Another Implication of Strikeouts

This post aims to make use of a few graphics to briefly consider strikeout rates and their potential impact on walk rates.  In short, does shirking pitching to contact for the sake of coveted strikeout rates lead to proportionately more walks?

Much has been made of the rising number of strikeouts league-wide, and for good reason.  The number of strikeouts has risen every year since 2005 and the early numbers from this March/April indicate this trend will continue into 2019.  

Given this decade-plus uninterrupted stretch of spiking strikeout totals, it only makes sense that there would be a corresponding drop off in balls put into play.  There are multiple implications there.  Longer individual at-bats, changing strategies for playing defense, and struggles with pace of play more broadly are all part of the far-reaching consequences of strikeout proliferation.

Here is a plot using 2018 data that juxtaposes K’s per 9 innings and BB’s per 9, the data for which is courtesy of FanGraphs’ database.  

The darkest, most vertically tilted, line represents a K:BB ratio of 6:1.  As you can see, few pitchers (with 50+ IP) eclipsed this considerable threshold.  The three remaining lines (from relatively vertical to more horizontal) represent a 3:1 K:BB ratio, 2:1, and a 1:1 ratio, respectively.  In 2018, there appears to be a pretty even distribution among the three areas between these four lines.  In short, pretty solid K:BB ratios to match a year of unprecedented K rates.  

That said, the visualization above does not clearly indicate a significant jump in walk rates to go along with more strikeouts; should there be a clear correlation between strikeout and walk rates, those data would generally be running diagonally toward the upper-right corner of your screen.  

For comparison’s sake, it’s worth examining the same rates from another point in time, to cross check whether a correlation appears more clearly in another timeframe, but also to see how rates have changed over time.  I chose similar data from 2010, intending to pit the unprecedented 2018 strikeout rates against a year dubbed “The Year of the Pitcher”, albeit really just a recycled term from the original “Year of the Pitcher” in 1968.  

2010 was a noteworthy year given more than a few incredible individual accomplishments, including 6 no-hitters, and a considerable drop off in runs scored.  Comparing last year’s data to a year in which pitchers were considered particularly dominant seemed like an interesting exercise.

Here is the exact same scatter plot visualized, but populated with 2010 data:

It might not be immediately clear, but this second plot does feature considerably lower strikeout rates overall. 

Pictures are good, but this is a case where numbers can be most helpful.  Despite 2010’s reputation as a pitching dominant year, the median K/9 rate was just 7.025 opposed to 8.704 in 2018.  A median jump of nearly 1.7 strikeouts per 9 across the board for pitchers, using a “pitching dominant” year as a baseline, is quite drastic.

Finally, to address the question posed at the top of this post.  Have pitchers paid for their strikeouts with walks? It does not appear so, at least not dramatically.  In 2010, the median walk rate of pitchers who threw 50+ innings was 3.157; for 2018 pitchers, that median marker was actually lower, with the median BB/9 pitcher of the cohort walking 3.085 batters per 9.

Comparing data from two seasons nearly a decade apart, of course, is not terribly empirical, nor is it the only way to address this question.  The following two plots compare 2010 and 2018 pitchers whose K/9 rates were either in the first or fourth quartile.  Horizontal blue lines have been superimposed in order to highlight the median walk rate for either group.

First, 2010:

Next, 2018:

Using this rather crude measure, it’s clear that there is at least some bump in walk rates for the higher K/9 group in either cohort.

In 2010, high strikeout pitchers walked batters 3.3 times per 9 innings opposed to low strikeout pitchers who walked batters just 3.0 times per 9.  Not a huge difference, but absolutely noteworthy.

In 2018, the median highest quartile strikeout pitcher walked batters 3.24 times per 9 innings whereas the lowest quartile’s median rate was 2.96 walks per 9.  Essentially, the high-strikeout group featured a median jump of ~1.5 K’s per 9 compared to the lower-strikeout group, but at the cost of a (less than proportional) ~0.3 rise in walks per 9.

Pinning down causality here isn’t feasible.  High strikeout pitchers might walk more batters simply due to throwing more pitches per plate appearance, on average, than their more contact-oriented coworkers.  High strikeout pitchers might reach those rates by pitching out of the zone intentionally with greater frequency.  Batters might be more inclined to take pitches against pitchers whose intent is to induce chases.

Regardless, there does appear to be a weak relationship between K/9 and BB/9, but not so drastic as one might have assumed.  Strikeouts are on the rise.  Home runs are on the rise too.  But in a world of FIP and Three True Outcomes, walks, while rising still, lag a bit behind.