Carlos Martinez is still experimenting

Ed. Note: Please welcome Matthew to VEB. He's written Fanposts here as well as Community posts at at Fangraphs. We are excited to have him.

Going for the sweep against the Colorado Rockies last Wednesday, the St. Louis Cardinals sent Carlos Martinez to the mound. The ace of the staff struck out 8 in a performance that was enough to get the win, but was otherwise unimpressive. The 6 inning, 5 run start closed out a month that was, by far, the worst of his season (and career). In July, opponents posted a batting average of .308 and an OPS of .928 against him--both season highs by a considerable margin.

Martinez has been dominant over long stretches, too. One only has to look at this past May to see that. A month where opponents posted a dismal BABIP of .202 and a dreadful batting average of .170, resulted in 4 wins and an ERA just above 2.00 for the 25 year-old. It's rare to see variance like this not separated by an injury. Now, I don't mean to suggest that Carlos has lost his touch--that he has regressed to mediocrity and is no longer the brightest light in an organization full of promising hurlers. He is still an ace in every sense of the word. However, the difference between his performance in May and in July warrants a closer look.

In mid-June, Joe Schwarz (@stlcupofjoe) wrote about Martinez's early success, especially against lefties. I'll take a similar approach in analyzing the differences between May and July.

First, let's look at Carlos's slider in the two months we are comparing (via BrooksBaseball.net and BaseballSavant.com).

Carlos Martinez, Slider versus LHH

Month	Frequency	Velocity (MPH)	Horizontal Movement (in)	Spin Rate (RPM)
May 2017	24.19%	86.10	4.90	2224
July 2017	17.14%	85.67	4.73	2158

From May to July, the frequency decreased slightly, which says more about trust in the pitch than it does about its efficacy considering the sample sizes are similar. The velocity and horizontal movement stayed virtually the same. However, the spin rate decreased. For breaking pitches, such as the slider, a higher spin rate is better. It essentially creates a firmer pitch, allowing the pitcher to locate it more consistently. While this difference isn't huge, it might have some significance.

Moving to location, on the left is Martinez's slider location to left-handed hitters in May and on the right is the slider location to left-handed hitters in July.

BaseballSavant

As you can see, these are two very different core locations. In May, Martinez was able to locate his slider low and away to lefties, painting the outside corner with a nearly unhittabe pitch. In July, he left the same pitch down-and-in. Not far enough down to create a positive outcome and not far enough in to jam the hitter. The result is the difference between plays like this:

BaseballSavant

And plays like this:

BaseballSavant

On this pitch from the July 7th game against the Mets, notice how Yadi sets up low and away (the May core location) to Bruce, but Martinez misses the spot and throws it closer to the July core location.

This pitch virtualization graph paints a clear picture of the change as well.

In this overlay, a left-handed hitter would stand below home-plate. The faded trajectory is from May, while the brighter one is from July. You can see how much more of the plate the slider caught throughout July than it did in May.

TexasLeaguers

Up to this point, the spin rate seemed to be at least somewhat responsible for the change in location. And it is probably one of a number of factors affecting Martinez's command. But from May to July there was also a change in the release point for his slider.

BaseballSavant

The May release points are on the left; the July release points are on the right. These are all pitches thrown to left-handed hitters. Carlos seems to have shifted his release point slightly (a few inches) to the right (his left)--closer to home plate. When a release point changes, it could be the result of a couple different adjustments. (1) The pitcher changed his delivery, releasing the ball at a different point. (2) The pitcher kept the same delivery, but shifted to a different part of the rubber, altering the release point.

It looks like Carlos is still experimenting with how he wants to attack lefties by changing his positioning on the mound. Here is a picture from a start he made in May.

BaseballSavant

Look, his back foot is a few inches from the first base side of the rubber. Now, compare this with a start he made in July.

BaseballSavant

Here, Martinez is all the way over on the first base side. The heal of his back foot is aligned with the far left edge of the rubber.

Before the season started, Jeff Sullivan wrote an article (Carlos Martinez Has One Issue Left) at Fangraphs discussing Martinez's struggle with left-handed hitters and how he is "still a work in progress" in that regard. He noted that changing positioning on the mound is done to alter attack angles.

So, if Carlos moved around on the mound to change his attack angle during May, why did he change it up again after such a dominant month? Your guess is as good as mine. It could be that he feels more confident throwing from the first base side of the rubber. It's also possible he made the switch to better locate the inside pitch to lefties. What's more probable is that he is still experimenting, searching for the most comfortable way to attack lefties. Although he was successful in May, he believes he can be even better.

Whatever the case may be, when he finally does figure it out, Carlos will have a whole new level to go to.

Credit to BrooksBaseball.net, BaseballSavant.com, and TexasLeaguers.com for the data and graphics used in this post.