Breaking Down the Swing: Best Hitters of 2012

Often players will be credited for being very efficient with their swings, or evaluators and coaches will praise a hitter for having tremendous bat speed.  Those who work with hitters and study the art of hitting on a regular basis know that it takes a lot more than being a good athlete or having fast hands to be a successful hitter.  I myself work with many amateur hitters at Carmen Fusco’s Pro Baseball & Softball Academy in New Cumberland, PA.  We use video analysis as an integral part of the learning process, and I spend many hours outside of work devoted to breaking down MLB, MiLB, and draft-eligible players’ swings and pitching deliveries.  In this study I have conducted, I wanted to collect data regarding the best Major League hitters’ swings to discern what actually matters and is worth commenting on from a mechanical perspective of a hitter.

Going into this project, I wanted it to be primarily a data-driven approach to what players do in the batter’s box.  This is a study of hitters’ mechanics at the Major League level, hopefully useful in producing predictive or at least somewhat comparative parameters to be applied to unproven professional or amateur players.  Many criticisms and compliments get heaped on hitters for how their swings work and the correlation to big league success.  However, I have not seen many of these thoughts backed up with hard evidence as proof or even fact-based suggestions that they are truly instrumental to a player’s results on the field.  I will mix in many of my own thoughts here and there as well, but this is meant to be used as an objective analysis of hitters’ mechanical processes.

I compiled a list of the top 50 hitters from the 2012 season according to Fangraphs’ Batting component of WAR.  I then looked at side views of each of these hitters from highlights of the 2012 season in which each player hit a homerun.  In the case of switch hitters, I used the side of the plate where they were most successful.  In all but Melky Cabrera’s 2012 stats, that described their left-handed swings.  I then took a series of over 60 discrete, objective measurements of each player’s swing using Don Slaught’s Right View Pro video analysis software to ascertain ranges of values.  I also wanted to see if any of these moves had even a slight correlation to the kind of hitters they were last season.  As an extension of this exercise (in a future study), I would like to apply this data to amateur players to see if there is any kind of predictive quality to be used for evaluation purposes.  Since the majority of scouting videos consist almost entirely of side views of players’ swings, finding out what information can be gleaned from videos in this manner could help further the understanding of player evaluation.

There are some obvious biases with this approach, sample size being among them.  Just because the top 50 hitters from last year execute their swings in a certain way does not mean it is the only way to have a successful swing.  This study also does not describe a hitter’s ability to consistently get into these positions or sequence in the same way every swing, or if that is even a desirable quality.  A hitter may also be able to hit one pitch location very well but struggle to cover a large portion of the strike zone.  These specific measurements of one swing from each batter’s 2012 season does not describe how much or how little a hitter can adjust to different pitches and zones.  However, I believe the fundamental components of hitters’ swings stay the same even over multiple seasons, something that can also be explored in another study.  To me there is enough data here to draw some meaningful conclusions, or at least to get the conversation started in some more complex portions of the swing.

As for my own thoughts, I do not believe in models when it comes to something as complex as a baseball swing or throw.  From my experience and the thoughts of those much smarter than I, the differences in body types and biomechanical processes are so great that no idiosyncrasies can be written off as wrong or detrimental to a hitter.  I also think that the ability to sync up to a pitcher’s delivery, a batter’s visual skills, pattern recognition, and the way the body sequences itself are much more important than a batter being in a certain position at any given part of the swing.  Most important is a player’s mental makeup, as this more than anything determines whether a hitter reaches his ultimate potential.  These attributes are largely outside the scope of this endeavor.  This project is an attempt to see how scientific side angle video analysis of hitters can be, whether that is by specific angles and sequences or more general ranges of movements.  I tried to focus as much as possible on objective measurements of body sequencing whenever possible instead of simple positions, though in some cases I was interested to see what these less complex measures could tell us about a hitter as well.

First things first, let’s tackle the common misconceptions I still hear all the time from players and even paid coaches and ex-professional players.  Forgive me for indulging in topics so obviously wrong to those who have studied hitters in the last 20-30 years with the advent of pause buttons on web browsers and DVR, but these ideas are still floating around some of the most expert circles.

SQUASH THE BUG

This is a fairly easy one to look at, with no actual measuring to do.  The common knowledge argument is that hitters are more powerful when they can keep all their weight on their back leg and do not transfer weight during the swing.  Coaches will demonstrate this by twisting their leg into the ground as if they wanted to squash a bug or put a cigarette butt out, thus enabling the hitter to use his hips more in the swing.  Never mind the fact that as you put more weight on your back leg the more rigid your hip joint becomes or that it puts you in an non-athletic position unsuitable for any move in any sport, let alone the other parts of baseball.  You throw a ball; you transfer your weight to the leg closest to your target.  You run your weight transfers.  You throw a punch, your weight transfers.  You throw a Frisbee, your weight transfers.  I decided to simply look at the position of our 50 hitters’ back feet at the point of contact to see whether their feet were exhibiting any signs of weight being put on them.  Here are the results:

Table 1. Back foot position at contact

As it turned out, 14 of the 50 hitters had their feet come completely off the ground and 12 dragged their toes across the ground; both of these moves require weight to be on your front foot for them to happen.  22 of them came up onto the tips of their shoes without moving one direction or another, and only 2 could be argued for having any measurable weight on their back foot with a slight bend noticeable.  Here are pictures of the two who stayed on the ground, Matt Kemp and Shin-Soo Choo (click to view images).

Versus some of the more extremes who came off the ground, Albert Pujols and Jose Bautista:

Not buying this one.  Onto the next then.

EXTEND AT CONTACT

This has been around for decades, and also includes such coaching cues as “alligator jaws” at contact and “roll the wrists through the ball”.  This is an easy measurement as well.  I measured the angle of the top arm (right for right-handed hitters, left for left) at contact for the 50 hitters and got the following results:

Table 2. Angle of top arm at contact

The closest to 180 degrees was 148, with an average of 120.7, so the data says otherwise.  Intuitively there seems to be no merit to this claim.  If you wanted to push someone, as if throwing a block in football perhaps, you would not want to make contact with straight arms to get the best results.  You would want some degree of bend to engage all the muscles in your arms.  Hitting a baseball should be no different, and in fact it appears not to be.  Hitters extend after contact as a sign that they stay through the ball, which also gives them more room for error on their timing.  A bent top arm is still more desirable.  Here are pictures of the highest- and lowest-angled hitters, Adam Jones and Matt Holliday, respectively, with Alex Gordon representing the average (121°).

KEEP YOUR HEAD STILL

This is sort of in the same vain as the squash the bug idea, the claim being that keeping a hitter from moving his head and weight forward will enable him to use his lower body to hit the ball harder.  I have also heard coaches say that a hitter hurts himself by moving his head forward since it “speeds up the pitch” as he inches closer to the pitcher.  Here’s how it bears out in the 50 hitter sample:

		%	Inches			%	Inches
Stance-Negative Move Vertical	Avg	-1.58%	-1.1	Negative Move-Foot Strike Horizontal	Avg	8.67%	6.1
	Std Dev	2.15%	1.5		Std Dev	4.29%	3.0
	High	4.29%	3.0		High	20.77%	14.5
	Low	-7.55%	-5.3		Low	0.55%	0.4
		%	Inches			%	Inches
Stance-Negative Move Horizontal	Avg	-2.43%	-1.7	Foot Strike-Contact Vertical	Avg	-0.87%	-0.6
	Std Dev	3.82%	2.7		Std Dev	2.54%	1.8
	High	3.54%	2.5		High	3.68%	2.6
	Low	-11.29%	-7.9		Low	-7.22%	-5.1
		%	Inches			%	Inches
Negative Move-Foot Strike Vertical	Avg	-4.73%	-3.3	Foot Strike-Contact Horizontal	Avg	-0.23%	-0.2
	Std Dev	3.12%	2.2		Std Dev	2.87%	2.0
	High	1.23%	0.9		High	6.60%	4.6
	Low	-12.30%	-8.6		Low	-7.60%	-5.3

Table 3. Head movement during stages leading up to contact

This table certainly needs some clarification.  I plotted the movement of the head in both the vertical and horizontal directions at different points in the swing.  The only type of measurement available was in the form of screen pixels, so I had to standardize them due to different zooms and camera distances from the hitters.  I decided to use the total height of each batter at foot strike as a reference point.  I then calculated the percentage of the batters’ heights for each head movement.  Finally to get some perspective on the abstract percentages, I simply assumed the average true height of the batters at foot strike to be 70 inches to get a distance measurement that makes some sense in the form of estimated inches.  Any time inches are mentioned in the rest of this piece, they have been calculated in this manner. 

Note that I am not claiming an exact figure for each hitter based on this estimation.  I only want to give an idea of what the percentages actually mean.  These are the two columns you see above for each move, and in looking at any individual hitter you are welcome to use the percentages to calculate a more exact figure.  As for classifications of the parts of the swing, I assigned “Negative Move” to the point when the head moved farthest back, or where the front leg was highest off the ground in the cases of those who only went forward.  “Foot Strike” in this section refers to the point when both of the batters’ feet were on the ground and right before the hands committed.  “Stance” and “Contact” I think are self-explanatory.

The average hitter in this group moves his head down 1.1 inches and back 1.7 inches, then moves down 3.3 inches and forward 6.1 inches before committing hands.  From this point the average head moves down 0.6 inches and backward 0.2 inches before contact.  Just looking at the common knowledge advice of not moving your head forward, every one of these hitters moves forward to some degree.  Joe Mauer is the lowest at 0.4 inches, while Torii Hunter moves forward 14.5 inches.  Selection bias be damned, I wondered if head movement in the horizontal direction correlated to any of the main offensive statistics, as in if power hitters moved their heads more or less.  Here are the results:

Table 4. Horizontal head movement correlated with triple slash batting stats

Next to no relationship here.  I think this one can be considered dead, simply based on the fact that all of them moved forward to some degree.

While we are at it, let’s expand on head movement a bit and see what conclusions we may be able to draw from this data.  Going from the negative move into foot strike, not only did every hitter move his head forward, but only three hitters’ heads went up; those three who came up (albeit less than one inch) were Billy Butler, Miguel Cabrera, and Yadier Molina for what it’s worth.  The pitch heights for these three homeruns were in the upper two-thirds of the sample, though no correlation was found between the total sample heights and vertical head movement during the move from negative move to foot strike.  Many amateur and some highly touted minor league players tend to want to come up out of the ground as they stride, so it should be interesting to see if this can be used in evaluation.

From foot strike to contact, the average hitter moved his head very little, though there were still ranges among the group of hitters.  In the vertical plane, head movement ranged from -5.1 to 2.6 inches, while in the horizontal direction these limits were -5.3 and 4.6 inches.  From personal experience, I know that many amateur players will change the position of their heads drastically as they approach contact with the ball, so this is something to remember for future prospect analysis.  At this point in the swing I decided to see how much vertical head movement was influenced by the height of the pitch relative to each hitter.  Remember this is the part of the swing where the foot is down, and the hands have started to commit leading up to contact.

I ran a linear regression of the two values, vertical movement and height of the pitch, and found an R² value of 0.2968, showing a significant positive relationship.  The x-intercept of the trendline was found to be 38.5 inches (or 55% of hitter height), conveniently located in the middle of the strike zone.  Pitches below this level were associated with heads moving down, and pitches above the opposite.  It would appear that the minimal change in vertical head position for these hitters can be partly explained by the height of the pitch. 

Presumably, this is because hitters are geared for the middle of the strike zone.  Once they have started to commit their hands is when they make an adjustment with their shoulders to get on the same level of the ball, moving their heads up or down accordingly.  This is an interesting finding, though this is perhaps a complicated observation.  Hitters whose heads move down as they approach contact may simply hit low pitches better, and the opposite for those whose heads stay stable or come up.  Nonetheless, this is still something to explore further in the prospect group as well.

In Part II of this article, I will explore how the head moves relative to body movement and posture changes.  I believe this to be more important to a hitter’s success, as well as being much more practical to measure.

STRIDE LENGTH

There are hundreds of ideas about what is the best way to stride, from no stride at all to leg kicks, toe taps, toe holds, etc.  The no-stride approach is usually taught in conjunction with limiting head movement, while those who advocate for leg kicks believe this leads to more rotational power.  I have never been able to find any scientific research regarding any advantage one has over the other, so all I can report is what MLB hitters do.  To classify each type of stride proved to be a futile attempt at simplification, since everyone had enough distinction in their strides to be described differently. 

I decided instead to look at stride length to see if there were any similarities.  I measured the distance between front and back foot during the stance and at foot strike.  Please note that the front foot went back before going forward to foot strike in some cases.  I only looked at start and endpoints for this calculation, since it was difficult to determine the width of hitters’ bases when their feet were in the air, as in a leg kick that goes back and then forward.  Here are the results, put into inches using the same estimation as previously described:

Table 5. Stride length—swing base at Stance subtracted from swing base at Foot Strike

Note that Edwin Encarnacion was not included in this sample due to my inability to find a video from the start of his swing.  This is a pretty wide range of values around an average of 9.4 inches.  The two hitters who had a net negative stride were Albert Pujols (-1.3 in) and Ryan Ludwick (-0.3 in), though as you can see, very slightly negative.  The two longest strides were from John Jaso (24.6 in) and Evan Longoria (23.0 in), who also started with very narrow bases in their starting stances.  So here again, as with head movement, essentially everyone moves forward to some degree, though no stride obviously does not mean a hitter cannot succeed.  This seems to be something that doesn’t have any effect on how a hitter will hit.

Maybe base width after the stride is something that shows more homogeneity.  Here is the data:

Table 6. Swing base after foot strike

This distribution is negatively skewed with most of the points on the lower end of the set with some outliers to the right.  The two widest-set hitters (again, in relation to their height) are Nick Swisher (78.6%) and Giancarlo Stanton (71.7%).  Meanwhile, the two narrowest hitters are Evan Longoria (43.5%) and Adam LaRoche (43.0%).  Out of curiosity I checked to see whether there was any correlation to the type of hitter these guys were last year, thinking that perhaps hitters who worked into the ground better hit for more power by being able to use their legs more, or for a better average by being able to buy time moving forward with a wider base.  Here are the results:

Table 7. Swing base after foot strike correlated to various offensive stats

Yet again, the data shows no real ability to tell what type of hitter these guys are from just looking at swing base.  Like the other information so far, maybe this will be useful when looking at a hitter who falls outside this distribution.

ELBOW UP OR DOWN? HANDS UP OR DOWN?

Coaches will often point out when a hitter’s hands are too high or too low where as a problem or reason for success in a player.  The elbow is another element of the swing that coaches will like to change in a hitter.  Some say it needs to be up; others say that a high elbow makes the swing longer so it should be down, bringing the hands closer to contact and shortening the swing by pushing the bat to the ball.  Should the hands start the swing or does the elbow attack first?  I like to think of it in terms of throwing a ball; you would never ask someone to keep their elbow down and push the ball to the target.  The arm acts like a whip with the elbow moving before the hands, leading to pitchers getting into crazy external rotation before delivering the ball to the plate:

Swinging a bat is very similar to a sidearm throw with a ball, though without quite the same ranges of motion due to having both hands on the bat.  If a hitter is forced to start with the elbow down, it is much more difficult to get the same whip-like action that pitchers get when they throw.  With this in mind, I decided to look at the position of the hands and the elbow during four positions leading up to contact for each of the 50 hitters.  An important point to note is that every one of these hitters did not start to move their hands forward to swing until after the front foot was completely on the ground, a sequence I know is not always found in amateur hitters.  Let’s take this in portions, starting with the vertical movement of the hands and elbows:

Table 8. Vertical movement of the hands and elbow relative to the top of the head at each point of the swing

On average, the hands start -8.8 inches below the top of the head in the Stance and move down to -12.5 inches on the Negative Move.  At Toe Touch they move back up to -10.7 inches and then down to -12.0 inches as they begin to attack the ball.  Note that Heal Plant was the first frame where the hands visibly started moving to the ball, so these values reflect the hands already in motion.  Meanwhile, the elbow begins -14.6 inches below the head and follow a similar pattern; down to -15.5 at Negative Move, up to -13.4 at Toe Touch and down again to -20.4 at Heal Plant. 

From the magnitude of the change between Toe Touch and Heal Plant for the hands and elbows, it would appear that the elbows begin the process of the arms swinging the bat.  Elbow height changes by 7.0 inches while the hands only go down 1.4 inches in the same amount of swing time.  Let’s make sure this is corroborated by the horizontal direction values as well, since the elbow should catch up to the hands horizontally if they are in fact the first to attack.

First, some pictures so we know what these position numbers actually mean.  Below represent the highest hands and the lowest hands relative to the top of the head in the sample (again minus Encarnacion).  They are none other than Carlos Quentin and Allen Craig (measurements in table):

These represent a pretty wide range of starting positions for the hands.  This does not appear to be much of a teaching point when it comes to getting everyone in a similar position to start.  How about the elbow position?

Yoenis Cespedes and Nick Swisher represent the two extremes of Stance elbow height.  These four pictures represent a nice sample of how different these MLB hitters look.  We have Cespedes with high hands and elbow, Swisher with low hands and an even lower elbow, Craig with low hands but a slightly higher elbow, and Quentin with high hands and a lower elbow.  Interestingly, Nick Swisher also has the lowest elbow relative to the position of his hands, while Allen Craig’s elbow is the highest above the hands.  Not much for evaluating or teaching purposes so far due to the spread in values.  How about horizontal positioning?

		%	Inches			%	Inches
Stance Hands-H	Average	-7.00%	-4.9	Stance Elbow-H	Average	-17.85%	-12.5
	Std Dev	4.80%	3.4		Std Dev	3.69%	2.6
	High	7.42%	5.2		High	-5.33%	-3.7
	Low	-16.40%	-11.5		Low	-24.62%	-17.2
		%	Inches			%	Inches
NM Hands-H	Average	-10.96%	-7.7	NM Elbow-H	Average	-16.95%	-11.9
	Std Dev	3.96%	2.8		Std Dev	4.22%	3.0
	High	1.89%	1.3		High	-5.62%	-3.9
	Low	-18.71%	-13.1		Low	-23.41%	-16.4
		%	Inches			%	Inches
TT Hands-H	Average	-14.06%	-9.8	TT Elbow-H	Average	-14.85%	-10.4
	Std Dev	2.62%	1.8		Std Dev	5.37%	3.8
	High	-8.11%	-5.7		High	-3.92%	-2.7
	Low	-20.35%	-14.2		Low	-23.19%	-16.2
		%	Inches			%	Inches
HP Hands-H	Average	-13.61%	-9.5	HP Elbow-H	Average	-12.28%	-8.6
	Std Dev	2.31%	1.6		Std Dev	4.90%	3.4
	High	-8.93%	-6.3		High	-3.28%	-2.3
	Low	-20.35%	-14.2		Low	-22.05%	-15.4

Table 8. Horizontal movement of the hands and elbow relative to the back of the head at each point of the swing

On average, the hands start -4.9 inches behind the head, move back to -7.7 inches at Negative Move, and then continue back to -9.8 at Toe Touch.  Not until the hands commit at Heal Plant do they move forward to -9.5 inches relative to the head.  The elbow starts -12.5 inches behind the head and progresses steadily through each stage of the swing closer to the head, going to -11.9 inches at Negative Move, and -10.4 and -8.6 at Toe Touch and Heal Plant.  This would definitely seem to indicate that the elbow initiates its movement to the ball before the hands do, and that they do not move in unison.

To summarize these averages, let’s look at some pictures of hitters who demonstrate what is going on here.  Only 2 hitters actually followed this exact pattern of hands and elbows throughout their swing:  Josh Willingham and Torii Hunter.  It may be best to open each up and look at them in succession to see the movements.  Remember, these are only the averages, and you can see how much the values vary over the whole sample.

Here they are in the Stance position:

The Torii Hunter picture is blurry from the fade from the last camera angle in the highlight, a necessary evil with a few of the videos I collected.  The hands move down and back, while the elbow moves down and slightly forward relative to the head going into Negative Move:

The hands come back up slightly but continue to move further back away from the head as they get into Toe Touch.  The elbow continues to move forward and loads upward slightly:

The hands start to descend and move forward again as the actual swing begins, and the elbow does the same at Heal Plant/Hands Commit, similar to the pictures at the beginning of this section depicting Maximum External Rotation:

So, starting position of the hands and elbow does not seem to have much of an effect on allowing a player to be a successful Major League hitter based on this data.  Remember that these pictures above are what the average looks like, though there is a vast amount of variation between each individual hitter.  The most useful piece of information to take from this section is that the elbow moves into the swing first and independent of the hands for a small amount of time, an important part of having a properly-sequenced swing.  Where the elbow and hands start is not as homogeneous as some coaches would have you believe.  As long as the batter has room for his elbow to slot in at the start of the swing, I believe the rest of it to be personal preference and comfort. 

Many hitters start the elbow high and swing right from there, as if they are preloaded to begin the swing.  Others may start lower or more forward and load up while they take a step.  All-in-all, the best way to describe this movement is to liken it to the way the throwing arm moves as a pitcher prepares to throw a ball to the plate; the elbow acts as the proximal part of a whip with the hand motion following in sequence.  Scapular load may be even better to refer to the movement of the elbow in the preparatory phases of the swing than elbow load; scapular load is a term used to describe the early portion of the pitching delivery as well. 

This could explain how in many hitters the elbow could get closer to the position of the head without the body turning.  The elbow did not load up or back in this case; it instead loaded behind the batter toward the middle of his back, pulling his shoulder blades closer together and adding another link to the whip.

As an interesting sidenote, I have found that this hands-elbow relationship also allows batters to buy time on offspeed pitches.  Once the hands start to swing, it is very difficult to make in-flight adjustments to the swing without slowing it down or altering it drastically.  If the elbow and hands go at the same time, the batter must start everything exactly on time to deliver a well-timed swing.  If the elbow starts to attack and the hands stay independent, the batter can wait fractions of a second longer before delivering the rest of the swing.  This is just one way batters show the ability to be on time for wide ranges of speeds, but I thought it important to mention.  Again, more will be explored in future articles.

In summary, we have shed serious doubt on the remaining proponents of squashing the bug and extending at contact.  The data also shows how these 50 hitters all moved their head forward, and in most cases moved down into their legs, as they progressed toward ball contact.  Head movement after the hands committed to the swing was found to be fairly well-correlated with the height of the ball relative to the middle of the strike zone. 

Stride length seems to be determined only by personal preference, though all of these hitters move forward to some degree.  Swing base also does not appear to have an effect on MLB success.  The elbows and hands form a complex movement that highly resembles the throwing action of pitchers as they go from breaking of the hands to release, and all seem to have their elbows up, though with much variation.  The elbow attacks before the hands start, and this entire process with the upper half starts after the the front foot has hit the ground. 

None of these moves have any predictive quality regarding what type of success the hitters had in 2012, though these ranges of motion should still be useful in future evaluations of other players.  Some of the future topics to be discussed are:  the angle of the bat from the side view and how it relates to the movement of the hands and elbow, lower half mechanics with the front and back legs, and posture changes as evidenced by the relationship between the head and the hitter’s center of mass.