I proceed to seek out Statcast’s bat monitoring information fascinating. I additionally proceed to seek out it overwhelming. Hitting is so advanced that I can’t fairly think about boiling it right down to only a few numbers. Even after I have a look at a few of the extra advanced displays of bat monitoring, like squared-up charge, I typically can’t fairly perceive what it means.
I’ll offer you an instance: after I seemed into Manny Machado’s early-season struggles final week, I discovered that he was squaring the ball up extra often when he hit grounders than when he put the ball within the air. That sounds unhealthy to me – arduous grounders don’t actually pay the payments. However I didn’t have a lot to match it to, except for league averages for these charges. And I didn’t have context for what shapes of squared-up charge work for varied completely different profitable batters.
What’s an analyst to do? Should you’re like me in 2024, there’s one most well-liked choice: ask my pleasant neighborhood massive language mannequin to assist me create a visible. I had an concept of what I needed to do. Primarily, I needed to create a chart that confirmed how a given hitter’s squared-up charge diverse by launch angle. There’s a distinction between squaring the ball up like Luis Arraez – line drives into the hole all day – and doing it like Machado. I hoped {that a} visible illustration would make that somewhat clearer.
First issues first: I downloaded each ball in play from this yr the place Statcast recorded a bat pace, pitch pace, launch angle, and exit velocity. Then I manually calculated whether or not every batted ball was squared up. As a refresher, a batted ball is squared up if the ball travels at 80% of its most theoretical velocity, as measured by a proxy system: 1.23 * bat pace + 0.23 * pitch pace at dwelling plate, which is roughly 92% of pitch pace at launch. Should you’re thinking about following together with me at dwelling, you’ll find that information right here. If not, bear with me, as a result of I’d like to indicate you some photos I made.
From there, Gemini (my LLM of alternative, although I’m certain others would find yourself in roughly the identical place) and I set to work. We calculated the squared-up charge of every hitter at every angle. I needed to make a number of selections right here about mixture information. I made a decision to bulk up each angle by in search of balls hit inside 10 levels of it both means, then threw out each bucket that didn’t have a minimum of 20 information factors after doing that bulking up. There’s some overlapping information this fashion, however pattern sizes are sufficiently small, and I believe that hitter intent is broad sufficient, that for those who’re questioning how often somebody squares up a batted ball at 15 levels, 5 levels and 25 levels are each helpful inputs.
These are the elements that I got here up with, however I wasn’t fairly certain flip that idea right into a program that might make graphs out of my concept. However that’s nothing I couldn’t remedy after a number of hours of developing with concepts, translating them into Python code utilizing generative AI, discovering issues with the code, developing with new concepts to resolve these issues, translating these new concepts into new code, discovering new issues… you get the thought.
Primarily, I needed a graph of how good Machado is at squaring up the ball relying on whether or not he’s hitting it down, flat, or up. Nice information. I acquired precisely that graph:
He’s squaring up a ton of his contact on the bottom, identical to we knew. He’s getting essentially the most out of his bat pace far much less often on the juicy launch angles within the 20 diploma vary. That doesn’t sound very similar to Arraez, the bat management god, in any respect. However what does Arraez’s graph seem like? It seems like what you’d count on:
As a aspect observe, the dimensions of the circles is proportional to the share of contact in that bucket. Arraez’s greatest circles are line drives of varied varieties. He hardly has any excessive grounders or excessive popups. That’s what superb bat management seems like.
How does that examine to Machado? After a spherical or two of dancing with Gemini, the software I constructed might help with that too:
You can just about guess this even earlier than this graph, however it’s nonetheless good to see it in photos. Machado is squaring up grounders on the similar charge, however his swing simply isn’t getting it performed within the air proper now. We will throw in a 3rd hitter to indicate what it seems like if you’re the alternative of Machado. Right here’s Bryce Harper, whose uppercut swing is etched into pitchers’ nightmares in all places:
Harper is the brand new sequence, in Philly purple. When he hits the ball on the bottom, he’s not often squaring it up. In different phrases, these are largely mishits; when he’s squaring the ball up, it’s typically within the air. He constantly beats Machado at squared-up charge within the air, and he hits extra fly balls as properly. He won’t sq. the ball up as often as Arraez, however he swings a lot more durable and connects usually sufficient. Maybe unsurprisingly, he’s mashing to date this yr.
For one more enjoyable comparability, let’s have a look at Aaron Decide and Juan Soto:
They’re each making pristine contact throughout the board. They’re at or above an 80% squared-up charge for just about all the pieces within the air, they usually’re each swinging arduous too. That’s a lethal mixture. Decide is even avoiding grounders; he doesn’t actually have a left tail to talk of. His greatest cluster of launch angles is essentially the most harmful one in baseball if you’re hitting the ball arduous. In different phrases, he’s swinging arduous, squaring the ball up often, and doing it on dwelling run trajectories. No marvel he’s slugging .703.
These two elite hitters are getting it performed in nearly the identical means. However it’s not the identical for everybody. The Dodgers’ three stars present some variation:
Mookie Betts has become an excessive fly ball hitter. Right here’s the graphical proof of why that’s working: He’s contacting the ball most squarely at round 30 levels of raise, and he’s hitting the ball within the air extremely often. To the extent that he has mishits, he’s getting too far below the ball and popping it up, which is smart given his general strategy. Shohei Ohtani, too, is squaring the ball up most often within the air. He isn’t hitting a ton of grounders, although greater than Betts. He’s additionally completely rifling low line drives — have a look at all these excessive blue circles within the 10-20 diploma band.
Then there’s Freddie Freeman. He hits all the pieces sq. at about the identical charge. His most frequent launch angles are principally all the pieces from 10-40 levels. There’s nearly no variation in his line; each Betts and Ohtani have greater highs and decrease lows. Freeman’s swing appears to be a chameleon; it simply adjustments to suit the contact sort. In numerous methods, he’s a burlier however much less exact Luis Arraez:
They each simply rake, plain and easy. Arraez hits it flush extra often, in fact, however Freeman swings 7 mph sooner. Arraez focuses extra on the 5-15 diploma band; Freeman faucets into his energy by hitting extra balls within the 25-35 diploma vary. However they’re each completely peppering all the pieces, whether or not within the air or on the bottom, they usually each hit a ton of line drives. These guys are unbelievable.
We will do extra. Need to see some younger American League shortstop dynamos? Check out Bobby Witt Jr. and Gunnar Henderson:
Witt has a promising contact form, however not an ideal one. It’s like Freeman’s, solely shifted down a bit and with extra grounders. There are some purple flags, like his comparatively low squared-up charge when he’s placing the ball within the air. To be trustworthy with you, although, I’m undecided how vital squared-up charge is in these small and cut-up samples. I’m extra thinking about form for now, and I’ll have time to do extra testing of how a lot the degrees matter later. The important thing half, for me, is that Witt’s most frequent outcomes are fly balls and line drives, however his most frequent square-ups happen on grounders. Make that correction, and much more upside may very well be out there.
Henderson, then again, looks like he was designed in a lab. He squares the ball up most often on the launch angles the place arduous contact is most advantageous. He doesn’t have sufficient popups to get any dots up there. His grounders are all mishits. Positive, possibly he may focus much more batted balls round his finest swings, however he’s doing precisely what I would like each hitter to do: hitting the ball flush when he elevates, and doing so with plus bat pace.
Right here’s a thriller that this information can remedy: Why does Henderson have 20 homers to Witt’s 11? Witt hits the ball more durable, hits fewer grounders, and even has a better barrel charge. However Henderson’s swing is designed to sq. the ball up within the air extra often, so he’s lined up high-value launch angles and high-value exit velocities higher than his Kansas Metropolis counterpart.
I believe this information will get much more attention-grabbing when we have now entry to a number of years of historical past. I’d like to know if Kyle Tucker’s swing form has modified together with his decrease groundball charge and otherworldly manufacturing. I’d be thinking about seeing how hitters who change their batted ball tendencies change their squared-up tendencies. I wish to see whether or not Nick Castellanos has at all times squared up the ball precisely like Bryce Harper, or whether or not he used to have a distinct form and the brand new one is correlated along with his downfall:
I haven’t fairly discovered what to do with all of this in the long term. I believe it’s extra of a storytelling software than one thing that may inform you who can be nice and who will battle. That stated, I like the tales! Arraez is nice within the methods you’d count on. Betts maxes out on energy along with his swing. Harper’s uppercut is cool to see in information. And the way about that uppercut in opposition to Yandy Díaz’s ground-friendly methods:
Yandy is smashing these grounders. You and I already knew that, however it’s cool for this information to confirm the attention check. That’s principally what that is to me; a means of changing some dry information factors right into a story.
The software I constructed isn’t stay on the pages of FanGraphs for a lot of causes. It’s hilariously rudimentary. It’s buggy. It’s programmed by me, a coding imbecile, slightly than by our group of fantastic builders. It won’t even be helpful in the long term.
So no, you’ll be able to’t simply click on on a single hyperlink and mess around with this to your coronary heart’s content material. However I’ve two issues to supply that can hopefully make it as much as you. First, that is an open supply undertaking. You will discover the Python script that generates these graphs right here, together with the underlying information. I’m definitely not assured that that is essentially the most environment friendly technique to do issues – I used to be constructing from scratch with out numerous expertise on this space. In case you have some enhancements or whatnot, let me know!
Second, I occur to have the code and the flexibility to submit photos to the web. So for those who’re thinking about a selected comparability, ask me beneath within the feedback. I’ll get to as many as I can for the following day or so, as a result of I perceive that “hey, simply discover ways to use this laptop programming language actual quick” isn’t precisely a technique to assure broad entry.
So, yeah. That’s the top! No actual conclusion right this moment, except for a) I believe this software is cool and b) listed here are some photos of it. I hope you prefer it, and I hope there can be extra bells and whistles earlier than lengthy.
