A Visual Deep Dive on Spray Angles
Introducing the BOSA Framework
There has been some discussion online as to whether or not the horizontal spray angle (HSA) should be included in certain descriptive and/or predictive metrics. The common example cited is Isaac Paredes who appears to produce surplus value from a focus on pulling fly balls, which would suggest to some, that we should be including this important variable when describing a batter’s hit quality.
I’m not going to be diving into that debate today. Rather, the focus today will be to deepen our understanding of how the 3 launch variables - exit velocity (EV), vertical launch angle (VLA) and horizontal spray angle (HSA) interact with each other, and how they interact with one of their key outcomes - launch distance (DIST).
DATA QUALITY CAVEAT
The HSA is calculated using the hit chart x and hit chart y variables. These data are less than perfect. However, they are all I have, and, they appear to work quite well. I would happily re-do this entire piece if MLB would publish the hit data details, including spin rate (in all 3 directions) and spray angle, and ideally the entire sample of data points for each batted ball.
Negative numbers = Pull Side
Positive numbers = Oppo Side
Horizontal Spray Angle Affects Distance
AKA Fly Balls Carry More to Center Field
If you’d like to understand why, Dr. Nathan has a detailed explanation, I’ll post his summary here verbatim, but I encourage you to read the whole article:
In this article, I have shown evidence from StatCast data that fly balls carry better to centerfield than to right or left field. I have also shown that the reason for reduced carry to left and right is sidespin. The reduced carry due to sidespin comes from two primary causes: the increased drag as the total spin increases and subtle changes to the Magnus force that reduce the lift. Finally I have shown that the curved path of a ball hit with sidespin contributes very little to the reduction of distance.
Source: http://baseball.physics.illinois.edu/carry-v2.pdf
Sweet Spot (8° to 32° VLA) & Distance per EV
We don’t quite see distance maximized at precisely 0°, but it’s pretty close. Keep in mind that I don’t have a ton of precision in the exact spray angle, so it could be shifted +/- 5 degrees. We can have a little more confidence in the data now that we’ve essentially reproduced Dr. Nathan’s findings. Note that Dr. Nathan used a much more specific sample than the one above.
Horizontal Spray Angle Affects Launch Angle
When I saw this chart, I began to question whether my data were broken somehow. I have zero high level baseball experience, so my intuition was based on the concept of “lift and pull” and that somehow those two went hand in hand. Turns out, it’s really hard to both pull the ball and lift it at the same time.
I was concerned that I had messed things up. So I asked Twitter. Amazingly, I got two responses right away. Here is Robert’s explanation:
I also received an explanation from an anonymous source, who is highly qualified to speak on the subject. I’ll post his comments verbatim (I’ve lightly edited for grammar).
The further out in front of the plate a batter makes contact… the more the bat extends away from him. When the top hand releases the barrel - it gets really really flat, making it virtually impossible to elevate the ball. This is what happens when batters are fooled by off speed pitches.
When the barrel releases the angle of the bat is always gonna be towards the pull side.
And that’s because when hitters guess wrong/get fooled they release the barrel, it points to 3B side (for RHB)
The flip side what happens is if a batter is super late on a FB… the bat is angled far behind him and he will “cut” under the ball too much and it backspins it so much creating a flare down the RF line (for RHB)
There’s a trade off with it. The more try to pull… the easier you can get fooled by off speed. The longer you wait… the more time you can have to decide, potentially improve ability to make contact. The smaller guys that have marginal power but cheat pull side usually have elite true contact/swing decision skills.
Drew Haugen showed a similar result in his excellent piece on spray angles:
Horizontal Spray Angle Affects Exit Velocity
We see a very strong relationship between the horizontal spray angle and exit velocity. If we assume a linear trend, each degree away from pure pull side costs about 0.09 to 0.1 MPH of exit velo, or roughly 10 MPH from -45 to 45, which is quite significant.
So Everything is Connected to Everything?
Yes. HSA affects VLA, EV and DIST, but EV and VLA also affect distance. If you pull the ball, you get more EV, but maybe you lose some VLA? We’re ignoring that ballparks often have short porches, so maybe if it all equals out, it’s still worth it?
Batter Optimal Spray Angle (BOSA)
We want to classify hitters by what spray angles they do best in. At first, my intuition was to classify it as an “approach”, but my anonymous advisor pointed out that we can’t actually know a player’s approach. They may have an opposite field approach that results in a ton of pull side contact. Rather than assume, we’ll simply try to figure out where they excel, and call it their BOSA. No I am not a 49ers fan. I used to root for whichever team Peyton Manning was on, and have since moved on to rooting for Pat Mahomes.
Methodology
Numbers are average HSA by trajectory. This is essentially the same chart as above which showed the relationship between HSA and VLA. For most batters, if they hit a ground ball, it’s not what they intended to do (this isn’t necessarily true for some batters, but for today we’ll pretend it’s mostly true). The same can be said for Pop Ups, it’s not what the batter wanted to do.
BOSA should capture the average spray angle when the batter makes good contact. How we define that is somewhat arbitrary, but we’ll look at things three ways and try to come up with a reasonable approach.
Option 1 - Sweet Spots = 8° to 32°
Generally speaking, good things happen when a batter generates contact in this VLA range. It of course varies based on how hard the batter hits the ball, but it is a generally accepted range of VLA that represent good contact.
Option 2 - Hard Hit Balls = 95 MPH+
Alternatively, we could only look at batted balls that are “hard hit”.
Option 3 - Hard Hit and Sweet Spots
This isn’t quite “barrels” but will do a great job of isolating when batters are making high quality contact.
Filter Out 2 Strike Counts
Some Batters will change their approach on a 2 strike count, so we’ll exclude all 2-strike counts for all 3 options.
BOSA - Sweet Spots (Min 50 BBE <2 Strikes)
We see a few batters where they would be classified differently. Jose Barerro and Joey Votto would be middle of the pack if we only looked at HH. Sal Frelick looks like a pure oppo hitter by sweet spot, but less so for HH. Steven Kwan is a complete oppo hitter by sweet spot, but more of a pull hitter by the other two metrics.
Isaac Paredes is an extreme pull hitter no matter how we slice it, but Kody Clemens has him beat.
BOSA - Hard Hit
This is interesting. The list of oppo hitters looks very consistent, however, the list of pull hitters is a little noisier, with a lot of low exit velo guys (I’m using EV80 since average is close to 100) looking like pull hitters. Is Adam Frazier a pull hitter? HH and HHSS angles would suggest that’s where he’s optimized for. Or maybe since a lot of his BIP are below 95 MPH, this is only capturing the times he pulls the ball (remember there is a direct correlation between HSA and EV).
We do have one definitive takeaway:
If your hardest hit balls are to the oppo side, you’re definitely an oppo hitter.
That makes sense, since you aren’t juicing your exit velos by pulling the ball.
BOSA - Hard Hit & Sweet Spot (Min 50 total BBE, sample of HHSS could be lower)
The name that stands out to me the most is Mike Trout. He’s a moderate pull hitter for sweet spots and hard hits, but becomes an extreme pull hitter for his HHSS batted balls. I don’t know what to make of David Fletcher. It’s highly subjective, but Isaac Paredes being 28th on this lens makes me trust it less.
BOSA
It’s a lot harder to hit a sweet spot VLA to the extreme pull side. It’s a lot harder to hit the ball hard to the oppo side. Therfore, we’re going to set the following logic for classifying BOSA:
Pull Hitter: Average HSA <= -9.5 on sweet spot batted ball events. This is roughly where Aaron Judge is and felt like a decent cut-off point.
Oppo Hitter: Average HSA >= -3.5 on hard hit balls. This is roughly where Christian Yelich is.
That’s the rough yearly breakdown. I wouldn’t read too much into the trends.
Batters Optimized for Pull Will Have a Sweet Life?
The above chart would suggest that pull hitters will have higher sweet spot ratios than other hitters since they spend more of their trend line in the sweet spot zone. However, this is not the case, since the actual distribution isn’t linear across HSA:
The numbers aren’t definitive, but it would suggest you don’t want to be an oppo-optimized hitter.
Batters Optimized for Middle Get More Backspin
This is sweet spot BBE since 2019. Formula is DIST/Average DIST given VLA&EV. While the chart might look compelling, we’re talking about 0.5% or so more distance than expected, so a rather small difference in the aggregate (a 350 foot fly ball would be a 351-352 foot fly ball). This is perhaps an interesting result, but not very meaningful.
Batters Optimized for Oppo Are More Resilient
If we assume a linear trend (not necessarily a valid assumption), Oppo-optimized hitters lose only 0.05 mph per degree of HSA, compared to 0.10 mph/degree for their counterparts. This is a very interesting result, and could have something to do with swing planes, or something else. I would caution against drawing conclusions about causal factors from any of these charts.
Exit Velocity
Pull-optimized hitters are getting better at hitting the ball harder, but it’s not a clear trend.
BOSA and Launch Angle
Pull-optimized hitters tend to average much higher launch angles than their counterparts. I don’t really have a good explanation for why, other than a theory that power hitters are vertical launch angle hitters are pull optimized hitters.
Slugging on Contact
A pull-optimized approach is associated with higher slugging on contact. It’s not clear to me what’s the cause and what’s the effect. As noted above, it might just be that this is reflecting a batter archetype rather than a causal effect of a pull-optimized approach. Sampling bias alert!
Swing And Miss
Pull hitters may just be swinging harder, sacrificing contact for power. Or maybe it’s the approach. I don’t have a strong opinion which is the cause and which is the effect.
BOSA and Hard Hit%
Naturally, we see hard-hit percents higher in pull regions for pull-optimized hitters and in oppo regions for opp-optimized hitters. This is also a byproduct of how we defined them.
BOSA and Sweet Spot%
Nothing surprising here.
Slugging on Contact and HSA
We see that either extreme pull or extreme push can be very very good. This is even more so if we just look at sweet spot BBE:
So let’s define a couple of metrics:
XPF% = Extreme pulled fly ball percent = % of FB with a HSA < -40
PFB% = Pulled fly ball % = % FB with a HSA < -25
The 2 metrics naturally correlate. But does that make you a better hitter?
Doesn’t look that way to me. Does it help you hit HRs?
PFB% does correlate with HR% (R^2 = .125 compared to 0.056 for XPF%). If you boost your PFB% you might hit more home runs, but it may or may not help you become a better hitter. For those of you who play fantasty baseball, pulled fly balls are relevant.
Concluding Thoughts
Each batter is unique. When we look at horizontal spray angle at the macro level, there is no clear evidence that one approach is superior to another. Pulled Fly Balls will result in more Home Runs, but depending on what kind of batter you are, it may not result in more damage. Having a pull-optimized approach is associated with more damage on contact, but it also comes with more swing and miss. I’ll have more to say on this subject, but this post was long enough as is. Stay tuned (and subscribed)!