Z-Contact% as a Function of Strictly a Pitcher’s Fastball

A couple of weeks ago, I investigated Justin Verlander’s resurgence. I found reasons to validate his hot streak but turned up additional question marks along the way.

One of them was his zone contact rate (Z-Contact%). At 79.7 percent, it would have been the second-lowest of his career by several percentage points (despite not performing “at peak”). However, I realize now, unfortunately, that I must have encountered a glitch in the leaderboards — his Z-Contact% as of August 21 (because the post, despite running the same day as his Aug. 26 start, was published prior to it) was 85.7 percent.

Regardless, it got me thinking what affects a pitcher’s zone contact rate because it correlates very strongly with strikeout rate (R-squared = .594). User DoubleJ speculated about the metric via comment on one of last week’s posts:

how do you feel about using Contact% vs ZContact% as an analytical tool? I like the [latter] as an intuitive proxy for factors like late break, ability to work the black, unpredictable pitch sequencing, etc.

I agree. Unfortunately, I’m not aware of metrics that capture late break, and Baseball Heat Maps’ Edge% leaderboards are only available for the seasons spanning 2008 through 2013. The effects of pitch sequencing can be quantified, but the results are not exclusive of the quality of a pitcher’s arsenal.

I think a pitcher’s fastball can reveal a lot about his Z-Contact% scores, though. A quick glance at the leaderboards (50-plus innings) shows Aroldis Chapman, Craig Kimbrel and Dellin Betances — aka dudes with some of the fastest fastballs in the game. Granted, they throw in short spurts; their usage patterns likely explain why they, along with 22 other relievers, own a majority of the 30 best zone contact rates.

Still, it seems fastball velocity plays a prominent role. I imagine if that’s true, then its movement and usage frequency does, too. I also think movement is codependent. It’s one thing for a fastball to run, and it’s another for it to rise (or sink). But to do so simultaneously is, I hypothesize, multiplicatively more effective.

It’s probable that these relationships differ among starters and relievers, so I concocted a multiple regression equation that samples all relief pitchers (as classified by FanGraphs) who threw at least 40 innings and another that samples all starting pitchers (ditto) who threw at least 100 innings in seasons spanning 2010 through 2014. The data includes fastball Z-Contact% (Z); horizontal movement, in inches (X); vertical movement, in inches (Z); velocity, in mph(V); and usage frequency (Pct). I also include a term that captures the interaction between horizontal and vertical movement (Mov).

Relievers

Strikeout and zone contact rates remain just as strongly correlated when the sample is stripped down to only relievers (R-squared = .576).

The equation for relief pitchers looks as such:

Z-Contact% = 1.19540 — .02416*Pct — .00310*V — .00315*X — .00610*Z + .00051*Mov
Adjusted R-squared: .146

The model is not exceptionally powerful, especially when compared to other regressions we (or other web sites) have presented. But for a cross-sectional analysis, it still captures a moderately sized correlation between Z-Contact% and fastball components.

All variables are statistically significant at 95-percent confidence except horizontal movement, which is significant at 90 percent.

Starters

Strikeout and zone contact rates remain strongly correlated when the sample is stripped down to only starters (R-squared = .493).

The equation for starting pitchers:

Z-Contact% = 1.0602 — .02712*Pct — .00110*V — .00809*X — .00690*Z + .00076*Mov
Adjusted R-squared: .141

Again, the model is not exceptionally powerful, but it still depicts a moderate correlation between Z-Contact% and fastball components. Horizontal movement is now statistically significant at 95-percent confidence.

Comparisons

The two equations aren’t hugely different, but there are some slight variations. It appears more frequent usage of a fastball is slightly more important for starters than it is for relievers; however, fastball velocity is three times more important for the latter group.

Moreover, horizontal movement also plays a much more important role in determining a starting pitcher’s Z-Contact% than it does a relief pitcher’s. This could be some sort of feeble insight into pitch sequencing and/or the diversification of an arsenal: fastball run may be more effective when a pitcher has additional pitches, each with its own movement, to offer up.

The Most Peculiar Starting Pitcher Outliers Who Aren’t Knuckleballers, for your Entertainment

Max Scherzer and Chris Sale are, like Kenley Jansen (were I to include the tables for relievers), masters at defying expectations.

There’s a pretty clear distinction in quality between the pitchers at the top and bottom of each table. I’d chalk up the remaining differences to things we would already expect, such as a better arsenal (both in quality and quantity of pitches) and pitch sequencing.

Epilogue

It’s wholly unwise to assume one pitch can define a pitcher’s success. But his fastball, in a vacuum, can lend insights into his overall effectiveness within the strike zone — which reflects upon his overall performance — when he throws the pitch for a strike.