Overthinking It: The 10 Phases of Phil Hughes, Compulsive Pitch Tinkerer

There's always adjustments that go on. From the time I came up, I can't even remember how many times I've had to try and change what I do.—Phil Hughes

You all have a milestone by which you mark the return of baseball: the day after the Super Bowl, Truck Day, pitchers and catchers report day, Kevin Towers Terrible Prediction Day, Final Qualifying Offer Free Agent Signing Day, Opening Day. Some of you can circle your milestone days on the calendar months in advance; others can’t pin them down precisely but know roughly when they’ll arrive.

I envy you. My milestone doesn’t happen on a predetermined day, and sometimes it takes until summer. All I can do is hope to hear the words that will tell me it’s time. And this year, those words were uttered early:

Phil Hughes Pitch Type Tinkering Day. Welcome back, baseball.

***

A few months before Hughes made his major-league debut, Kevin Goldstein ranked him the top prospect in the Yankees’ system (and second overall). Here’s how Kevin described Hughes’ repertoire:

His 92-96 mph fastball has good movement to go along with outstanding location, and his hard curveball gives him a second major-league-quality out pitch. His change-up is at least average, and has nice fade and deception.

Sounds simple: fastball, curve, and change. But that wasn’t the original Hughes; he was closer to his purest form a few years before that. When Hughes was drafted in 2004, he threw a slider, not a curve. Yankees minor league pitching coordinator Nardi Contreras believed that high school draftees should throw curveballs, not sliders, however, and so Hughes developed a good curve (with some help from Mike Mussina) and abandoned his original breaking ball—for a while. By the time he’d mastered Double-A, he was already recidivous. This quote comes from seven springs ago:

This spring training, he wants to reinvent his high school slider to give it a tight, darting action.

“Not like a huge breaking ball,” Hughes said. “I don’t really want two of those.”

Maybe the high school Hughes was the real him—the one who went fastball-slider with an occasional change. Or maybe deep down he never knew what he wanted to throw when he grew up; if we could dig up old Hughes home videos, we might find footage of him agonizing over his arsenal in Little League. What we do know is that ever since he made it to the majors, Hughes has been ping-ponging from one pitch to the next, jettisoning whatever isn’t working and searching for something else that might—even if it’s an offering that has failed him before.

With the help of Brooks Baseball’s game logs, custom pitch-type classifications provided by Pitch Info, and Harry Pavlidis himself, I’ve identified nine distinct phases in Hughes’ career. In each of them, he significantly altered (or eliminated) his usage of a particular pitch over a sustained period before switching things up yet again. And in most cases, he explained exactly what he was up to.

As we work our way through Hughes’ seven-season journey, you can compare his performance to the MLB baseline by referring to this table, which displays league-average results for each pitch type Hughes has thrown from 2007-13. (SLGCon is slugging percentage on contact; LD% is line drives per line drive + fly ball; HR/OFBIA% is homers divided by outfield fly balls.)

Phase One: The Pre-Cutter Period
5/1/2007–4/29/2008

Relevant quote:

I’ve never tried to throw a cutter. People are under the notion that you can just pick up any pitch and go ahead and use it and it will be as good as the best in the league. Mariano has the unique ability to throw a really good cutter and a lot of guys do throw cutters. But they aren’t Mariano Rivera cutters. I’ve never attempted to throw one. Maybe sometime down the line I’ll mess around with it. But not everybody has Mo’s cutter.—2/29/2008

Hughes’ second start—the one in which he held the Rangers hitless for 6 1/3 before hurting his hamstring—is the first for which we have PITCHf/x data. From that outing through the following April, Hughes threw what his early quotes and scouting reports suggested he would: a lot of fastballs and curves, with the occasional changeup and slider mixed in for good measure.

When he did throw the slider, it didn’t induce many swings, and too many of those swings made hard contact (low whiff rate, high line-drive rate, and above-average SLGCon). Those results—or maybe even the question about the cutter that lead to that quote—may have given Hughes the idea for Phase Two.

Phase Two: Enter the Cutter, Exit the Slider
9/17/08–6/10/09

Relevant quote:

While Hughes bulked up his innings total, he also used the time to work on his cutter, a pitch he started to develop late in the season after he decided to scrap his slider altogether.

"My slider wasn't working at all," Hughes said. "I worked a lot this fall on my cutter and my changeup, and both have come a long way."—11/28/2008

Hughes’ slider was ineffective early on, so he ditched it and picked up a cutter, throwing it 23 times against the White Sox in September 2008 after not having thrown it once in his first 14 career starts. The pitch got whiffs and groundballs, though when it went in the air it tended to leave the park.

Phase Three: Farewell, Changeup
6/14/09–10/4/09

Hughes’ changeup command in Phases One and Two left a lot to be desired: He threw it in the strike zone less than 20 percent of the time, compared to the 30-plus-percent league average, and he wasn’t getting whiffs. So when he went to the bullpen, he put it on ice…

Phase Four: Welcome Back, Changeup
4/15/10–10/2/10

…and then brought it back when he returned to the rotation, albeit only on occasion. This time it did a bit better.

Phase Five: The Slider Returns
4/3/11–8/2/11

Relevant quote:

Disappointed in the pitch this spring, Hughes tweaked his cutter grip and turned the pitch into more of a slider, something slightly slower and bigger.

“It’s probably technically more slider now,” Hughes said. “But I’ll still call it a cutter because I don’t want to get in the mode of getting around it and lazy with it. If I just tell myself it’s a cutter, I’ll throw it with conviction.
…
“It’s bigger so I assume it has to lose a little velocity to get that,” he said. “I don’t think it’s something that’s slow enough that they recognize it… I have to give it my fastball arm speed and not get lazy with it. If I do that, I don’t think it will fall in the same mode I was when I was 16 years old throwing my slider, because I didn’t really know what I was doing (back then).”—3/23/11

Hughes threw two sliders against the Tigers on April 29, 2008. The next time he threw one: April 3, 2011, also against Detroit. He used the slider (which wasn’t actually any slower) very sparingly, but when he threw it, he paid the price. Opposing batters put the pitch in play four times, with three fly balls and two home runs to show for it. Of course, we’re talking about 2011, which means that almost everything Hughes threw was hit hard.

Phase Six: The Slider Leaves Again
8/7/11–5/28/12

Over the next four months of regular season, Hughes hardly threw the slider at all, keeping his curveball rate steady at 20 percent.

Phase Seven: No Cutter, No Slider
6/3/12–8/22/12

Relevant quote:

I got tired of getting beat with [the cutter]…It’s basically become a non-option.”—7/12/2012

And another:

“I’m not throwing a cutter or a slider any more. I’m just fastball-curveball-change now. I felt that with a better curveball—and being able to throw strikes with it, and having a strike breaking ball and a harder breaking ball—I drop my arm angle a little bit just to get a little more sweeping action on the back foot to a lefty. I just felt that I could do the same thing with the same pitch. I basically threw those other pitches out.”—7/27/2012

For a few months after Phase Six, Hughes cut back to three pitches again, getting rid of the cutter, which had been burned to the tune of a 1.048 SLGCon in Phase Six.

Notice a pattern here?

Hughes loves cutters in April, then sours on them by the end of the season. (He also loves them in April a little less every time.)

Phase Eight: Slider, but No Cutter
8/28/12–8/10/13

Relevant quote:

"[The slider is] something I'm messing with," Hughes said. "It's only been two or three starts, it's a little something to keep them off balance, especially to right-handers."—9/14/2012

And another:

"[The slider has] given me a different look, especially when I have my good fastball," Hughes said. "I think guys have to commit to one or the other. With my curveball they could sort of see it out of my hand. And the changeup isn't really that effective to right-handers. I think the slider has given me a way to combat right-handed hitters, which really gave me trouble last year. It was unusual because right-handed hitters hit close to .300 (.308), and lefties were really under that (.211). I felt the slider as it’s developed has given me a weapon, especially against right-handed batters."—5/10/2013

In Phase Eight, the slider came back, and not just by a little bit: Suddenly Hughes was throwing it far more often than he had in any previous phase, with somewhat improved results. To make room in his repertoire, he eased up on the curveball and continued to ignore the cutter.

On the morning of the day that the article that second quote came from was published, Hughes had a 3.60 ERA and almost a strikeout per inning, with a K:BB ratio close to five.

“Since adopting the slider, Hughes has become a vastly different pitcher,” the article said. “He is finally living up to the potential placed on him, and he's setting himself up for a possible big pay[sic] at the end of the season when he becomes a free agent.”

Later that day, he gave up seven runs (six earned) on seven hits, two walks, and two homers in 5 2/3. His ERA from that day through the end of the season was 5.69. Bringing back the same slider he’d tried and grown disillusioned with before wasn’t the secret to success.

Phase Nine: Splitter Replaces Changeup
8/15/13–9/25/13

Over his final eight games (and seven starts) of last season, Hughes tried something new: a splitter. Over 60 percent of the splitters were balls, and few were close enough to the zone to get swings. Hughes also upped his slider usage further and got a good swing rate with the pitch. Which takes us to…

Phase 10: So Long, Slider (Again)
Present–?

Counting the early instance when he quit because of Contreras, this is the fourth time that Hughes—who’s still only 27 years old—has decided to stop using the slider. Hughes has tinkered with his mechanics, too, and been in and out of the Best Shape of His Life (and in and out of the bullpen). We don’t quite have a handle on him, but maybe that’s not the fans’ and analysts’ fault. Maybe the problem is that Hughes, like Mitch Hedberg’s Bigfoot, is just blurry. He’s a large, out-of-focus starter roaming the countryside.

From afar, it’s natural to wonder why Hughes can’t make up his mind and stick to the same pitch selection. Did he OD on True Detective? Is he just trying to keep pace with his opponents in what he sees as “a constant game of adjustments”? Or is it a lack of conviction? Does he not trust his stuff, and do we have to downgrade his makeup and adjust our projections accordingly?

I asked a scout: Is it a red flag if a pitcher can’t seem to decide what works for him?

“No,” he said. “It’s a red flag if nothing is workable.”

Ultimately, that’s the problem: No matter what changes Hughes makes, his fastball doesn’t regain its lost buzz, and his curveball doesn’t recover its lost bite. He’s not going to find a combination of pitches that turns him back into the second-best prospect in baseball.

That doesn’t mean that he can’t be a solid starter; as the same scout said, “Hughes has the requisite array of pitches,” even if we’re never totally sure which ones he’ll take to the mound. I still think the Twins’ decision to sign him was one of the smartest of the winter, and I’m fully on board with a Minnesota bounceback in a more fly ball friendly park. But I’m not even going to guess whether that bounceback comes with more slider, cutter, or curve.

Thanks to Morris Greenberg, Chris Mosch, and Harry Pavlidis for research assistance.