It seems to me that the timing of a more acute swing is really important. If a player has a 45-50 degree swing, and the ball makes contact with the bat before it gets around, wouldn’t it hit the dirt in front of the plate?
The flip side, if a 45-50 swing makes contact with the ball late, or out in front of the zone, wouldn’t it be popped up foul?
It seems like swing angle is a part of the piece to this puzzle, and my instincts tell me that it is probably only the extreme cases that have a significant influence on result.
That said, I love the idea of examining swing mechanics as the rawest skillset a batter has. It is equivalent to examining a pitcher’s pitch arsenal for velocity and movement. One other point on swing mechanics I’ve always wondered about: If F-m*a, wouldn’t bat acceleration be more important than bat speed?
I think it’s more about conservation of momentum during the (nearly) elastic collision between the bat and the ball, where momentum = m*v.
Comment by Jimmy Johnson — August 30, 2012 @ 11:59 am
I actually spent a little time looking at Votto — for that same humorous and digestive purpose. If I recall correctly, his swings varied considerable (as we might expect given his .408/.418/.422 GB/NU/FB split), but that his overall swing looked a bit uppercuttish, in the low 50° — which is a bit opposite of expectations.
A Reds fan would be better equipped, however, to speak on his swinging motion than me. I only saw 10 or so swings from him. And given how good his numbers are anyway, it’s very possible it’s just random variation that his GB/FB splits are 14 points off.
A tremendous bouquet of thank yous for this intriguing article.
Question: would you expect this swing angle analysis to tell a similar story in success-by-pitch-location? I.e. flatter swings performing better high in the zone, uppercuts better low in the zone. It seems like a given, but it would be neat to see which hitters have the greatest discrepancy in high vs. low performance, and whether or not this tracks to a similar swing analysis. Is this possible?
This is so great Bradley. The only thing I wonder about is if the placement of the ball determines so much of the swing angle the way you’re figuring it. I like your analysis of the follow-through. Maybe that’s the place to focus in order to determine an uppercut vs a level swing plane in a statistical fashion?
Yeah, there is definitely an interplay of location and even pitch selection when it comes to the swing angles. A low ball /uppercut hitter may just pass on the high stuff until he gets something low to swing at — thereby never showing us a more level angle.
At the some time though, there some guys you can watch — like David Murphy, IMHO — and see they take the same approach on every swing, dipping their back shoulder and swinging upwards even at high pitches.
In terms of statistically establishing a method of determining uppercuttiness, I definitely think the follow-through merits more attention.
I think — though I’ve not thought tested it just yet — that high-ball hitters and level-swingers ought to be the same person. But at the same time, I would not be surprised to see an absence of correlation.
For instance, Ben Zobrist hits fly-ballers better, but he also goes after low-zone pitches pretty well, too, dropping his hands down low and maintaining a level swing even in the bottom of the zone. When you have vision like Zobrist, you can probably match a level swing to a ball anywhere in the zone, so maybe vision is the component that determines high-low frequencies moreso than swing plane? I dunno, it’s work studying.
I think a 45-50° swing will have just as much need for timing as a 60-70° swing. A ball hit directly down has a decent shot of going foul, but a ball hit late or early from a level swing would go foul just as much.
Still, I think there is perhaps something to that line of thinking. Maybe uppercut hitters are more prone to bat speed problems and timing issues? Likewise, maybe vision plays a bigger role for level hitters?
Another avenue to take this down might even include pitch type success. A level swing out to excel against four-seemers, sliders and some changeups, while a vertical swing should excel against curves, sinkers and two-seamers, right? I don’t, but it’s worth exploring.
This is great stuff, Bradley, I really enjoyed it. Thanks. It looks like most guys have a larger PF on FB pitchers than GB pitchers, even though many of them have higher wOBA vs GB pitchers. I would assume this is because home runs and doubles are going to occur mostly on fly balls, so a fly ball pitcher is more likely to induce extra base hits even to hitters who have an uppercut swing that is better suited for GB pitchers. There are some exceptions (Murphy and Heyward, for example) but it looks like it might be a general pattern.
Great analysis. I would guess that two other influencing factors would be where in their stance contact (out in front or back) is made and the tendancies of hitters to hit a certain hemisphere (top or bottom) of the ball therby creating the ball carrying backspin or grounder topspin. I suspect that the latter data would be difficult to determine or would be a statistical wash.
So this is an intriguing idea, but unfortunately you’re not measuring an uppercut correctly. Pitch height determines barrel position relative to a hitter’s hands, which in turn completely defines the angle you are using to determine “uppercut” vs. a “flat swing”. I hate to say it, but if you watch a video of Jason Heyward from a side view you’ll see that he has one of the flatter swings in baseball. His relatively low finish is another indicator of this. Finish isn’t the best way to define swing plane either, though, because some guys get to a really high finish by just being really steep (think of a swing path that looks more like a V than a flat U).
The first place to start in defining an uppercut would be looking at the angle between a HORIZONTAL line drawn at the lowest point in a swing and the barrel path after this point. This would still say that guys with steep swing paths have uppercuts, so in my opinion you would want to look at a second factor. Using some definition of the hitting zone (the horizontal area from a side view of the hitter in which a hitter can reasonably make contact within his swing path) measure the percent of that zone during which the barrel is moving up versus down or perfectly horizontal.
Guys with flat swings by definition stay in the hitting zone longer, which is a good thing because it gives them more margin for error when their timing is off. You would also probably find that these guys spend a larger portion of their swing path with the barrel moving up, but it’s at a very slight angle. Therefore, they would look like a hitter with an uppercut on the second measure, but not at all on the first.
Guys with uppercuts OR steeper swings generate more power, generally speaking, because their swing planes generate more loft. Both of these types of hitters will have more extreme angles on the first measure in order to finish high, but hitters that just have steep swings spend a good portion of the swing path moving down into the hitting zone before quickly going up to generate the high finish. Therefore, my guess is that they will have a large angle like a hitter with an uppercut on the first measure, but they will not grade out as having an uppercut on the second measure.
Sorry for such a long post on hitting mechanics, but I think it’s an interesting idea if you define swing plane correctly. Obviously finding side views of hitters hitting balls at similar heights and then measuring all of this may not be doable, but right now saying that Jason Heyward has an Adam Dunn-like uppercut and this is why they both hit GB pitchers much better than they do FB pitchers likely won’t hold too much weight in baseball circles. Here’s a couple of youtube videos on each guy to make the point:
Dunn defines an uppercut and peripherally we see this in huge FB, HR, and K rates because he generates loft, but doesn’t keep his barrel in the zone long enough to make contact when he’s fooled. On the other end of the spectrum Martin Prado has opposite batted ball and contact numbers and opposite swing mechanics, but also finds himself near the top of the list for GB – FB pitcher performance difference.
Definitely worth looking into further and I like the idea with a better measure of swing plane.
Comment by Graham Tyler — August 30, 2012 @ 5:23 pm
Great stuff Woodrum, but this going to take a while to digest. As an aside, I’ve heard it rumored that this kind of analysis went into Beane’s choice process in his A’s acquisitions that led to this 2012 team. I wonder if your data/insight is at all revealing with regard to that.
Comment by channelclemente — August 30, 2012 @ 7:29 pm
I knew going into the project that I would be redefining swing planes — or at least measuring it different than the traditional hitting coach’;s / scout’s method. I think you and I might have to disagree on Heyward. I do think he is capable of level swings — and he’s a good enough hitter to balance out his GB/FB split over time (I would not at all be surprised to see him perfectly neutral in a few years).
But his follow-through and his typical swing angle (both in his shoulders and his hand-to-barrel as I measure it above) to me look like an uppercut. After reading your comment, I went and looked at a few more of his swings at random, and I still think he’s a bit — if not a lot — uppercuttish according to my definition.
I do agree that side camera angles could vastly improve the discourse here, but unfortunately to my knowledge, there is not a rich, accessible database for those ends. We can find a swing or two here or there, but I am already unimpressed with my own sample size of visual data, so I’m uncomfortable of looking at one swing and making significant pronouncements — as is a bit too common in video analysis, IMHO.
Very interesting article. Just out of curiosity, did you happen to look at Alfonso Soriano’s swing? Seeing as how he has tended to have more power on balls low in the zone, but has a split that tends toward neutral, then flyball, and then groundball pitching… Could it be that he only functions against mistake pitches by lefties? I say this semi-sarcastically as a Cubs fan who doesn’t hate Soriano as much as most folks.
Comment by Dr_Caligari — August 30, 2012 @ 10:05 pm
Fascinating stuff. Thanks for the effort you put into it!
I definitely agree that right now there’s no way to get a large enough sample of side angles to be useful, which is unfortunate because I think it would really help.
In talking about Heyward I would just make a couple of points. The first and main one is that pitch location drives a TON of any sort of swing plane analysis you’re going to do, but especially barrel-hand angle at contact. Regardless of where a guy starts from and ends, his barrel path and position at contact on a ball down and in will be drastically different than on a pitch up and away. To the extent that two guys have different barrel-hand angles at contact on the same pitch it probably has a lot more to do with posture than with swing plane.
The second point is more about semantics than anything else, but typically when you think of an uppercut a guy necessarily has to start his hands low and then finish high. A guy who starts high and finishes high on a low pitch wouldn’t uppercut. That being said, there may be certain points in the swing where the barrel path for a guy who starts higher and finishes high looks the same as a guy who starts low, uppercuts, and finishes high. So while uppercut may not be the right word, this may be the quality you’re looking for.
The last thing I’ll say is that I think if you wanted to keep going with this there are definitely other factors that might be possible to analyze that could be pretty interesting. In considering the effect of the pitcher I’d be interested to know whether certain traits common to most GB pitchers are driving these splits for hitters. In other words, is it something about the movement, velo, pitch mix, or location common to most GB pitchers that these hitters handle especially well? The first place to start would be to figure out among the pitchers if and what the common characteristics are (I’m assuming less FB velo, more offspeed, obviously more vertical movement, and more pitches down in the zone). It may not necessarily be one hitting trait that drives the splits, but instead one (or more) pitching one. For example, Heyward has long arms, gets pretty low into a wide base at contact, and generally has his head and torso out over the plate relatively far at contact, which all may allow him to dig out low pitches pretty well for a tall guy. So maybe the split has more to do with the general location of what he sees than with one particular swing trait.
On the hitting side maybe simplifying it to look at guys with low hands, high finishes, something you can clearly see from a front angle may allow you to get deeper into it. Again, just ideas if you wanted to go further with it because I appreciate all the work you put in and I think it’s really interesting stuff.
Comment by Graham Tyler — August 31, 2012 @ 12:48 pm
Great post and great comments. This is the kind of thing I come to FanGraphs for.
I’d give my testicles to see how the Rays’ proprietary lineup analysis works.
Zobrist may drop his hands a bit more than some, which makes his swing slightly more level than average on balls down in the strike zone, but it’s far from level in absolute terms.
Also, Zobrist hinges his back knee more than some, and similar to Pedroia and Beltre, which lets him hit lower pitches higher relative to his body than some. However, that says more about what his back knee is doing, and how he adjusts to hit pitches down in the strike zone, than it does about his swing plane.
How about Alfonso Soriano for an uppercut swing? He always seems to be golfing out low balls. Albert Pujols seems to murder high fastballs with a nice level swing and also seems to ground into a lot of double plays.
The question of ‘level, or ‘not level’ is most often interpreted in the plane of the balls arrival, not the surface of the earth.
Comment by channelclemente — August 31, 2012 @ 10:03 pm
There seems to be at least three determinants of success against a GB or FB pitcher. Swing plane as viewed from the side, High/Low ball location preference which would be viewed from the pitcher and seen through bat angle. All hitters will adjust but some are more comfortable with a steeper bat angle (ie. lower ball preference). The third and I believe this may prove out to have the highest correlation (but have only looked at it briefly) is ball contact preference – ie. top, center, bottom as can be seen through GB/FB ratio.
There would seem to be some independence in the variables. For example, Joe Mauer has a slight uppercut (flat to the plane of a pitch, say 6-10 degrees). This is seen via bat through the zone from the side as well as posture established at foot plant (ie how much lean towards the catcher) until after contact. Ball contact preference is likely slightly north of center giving him a GB/FB of 2.5 but a high LD rate, low IFB% and Swinging Strike rate (last two indicative of matching the plane of the pitch.
I don’t doubt that swing plane is a factor but I’m guessing that other factors would prove to be meaningful as well.
Comment by Swingdoc — September 1, 2012 @ 10:59 am
Any chance for an updated spreadsheet?
Comment by Daniel Steinberg — May 7, 2015 @ 4:41 pm