Analysis
Exit velocity and launch angle explained, and what they miss
Exit velocity and launch angle explained: the 95 mph line, the 8 to 32 degree sweet spot, how a barrel is defined, and what both numbers still miss.
By CricketTaken EditorialPublished Analysis20 min read
A hitter turns on a fastball and destroys it. The ball comes off at something enormous, the crowd stands up, and the left fielder catches it on the warning track without moving his feet. Everyone agrees he hit it well. Nobody can say how well, or whether hitting it that well again would be a good idea.
Exit velocity and launch angle explained properly are an attempt to make that sentence mean something. Two numbers, taken at the moment the ball leaves the bat: how fast it is going, and at what angle. Neither is interesting alone. Together they turn out to describe the outcome of a batted ball better than the outcome itself does, and that is the whole reason they took over the sport.
The trouble is that the pair have become slogans. Hard-hit rate gets quoted like a batting average. Launch angle gets used as a personality type. Both are more specific and more limited than the way they are talked about, and the limits are where the interesting part is.
What the two numbers actually are
Exit velocity is the speed of the ball as it leaves the bat, in miles per hour, measured immediately after contact rather than when it lands. It is recorded on every batted ball event: hits, outs, errors, the lot. A hitter's average exit velocity is the sum of those readings divided by the number of batted balls, which means it is a measure of typical contact quality and not of how often he makes contact at all.
Launch angle is the vertical angle at which the ball leaves the bat, in degrees, relative to the ground at home plate. Zero degrees is a ball travelling flat. Positive numbers go up. Negative numbers point into the dirt, and a large negative number is a ball chopped straight down.
Both come from tracking hardware installed in every major league ballpark from 2015 onwards, which follows the ball through the hitting zone and out of it. The measurement is of the ball, not of the swing, which matters more than it sounds. A hitter does not have an exit velocity. He has a distribution of them, produced by the interaction of his swing with pitches he did not choose, and the number quoted as "his" exit velocity is a summary of that distribution.
One more definition, because it gets skipped. The vertical angle is only half of the direction information. A ball hit at 103 mph and 28 degrees down the left-field line and the same ball hit to straightaway centre are the same two numbers and, in most parks, different events. The headline pair say nothing about where the ball went horizontally.
- 95Exit velocity, in mph, that counts as hard-hit
- 98Minimum exit velocity, in mph, for a barrel
- 24Degrees of launch angle inside the sweet spot
- 116Exit velocity, in mph, at which the barrel band stops widening
Published definitions from Major League Baseball's Statcast glossary. The sweet spot runs from 8 to 32 degrees of launch angle, a 24-degree window.
Neither number means anything without the other
This is the point everything else hangs on, and it is why the two are always quoted together.
Take exit velocity alone. A ball hit at 105 mph straight into the ground is an easy out. A ball hit at 105 mph at 45 degrees is a long fly out unless the park is small. A ball hit at 105 mph at 27 degrees is very often a home run. Same speed, three completely different events, and no amount of extra precision on the speed reading tells you which one happened.
Now take launch angle alone. A ball leaving at 27 degrees is a home run if it was hit at 105 mph and a routine fly ball if it was hit at 82. The angle is a description of trajectory shape, and the trajectory only carries if there is energy behind it.
So the unit of analysis is the pair. That sounds obvious written down and it is routinely ignored in practice, most often by people quoting a hitter's average launch angle as though it were a strategy.
The relationship between the two is not symmetrical, either. Exit velocity is much more clearly a skill: hitters who hit the ball hard tend to keep hitting the ball hard, and the year-to-year stability of average exit velocity is high enough that a single season tells you something real. Launch angle is more a description of approach, and it is much easier to change deliberately. That asymmetry is why clubs treat raw contact quality as a scouting input and angle as a coaching input.
Why 95 mph is the line
The threshold for a hard-hit ball is 95 mph off the bat. A hitter's hard-hit rate is the share of his batted balls that reach it, and a pitcher's hard-hit rate against is the same thing measured from the other side.
The number is not arbitrary and it is not physics. It is the point at which the league's own production data separates cleanly, and the league has published what the split looks like.
Read the slugging bars against each other and the case for the threshold makes itself. The difference is not a nudge. It is the difference between an offence and a rally-killing machine, produced by a single binary cut on one measurement.
Two cautions come with it, and neither is usually attached when the figure is quoted.
The split is a comparison of batted balls, not of plate appearances. It says nothing about how often a hitter puts the ball in play at all. A hitter who reaches 95 mph on every ball he hits and strikes out constantly may be a worse hitter than one who never does and never misses. Hard-hit rate is a denominator trap and the denominator is batted balls.
And 95 mph is a floor, not a target. It counts a 96 mph ground ball into the shift and a 112 mph line drive as the same event, which they very obviously are not. Hard-hit rate is a coarse instrument, deliberately, because coarse instruments are stable. It is a first sort, not an answer.
The launch angle bands, and where the sweet spot actually sits
The league publishes a contact-type classification built purely on launch angle, and it is worth knowing because a great deal of older analysis is phrased in these terms.
| Contact type | Launch angle |
|---|---|
| Ground ball | below 10 degrees |
| Line drive | 10 to 25 degrees |
| Fly ball | 25 to 50 degrees |
| Pop up | above 50 degrees |
Separately, the sweet spot is defined as a batted ball leaving the bat between 8 and 32 degrees, and sweet-spot percentage is the share of a hitter's batted balls that land inside that window.
Notice that the sweet spot does not line up with the bands. It starts two degrees below the line-drive floor and stops eighteen degrees short of the fly-ball ceiling. That mismatch is the interesting part, and it is doing something specific.
- Hard ground balls, 8 to 10 degrees2deg
- Line drives, 10 to 25 degrees15deg
- Low fly balls, 25 to 32 degrees7deg
The sweet spot spans 24 degrees. Broken down against the league's published contact-type bands, most of it is line-drive territory, with a two-degree reach down into hard ground balls and a seven-degree reach up into low fly balls.
Show the numbers
| Item | Value |
|---|---|
| Hard ground balls, 8 to 10 degrees | 2deg |
| Line drives, 10 to 25 degrees | 15deg |
| Low fly balls, 25 to 32 degrees | 7deg |
The window is mostly line drive, which is unsurprising, and then it does two deliberate things at the edges. It reaches down two degrees below the line-drive floor, because a scorched ball at nine degrees is a base hit through the infield rather than a grounder to short. And it stops at 32 degrees, well below the top of the fly-ball band, because everything above that is where fly balls turn into outs unless they were struck extremely hard.
That upper cut-off is the honest version of the launch angle argument. The fly-ball band is 25 degrees wide. The productive part of it is about seven. Most of the fly-ball band is bad news.
A barrel is defined by its results, not by physics
The barrel is the single most useful batted-ball classification and the one most often described wrongly. It is not a measure of how well the ball was struck in any mechanical sense. It is a lookup table built backwards from outcomes.
A barrel is a batted ball whose combination of exit velocity and launch angle has historically produced at least a .500 batting average and a 1.500 slugging percentage. That is the definition. Everything else is the shape of the region that satisfies it.
The floor is 98 mph. Below that no combination of angles qualifies, whatever the trajectory looks like. At exactly 98 mph the qualifying window is narrow: launch angles from 26 to 30 degrees, and nothing else. Add a single mile per hour and the window opens to 25 to 31. Another one and it opens to 24 to 33. Each further mph adds another two or three degrees, until at 116 mph almost anything counts, from 8 degrees to 50.
The shape of that chart is the whole idea. Precision and power are substitutes. A hitter who can only reach 98 mph has to be almost exactly right about the angle to produce elite damage. A hitter who can reach 112 can be several degrees off in either direction and still hit the ball out. That is what people mean, without knowing it, when they say a hitter has margin for error.
It also explains why the barrel is a better single number than either input. It weighs speed and angle against each other at the correct exchange rate, rather than treating them as two separate boxes to tick.
Two things to keep in mind about the classification. First, it was calibrated on the outcomes of batted balls since tracking went league-wide, in the parks and conditions of that period. It is an empirical region, not a law, and if the ball or the parks or the defensive alignment change enough, the region drifts underneath it. Second, actual barrels have historically produced far better than the .500 and 1.500 thresholds used to define them, because the thresholds are the boundary of the region and most barrels sit well inside it.
From a struck ball to a number on a broadcast
The pipeline from contact to the figure on screen has more steps than most people realise, and each one is a place where a judgement was made.
- The ball leaves the batTracking hardware in the park follows it out of the hitting zone and records its speed and its vertical angle at the moment of separation, along with its horizontal direction.
- Speed becomes exit velocityReported in miles per hour, on every batted ball including outs and errors, which is what makes rate statistics possible rather than just highlight numbers.
- Angle becomes launch angleReported in degrees against the ground at home plate. Zero is flat, negative is into the dirt, and the reading is taken at contact rather than inferred from where the ball landed.
- Sort by angle aloneBelow 10 degrees is a ground ball, 10 to 25 a line drive, 25 to 50 a fly ball, above 50 a pop up. This is the old batted-ball vocabulary rebuilt on measurement.
- Sort by speed aloneAt or above 95 mph the ball is hard-hit. A hitter's hard-hit rate is the share of his batted balls that clear the line, and it says nothing about how often he puts a ball in play.
- Test the pair against the barrel region98 mph minimum, with a qualifying angle band that widens by two or three degrees for every additional mph. This is where speed and angle are finally traded off against each other.
- Look up what balls like this have done beforeThe combination is matched against every comparable batted ball since tracking went league-wide, and the historical hit and total-base rates for that combination are returned.
- Read out the expected statisticsExpected batting average and expected weighted on-base average come from that lookup, with sprint speed added for weakly topped balls where the runner's legs decide the outcome.
- Compare with what actually happenedThe gap between expected and actual is the part attributable to the fielders, the park, the wind, the positioning and the luck. It is a residual, not a verdict.
The sequence used to produce hard-hit rate, sweet-spot percentage, barrel rate and the expected statistics. Steps six and seven are the parts most often quoted and least often understood.
The expected statistics at the end of that chain are where the two headline numbers earn their keep. Expected batting average asks what share of historically comparable batted balls became hits. A line drive with an expected batting average of .700 gets that figure because balls hit at that speed and that angle have fallen in seven times out of ten. Expected weighted on-base average does the same across the full range of outcomes rather than the hit-or-not binary, and folds in the strikeouts and walks that never became batted balls at all.
Sprint speed enters at exactly one place, on weakly topped balls, and the reason is that on a slow chopper the runner's legs really are the deciding variable. Adding it anywhere else would be noise.
The launch angle revolution was misdescribed from the start
The popular version of the story is that hitters worked out that fly balls are better than ground balls, started swinging up, and the home run rate went with them. That is not wrong, and it is a bad summary of what actually changed.
What changed is that hitters stopped optimising the average and started optimising the distribution.
An average launch angle is close to useless as a target, because two very different hitters can share one. A hitter who produces a steady stream of low line drives and a hitter who alternates between choppers to third and towering pop-ups can have identical average launch angles and wildly different outcomes. The average sits in the middle of a distribution that never actually visits the middle. Sweet-spot percentage exists precisely because it counts the balls inside a useful window instead of averaging across everything, and it is a much better number to look at for that reason.
The second misdescription is that the goal is more air. It is not. The goal is fewer balls at the extremes. The bottom of the distribution, balls hit into the ground, produces almost nothing without exceptional speed on the ball. The top of the distribution, balls above the fly-ball band, produces essentially nothing at all: a pop-up is an out at any exit velocity anybody has ever recorded, which makes it the only batted-ball outcome that hitting the ball harder does not improve.
So the actual instruction, when it is being given properly by a hitting coach rather than by a broadcast graphic, is to squeeze the distribution towards the middle band from both ends. Fewer grounders. Fewer pop-ups. Not "hit it higher".
The third thing the revolution story leaves out is the cost, and there is always a cost.
Swinging for angle is not free
A swing that lifts the ball more consistently generally has a steeper upward path through the hitting zone. That path intersects the flight of the pitch over a shorter distance, which means the timing window narrows. Narrow the timing window and the swing misses more often, and the misses arrive as strikeouts.
There is a second cost that gets less attention. A steeper swing that is slightly late or slightly under produces the one batted-ball outcome with no value whatsoever, the pop-up, and it produces more of them than a flatter swing does. Trading ground-ball outs for fly-ball outs is roughly neutral. Trading ground-ball outs for pop-ups is a loss.
This is why the sensible version of the argument is individual rather than universal. A hitter with the raw power to reach the wide part of the barrel region gains a lot from lifting the ball, because his margin for error at those speeds is generous. A hitter without it gains much less, because at 95 mph off the bat the qualifying window is not merely narrow, it does not exist at all: the barrel floor is 98. For that hitter, hard line drives and using the whole field is not a consolation strategy, it is the correct one.
The same logic runs the other way for pitchers, which is why designing a pitch now means designing the batted balls it produces rather than only the swings it misses. A pitch that generates contact at 4 degrees is doing a different job from one that generates contact at 40, and both are doing a better job than one that generates contact at 26 degrees and 104 mph.
Attack angle is not launch angle
These two get conflated in almost every discussion, and separating them clears up most of the confusion about what a hitter can actually control.
Attack angle is a property of the swing: the upward or downward tilt of the bat's path through the hitting zone at the moment of contact. Launch angle is a property of the ball afterwards. They are related and they are not the same thing, and the gap between them is where the skill lives.
The reason they differ is that the bat is round and the ball is round, so the angle the ball leaves at depends on where on the vertical face of the ball the bat arrives. Meet the ball slightly below its centre and it leaves higher than the bat's path. Meet it slightly above and it leaves lower, sometimes into the ground. A single swing, repeated identically, produces a spread of launch angles depending on a contact point measured in fractions of an inch.
Add the pitch to that picture. A ball arriving from a downward plane and a ball arriving flat present different geometry to the same bat path, which is why the same swing yields different launch angles against a high fastball and a low breaking ball. The hitter's approach is fixed for the duration of the swing. The input is not.
Three consequences follow, and they are the useful part.
A hitter cannot choose a launch angle. He can choose a bat path and a target contact point, and then he gets a distribution around it. Anybody promising to teach a specific number is selling something.
The width of that distribution is itself a skill, and it is not the same skill as bat speed. A hitter who consistently arrives at the same part of the ball produces a tight spread of launch angles, which is worth a great deal even at moderate exit velocities, because a tight spread inside a productive window beats a wide spread centred on the same average.
And a change in average launch angle can happen without any change to the swing at all. A hitter who starts seeing more pitches at the bottom of the zone will hit more grounders with an unchanged bat path. Reading a shift in his numbers as a change in approach, when it was a change in what pitchers were doing to him, is a mistake that is very easy to make from the outside and very hard to make with the pitch data in front of you.
What clubs do with it that the public cannot
The public leaderboards are a small window onto a much larger apparatus, and the difference is worth understanding before drawing conclusions from what is visible.
Clubs collect this data everywhere, not only in major league parks. Contact quality is tracked through the minor leagues, in college and amateur environments where a club has installed hardware, and in every cage session and batting practice round a club chooses to instrument. That changes what the numbers are for. In public, exit velocity describes what a hitter did. Internally, it is used to ask whether a hitter can do something he has not done yet, on far more repetitions than a season of games could ever supply.
The acquisition use is the obvious one. A hitter whose contact quality is much better than his results is a buying opportunity, and a hitter whose results are much better than his contact quality is a selling one. That trade has been available to anybody with the public data for years, which means it is well arbitraged and the easy money has gone.
The development use is where the advantage still sits. A club can watch a hitter's launch angle distribution respond to a swing change in real time, in a cage, over hundreds of repetitions, and know within days whether the change is taking hold rather than waiting for a season of results. It can also run the same measurement on a pitcher's bullpen sessions and see what contact his new grip is producing before anybody faces it in a game.
The gap between those two uses is the reason a club's evaluation of a player and the public's can diverge sharply while both sides are looking at honest numbers. One side is reading a season. The other is reading a process.
What the pair cannot see
Everything above is what the numbers do. Here is what they leave out, which is the part that turns a good analyst into a careful one.
Horizontal direction. The headline pair are vertical only. Two identical readings pulled down the line and hit to the deepest part of centre are different events in almost every park, and the difference is not small. Expected statistics built on speed and angle alone treat them as the same ball.
The park. A 100 mph fly ball at 30 degrees is a home run in some venues and a long out in others, and the gap between those venues is systematic rather than random. That is exactly what park factors exist to measure, and the expected statistics do not include them by default. Comparing two hitters' expected numbers without asking where they play is a mistake people make constantly.
The air. Temperature, altitude and wind all change how far a given batted ball travels, and none of them are in the two numbers. A ball struck identically in a cold April night game and a hot July afternoon does not go the same distance.
The defence. Where the fielders were standing decides a large share of batted-ball outcomes, and expected statistics deliberately exclude it, which is the point of them. Excluding it is useful for evaluating the hitter and misleading if you forget you have done it. The gap between a hitter's actual and expected numbers is not a measure of luck alone. It contains positioning, opposition defensive quality, his own speed on balls where speed was not modelled, and genuine randomness, in unknown proportions.
Spin. Two balls can leave the bat at the same speed and the same angle with different backspin and travel meaningfully different distances. The spin imparted at contact is a real physical variable and it is not in the headline pair. This is the batted-ball equivalent of the point that gets made about spin rate on the pitching side, where the raw number matters far less than what the spin is actually doing to the flight.
The pitch. A 108 mph line drive off a 70 mph curveball and the same reading off a 99 mph fastball represent different achievements, because the hitter supplied a different share of the energy. Exit velocity is a measurement of the ball, and the ball arrived with some of that speed already on it.
None of those omissions are defects. A measurement that tried to include everything would measure nothing. But every one of them is a reason that the residual between expected and actual outcomes is not a clean read on fortune, and treating it as one is the single most common misuse of the whole family of statistics.
What this changes for pitchers
The batted-ball numbers arrived in hitting analysis and quietly rewrote pitching analysis on the way through.
The old question was whether pitchers control what happens to balls in play at all. The classic answer, which built an entire branch of the subject, was that they control it far less than anyone assumed, and that year-to-year variation in the rate at which balls in play become hits is mostly noise and defence. That answer is still broadly right, and it was reached without any measurement of how hard the ball was hit.
Contact-quality data lets the question be asked more precisely. A pitcher's exit velocity against, hard-hit rate against and barrel rate against are all measurable, and they are more stable than the outcomes they produce. Suppressing hard contact turns out to be a real skill that some pitchers have, which is a genuine amendment to the older view rather than a refutation of it: they control quality of contact somewhat, and they control what the fielders then do about it hardly at all.
That distinction matters for how the fielding-independent family of statistics is read. The metric that strips out balls in play entirely does so on the grounds that the pitcher's influence over them is small and slow to establish. Contact-quality data does not overturn that, but it does explain why some pitchers persistently beat their fielding-independent numbers and others persistently fall short of them, which was previously filed under mystery.
Using the numbers without embarrassing yourself
Four habits, which cover most of the errors.
Ask what the denominator is. Hard-hit rate, sweet-spot percentage and barrel rate are all shares of batted balls. None of them know anything about strikeouts. A hitter can lead a league in barrel rate and be a below-average hitter if he rarely puts a ball in play, and quoting one of these rates without the contact context is how that gets hidden.
Separate maximum from average. Maximum exit velocity is a scouting number. It tells you what the hitter is physically capable of on his best swing, it stabilises very quickly because it only takes one ball, and it says almost nothing about production. Average exit velocity is a performance number, needs many more batted balls to mean anything, and is the one that shows up in results.
Do not compare expected numbers across parks without saying so. The expected statistics do not adjust for venue. Two hitters with the same expected weighted on-base average in very different ballparks have not had the same season, and neither have their pitchers.
Treat the residual as a question. When a hitter's actual production sits well below his expected production, the useful response is to go and look for the reason: pull-heavy contact into a shifted defence, a home park that swallows fly balls, a slow runner on topped balls, a run of hard line drives at people. Sometimes the answer really is that nothing went his way. Often it is something structural that the model was never asked to see, and that a valuation framework like wins above replacement will eventually have to account for anyway.
The two numbers are the best description of contact quality the sport has ever had. They are also two numbers, taken at one instant, about a ball that then has to travel through real air, over real grass, past real fielders, in a park with its own dimensions. Hold both of those thoughts and the data is enormously useful. Hold only the first and you will spend a season being surprised by results that were entirely predictable.
Common questions
What is a good exit velocity in baseball?
Major League Baseball's published threshold for a hard-hit ball is 95 mph off the bat, and a hitter's hard-hit rate is the share of his batted balls that reach it. That line is the useful anchor because the league has published the production split either side of it, and the gap is enormous. Individual leaderboard figures for average and maximum exit velocity are published each season and move with the run environment, so the threshold is the durable number rather than any particular player's average.
What is a barrel in baseball?
A barrel is a batted ball whose combination of exit velocity and launch angle has historically produced at least a .500 batting average and a 1.500 slugging percentage. It requires a minimum of 98 mph off the bat, and at that speed only launch angles between 26 and 30 degrees qualify. Every additional mph widens the band of qualifying angles, until at 116 mph anything hit between 8 and 50 degrees counts.
What is the ideal launch angle for a hitter?
The league defines the sweet spot as a batted ball leaving the bat between 8 and 32 degrees, and sweet-spot percentage is the share of a hitter's batted balls inside that window. There is no single ideal angle, because the angle that works depends on how hard the ball was hit: a 108 mph ball rewards a much wider range of angles than a 92 mph one. Aiming for an average launch angle is the wrong target anyway, since the average can look correct while the underlying distribution is mostly grounders and pop-ups.
Does a higher launch angle mean more home runs?
Only up to a point, and only with the exit velocity to support it. Balls above 50 degrees are classified as pop-ups and are close to automatic outs regardless of how hard they were hit, while the fly ball band runs from 25 to 50 degrees. Lifting the ball more without hitting it harder mostly converts ground-ball outs into fly-ball outs, which is why the trade-off between angle and contact quality is the whole argument.
How does Statcast measure exit velocity and launch angle?
Tracking hardware installed in every major league ballpark since 2015 follows the ball out of the hitting zone and records its speed and its vertical angle relative to the ground at the moment it leaves the bat. Exit velocity is reported in miles per hour and launch angle in degrees, where zero is parallel to the ground and negative values point down into the dirt. The same measurements are recorded for every batted ball, including outs and errors, which is what makes rate statistics built on them possible.
Filed under Baseball·baseball · mlb · baseball statistics · analytics · statcast · hitting