### Introduction

It wasn’t that long ago that folks focused about strikeout rates and walk rates. Since we now have much more data on pitch locations and outcomes, we hear more about swing and miss rates. Here we’ll explore swing and miss rates for 2018 hitters, see how these rates vary by pitch type and count, and use heat maps to explore patterns of these rates about the zone. In general, we define **a swing rate** as the percentage of pitches that are swung at, and a **miss rate** is the percentage of swinging strikes among those pitches that are swung at. For point of reference, the overall swing rate in 2018 was 46% and the overall miss rate was 23%.

### Swing and Miss Rates of Regular Hitters

To start, I collect and graph the swing and miss rates for all 2018 hitters who have seen at least 1000 pitches. There is a slight positive trend in the graph indicating that players who have high swing rates tend to have high miss rates. I label some interesting players — Jorge Alfaro had an unusually high swing rate and high miss rate. Joey Gallo, in contrast, had a moderate swing rate but a high miss rate. Mookie Betts, the likely MVP in the American League in 2018, is very low on both rates. Three players, David Fletcher, Joe Panik, and Michael Brantley had miss rates under 10 percent.

### Swing and Miss Rates of 2018 World Series Players

Since we are watching the 2018 World Series, it is interesting to plot the swing and miss rates for the Dodgers (blue) and Red Sox (red). One interesting thing is that the Dodgers tend to have low swing and high miss rates, and the Red Sox tend to have high swing rates and low miss rates. In other words, the Red Sox appear to be more aggressive (high swing rates), and more successful in their swings (low miss rates).

### Swing and Miss Rates for Different Pitch Types

How do these rates depend on the pitch type? Here I graph the rates where each plotted point corresponds to a particular type of pitch. Batters tend to swing at splitters (FS), change ups (CH) and cutters (FC), and are less likely to swing at curve balls (CU). With respect to miss rates, batters are less likely to miss sinkers (SI) and two-seam fastballs (FT) and more likely to miss splitters (FS), sliders (SL), knuckle curves (KC) and curve balls (CU).

### Swing and Miss Rates for Different Counts

How does the count impact swinging and missing? Here I plot the rates for each count where I color the label by the type of count (favoring the pitcher, favoring the batter, or neutral). Generally, batters are more likely to swing at deeper counts — most likely will swing on a full count (3-2). Batters are most likely to miss the pitch on a 0-2 count and less likely on a 3-0 count. Generally the pitcher counts (0-2, 1-2, 2-2) are associated with high swing rates and high miss rates. This makes sense since the pitcher has the advantage and can strike the batter out with a pitch out of the zone.

### Swing and Miss Rates and Pitch Location

How does a batter’s propensity to swing or miss depend on the location of the pitch? I fit a generalized additive model where the outcome is 1 (swing) or 0 (no swing) and the logit of the probability of a swing is a smooth function of the pitch location. Here is a heat map of the fitted probability of swinging. This graph isn’t that exciting — it just says that a batter is most likely to swing in the zone.

I fit a second model where I only consider the pitches where the batter swings and the response is 1 is the batter misses the pitch and 0 otherwise. I fit a GAM where the logit of the probability of miss depends smoothly on the location. This graph is more interesting — it is common to miss low and high pitches (red area), and it is uncommon to miss pitches (blue) in a wide area across the zone.

I would suspect that the miss region would depend on the batting side. So I fit this model twice — once to the batters on the left side and one to the batters on the right side. The corresponding heat maps are shown below. The miss regions (red) look similar to me, but I notice differences in the not-miss regions (blue). Generally it can be said that batters are likely to make contact with balls on the inside of the zone and less likely to make contact with balls on the outside of the zone.

### Wrapping Up

- Although I have shown general tendencies of swing and miss rate, I would think that teams would be very interested in understanding these rates for individual batters. For example, here are heat maps for the probability of missing for Mookie Betts and Max Muncy. Betts’s sweet spot (in terms of not missing the pitch) is large, while Muncy can miss pitches in the outer region of the zone.

- Also I would think a team would be interested in these missing and swinging rates for individual batters against different types of pitches, different count situations (pitcher and hitter counts), etc. Basically, one is interested in learning about all of the possible variables that can contribute to a batter missing the pitch.

### Data and Code

All of the data was collected from the baseballr package — I have a data frame containing information about all 721,188 pitches in the 2018 season. On my GithubGist site, I show all of the R code that I use for this particular blog post.

### Related Posts

I’ve written previous posts on swing and miss rates.

- A New “Swing and Miss” Stat proposes a different way of understanding miss rates for pitchers.
- WS Games 3 & 4 — Swing and Miss — I explore this type of data for a couple of World Series games in 2016.
- World Series Special – Swing and Miss Rates — I use 50% contour plots to show the likely swing regions in the zone for special players in the 2017 World Series.