Since he’s become a regular starter, it’s tempting to look at Zack Greinke‘s career and say that he’s been about the same pitcher with an outstanding peak in 2009. Other than that wonderful year, he’s managed an FIP in the mid-threes by striking batters out at a slightly-above-average rate for a starter and showing excellent control of his four-pitch mix. He’s no metronome – his ground ball to fly ball mix has oscillated – but he seems pretty stable.
But zoom in on just the last two years, and the change seems startling. He went from striking out 9.5 batters per nine to a mere 7.4. He added a run of FIP. His swinging strike percentage dove two percent (from above to below average). And his ground-ball rate changed, as is his wont (from dead-on average (40%) to above-average (46%)). His pitching mix changed fairly significantly, too, dropping almost four percent of his curves and five percent of his sliders, while almost doubling his changeup usage.
And yet, with these changes in his mix, many aspects of his repertoire stayed the same. Looking at his Pitch Type Values, he still had an excellent fastball (10.1 runs above average, +39.8 over four years) and slider (12.3 runs above average, +49 over four years). His changeup was still meh (2.5 runs below average, -13 over four years). Since these numbers include luck and only measure actual results measured by linear weights, we do want to treat these numbers with a grain of salt, but sustained numbers in either direction seem to suggest that his fastball and slider are strong pitches.
So what happened to his curve last year? With the above caveat in mind, his curve went from a scratch-or-better pitch to below average in four years, like this: 2007: +1.5; 2008: +0.5; 2009: +2.4; 2010: -7.0. That last number could have been due to luck – a few curves might have been smacked past a diving infielder, or dunked into short center – but could there have been more going on? The velocity didn’t change much, and though the curve had less movement, there’s been some variation on all of his pitches over the course of his career.
Let’s check the heat maps. It’s why we have these toys. Take particular notice of the top quarter of the strike zone in the maps below.
It certainly looks like he threw fewer high curves to righties in 2010, doesn’t it? If you check back further, you see that those same high curves to righties were ‘missing’ in 2008, too. Now, curves have good whiff percentages when compared to fastballs, but we can’t blame all of his missing strikeouts on his reluctance to throw the high curve. And this still could just be the normal variation that a pitcher goes through as he adjusts to the league and the league adjusts back.
Looking back at Dave Allen’s excellent piece on Tommy Hunter‘s high curves, we find this prescient line, given the fact that Greinke got fewer strikes and more grounders when he used the high curve less in 2010:
It is an interesting pitch: while the average curve gets lots of called strikes and grounders, Hunter’s gets many more swings and then flyballs.
I’m no pitch f/x guru, but I know some. Lucas Apostoleris, from Beyond the Box Score and Don’t Bring In the Lefty, has worked up Greinke’s 2010 pitch f/x data before and was kind enough to run the numbers. It turns out that yes, with the caveat that Greinke has as many as three different kinds of curve balls, his “curve” did get more swings in 2009 when it was higher (41.3% swings in 2009 vs. 34.6% in 2010). Alas for our comparison to Tommy Hunter’s high curves, Greinke got fewer fly balls on the pitch in 2009 (29.5% in 2009, 39.7% in 2010) while his groundball rate remained virtually unchanged. Instead, his curve got more whiffs when he threw the high curve (28.2% whiffs in 2009, 19.9% in 2010). Not quite the same thing that happens with Hunter’s high curves.
And before we get all caught up in the lack of the high curve, there is indeed some luck at play. Josh Smolow, Garik16 at Beyond the Box Score and twitter, said:
His curve by my database has a run value of +4.0077 (positive = bad). But if you use expected run values, which are not BABIP or results dependent, the Curve is NEGATIVE 3.7156, a difference of over 7.
So what we have here is a minor change in Greinke’s curveball philosophy followed by some poor luck with the new approach. Given that the swing and whiff rates were higher when he threw the higher curve, it wouldn’t be terrible for him to return to his 2009 approach. That won’t mean he’ll win another Cy Young, but we shouldn’t discount the possibility that throwing some high curves to righties this season will serve Greinke well.