Jordan Zimmermann’s last start, a dominating 6-0 win against the San Diego Padres, was one for the Nationals record books. Perfect through 5.1 innings, the righthander ended his night with a two-hit, twelve-strikeout, no-walk shutout, good for a game score of 95, besting the previous Nats record of 90 by John Patterson in 2005 as well as Josh Beckett’s no-hitter earlier in the season.
A man of few words, Zimmermann was decidedly chatty after game, with much of the discourse surrounding the run support provided him; in particular, Zimmermann discussed his approach, with the added luxury of runs and an expansive ballpark:
Pour strikes in the zone…It’s a big ballpark. Just let them hit the ball.
And throw strikes he did, with 73 percent of his 114 pitches during the complete game effort being strikes. a five percent hike in strike rate compared to his other 2014 starts. However, the propensity to throw strikes from a guy known for his ability to effortlessly pound the zone isn’t necessarily news, or even interesting. What does raise an eyebrow is Zimmermann’s quote regarding just letting hitters hit the ball, given the lead he was afforded.
Did his willingness to throw strikes and ‘let them hit the ball’ reflect in his pitch data? With a lead of varying size, does Zimmermann become a more flippant pitcher, relaxing his normally dogged approach to getting hitter out?
For this exercise, we are looking at how Zimmermann’s approach, both in terms of pitch selection and location, changes, given the lead he has in a given inning. Right away, I must provide some discussion to the methods of data collection, as I have admittedly taken a shortcut, one thoughtfully provided by Zimmermann this year in the form of 27.1 shutout innings. We will discuss the potential relaxed approach through the lens of these innings, as it involves fewer data collection and management gymnastics, simply due to the fact that in these innings, he never pitched behind, runs-wise.
With this important caveat in mind, what do we have in these 27.1 innings? Overall, we have range of data where Zimmermann had a lead of one through seven runs; I have also included first inning data, where he had no lead, as a sort of control variable, assuming that these data could be thought of as his baseline approach with respect to concentration and intensity.
First, let’s look at pitch type and frequency; percentages of each pitch thrown are on the y-axis, while the number of runs ahead are on the x-axis:
By the looks of this graph, Zimmermann is essentially the same guy with respect to pitch selection regardless of lead, pounding hitters with a great fastball. The bigger the lead, however, we do see him going to the slider more frequently than in lower leads, where he is more wont to mix in all of his four pitch repertoire.
Let’s now shift attention to the results of these pitches. Here, I have categorized outcomes into a handful of ‘bins': ball, called strike, swinging strike, foul balls, and balls in play. If we take Zimmermann’s comment at face value, we should see more balls in play, the larger the lead:
Lo and behold, we do see a big hike in balls in play, the bigger the lead; there’s a great deal of variability of how many called strikes Zimmermann gets given the lead, but this is offset somewhat by a relatively stable swinging strike rate. We also find JZimm having a bit of an aberrant spike in balls with a four run lead, but settles down with larger leads.
And that fantastic command of the strike zone—how does that change, if at all, across runs ahead and pitch result?
The bigger the lead, the more we find Zimmermann pepper the heart of the plate with pitches, with most of them resulting in a ball in play, just as he had described after his shutout. The more an opposing team is with respect to tying or going ahead, the more likely he is to paint the corners, per his usual approach.
One last graphic, this time, taking into account pitch velocity as well as inning; there’s a slight chance that the additional variable of inning to go along with runs ahead metric might shed extra light on the data.
Overall, Zimmermann’s velocity remains consistent, regardless of inning or size of lead, with perhaps some slight increase in velocity with a bigger lead in later innings. However, the effect isn’t huge, and in general, we see no dramatic trends in velocity.
While the data as sampled do have some limitations, they do show Zimmermann as being true to his words, showing a propensity to put the fate of his ability to get outs in games where he has a large lead into the hands of the batter. He does this by throwing almost exclusively fastballs while also throttling back on his desire to paint corners, allowing his pitches to get more of the plate than in situations where his lead could suddenly become a defecit in a swing of the bat. It’s an approach that is simple in theory, but difficult to carry out with any consistent success, making Zimmermann’s mastery of the approach all the more enjoyable to watch.