Shelby Miller just started his tenth game of the season. He has, to his name, all of one single quality start. It came a few weeks ago, in Atlanta, where Miller went to work against one of the worst team offenses in recent baseball history. Miller was removed after the six-inning minimum. He racked up one strikeout, to go with a pair of walks. He also hit a guy. That guy was Erick Aybar, who has a .423 OPS. In Miller’s one quality start, he was statistically bad. Then there are the nine other starts.

In an era of fair and balanced transactions, no offseason move got even a fraction of the criticism of Arizona’s Shelby Miller trade. Those opposed to the move believed the Diamondbacks overpaid for a non-elite starting pitcher. FanGraphs, of course, figured the Braves made out like bandits, and that also happened to be the industry consensus. But to be absolutely clear, no one back then thought that Miller was anything less than a legitimate No. 3. The criticism then had to do with Ender Inciarte and Dansby Swanson. If anything, there were indications Miller might’ve been on the verge of breaking out. At the moment, he’s a shell of himself. Miller has gone completely awry, and he and the Diamondbacks are suffering.

What does rock bottom look like for a big-league starting pitcher? I guess it would look like losing your job. Getting sent to the bullpen, or, worse, getting sent to the minors, or dropped outright. Miller, for the time being, remains a part of the Arizona rotation, but he’s at sort of a relative rock bottom. Tuesday, he was awful against the Pirates. Now here’s his season line, converted to percentile rankings in various important categories. This is out of all pitchers in the majors with at least 20 innings.

Miller has been horrible across the board. People have been worried about Chris Archer, but at least he has his strikeouts. When people were worried about David Price, he had his strikeouts and walks under control. Miller hasn’t struck enough batters out. He’s walked entirely too many — he’s at 29, to go with 30 whiffs. And the batted balls have gotten pounded, so there hasn’t even been modest contact, which has previously been a Miller specialty. If you don’t like hard-hit rate, Statcast agrees that Miller has been hit hard. Everything bad you could imagine — statistically, it’s all been happening. Miller is completely wrong.

It’s not necessarily his first time going through this. That provides some hope; Miller has scuffled before. Back in 2014, there was a 16-start stretch during which Miller had 45 walks and 59 strikeouts. His hard-hit rate was up, and Miller just appeared to be wonky. He got better from there; eventually, he was deemed worth the Arizona trade package. But Miller didn’t struggle like this. This is the lowest point. And what’s funny is, at a glance, it’s hard to see what’s going on. Here’s Miller from last year, when he was fine:

Here’s Miller from Tuesday, in a bad start in a bad month in a bad season:

One of those pitchers got a blockbuster trade return. One of those pitchers is living a nightmare. In case you ever lose perspective, this is how tough it is to be successful in the major leagues. Somewhere in here is the line between being a good starter and being one of the very worst starters at the level.

When something like this is happening, it’s not just a problem with pitch selection. And while Miller’s velocity is down, it’s not way down, and look at that Pirates clip above. Miller actually just had good velocity on Tuesday. He still got knocked around, suggesting a problem with location, suggesting a problem with mechanics. I don’t think that Miller is hurt. I’m sure he’s beat up, psychologically, and that can wear on a guy, but I assume Miller just doesn’t have his comfortable delivery.

It’s time to speculate about that! It’s time for an amateur dive into pitching mechanics. All the usual caveats apply here. What follows might very well not be insightful, but I feel enough like I might be on to something that I’m going forward with this. We’ll see Miller throwing fastballs in 2015 and 2016. On the left, last year; on the right, Tuesday. The angles are almost identical.

This is Miller at his maximum, I don’t know, reach-back. I’ve added some red lines to show how Miller collapses on his right leg. Everyone collapses on the back leg, but there is something here, something I observed in several other clips. The visual angle from last year is slightly obtuse. The visual angle from this year is slightly acute. That means that Miller is either collapsing more, or he’s collapsing earlier. Either one would be a meaningful change. Collapsing earlier would throw off his timing. Collapsing more would alter the plane, and change how Miller generates force.

Here’s the same set of pictures, but looking at something else:

Last year, Miller reached more behind his back. You don’t see that on the right — his hand and the ball are in front of his hip, visually, whereas on the left, there’s space between them and Miller’s butt. The left arm looks about the same, so it’s not an issue of full shoulder rotation. It’s more like partial rotation. Last year, there was more time for Miller’s arm to generate speed. In the image on the right, there’s just less distance to go between the ball and the eventual release point. Seems like it could be compensating for a timing problem.

Here’s our last image pair, showing Miller at release:

The release points themselves aren’t too dissimilar. Visually, you see the ball a little above home plate, to the right side. What’s different is how the ball is getting to the release point. Miller on the left has a lesser arm angle than Miller on the right. On the right, Miller is reaching higher, which could have to do with collapsing so low. It might be a way to try to regain lost height. Maybe this is just something that happens when Miller’s delivery is slightly rushed. If he’s collapsing early, that kicks everything off, and then the arm isn’t totally in sync. When the arm’s out of sync, the pitches are out of sync, and baseballs don’t go where they’re supposed to.

Mechanically, I’m an idiot. I don’t know Shelby Miller’s mechanics like Shelby Miller, nor do I know them like Miller’s coaches. They’ve been working on him for weeks, and they could have their own suspicions. What is evident is that Miller isn’t yet fixed. They’ll keep trying, but the season keeps going, and the numbers keep getting worse. You wonder if they could even get any worse.

That’s what’s known: performance-wise, Shelby Miller is at rock bottom. My guess is what’s featured above; I think his delivery might be off from the start, and that would leave him out of whack. If this is an issue with Miller’s right leg, that seems like the kind of thing that could and should be fixed. There’s not actually all that great a distance between this version of Shelby Miller and a successful one. You could forgive Miller if he feels like he’s a million miles away.