What's up with Carlos Martinez's changeup?

Through (technically) seven starts in 2016, Carlos Martinez has not yet been able to replicate his performance from 2015, a year in which he deservedly received the honor of being named to the National League All-Star team. While Martinez started the season with a sterling 4-0 record and a shiny sub-2.00 ERA, he simply was not missing bats as frequently as one would project given past performance (24.4% strikeout rate and 10.5% swinging strike rate in 2015) and the further development of his already electric repertoire. Yet, since starting 4-0, Martinez has taken the loss in each of his next three starts, and he remains below both his projections and career average on strikeout rate.

2016 Statistics

GS	IP	K%	BB%	ERA	FIP	fWAR
7	43.0	18.5%	8.1%	3.14	4.22	0.3

Instead of providing the borderline lazy "small sample size" reason for Martinez's struggles thus far in 2016, I intend on taking a closer look at a pitch that led to so much success for the 24-year-old right-hander last season. The pitch is his changeup, and according to the offseason poll I included in my "Repertoires in Review" series, Martinez possesses the best changeup on the Cardinals (as he received 30% of the 466 total votes; for perspective, Michael Wacha was second with 20%). First and foremost, is the actual makeup of the changeup any different from what it was last season? Let's take a look, using PitchF/x data:

PitchF/x Basics (Via BrooksBaseball.net)

Remember: Regarding horizontal movement in right-handed pitchers, a negative value means arm-side movement, whereas a positive value means glove-side movement.

Year	Frequency	Velocity (MPH)	Dragless Horizontal Mov. (in.)	Dragless Vertical Mov. + Gravity (in.)
2015	17.18%	87.69	-13.61	-28.33
2016	17.71%	88.04	-13.35	-28.35

As you can see, from a basic PitchF/x standpoint, Martinez's 2016 changeup has been virtually identical to his 2015 changeup. Both horizontally and vertically, the pitch is following essentially the same path from his throwing hand to Yadier Molina's mitt. If it makes it there, that is. In 2015, the changeup garnered whiffs on 42.8% of swings, but thus far in 2016, the pitch has resulted in whiffs on only 22.41% of swings, an eye-popping drop of ~22 percentage points. Even further, Martinez is using the pitch in 2016 with basically the same frequency as 2015, so it wouldn't be fair to state that opposing hitters have been able to "sit on the pitch," an unlikelihood even if he was throwing it more considering the threat of Martinez's other three pitches.

Changeup Location (Via baseballsavant.mlb.com)

While the overall locations represented in both heatmaps are generally pretty good, it is clear that Martinez located his changeup lower more frequently in 2015 than in 2016 (as shown by the fact that the core location for 2015 has a y-value of 1.25 versus a y-value of 1.5 for 2016 thus far). Plus, and this is partly due to a larger sample size of pitches, Martinez's heatmap is much tighter in 2015 (the left) and does not stray toward the top of the zone like we see from in the 2016 heatmap. Thus, location is almost certainly part of the problem with the pitch. When the changeup is down, it is nearly impossible to make solid contact and often leads to a multitude of swings and misses. When it's up, it's whiffed on much less frequently and possesses the potential to lead to dangerous results.

Sequencing and a Potential Issue as represented by this Bases-Loaded Duel with Corey Seager

101.7 MPH Fourseamer: Strike Two

While Martinez missed Molina's spot by over a foot, the 101.7 MPH fourseamer was up and in on the hands of Seager, and from a sequencing standpoint, the set-up pitch opened his put-away pitch tool bag wide open. Martinez could put the 22-year-old shortstop away with a changeup down in the zone, a nasty slider (either backdoor or down and in as Jim Edmonds suggested during the broadcast), or even a sinker down and away in hopes of inducing weak contact and possibly a game-changing double play.

91.6 MPH Changeup: Ball Two

Given the title of this article, you likely guessed correctly: Martinez went with a changeup for his put-away pitch with two strikes. While the changeup was always down, starting just above Seager's knees out of the hand, it looked like a strike long enough (and possessed a great deal of downward movement) to tempt (or fool) a hitter behind in the count. Plus, it was in a location that has led to a lot of whiffs for Martinez over the years, especially with two strikes in the count.

Yet, despite tremendous utilization of pitch sequencing, Seager did not even flinch. Either Seager has amazing pitch recognition skills (and he might, but let's collect some more MLB data before making such an assertion) or he laid off the pitch because he knew he would not have been able to hit it even if he swung. Or, and this is much more difficult to prove given that we only have one camera angle to work with, he was tipped off and knew the pitch was coming. If a batter knows a changeup is coming, then he has enough experience to also know that a changeup starting down in the zone will almost certainly drop out of the zone before traveling the entire 60 feet, six inches.

And no, let me clear that I am not, in even the least bit, accusing the Dodgers of stealing signs from second base. Rather, I ponder the possibility that hitters may be picking up on some sort of mechanical difference between Martinez's fastballs and his changeup. I set up the GIFs side-by-side and clicked through frame-by-frame and at first could not pick up on anything as Martinez's arm followed a nearly identical path from the glove, around his shoulder, and through the pitch. However, I did spot something at the very beginning of his motion, that, to us, may seem minute, but to a professional hitter, may be just enough to help him adjust and/or know what is coming. See below:

On the left, a fourseamer, Martinez's hands break (meaning his hand, holding the ball, leaves the glove) as he is coming down from his apex and just before he begins to drive his lower body toward home plate. On the right, a changeup, his hands break almost immediately after his knee reaches its apex (and you can clearly tell he is much more straight-up when his hands break). As I said above, the difference is so small, but it comes at a crucial point in the delivery, a point that can help tip the batter off as to what is coming next.

Hands break later in the motion? Fastball. Hands break earlier in the motion? Changeup. This can be a very useful tool for hitters given the significant difference in velocity between the two pitches. Is this anything? Maybe, maybe not, but it is definitely something I hope pitching coach Derek Lilliquist and Martinez discuss in detail, using video (and ideally, more angles), before his next start.

Post-Publish Edit (6:50 PM CST, 5/16/16)

Upon even further review (and through the help of another set of eyes), I would like to write about another difference spotted between Martinez's fourseamer and changeup mechanics. As with the example above, Martinez is throwing his fourseamer on the left, and you can clearly see some daylight between his glove and his torso. On the flip side, there is noticeably less distance between Martinez's glove and torso when throwing his changeup.

Admittedly, the batter, standing at home plate (and not having access to the center field view), may have a little difficulty picking up on the different times Martinez breaks his hands in the first flaw I pointed out, unless he knows exactly what to look for and his line of sight isn't blocked off too much by Martinez's left knee. Yet, I have zero doubt that a Major League hitter can pick up the differences in glove distance from the body. As with the first flaw discussed, the difference may be very small, but with the advancement of technology and hitters and hitting coaches doing whatever they can to get an edge on pitchers, I am certain hitters are able to pick up on the difference. Putting it simply, glove out, away from the torso? Fourseamer. Glove tucked closer into the torso? Changeup.

92.0 MPH Changeup: RBI Single

Finally, tipped off or not, hanging a changeup near the middle of the plate doesn't usually end well. It certainly didn't here as Seager deposited the pitch into right-field for the go-ahead single.

Bottom Line

Carlos Martinez has a filthy repertoire. No one can reasonably deny this. That being said, he needs to do himself a favor and get his mechanics tightened up because frankly, he should never lose a battle when he is able to sequence pitches as well as he did in the example above.

Credit to PitcherList.com for the GIFs used in this post.