The Unpredictability of Rich Hill

I don't think there is a more unpredictable player in all of baseball this year than Rich Hill. One of the newer Oakland Athletics, Hill enters 2016 with job security but not many reasons for it.

Since 2009 he's made just four major league starts and in each of the past two seasons he's thrown less than 100 innings across all levels. Typically that recent history doesn't equal a guaranteed spot in a major league rotation.

However, unpredictability can be a good thing — even a great thing. Brandon Moss and Yoenis Cespedes were unpredictable. Brandon McCarthy, Bartolo Colon, and Scott Kazmir were also unpredictable. It's that unpredictability that made Hill a relative bargain and I believe it's what can make him the A's best starter in 2016.

The Good

In somewhat of a rebirth, Hill enjoyed a four-game stint of Kershawian proportions in 2015. He threw 29 innings for the Boston Red Sox, struck out a whopping 34% of batters he faced while walking just 4.7%, and he allowed only 5 earned runs. His 2.27 FIP and 2.50 xFIP point to his 1.55 ERA having a degree of legitimacy.

The Bad

Hill's 2015 lasted just 29 innings and four starts. His BABIP was crazy low at .197 and his LOB percentage was far above average at 87.9% (league average: 72.9%). Nothing in his statistical profile points to his 2015 results being the expectation for 2016.

Ways Hill Has Changed

One could easily dismiss Hill's 2015 success as a small sample of innings or that he came back from basically a two season absence with a little more gas and caught the league off-guard.

I think Hill made an adjustment that fueled his late-season surge and that can carry over into 2016.

Pitchf/x, courtesy of Fangraphs, tells us his repertoire has always included two fastballs, a slider, a curveball and a changeup. For the first time since 2010 Hill went to his four seam fastball regularly and he all but dumped his slider. In seasons in which he threw 30 or more innings he threw his slider 2.8%, 2.0% and 9% of the time, respectively. In 2015 he threw his slider less than 1% of the time.

However, the biggest change to his repertoire came from his increased changeup usage. From 2007-2010 he threw his changeup 4.6%, 2.4%, 5.8%, and 4.2%, respectively. Then he basically stopped throwing it until 2015, when he threw it 8.1%.

Not only did he throw his changeup more often, but it was much more effective too. After only ever posting negative runs-above-average values with his changeup he came in at 1.7 runs-above-average in 2015. Additionally his changeup, per 100 pitches, rated better than his curveball.

His changeup isn't necessarily a strikeout pitch but it doesn't have to be for him to be successful. You can call it small sample size, because he did only throw 29 major league innings in 2015. However, I think there is something to this uptick in changeup usage.

Add It All Up

Hill isn't unpredictable in quality so much as he is in quantity. Projection systems haven't quite made up their mind on him. Steamer expects more than 180 innings while ZiPS forecasts just 84.

The easy comparison can be made between Hill and Kazmir. Both were in and out of major league baseball for two seasons before their resurgence, with Kazmir's coming in 2013 with Cleveland. That season Kazmir tossed just over 150 innings with a 4.04 ERA, 3.51 FIP, and a strong 17.1 K-BB%.

An optimist might say that Hill has saved his bullets by not pitching much in the past two years. A realist probably says he's turning 36 and hasn't had a full major league season in years. If I had to guess I'd say Hill will give the A's 20 starts and somewhere between 110 and 120 innings.

I think Steamer is right on the money by projecting an ERA and FIP around 4.00 and nearly a strikeout per inning. However, given the signs that he has changed as a pitcher there is potential for much better performance. I would not at all be surprised to see Hill post a low-3 ERA and be worth nearly 4 fWAR.