A little under a week ago, reader Mike Ford sent me an email regarding Felix's mechanics. The gist was that, while he agreed with the general consensus that Felix's throwing motion is violent and therefore something of a risk, he'd stumbled across this writeup (scroll down) that's much higher of Felix's delivery, and was curious to know what I thought.

I sat on it for a few days, planning to come back to it and do a while big post breaking down his entire delivery when I had a few hours. I'd already gone through a lot of video and taken most of my necessary screengrabs when I saw this piece by MGL on The Book blog. That's when I decided to change my approach, because I agreed with what he had to say:

For something as popular to analyze as pitching mechanics, there's a lot of conjecture out there (I've been guilty of this a million times), but a startling lack of evidence. Why should we take somebody at their word? What has anyone done in the field to earn that kind of trust? If you're going to broach the subject, you should come to the table with some measure of support other than your own personal opinion.

So that's when I decided to delve into the primary literature. Not looking for anything exhaustive, but for something with substance. And, starting at ASMI, I eventually made my way to this paper by Matsuo, Fleisig, Zheng, and Andrews (J. Appl. Biomech., 2006). It deals with the relationship between elbow torque (more = bad) and both trunk tilt and shoulder abduction. For those unfamiliar with the terms, much like myself a few days ago, "trunk tilt" essentially refers to the angle between the upper body and a vertical line perpendicular to the hips, while "shoulder abduction" refers to the height of the throwing elbow (or the angle between the upper arm during release and a vertical line drawn up the side of the trunk, as if the arm were just hanging down).

That's probably unclear. Elucidation of scientific terminology isn't one of my strong suits. You should really just go ahead and read the article - not only because it'll help you understand what I'm talking about, but also because it's a pretty easy read for a piece out of an academic journal.

Anyway, here's where I'm going with this. A picture of Felix in the middle of his delivery (source: CNN):

The first things you notice, once you get past the chin: a high right elbow and a distinct lean towards first base. After applying a little MS Paint:

With a little measurement, you can come up with the following values:

Trunk Tilt: 27 degrees, contralateral (towards non-throwing arm side)
Shoulder Abduction: 110 degrees

(Give or take a few degrees, since MS Paint is hardly an exact science.)

By themselves, without any frame of reference, those numbers are meaningless. But in looking at the paper, we can see that Felix has a moderately high elbow and a substantial amount of lateral tilt. Of course, we've known this for a while - it's always been a part of Felix's delivery, as he has to lean left to clear his throwing arm.

Where this gets more interesting, though, is when you look at the relation between Felix's numbers and the amount of torque such angles put on the elbow. Scroll to Figure 2b (p. 4) and you can see that a guy with a trunk tilt of 30 degrees and a shoulder abduction angle of 110 degrees puts nearly 66% more torque on his elbow than someone at 10/100 degrees. For someone with a trunk tilt of 20 degrees, the torque is still up about 40-50%. And so on and so forth. According to this study, the "safest" delivery in terms of least torque on the elbow involves a lean of 10 degrees to the side, with 100 degree abduction. That's not Felix. He's got a more exaggerated lean to the side, and a higher elbow.

What does all this mean? Possibly nothing, as the authors are careful not to state any concrete conclusions based on their study. There are a few variables that're virtually impossible to account for when it comes to experimenting on living humans, and that clouds the picture a little bit.

That said, this paper suggests a possible issue, giving us some evidence based on scientific investigation that Felix's delivery may put added stress on his elbow, and a potentially considerable amount, at that. It could be nothing, or it could be Felix's "natural" (and therefore uniquely safest) body position when throwing, but let's put it this way - if I were a high school pitching coach, and I had a young pitcher who threw with Felix's degree of tilt and elbow height, I'd work with him to change it by making his body a little more upright, and bringing his elbow down closer to perpendicular to his side. Just because it might not pose a problem doesn't mean you shouldn't still try to play the odds.

Of course, with Felix, it's a little different. When you have a guy as flipping extraordinary as Felix, you don't screw around with his delivery, not when he's already established himself in the Majors. The potential benefit of better health just isn't worth the potential cost of Felix losing his identity as a young phenom. So, in situations like this, you sit back, enjoy the performance, and cross your fingers that nothing gets wonky in the elbow or shoulder.

So, is Felix at risk? Subjectively, we can say that his delivery looks a little rushed and violent, and that the way he puts everything he has into every pitch (watch his follow-through when he comes back) increases the amount of stress on his joints. Objectively, though, we're limited to things like this, where Felix's tilt and elbow height might raise some warning signs. Other possible issues either lack solid evidence or need to show proper citation to the primary literature, because otherwise we don't know where they're coming from.

Here's what it comes down to - if we're going to take anyone for their word as far as pitching mechanics are concerned, it ought to be the people with a keen physiological understanding of what goes on a pitcher's body when he throws. There's nothing wrong with going through the literature and using it to identify potential problems with a guy's delivery, but the minute clowns like me start distancing ourselves in the discussion, then the chances of inadvertently spreading misinformation go up exponentially. And that's not good for anyone.

Breaking down a pitcher's mechanics is a ton of fun, if you're into that sort of thing, but all too often people cross the line from "this looks like" to "this is" without any legitimate backing. I've done it myself, even when I knew I shouldn't have. MGL is right. We've been demanding evidence for performance-related ideas and theories for as long as baseball's existed. It's time to start taking the same approach to biomechanical evaluation. Giving people the benefit of the doubt for no reason at all is just too risky to allow.