Rebuilding Salazar: Never a Groundball Pitcher

May 30, 2014 by The Zach Walters Appreciation Guild 2 Comments

During his 2013 redemption campaign, Ubaldo Jimenez broke out in a way that shocked the Cleveland Indians’ fandom. While his walk rate remained high, and in spite of reduced fastball velocity, Jimenez struck out batters at a better rate than he ever had during his career. When asked about the role of Mickey Callaway, Jimenez stated, “When you have a pitching coach that is only telling you what to do and isn’t listening, it’s hard. Mickey has a lot of knowledge, but he also listens. He’s always trying to find out what you think and how you feel you need to improve.”

The lesson taught by this interaction between Callaway and Jimenez, that an instructor is at his best when he recognizes the pitcher’s strengths, is one that applies very directly to recently-excellent, but even more recently-subpar and yet more recently-demoted, Cleveland Indians’ pitcher Danny Salazar. Namely, Salazar’s skill is in generating swinging strikes and in working up in the zone, and to ignore this fact is to ignore the lessons that Callaway espoused only a year ago.

Central to this discussion will be the Heat Maps that have earlier this week debuted on FanGraphs. The advent of this feature is useful not because it’s an innovation, necessarily; Brooks Baseball, for instance, had had zone profiles for a while. The advantages of the FanGraphs Heat Maps are, on one hand, their greater ease in navigation, and on the other hand, their inclusion of league average Heat Maps.

For instance, the following is a chart of league-average runs-above-average based on pitch distribution, discussed by Dave Cameron in one of FanGraphs’s heat map roll-out pieces.

Credit: FanGraphs

As Prime Minister Cameron notes, this chart confirms intuition: the wilder a pitch, the worse it is for a pitcher. Pitches thrown inside the strike zone result in fewer runs than those outside the strike zone. In the case of Salazar, this ideally uncontroversial premise holds – in 2014, Salazar’s walk rate increased, his first-pitch strike rate decreased substantially (2013 67% to 57% 2014) and his effectiveness decreased. While he didn’t throw noticeably fewer in-zone pitches on the aggregate, batters chased his out-of-zone pitches at nowhere near the same frequency – resulting in an extremely undesirable trade-off of Swinging Strikes for balls. However uninteresting and obvious these points may appear, they are statements both intuitive and true. The confluence of these two traits should be noted, since intuitive and true, from this point on, are not inherently intertwined.

As a premise, it’s absolutely true that Salazar pitched high in the zone. For instance, another glorious heat map, showing pitch distribution by location, shows that when Salazar does pitch within the strike zone, it’s relatively high – for comparison, the league average pitch distribution is included.

Salazar Pitch Distribution (Credit: FanGraphs)

League Average Pitch Distribution (Credit: FanGraphs)

When Salazar misses the zone, it’s typically rather egregious (more egregious than last year), and when he does pitch within the zone, it’s higher within the zone than the league average pitch. These assertions are true; some of the conclusions that stem from these assertions, moreover, are also true.

As noted, the wildness has resulted in a deleterious walk rate. Moreover, while his Zone% had not changed perceptibly from 2013 to 2014, the balls he does throw are further from the strike zone. Thus far, the prevailing narrative about why Salazar isn’t inducing the same rate of whiffs as he did in 2013 is because hitters are making adjustments to Salazar. However, if Salazar were indeed missing the zone by larger amounts, it could indicate not that hitters have figured Salazar out but that his pitches are so far out of the zone that even A.J. Pierzynski wouldn’t chase them. In this situation, the marked rise in walk rate and, potentially, decline in whiff rate might both stem from the same cause – a decline in control not reflected in binary statistics line Zone%. Such speculation regarding Salazar’s decline in O-Swing%, however, is merely speculation and not an assertion this particular argument seeks to aggressively champion.

High In The Zone: Net Negative?

A third assertion, regarding his in-zone tendencies, is that his tendency to pitch high in the zone resulted in batters deciding all they could do was ding a ding dang their ding a long ling long – or for those not familiar with Ministry’s testament to lyrical nihilism, the assertion is that Salazar pitching high in the zone has resulted in more home runs and is therefore a net negative.  The argument continues that Salazar would be better-served keeping the ball low, as he would be able to elicit more ground balls under those circumstances. So common is this argument – indeed, so intuitive is it – that it’s accepted as fact that Salazar’s tendency to pitch high in the zone has a natural consequence of getting hit hard with unparalleled frequency. More explicitly, the argument would run:

1. Salazar has pitched high in the zone, resulting in more home runs
2. Pitching high in the zone is a net negative
3. Salazar would be better served to pitch low in the zone

 

The first assertion is true: pitching high in the zone leads to a greater aggregate of home runs, which Eno Sarris of FanGraphs, in a Thursday monograph of Chris Young, conceded. Up-in-the-zone pitches are more likely to be fly balls, and fly balls are the singular pitcher-dependent precondition for home runs. So in the sense that Salazar surrenders a great deal of fly balls, it should be expected that he would also surrender a great deal of home runs. Of course, given that HR/FB ratio is typically a widely-fluctuating stat (such that even former Cy Young winner David Price has an exceptionally high HR/FB ratio), and given that Salazar’s has been substantially above-average in both 2013 and 2014, it seems his HR/FB ratio should normalize to league average: this almost universally happens within the body of Major League pitchers. However, one might also argue that this line of reasoning cannot be applied to Salazar, as he is, very literally, not a major league pitcher at this point. This particular point seems more wordplay than actual argumentation, but the question of whether Salazar’s HR/FB ratio can be evaluated relative to the body of MLB pitchers is not central to the evaluation of whether pitches up in the zone are inherently harmful.

While it is true that Salazar allows a great deal of home runs, the subsequent assertion is that his pitching up in the zone is a net negative; it’s unclear that this is true. It results in a higher rate of fly balls, yes, but fly ball pitching is not inherently harmful – Clayton Kershaw, for instance, derives his perennially low BABIP directly from being a flyballer. Even less transcendent figures like Josh Tomlin, Jeremy Guthrie, Jered Weaver, and Marco Estrada have parlayed high career fly ball rates into depressed career BABIPs. League-wide, fly balls have a lower BABIP on average than ground balls. Salazar, as a fly-ball pitcher, should therefore have some manifest BABIP advantage on this front.

Instead of being an asset, however, Salazar’s .369 BABIP in 2014 has worked to his unequivocal detriment, being 4th-highest among pitchers with 40+ IP. Typically, BABIP is a function of luck, defense, and batted ball profile; his profile, as noted, lends itself to low BABIPs, and his Line Drive rate, the prime mover in high BABIPs, is merely league average – in other words, Salazar’s batted ball mix suggested that he should be running an average-or-below BABIP. Add to this the fact that the Cleveland defense has been the worst in the league – by a fairly substantial margin – it appears as though Cleveland’s ‘Leading in Absolute Value‘ defense, coupled with poor luck, is entirely sufficient to explain the extremely inflated BABIP. In terms of BABIP, at least, it’s not clear how working high in the zone, functioning as a flyball pitcher, would have harmed him.

Yet Salazar’s signature has always been the swinging strike, and working up in the zone has all been an avenue to the whiff. The following heat map, not from FanGraphs but rather from Brooks Baseball, documents his 2014 whiff rates by zone (N.B.: chart is from the catcher’s point of view).

Salazar Whiff Rate (Credit: BrooksBaseball.net)

There are risks to operating high in the zone – the aforementioned home run risks, for instance -but the swinging strike rate Salazar has generated high in the zone even given the substantial decline in whiffs from 2013 makes operating high in the zone a proposition with substantial reward opportunities. A swinging strike rate over 20% would be regarded as very good in any sector, but when that sector was on the plate already, it transcends merely good to terrifying.

Bearing in mind, however, the manifold risks of working high in the zone, the question is then not whether it has benefits or risks, since it clearly has both; rather, the question is whether the benefits outweigh the risks. Returning to the chart at the top, one determined the net benefit of pitch locations as a whole; in Salazar’s case, one can do the same, but sample size limits both certainty and precision. What follows are presentations of the data in four separate charts, which, from left to right, are a 5×5 batting area, followed by a 10×10 with radius-2 smoothing, a 10×10 with radius-1 smoothing, and 10×10 with no smoothing. The smoothing regards the sectors not as distinct granula, but as overlapping regions, with the center of each region being the most heavily-weighted. Blue means that a pitcher fared well in that sector; red means that a pitcher fared poorly in that sector – and this chart, unlike the above chart, is from the pitcher’s perspective.

Runs Above Average per 100 pitches (Credit: FanGraphs)

One can use any of these charts to draw their conclusions – the no-smoothing chart is the most precise, and the 5×5 chart helps compensate for the small sample size. however, given the small sample size, the 2-radius smooth, 10×10 chart is likely the best evaluative method, as it recognizes the ambiguity inherent in the small sample size and compensates for it in a way that even the 5×5 cannot.

Inverting the Strike Zone

If there exist takeaways from a sample size as small as Salazar’s 2014, those takeaways are threefold.

First, Salazar had tremendous run suppression in those regions outside the zone. There’s no uniform explanation for this – below the zone, there are high whiff rates; above the zone, he induced both whiffs and, high and outside to one LHH, a ground-out. It’s likely more efficient to simply hand-wave it as a fluke of sample-size, if one is unwilling to credit it to Salazar as a function of his bat-missing.

Second, Salazar working high in the zone has wildly divergent results, as one might expect from a high-whiff flyballer. On the whole, however, operating high in the zone appears to not be obviously detrimental. Perhaps this, too, is as the result of small sample size, yet to concede that half of the strike zone is too small a sample size is to concede that no conclusions can possibly be drawn about Danny Salazar’s directional pitching. However reasonable that concession may be, Salazar was sent down to AAA, according to Francona, in large part because he was unable to keep his fastball down. Regardless of whether one can responsibly draw conclusions about directional pitching, the organization concluded that his fastball up was a problem and made decisions on the assumption that conclusions could be made. Yet it seems, on the net, that Salazar’s pitching was league-average-or-better when working up in the zone, even in spite of the home run problems. This chart contradicts every intuition, every frustrated ‘this one’s gone too?‘ moment one had when Salazar pitched – but the runs-above-average for each segment of the plate suggests that it wasn’t high where Salazar’s troubles arose.

Finally, perhaps most perturbing, is that while Salazar did not have apparent trouble high, he did have trouble low in the zone, with a positive RAA for much of the low portion of the zone. This offense prevention was not as a result of any inability to garner ground-outs on low pitches, however. The chart below demonstrates the rate of ground balls per pitch:

Salazar Ground Ball Rate per Pitch (Credit: FanGraphs)

Those regions where Salazar’s offense prevention was at its weakest were those regions that most effectively induced grounders. A causative element cannot be proven between the two, nor is it clear what such a causative element might entail – but this relationship is wholly puzzling.

In reality, while the team and an overwhelming number of opinions suggest that it was Salazar’s tendency to pitch high that caused him problems, it appears to have been precisely the opposite: namely, Salazar was on the whole fine when he pitched high in the zone, but it was when he threw low in the zone or below the zone that he got into trouble.

The Path Forward

None of this is to say that Danny Salazar is a finished product or even close – his 9.2% walk rate ranked 32nd-worst of the 132 pitchers with 40+ IP. Moreover, there were very serious concerns about his mechanics, with the revered Doug Thorburn of Baseball Prospectus stating, to paraphrase, that Salazar’s core strength appears to have substantially declined. Salazar’s previously plus control has deteriorated, and this has harmed his ability to hit the zone as well as, potentially, his ability to get batters to chase. There are a great many reasons to be concerned about Danny Salazar.

However, Salazar’s inability to get the ball down is not among them. Danny Salazar will never be the ideal groundball pitcher – he may never be a merely good groundball pitcher. However, the question has been only rarely asked – if at all – whether Salazar can work high in the zone as a flyball pitcher. Given the rise to his fastball, it seems very possible. Salazar is by no means an extreme case in terms of fastball rise, but his 9.8 inch rise (relative to a normal gravitational parabola) is approximately an inch greater than the average four-seam. The following two charts, moreover, document the correlation between fastball rise and fly ball/ground ball rates since PITCH f/x was made public. The correlation is not nearly as strong as Whiff Rate to K rate or Walks to On-Base Percentage, but even at first glance, it’s unmistakeable:

Fastball Break vs. FB% (Credit: FanGraphs)

Fastball Break vs. GB% (Credit: FanGraphs)

Fastball rise is correlated to flyball rate positively and to groundball rate negatively. Salazar has worked high in the zone before, and he has made it work. As a result of this pitching style, Salazar shocked the majors in 2013, putting up ace-like strikeout numbers, very good walk numbers, and excellent run prevention in front of a defense that was, even in 2013, one of the worst in the league. There are absolutely aspects that need tweaking, and perhaps the mechanical issues run too deep to fix in a few weeks or even months.

Despite this, one hopes that the organization recognizes Salazar’s ability, his success working up in the zone, and does not attempt to force him to become something he clearly is not. In 2013, Salazar’s split-change was an excellent whiff pitch – not a groundball pitch. The same applies to his fastball. Salazar’s arsenal is a strikeout-first arsenal. If Callaway’s strength is indeed working to the pitcher’s strength, one hopes that he recognizes this fact.

Advertisement