Pitch Location and Swing Angles: Dunn and Bruce

Last Thursday, I took a stab at predicting how batter’s swing influences their ground ball / fly ball splits. One of the most important retorts to the research (a retort made both in the comments and on The Book blog) was that pitch location was the determining factor of bat angles — what I was attributing to hitter tendencies (at least for hitters who have big GB/FB platoon splits).

Consider today’s offering a second puzzle piece — hopefully an edge piece — in what is a 1000-piece puzzle of understanding GB/FB splits. Today I offer the case study of two (essentially randomly picked) hitters with large GB/FB platoon — Adam Dunn and Jay Bruce.

The results surprised me, twice, and in the end, it appears these two hitters employ different swing patterns, suggesting there may be traction with my original theory, even though pitch location does have a considerable affect on swing angles.

Why Bruce and Dunn? Well, if we refer to the dataset Jeff Zimmerman helped me put together in the first article — and limit the players to those with 300 PA — we find there are two obvious candidates for this study: Adam Dunn and Derek Jeter. But Jeter is a no go.

Dunn hits ground ballers well — even if we cross-reference my numbers with Baseball-Reference GB/FB splits — and so we should expect an uppercut or at least uppercuttish swing from him. He has a .400 wOBA against ground ballers over the last three years, a .315 wOBA against neutral pitchers and a .327 wOBA against fly ball pitchers.

Jeter has a .302 wOBA, .330 wOBA and .353 wOBA against GB, neutral and FB pitchers — a perfect balance to Dunn!

But here’s the deal; this is a plot of Jeter’s home runs this year:

Yeah, Jeter’s results are almost definitely going to be rich with large, level angles. But that is because Jeter is clearly a high-ball hitter, not necessarily a hitter stuck using one swing angle / swing plane*, like my theory suggest some players are. So we have to enter this bit of information into our calculus, and keep moving because he is essentially useless for today’s purposes.

* I used the term “swing planes” in the original article, but since I was essentially redefining the means of measuring a swing plane, I think it is useful for clarity purposes to call it “swing angles” — so please, dear friends, do not worry about telling me how to measure swing planes; I know how to do it; I’m not doing it the old way on purpose.

That is not to say we throw out the Jeter data. No, we enter it into our mind computers and write down in pencil: Some guys may succeed against fly ball pitchers because they are high-ball hitters. This is neither a surprising suggestion or even a suggesting that damns my initial theory of swing angles.

But it does mean we need a different hitter. So we move on to Jay Bruce of the Reds. Bruce has a .338/.337/.384 wOBA GB/NU/FB distribution, making him another strong fly ball hitter with a lot of evidence to suggest its a real, legit skill for him. And, more importantly, he has not only home runs all ovah the place, but he has home runs aplenty.

So let us start with Bruce. Here are his home runs (as of Tuesday):

And here are the angles I recorded for his highest and lowest homers (6 high, 7 low):

Homer Height (pz) Bat Angle HR Date
1 3.81 62° 4/26/2012
2 3.30 67° 4/28/2012
3 3.22 64° 9/2/2012
4 3.13 59° 8/15/2012
5 2.88 55° 8/12/2012
6 2.85 65° 5/4/2012
7 2.81 58° 9/3/2012
8 2.15 45° 5/27/2012
9 2.00 52° 5/7/2012
10 1.97 50° 6/16/2012
11 1.94 50° 4/8/2012
12 1.84 53° 9/4/2012
13 1.63 55° 4/8/2012

Ideally, I would have all 32 of Bruce’s homers in that chart — heck, all 132 homers of his career — but like I said, this is just another piece of the puzzle. I heartily encourage anyone interested in filling in data where my spreadsheets are left blank.

But for today, this 13 homers should be good enough.

One of the key components of understanding my swing angle theory is that we need to look at the full, typical swing, not just the bat angle at the point of contact. As I noted with A.J. Pierzynski in the first article, some bat angles may be low, but the follow-through can still be unusually level — or vice versa, a level swing can finish with an uppercut flourish.

So it is important to actually see the swings of these players in action. Here is Jay Bruce crushing a high pitch:

And here he is lasering at a low pitch:

Honestly, Bruce is not a guy I would pick out of a lineup as having a level follow-through. I would not call it an uppercut swing, but he does seem to let the bat travel upwards more than a few of the other previously examined players — such as the aforementioned Pierzynski.

So let us compare him to Dunn. This is Adam swinging at a high pitch:

And this is Dunn on a low pitch:

Dunn’s swing is closer to what I might define as a prototypical uppercut swing. That high pitch swing looks much different than his low-pitch swing, but when I think of Adam Dunn, I do not typically think of a level bat.

Side note: After watching over a dozen of Dunn’s homers many, many times, I found myself surprised at how much he appears to not get his legs involved at all in the swing. It’s like his whole swing is coming from the upper body — but at the same time, I have never had as much difficulty taking pictures of ball-on-bat moments as I did with Dunn. Either the CSN cameras are garbage, or Dunn’s bat speed makes his bat go invisible in the zone. I’m leaning towards monster bat speed.

So here are Dunn’s homers this season:

And here are the swing angles I record for 13 of his jacks (6 high, 6 high and 1 in the middle):

Homer Height (pz) Bat Angle HR Date
1 3.9 68° 6/15/2012
2 3.3 64° 4/29/2012
3 3.2 59° 8/18/2012
4 3.1 62° 7/23/2012
5 3.1 67° 6/10/2012
6 3.1 60° 6/9/2012
7 2.7 55° 6/5/2012
8 2.2 48° 5/19/2012
9 2.2 48° 5/3/2012
10 2.2 46° 8/24/2012
11 2.1 45° 8/15/2012
12 1.9 45° 5/2/2012
13 1.8 57° 7/24/2012

First of all, in a stroke of magic, both Dunn and Bruce homered on a pitch nearly 4 feet off the ground. Dunn put a 68° swing on a pitch 3.9 feet high, and Bruce put a clobbering 62° hack on a pitch 3.8 feet above the plate. So far, the idea that Dunn is a primary uppercut swinger does not look good.

But if we compare the home runs in the lower portion of the zone — and Bruce actually has a homer beneath the 1.75-foot line, which is the typical bottom of the zone (though he is 3 inches short than the Sears Tower Adam Dunn) — we can see Bruce only dipped beneath 50° one time, and that was on a pitch down and in. Meanwhile, Dunn went above 50° on only one low pitch, and that was his lowest homer, which was a breaking ball away. The remainder of Dunn’s swings were at or beneath 48°.

It appears that — given these few swings — Bruce has less variation in his swing angle. He sticks around the mid-50° to the mid-60°, while Dunn varies dramatically, going as low as 45° (which is the lowest I’ve ever seen) to 68° (which is almost as high as I’ve yet seen).

If we put these two datasets together, we can see the pattern pretty clearly:

In other words, Dunn has shown vertical location affects his swing angle significantly, while Bruce puts just about the same kind of swing on each pitch. Think back to the GIFs above — both of Bruce’s high and low swings look pretty similar, while Dunn’s high and low swings look considerably different from each other.

Dunn’s ability to change his swing angle is something I would have expected from a hitter with small-split hitting talents (a guy who hits GB and FB pitchers equally well). But the book on this is still open. (We will discuss that in a moment.)

I found the low R-squared in Bruce’s data interesting. It appears horizontal and vertical location combined had the most important effect on his swing angle — which varied little by comparison — whereas horizontal location was by far the chief determinate in Dunn’s swing angle.

I regressed Bruce’s swing angle against both horizontal and vertical location, and got this equation: 38.75 + -7.80(px) + 7.15(pz) with a great P-values and a 0.844 R-squared. That’s for you nerds out there. For the non-nerd, that means that Bruce’s subtle changes in angle have more to do with vertical location than horizontal location, but altogether, the combo of the two is by far the best determinate for his swing angle.

One of the important questions we must ask here is: Why home runs? Why did I look at home run swings instead of just a bunch of general swings? Or maybe swinging strikes?

The answer is: I probably will look at those things. In fact, I think by leaving swinging strikes out, we are making Adam Dunn look more competent against high-ball pitchers than he really is. Dunn has a 1.05 power factor (PF) against ground ball pitchers, but a 1.31 PF against fly ball pitchers — despite having a way worse wOBA against fly ball pitchers.

That means when Dunn makes solid contact against a fly ball pitcher — quite possibly in the upper reaches of the zone — the ball does not usually land. Ever. But there is a good chance that he is swinging at a 45° to 50° angle against high-zone pitches, but because we are not looking at swinging strikes or popups, we see only his successful, 65° swings.

Again, I encourage the readers to speak their minds on where the data, research and analysis is lacking. I am beyond pleased with the dialogue from the first article, and would not have been able to progress this theory further without keen, critical perspectives from strange, unknown Internet People like you.

For those too lazy or too brilliant to read the first article, I measured the swing angles from a point originating at the hands / butt of the bat, extending to the barrel, apposed to an imaginary vertical line perpendicular to the ground. Because I love my work being double checked (discount or nay), I have included all 26 home run moments for these two players. So if there is any doubt about the angles I recorded, please use these screen captures as a reference. They are named according to their date, which you will find in the corresponding tables above.

Print This Post

Bradley writes for FanGraphs and The Hardball Times. Follow him on Twitter @BradleyWoodrum.

17 Responses to “Pitch Location and Swing Angles: Dunn and Bruce”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. baty says:

    it’s hard to tell without a side view of these swings, but between Bruce and Dunn, it looks like Bruce might have a more vertical posture through the point of contact. Even with the lower pitch swings, Bruce seems to keep his upper body over the ball, while Dunn’s upper body arcs back into his stance earlier so his chest rotates towards the sky as he makes contact.

    Wouldn’t it be awesome to see a selection of swing images like this per player profile page in a batted ball section? :)

    Vote -1 Vote +1

  2. GlennBraggsSwingAndMissBrokenBat says:

    This is neat! Thanks!

    Vote -1 Vote +1

  3. Captismo says:

    The legs aren’t expected to very much of the work anyway (in regards to Dunn). The power mostly comes from hip/torso separation just like pitching. But still, good work.

    Vote -1 Vote +1

    • Thanks!

      And yeah, Dunn’s definitely generating oomph from his midsection. I think maybe after watching him swing a hundred times and then watching, say, Adrian Beltre, it kind of feels like Dunn is using cheat codes.

      Vote -1 Vote +1

  4. Phantom Stranger says:

    Really interesting stuff, I’ve heard rumors that certain MLB teams (the Yankees being one of them) have top-secret data about the optimal swing planes to produce home runs or improve line-drive rates for each player. Supposedly some teams have taken their players to motion tracking facilities and attached sensors to the bat, trying to improve bat speed and the angle for maximum loft.

    Makes one wonder if the hitting coaches will eventually be replaced by software…

    Vote -1 Vote +1

  5. Baltar says:

    The definition of a swing angle you give above, “from a point originating at the hands / butt of the bat, extending to the barrel, apposed to an imaginary vertical line perpendicular to the ground,” is a definition of only one side of the angle. What is the other side?
    I still don’t understand what an angle you can measure from the front, no matter how you define it, can determine whether the swing was an uppercut or a level swing. That would be a measurement taken from the side, showing whether the swing slanted upward or remained level from beginning to end.

    Vote -1 Vote +1

    • DD says:

      basically, drawn a line from the hands as the swing crosses the plate straight down. The distance between the barrel and this line is the angle measured.

      Vote -1 Vote +1

  6. FredR says:

    Interesting follow-up, o gracious writer of baseball.

    I took a brief look at the data from your first article on swing angles, took the top 10 hitters with the largest wOBA split between GB and FB pitchers (at both the (+) and (-) ends of the spectrum), dug up some of their more basic stats, and discovered the following unsurprising things:

    The top 10 highest (wOBA GB pitchers – wOBA FB pitchers) produce the following average line:

    0.304 1.118 22.66% 40.11% 37.22% 9.17% 13.20%

    While the average line of the 10 players with the most negative (wOBA GB pitchers – wOBA FB pitchers) is:

    0.293 1.56 20.42% 47.74% 31.84% 9.63% 13.73%

    BABIP, LD%, IFFB% and HR/FB look pretty equivalent, but the 10 hitters that produce the best against GB pitchers just appear to hit fewer groundballs on the whole (exchanging them for fly balls). This is perhaps more than obvious. As groundballs are often outs and some fly balls are homeruns, the effect on wOBA for these players makes plenty of sense. I’m not sure how this trend will work out on a case by case basis but it is possible that you might be able to correlate your swing angle theory wtih GB% and FB% in general, rather than making the extra effort to find players that are successful vs. GB or FB pitchers. It could increase your pool of players/data considerably!

    Vote -1 Vote +1

  7. channelclemente says:

    Mr. Woodrum, it would be interesting to take a look at Josh Reddick from your perspective. I have heard rumors the A’s have explored the sort of model you are suggesting.
    Very interesting stuff.

    Vote -1 Vote +1

  8. ugglas arms says:

    Cpl things I notice…not sure how they factor in. It looks like dunn keeps a more consistent angle between his top and bottom portions of his body suggesting he makes the angular swing change with his wrists and hands. While bruce uses his arms and torso for the change…
    Secondly, it looks like dunn maintains steadier angles between his front arm and torso. This allows him to create more torque the closer his arms are to perpendicular to his torso and maybe this is why he appears to use less legs.
    I consider dunn to hit longer higher homers off the top of my head and if you were to draw planes that follwed his arms through the swing and one that follows his bat, they would be closer to parallel than bruces
    You would maximize torque and angular momentum if the arm and bat planes were parallel to each other and perpendicular to the torso throughout the swing

    Vote -1 Vote +1

  9. Kenny says:

    Um, might it be that Jeter is a high ball hitter as you say because he has a level swing rather than an uppercut swing? Basically, I don’t understand your argument for not looking at Jeter further.

    Vote -1 Vote +1

    • Kenny says:

      Nevermind, I read the rest of the article and I get why you left Jeter out – no homers on low pitches…

      Vote -1 Vote +1

      • Kenny says:

        Sorry about the Jeter obsession, but how about some gifs of Jeter swinging at a low pitch? What angle does he use to swing at low pitches? I would imagine it’s still pretty flat, so he hits a lot of grounders and line drives on low pitches, but has trouble hitting fly balls, and therefore home runs.

        Also, an interesting difference between Bruce and Dunn’s swings are the follow throughs. On both high and low pitches, Bruce ends with his hands at shoulder height. Dunn always ends with his hands above shoulder level. Maybe looking where a batters hands end is a good indication of having an uppercut swing or not.

        Vote -1 Vote +1

  10. Mike Green says:

    Both of the low pitch homer gifs were on pitches down and in. Productive bat-angles likely depend on not only on the vertical height of the pitch, but on the in/out location. You will notice that Bruce’s homers on the pitches up and in (4 and 5) have much lower batting angles than those on pitches up and away (2, 6 and 7), and that Dunn has hit no homers on pitches up and in (with the closest #7 having a low batting angle). The comparison between Bruce’s homers 4, 5 and 6 is especially interesting.

    It’s a complicated topic. Probably the place to start would be to divide the hitting area into a 3 X 3 grid- low and away, away, up and away, low, heart of the plate, up, low and in, in and up and in, and then compile league average bat-angles on pitches in each location. It might be that you would have to go 4 X 4 to capture nuances. You could then study individual players.

    Vote -1 Vote +1

  11. Tom the bomb says:

    Balter is on the right path. To determine whether the swing is an uppercut, level or down swing would be determined by the angle of the bat when looking at the batter from the opposite side of the plate and perpendicular to the general path of the ball trajectory. The angle you discuss is merely the batter locating the sweet spot of the bat to the location of the ball. Your angle is a function of ball location /flight and whether the batter made an on the fly correction to align the sweet spot to ball location at impact. In general your angle will be smaller on low inside pitches and larger for pitches high and away,as defined by you above. While the angle you try to analyze is critical to successful hitting as the bat must find the ball, the angle of the bat perpendicular to the ball flight when facing the batter on the opposite side of the plate is the angle the batting coach needs to teach the batters to copy.

    Vote -1 Vote +1

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>