Driveline Mechanics

Pitcher Analysis: Andrew Miller

May 4, 2008 – 11:46 pm

As requested by one of our readers (David), I’ve written a pitcher analysis piece about Andrew Miller. Miller, a dominant starter for the North Carolina Tarheels, hasn’t had much success in the big leagues and is currently getting destroyed with the Florida Marlins. Miller is the unhappy owner of a 9+ ERA and a WHIP over 2. (Ed Note: all stats in this article will be compiled from data prior to his May 4th start vs. San Diego)

Many feel he was rushed to the big leagues, getting called up to the MLB after appearing in just three games in High-A Lakeland in his rookie year. Miller would throw only 85 innings in the minor leagues before being called up and left in the big leagues, despite his high walk rates and mediocre strikeout totals…

To me, the problems can be very easily seen in this image:

I don’t think there is any question that Miller was rushed to the big leagues. I have omitted ERA from this image for a reason, since peripherals can tell us a much better story of how good a pitcher really is, while ERA can cloud it. As you can see, Miller was nigh-unhittable in college, posting very low hit rates (and extremely low HR9 rates), and striking out more than a batter per inning. However Miller, like most college power pitchers, struggled with command and walked a very high amount of batters until his senior year of college, where he got it under 3.0.

In 2007, Miller spent 43 innings at High-A Lakeland with a more troublesome walk rate in the minors (expected to only go up as the competition gets tougher) and an unimpressive strikeout rate. However, he gave up few home runs, and with a bit of luck, his ERA was just above three. Promoted to AA, Miller refined his control, increased his strikeout rate, and depressed his hit rate to absurdly low levels. Though we don’t have BABIP data for the minor leagues, there is basically no chance Miller’s expected hit rate even close to that low. In other words, he got very lucky. With a sub-1 ERA, the Tigers promoted him to AAA where he spent just 6 innings before being called up to the big club and was promptly given a dose of reality: Big leaguers don’t swing at marginal pitches; they are tough to strike out; and they hit for a lot of power.

Detroit traded the power lefty to Florida, where it was obvious that he would not be sent down, since the Marlins operate on a budget that makes your local McDonald’s franchise laugh. Unfortunately for Miller, this won’t help his development, as he will be expected to produce at the big league level without experimenting with mechanical tweaks or throwing a third pitch over 50% of the time to test it out - both things that can be done in the minors.

In 2008, Miller might as well be the king of unluckiness. He has posted a Batting Average on Balls in Play (BABIP) of .453. That’s ridiculous. For reference, the average MLB pitcher’s BABIP in 2008 so far is about .298. Miller’s opponents are batting more than 150 points higher on contact off him, and while he is giving up 4% more line drives and getting 4% less ground balls than 2007, that doesn’t mean he should be getting shelled out there. He has given up 48 hits in 25 innings pitched, which is simply preposterous. Furthermore, the average pitcher leaves about 70% of all runners on base. Andrew Miller has been leaving just over 60%. If you plug in his peripherals into a formula that gives you a Fielding Independent Pitching (FIP) stat, we would expect Andrew Miller to give up about a 4.95 ERA. Simply put, Andrew Miller is getting extremely unlucky.

When you’re unlucky and unprepared for the major leagues, that typically leads to drastic efforts to stop the bleeding. Miller is throwing 19% sliders in 2008, up from just 14.1% in 2007. He has decreased the rate he has thrown his fastball and also increased the frequency he throws his changeup. Basically, Miller is trying everything in order to stop this onslaught against him. It probably doesn’t help that he has a bad defense behind him either; the Marlins are tied for 20th with a .695 Defensive Efficiency Rate (DER). Even then, the Marlins pitchers should expect to have a BABIP of .305 (BABIP is approximately 1 - DER, since DER is simply a rate of balls in play turned into outs), and Miller has that crushed.

Miller’s 2007 Pitch f/x data looks like this:

type	Speed (MPH)	Break x (inches)	Break z (inches)	Balls	Strikes Called	Strikes Swinging	Foul/Foul tip	In play outs	Singles	Doubles	Triples	Home Runs
Fastball	92.66	8.78	6.73	191	97	30	106	49	31	2	1	3
Slider	78.74	-4.41	-3.2	39	15	6	24	6	5	1	0	1

Fangraphs reports that he threw his changeup less than 3% of the time, and Pitch f/x would obviously agree. Miller is quite obviously a fastball-first pitcher, complementing it with a slow slider and not much else. We would expect such a pitcher to suffer from drastic platoon splits, getting crushed by righties but holding lefties to a reasonable line with a fastball-slider combination, and Miller is no exception. Over his limited three year career in the big leagues, Miller holds lefties to a .197/.356/.258 line with no home runs, but righties tee off on him at a rate of .293/.396/.461 with all 8 home runs he’s given up. Basically when Miller is pitching to a righty, the average hitter turns into Derek Jeter. Does that sound like a successful starting pitcher to you?

Pitching Mechanics

The first thing I noticed was that Andrew Miller “hooks” the ball during his arm swing, much like Barry Zito:

Chris O’ Leary doesn’t like this since it implies tension in the wrist. I don’t like it because I think it will cause a timing flaw later in the delivery, and it also suggests that the pitcher does not have a smooth pendulum swing-type arm action.

Arm Action / Tempo

Miller is about 17-19 frames from maximal leg lift to footplant, which is good (I like anything under 20). However, he is just late at footstrike:

Miller’s arm has just barely passed through the horizontal axis at footstrike, which will cause his arm to lay back more violently in the external rotation phase, further stressing the subscapularis, supraspinatus, and UCL (two pieces of the rotator cuff and the ligament replaced in Tommy John surgery, respectively).

Furthermore, if you watch the animated image of Miller closely, you will see that I paused it for one second per frame in frames 12 through 15. This is a classic case of reverse rotation, as you can see very clearly that he intently takes the arm behind his back and laterally behind his body. This unnecessarily taxes the anterior shoulder structures including the labrum, pectoralis major, and the biceps. Additionally, reverse rotating the shoulders and taking the ball laterally behind the body gives the hitter a better chance to pick up and track the ball - by simply taking the ball towards second base and keeping it hidden behind the body, a pitcher keeps it hidden from the batter’s view, adding deception to his delivery. (Greg Maddux is a great example of this; it is also called being “sneaky fast.”)

Miller does do a good job of keeping his elbow below or in line with his shoulder which will avoid impingement concerns in his shoulder. However, when you look at the whole picture, Miller’s arm action isn’t good. I rate it as Poor.
Ball Release / Followthrough

Miller has great followthrough - he points his shoulder at the target, his glove is up at his shoulder and firmed up, and he lets his arm decelerate through without stopping it early. I have nothing else to say about this part of his mechanics; I grade them Excellent.

In conclusion, I think that Miller has been severely rushed to the majors. He lacks a third pitch, and as a result, he will always struggle against righties with only a fastball-slider combination. Sliders from pitchers who are losing the platoon battle get crushed while curves typically fare better; the lateral movement is easier to track in as opposed to out, which is why a changeup with tailing action is very necessary to be a complete starting pitcher. Andrew Miller needs time in the minors to master a changeup or he’ll be nothing more than a useful LOOGY (Lefty One Out Guy) or a fifth starter who gets crushed by lineups with lots of right-handed hitters. Additionally, his mechanics are not ideal and could use some tinkering, specifically the extreme reverse rotation which should clear up his poor timing at footstrike. Unfortunately the Marlins seem not to want to option him to the minors, so he’ll continue to get crushed unless he can learn on the job, and fast.

Help spread the word about Driveline Mechanics. Submit this story to Ballhype or Baseball Think Factory.