clock menu more-arrow no yes

Filed under:

Has Carlos Martinez cleared the final hurdle?

New, comments

The sample size doesn’t give us a definitive answer, but repertoire adjustments have been key so far in 2017.

Chicago Cubs v St Louis Cardinals Photo by Dilip Vishwanat/Getty Images

During spring training, FanGraphs’ Jeff Sullivan wrote “Carlos Martinez Has One Issue Left,” and the premise behind his article was essentially the same as my January 9th post titled “The final hurdle for Carlos Martinez” (though, as you’d expect, Sullivan’s was considerably better than mine overall). Whether you call it an “issue” or a “hurdle” doesn’t really matter, but the fact was clear that the right-handed throwing Martinez, through the 2016 season, struggled versus left-handed hitters (.329 wOBA against in 2015, .322 in 2016). And as a full-time starting pitcher, this deficiency ultimately limited the height of Martinez’s ceiling. Of course, pitchers tend to provide most of their value by shutting down batters of the same hand, but for the sake of overall consistency, they are still able to manage at least average numbers against batters of the opposite hand.

This hadn’t been the case for Martinez in his first two seasons as a starting pitcher:

Carlos Martinez versus left-handed hitters, 2015-2017

Year TBF AVG OBP SLG wOBA
Year TBF AVG OBP SLG wOBA
2015 365 .257 .339 .417 .329
2016 441 .256 .342 .387 .322
2017 172 .200 .304 .318 .280

I understand that we have not yet reached an adequate sample size to graduate Martinez from the class of “Pitching Successfully to Left-Handed Hitters 101.” Though 172 batters faced isn’t necessarily a tiny sample size, it is still a long way away from these hitting statistics’ respective stabilization points. However, when you consider one significant repertoire adjustment Martinez has made when facing lefties, the talk of a sample size and results doesn’t matter as much. Why? Because when it comes to analysis, especially of pitching, the process is just as important as the result. And as you will see below, Martinez appears to have fine-tuned his process versus lefties, to the point that it is not surprising to also see corresponding positive, though small-sample-sized, results.

Versus lefties, in 2015 and 2016, Martinez threw his slider 15.19% and 14.65% of the time. So far in 2017, Martinez has gone to the slider 20.09% of the time — a ~33% increase from previous seasons. The pitch may be exhibiting less horizontal movement than it did in 2015 and 2016, but a tighter spin rate (averaging 2199 rpm in 2017, versus 2131 rpm in 2016 and 1907 rpm in 2015) has led to a firmer pitch that he has been able locate more consistently. On the left, you’ll find Martinez’s slider location to lefties in 2016, and on the right is where the pitch has landed thus far in 2017:

Always be wary of heatmaps possessing two core locations. Sometimes it means the pitcher has two distinct approaches for that hitter (which isn’t necessarily a bad thing), but more often than not, it means the pitcher is simply struggling with command of the pitch. Regardless of which one is taking place here, neither core location in 2016 is ideal. Hanging a slider up in the zone (core number one) — no matter how much movement the pitch may get — is rarely a good thing, and “down and in” (core number two) has long been considered a hot zone for many left-handed hitters. This doesn’t even begin to mention red-orange concentrated zone surrounding the always dangerous middle-middle.

Well, so far in 2017, Martinez’s slider location has been perfect. Yes, perfect. Its core location — carrying a similar sample to the cores seen in the 2016 heatmap (despite a smaller overall sample size: 141 sliders in 2017 versus 240 in 2016) — paints the low and outside corner. Given the pitch’s ball flight (right to left and down) — though less in magnitude from previous seasons, the pitch still carries considerable horizontal movement (8.77 inches) — left-handed hitters will often “give up” on this pitch, thinking there is no chance it makes it back to catch the corner. This pitch is commonly called the “backdoor slider.”

Based on the 2017 heatmap, it is clear that Martinez has focused on perfecting the backdoor slider to lefties — something he had not been able to do in previous seasons. It’s a pitch that can lead to weak contact, but usually leads to called strikes. Frankly, no matter the result of the pitch, it typically leads to an, at worst, neutral (a called ball) outcome for the pitcher. The Los Angeles Dodgers stack their lineup with potent lefties, and yet, Martinez performed admirably against them on May 31st. Here are four examples of Martinez utilizing the backdoor slider, courtesy of the always wonderful @cardinalsgifs:

First-inning slider to Adrian Gonzalez (BrooksBaseball At Bat)

Fourth-inning slider to Yasmani Grandal (BrooksBaseball At Bat)

Eighth-inning slider to Corey Seager (BrooksBaseball At Bat)

Eighth-inning slider to Yasmani Grandal (BrooksBaseball At Bat)

This pitch simply wasn’t fair as it looked like a ball for almost its entire flight. If you look at BrooksBaseball’s strike zone map (linked to above), you’ll see that the pitch actually landed inside the outside corner. Even if Grandal would have swung the bat, the result had very little chance of being positive. Notice the pitch count and Martinez’s follow-through. He gave it his all on this called strike two.

Just in case you are worried all of this focus on the slider will negatively affect his ability to pinpoint the fourseamer, look no further than this 96.2 MPH heater to Cody Bellinger:

Bottom line, Martinez has always possessed the raw stuff necessary to be effective versus left-handed hitters. With four plus pitches in his back pocket, and three of them moving away from lefties, it’s comforting to see it finally come together. Here’s to hoping he can build on his early success as he has shown to be one of the very best starting pitchers in all of baseball through 13 starts.

As always credit to @cardinalsgifs for the beautiful GIFs, as well as BrooksBaseball.net, BaseballSavant.com, and FanGraphs.com for data used in this post.