Jose Fernandez Has Been Unhittable, and So Hittable

Just as much as we’ve trained ourselves not to take Spring Training stats at face value, we should also by now have trained ourselves not to take player and manager Spring Training quotes at face value. Words are words, and until those words become actions, they’re not super fun to consume or analyze. Like, here’s a few examples pulled from a preseason story by the excellent Clark Spencer, with expectations for the upcoming season inferred from comments by star pitcher Jose Fernandez and manager Don Mattingly.

Spring Training Report #1

• Expectation: “He just might not resort to his fastball quite as much…”
• Reality: Fernandez’s fastball rate is mostly unchanged, and in fact is slightly above (52%) his previous career-high (51%).

Spring Training Report #2

• Expectation: “We want to see him … not having to have the mentality that I’ve got to strike everybody out.”
• Reality: Fernandez is striking everybody out. His 36% strikeout rate leads baseball, and is a career-high. He’s leaning on the strikeout more than ever before.

Spring Training Report #3

• Expectation: “There’s nothing wrong with having guy [sic] hit a ground ball early in the count.”
• Reality: Fernandez’s 41% ground-ball rate is fractions of a percent off his career-low, and ranks in the bottom-quarter of all qualified starting pitchers.

Spring Training Report #4

• Expectation: “We just want him to understand that he can stay under control and be effective.”
• Reality: Fernandez has not stayed under control (career-high 9.4% walk rate, upper-quarter of all qualified starters) yet has remained very effective (2.52 ERA).

That’s literally all from one story, from less than three months ago. The Marlins submitted the new Fernandez as an offspeed-reliant, strike-throwing, ground-ball-getting control artist, and instead what they got was a more fastball-reliant, out-of-the-zone-working, fly-ball-inducing strikeout machine. At least they got the name right.

So, that’s all sort of bizarre. But just as curious as is the dichotomy between expectations and reality, Fernandez’s season when viewed on its own is equally puzzling. To gain a sense of its unique nature, one must look no further than this image, plotting two important, yet largely unrelated pitching traits against one another:

What immediately jumps out about that plot is that Fernandez is running just an absurd contact rate of 65%. The next-closest starter is Clayton Kershaw, at 69%. So, Fernandez has the contact rate of a dominant reliever, but when balls have been put in play against him, they’ve come off the bat like those against Bud Norris and Mike Leake. Fernandez’s 93-mph average exit velocity allowed is second-highest among starters and would seem to suggest his .329 batting average on balls in play this season is less the product of poor fortune and more the product of allowing consistently hard contact.

But how can that be? How can one pitcher be so impossible to touch, yet so easy to square up?

Let’s start with the fun stuff: the whiffs. The whiffs have always been there for Fernandez. It’s not like he’s a contact pitcher that’s transformed himself into a strikeout artist. He’s a strikeout artist who’s lately taken things to a new level. And you’ll never guess how he’s done it! Location and movement. A true visionary.

Fernandez is known for his breaking ball, and it’s the breaking ball which is (a) most changed and (b) most responsible for the uptick in whiffs. How it’s changed over time:

In late April, Eno Sarris wrote how Fernandez actually throws two variations of his breaking ball — one with more traditional curve break, and one that more resembles a slider. The slider version comes in a bit harder, is more firm, and gets more whiffs. Over the years, Fernandez has more and more often gone to that slider version of his breaking ball, to the point where his average “curveball” this season has nearly identical velocity and movement to the “slider” shape that Sarris identified in his piece.

Whereas the average shape of a Fernandez breaker in 2013 might’ve looked like this:

It now more often moves like this:

So, the movement on Fernandez’s best swing-and-miss pitch has changed, and changed in the way that makes it more conducive to whiffs. But the location of his entire arsenal has changed, too, in a way that helps explain both the uptick in whiffs and the uptick in walks:

All three pitches in Fernandez’s arsenal have moved away from the center of the plate and more toward the edges and outsides of the zone. The fastball is more inside to righties, and away to lefties. The curveball, with its new slider-like shape, is more buried. The changeup, upon which Fernandez is relying more against lefties, perhaps to counteract the more slider-like shape of his breaking ball, is better spotted on the outer-edge of the plate.

This shift in location makes it more likely that any given swing results in a whiff, but also more likely that any given pitch results in a ball. This goes a long way toward explaining the strikeout and walk spikes. But what about that exit velocity? Exit velo is a complicated beast, one which we’re far from understanding, but there’s something I’ve been spending a lot of time thinking about lately, and it seems to apply here.

Fernandez’s rate of contact allowed on pitches outside the strike zone has dropped substantially, from 56% before this year to 38% now — the lowest figure in the majors. His in-zone contact rate has dropped too, but only slightly. What that means is, Fernandez’s distribution of batted balls allowed has skewed heavily toward those that come against pitches inside the strike zone. In fact, a higher percentage of Fernandez’s batted balls have come against in-zone pitches than anyone else in baseball:

Ratio of in-zone contact% vs. out-of-zone contact%

1. Jose Fernandez, 2.1, .329 BABIP
2. Clayton Kershaw, 2.0, .242 BABIP
3. Nicholas Tropeano, 1.8, .327 BABIP
4. Chris Archer, 1.7, .324 BABIP
5. Michael Pineda, 1.7, .390 BABIP

Most pitchers generate their weak contact outside the zone. Against Fernandez, however, nobody is able to make contact out of the zone, so the only time a hitter puts the ball in play against Fernandez is when it comes against a hittable pitch.

Admittedly, there’s no correlation between the in-zone vs. out-of-zone contact ratio I presented above and overall BABIP or exit velocity. But you’ll notice that apart from exception-to-all-rules Clayton Kershaw, the other pitchers with the highest ratio of in-zone contacts vs. out-of-zone contacts are running exceptionally high BABIPs. On the other end of the spectrum — guys whose distribution of contact rate skews towards out-of-zone contacts — one finds noted batted-ball suppressors like Hector Santiago, Chris Tillman, and Jered Weaver. This feels like something. One end of the spectrum isn’t inherently better than the other, but it’s another interesting way to think about how certain pitchers might have the ability to control their contact, and one possible explanation for Fernandez’s (seemingly) alarmingly high BABIP and exit velocity.

Fernandez has definitely changed the shape of his breaking ball, which helps explain the rise in strikeouts, and he’s definitely changed his overall pitch location, which helps explain the rise in walks. The batted-ball stuff is a little more murky, but it’s certainly possible that the changes he’s made to induce more whiffs also make it more likely that when batters do make contact, they’ve got a better chance of squaring him up. The whiff rate and the contact management almost appear to exist on a sliding scale in this regard, and this is just the way that Fernandez’s adjustments have moved that scale. The good news for Miami is, every version they’ve seen of Fernandez has been incredible.