The Ever Expanding Strikeouts Game

Spin Rate vs Swinging Strike

Strike outs have been a recurrent topic the last few years at Major League Baseball. Since the introduction of statcast in 2015 the total strike out numbers skyrocket dramatically year after year setting all time records.

The new era of analytics has condensed the “old school” two strike approach to just a handful of players, but hey! who cares, is it not better to see your favourite player hit a +450 ft homer even though he strikes out the next 9 at bats?

Last year MLB teams combined for a total of 6,776 home runs, setting an all time record, but that came to the expense of 42,823 strikeouts, also an all time record. Using simple maths, for each home run we enjoyed, we had to endure 16 strikeouts, is that a lot? is that the type of game that we want to watch?

New technologies have provided measurements like launch angle and exit velocity who have completely re-shaped the hitters approach, but there are other measures not visible to the naked eye, such as the spin rate which may hold the key to quantifying pitch movement, and together with release speed can become the two primary attributes to evaluate pitchers.

Generally speaking we can say that there are three main skills that a pitcher can use to be effective: control, movement and velocity, although pitchers do not need to have all of them to be elite. Many times we have seen that pitchers can be successful with just a couple of these skills.

The goal of this analysis is to examine if there is any relationship between spin rate and pitch velocity and whether or not they have a real effect on swings and miss. Does a higher spin rate correlate to more swinging strikes?

For this study I have used statcast data from the 2019 season. I compiled all pitches thrown and removed all the pitches with less than 1000 instances. I also combined pitch types that behave similarly, like 2-Seam Fastballs and Sinkers, and Split Finger and Forkball.

To begin with, I have plotted the average spin rate and average velocity for all pitches. As you can see there are too many data points on this plot to make specific conclusions, although it is possible to make broad inferences, like the 4-Seam Fastball and Sinker on average have greater velocity or that curveballs possess on average a greater spin rate.

Since the previous plot did not provide any specific conclusions, I wanted to add some more granularity, so I have grouped all pitch types into three categories: Fastballs, Breaking Balls and Offspeed Pitches. And I have plotted each pitch type individually.

But even when broken out by pitch type it is hard to see any type of linear trend between velocity vs. spin rate. Then I ran a regression analysis for each pitch type with very low r2 scores.

Although this should not come as a total surprise, at the end all pitch types are somehow different depending on the pitcher as they all have different repertoires, their delivery angle and release point varies from one to another. In other words, a Curveball from Shane Bieber looks very different to the one thrown by Gerrit Cole.

With this in mind, I focused on the relationship between the spin rate for swinging strikes. In the table below I’ve listed out each pitch type’s sample size, mean and standard deviation for both spin rate and release speed.

Then I splitted up all the outcomes that each pitch type produced between in play or swinging strike and I performed a hypothesis testing using scipy ttest_ind to calculate the T-test for the means of the two independent samples.

Assuming the null hypothesis:

The spin rate of balls in play and the spin rates of swinging strike are equal.

The resulting values were quite shocking with an extremely low p-value and small confidence intervals for both swinging strike and in play outcomes.

  • p-value=3.44 e-167
  • Swinging strike confidence interval (2264.58, 2269.41)
  • In play confidence interval (2223.50, 2227.12)

To further the analysis, I plotted the average spin rate by outcome and pitch type as shown above. At a quick glance, we can see that almost all pitch types have slightly higher spin rate for swinging strikes than the pitches that did not, the most notorious being the 4-Seam Fastball with 40 rpm higher and the curveball with 38 rpm.

With the results obtained, I am confident to say that the null hypothesis can be rejected. As there are enough evidence for my alternative hypothesis that the balls with a higher spin rate produce more swinging strikes.

This concludes the spin rate and swing and a miss analysis. As you can see, spin rate plays a role on inducing swinging strikes, now, how big is that role? well, that is difficult to quantify as there are many other parameters that get into play when trying to define the reasons behind the swings and a miss. Without going any further, the focus that hitters have put on the launch angle and teams not seeing strikeouts as something worrying when assessing the players performance, has contributed to an even higher degree to the ever increasing pile of strikeouts.

Data Analyst with business background and over 7 years of experience on project management. Specialized in sports analytics.