Seeking Strikeouts, Pitchers Have Left The Zone

If you are one to find yourself reading obscure baseball blogs, such as this one, you’re likely aware that strikeouts are on the rise in MLB. Pitcher value has increasingly been tied to strikeout percentages, while hitters have seemingly responded by embracing the three true outcomes; these themes have been touched on here before. There are a handful of narratives around this shift in strikeouts, from teams increasingly seeking pitchers with swing-and-miss pitch characteristics to pitchers increasingly being taught to develop, or emphasize, those characteristics in order to be sought after. There is an underlying “chicken and the egg” theme here. While pitcher velocities are rising and spin rates are tracked religiously, pitch locations haven’t received quite as much fanfare (despite high-spin four-seams more often finding the upper third of the strike zone).

What follows is a visual breakdown of trends relating to pitch locations, specifically distinguishing between whether or not pitches are inside the strike zone. To create some context, the table below offers average pitch location/plate discipline metrics for qualifying pitchers, via FanGraphs. Sparklines have been appended to the bottom of each column to generally track how those metrics have shifted since 2002 (the first year this data was available).

This table confirms some general assumptions. For one, the relationship between In-Zone % (third from right) and Swinging Strikes (rightmost) is negative, and fairly strong. A pitchers’ contact percentage, as some might expect, also correlates negatively with the percentage of pitches in the strike zone. Some of this may have come as a bit of a surprise though; as pitches have left the strike zone more, on average, swing rates (third from left) have conversely risen. Despite fewer pitches being thrown in the strike zone, first pitch strikes (avg F-Strike) are increasingly common.

Pitchers at the extreme ends of In-Zone % have not been exempt from this phenomenon either. Below is a table of each season’s qualifying pitcher who threw the lowest percentage of his pitches in the strike zone, dating back to 2002. Another sparkline has been appended to map these percentages over time.

In the 2010s, no pitcher on the above list pitched in the zone more than 40% of the time. In 2018, Patrick Corbin pitched in the zone only just barely two-thirds of the time, and was rewarded with a 11.1 K/9 figure and a nine-figure contract. In 2019, Luis Castillo pitched outside the zone with similar frequency, was a deserving All Star, and struck out 10.7 batters per nine. This list does not have an overwhelming “serial offender,” but Kenny Rogers and Francisco Liriano each, more than once, led qualifying pitchers in lowest zone percentages. Next, the list sorted backward, to feature those pitchers who filled the zone most often.

Gone, seemingly, are the days of the Paul Byrds and Cliff Lees. Those two pitchers combined to top this list 8 times in the past 18 seasons. These two tables really drive home how much has changed: Lucas Giolito, who lived in the zone the most of all qualifying pitchers in 2019, would have pitched in it the very least in 2002 (where leader Matt Clement threw strikes 48.8% of the time). Giolito had a fantastic breakout campaign in 2019, and struck out 11.6 batters per nine along the way, but one has to wonder how many more swings and misses he might have induced pitching outside of the zone even a bit closer to the league average rate.

Another important consideration when evaluating pitch locations is the length of games, which has been increasingly scrutinized and grappled with. As the table below indicates, there is an obvious negative relationship between In-Zone % and the length of games.

To get a bit more granular, below are two scatterplot matrices. The first captures the average value among each season’s qualifying cohort, while the second includes all qualifying pitcher data from 2002-2019. Scatterplot matrices display the relationship between more than two variables at once. The distribution of each metric is given in the diagonal line from the upper left to lower right. Meanwhile, correlations between two metrics are given in the upper right and scatterplots visualize how those metrics interact in the lower left.

While the matrix above is interesting and, hopefully, a bit intuitive, the matrix below is more robust in that each pitcher-season represents its own data point, as opposed to each season’s average.

The matrix above presents some interesting relationships. With zone percentages in mind, for instance, we can see that there is no relationship between how much a pitcher throws in the zone and how much contact hitters are making on pitches in the zone (Z-Contact). The more pitchers work in the zone though, the more hitters are swinging (0.31 correlation).

In summation, below is a scatterplot of each pitcher’s In-Zone % for each season they qualified.

This final scatterplot should underscore the significant and ongoing trend of pitchers vacating the strike zone in order to adapt to today’s iteration of baseball. The ability to generate swings-and-missed is valued higher than ever and thus has, at least in part, incentivized this trend. What seems to compound this development is that hitters continue to swing (and in some cases swing more) in search of the extra base hits that will salvage stat lines that feature seemingly endemic swing-and-miss levels. Players will adapt, as they always do, but for now there does not appear to be an obvious balancing force in sight.