | 0.50 |
Yusei Kikuchi | 92 | 1.8% | 5,202 | 2.2% | 0.0% | 0.00 |
Chris Bassitt | 115 | 1.7% | 6,593 | 12.2% | 11.3% | 0.93 |
Joe Ryan | 94 | 1.7% | 5,526 | 2.1% | 2.1% | 1.00 |
Jordan Lyles | 94 | 1.6% | 5,877 | 0.0% | 0.0% | n/a |
Logan Gilbert | 92 | 1.4% | 6,401 | 0.0% | 3.3% | n/a |
Sonny Gray | 62 | 1.2% | 5,011 | 9.7% | 6.5% | 0.67 |
Framber Valdez | 74 | 1.2% | 6,141 | 0.0% | 0.0% | n/a |
Ranger Suárez | 54 | 1.0% | 5,145 | 3.7% | 1.9% | 0.50 |
Jameson Taillon | 53 | 1.0% | 5,500 | 0.0% | 0.0% | n/a |
Pablo López | 52 | 0.8% | 6,374 | 0.0% | 0.0% | n/a |
Corbin Burnes | 56 | 0.8% | 6,918 | 1.8% | 0.0% | 0.00 |
Cristian Javier | 41 | 0.7% | 5,805 | 0.0% | 0.0% | n/a |
Zack Wheeler | 43 | 0.7% | 6,111 | 2.3% | 2.3% | 1.00 |
Merrill Kelly | 43 | 0.7% | 6,235 | 0.0% | 0.0% | n/a |
Brady Singer | 36 | 0.7% | 5,538 | 0.0% | 0.0% | n/a |
Aaron Nola | 43 | 0.6% | 6,693 | 2.3% | 0.0% | 0.00 |
Josiah Gray | 30 | 0.5% | 5,585 | 0.0% | 0.0% | n/a |
George Kirby | 23 | 0.4% | 5,447 | 0.0% | 0.0% | n/a |
Kyle Freeland | 18 | 0.3% | 5,566 | 0.0% | 5.6% | n/a |
Martín Pérez | 5 | 0.1% | 5,731 | 0.0% | 0.0% | n/a |
Logan Webb | 1 | 0.0% | 6,770 | 0.0% | 0.0% | n/a |
IOWP = a pitch with release angles that, on average, have a zone rate of 20% or lower
Zone/IOWP = the percentage of IOWPs that actually find the zone
P.S. You can find release angle data on the Pitch Leaderboard.
Super interesting. I’m struggling to wrap my head around the shape of that initial heat map.
I can guess that the lower red clusters represent lefties and righties throwing… breaking balls?
And the top cluster (which skews to one side, because more righties?) represents fastballs?
And then I love the light shaded red dot near the top right — does that mean there were multiple pitches released at that angle?
I would have to parse the heat map but I think it may be the opposite. Pitchers tend to their fastballs more steeply (despite often locating up, i.e., for four-seamers) because (I’m guessing) they are compensating for the downward shape of their breaking pitches. So, they tend to actually elevate their vertical release angles.
And, haha, yes, those individual dots do theoretically indicate pitches release at those angles, although it’s possible they’re Statcast misreads, especially if they’re colored red.