Jakob Junis Analysis Report
Kitae Kim
Overview
In this report, we investigated the performance change of Jacob Junis, MLB pitcher at Kansas City Royals. Unlike the 2019 season, his slider was not showing good performance in 2020 and we attempted to find the reason using a numerical approach. In the end, we suggested a possible short-term and long-term solution.
We obtained MLB data from https://baseballsavant.mlb.com/
Pitch Analysis 2019 vs 2020
The most significant difference between Junis’ 2019 and 2020 performance is that his strike swing percentage dropped from 17.3% to 10.5%. We decided to investigate this change by performing analysis on his pitches.
Pitch Usage 2019 vs 2020
In 2020, Junis decided to drop his curveball, which did not have promising performance and instead increased the slider’s usage, which showed excellent performance in 2019. Slider usage in 2020 is 41.1%.; nearly equivalent to the sum of slider and curveball usage in 2019, which was 29.7% and 14.1%, respectively.
Pitch usage in 2 strike 2019 vs 2020
Primarily, Junis used his slider in 2 strike situations. He even increased slider usage in 2 strike situations from 2019 to 2020.
Slider Performance in 2 strike 2019 vs 2020
However, the performance of his sliders in two-strike situations declined significantly from 2019 to 2020. The swinging strike percentage dropped from 19.5% to 10.8% and the called strike percentage dropped from 6.7% from 2.7%. By taking it a glance at it, the leading cause of Junis’ performance change is due to the lower success of his slider, especially in two-strike situations.
Slider called strike + swing strike Percentage
| strikes | Total Pitch | Swing | Called Strike | Strike Percentage |
|---|---|---|---|---|
| 0 | 59 | 5 | 23 | 0.4745763 |
| 1 | 108 | 19 | 21 | 0.3703704 |
| 2 | 231 | 42 | 18 | 0.2597403 |
| strikes | Total Pitch | Swing | Called Strike | Strike Percentage |
|---|---|---|---|---|
| 0 | 136 | 24 | 43 | 0.4926471 |
| 1 | 134 | 18 | 19 | 0.2761194 |
| 2 | 199 | 42 | 11 | 0.2663317 |
| strikes | Total Pitch | Swing | Called Strike | Strike Percentage |
|---|---|---|---|---|
| 0 | 14 | 1 | 3 | 0.2857143 |
| 1 | 10 | 0 | 2 | 0.2000000 |
| 2 | 15 | 3 | 0 | 0.2000000 |
| strikes | Total Pitch | Swing | Called Strike | Strike Percentage |
|---|---|---|---|---|
| 0 | 11 | 3 | 3 | 0.5454545 |
| 1 | 14 | 1 | 4 | 0.3571429 |
| 2 | 22 | 1 | 1 | 0.0909091 |
Even though the sample size (total number of sliders pitched) in 2020 is small, we still can make some inferences based on what is presented. Overall, called strike + swing strike percentage of slider dropped from 2019 to 2020. Especially called strike + swing strike percentage in 2 strike situations against right-handed hitters dropped by one third. Considering the slider being the go-to pitch in 2 strike situations, it is a huge negative to Junis’ performance. We investigated the slider further to solve this issue.
Slider Movement 2019 vs 2020
The movement of sliders did not change much. For the reference, green dots represent strike/strike-swings while red dots represent hit-into-play. It is not shown in this graph; however, blue dots represent the ball-called.
Slider Velocity & Spin Rate 2019 vs 2020
From plots, it is clearly shown that the speed and spin rate of sliders dropped significantly from 2019 to 2020.
Velocity per Pitch Type
| pitch_name | mean | median |
|---|---|---|
| 4-Seam Fastball | 91.68069 | 91.7 |
| Changeup | 84.63376 | 84.7 |
| Curveball | 79.34623 | 79.5 |
| Sinker | 91.04099 | 91.1 |
| Slider | 82.42999 | 82.3 |
| pitch_name | mean | median |
|---|---|---|
| 4-Seam Fastball | 90.86081 | 90.65 |
| Changeup | 84.11818 | 84.30 |
| Sinker | 90.50263 | 90.60 |
| Slider | 79.72907 | 79.75 |
The velocity of other pitch types did not change much; meanwhile, the slider’s average speed dropped by about 3mph. Interestingly, Junis’s 2020 slider is as fast as his 2019 curveball, which he wanted to remove.
Slider and Curveball 2019 & Slider 2020
| pitch_name | game_year | Average Speed | Average Spin Rate |
|---|---|---|---|
| Curveball | 2019 | 79.34324 | 2421.054 |
| Slider | 2019 | 82.42809 | 2522.114 |
| Slider | 2020 | 79.72907 | 2408.314 |
Once we noticed that Junis’ 2020 slider is as fast as his 2019 curveball, we compared his 2020 slider with the 2019 curveball and slider. The average speed and the spin rate of Junis’ 2020 slider indicated that the 2020 slider shows more resemblance to the 2019 curveball than the 2019 slider.
Slider Pitch Location 2019 vs 2020
Slider Pitch Location After 2 strike 2019 vs 2020
The location plots showed that Junis throws his slider to the bottom right corner of the strike-zone to get strike-swings. It was a good strategy in 2019 as many green dots in that location tell us. However, he is not getting nearly as many strike-swings in 2020 using the same strategy. In 2020, blue dots(ball called) are more dominant at the bottom right corner. Hitters are not falling for his slider as much as they used to since the 2020 slider is different from the 2019 slider.
Swing Strike Percentage
| Velocity Range | Total Pitch | Swing | Swing Strike Percentage |
|---|---|---|---|
| (80,81] | 49 | 4 | 0.0816327 |
| (81,82] | 112 | 12 | 0.1071429 |
| (82,83] | 130 | 29 | 0.2230769 |
| (83,84] | 70 | 14 | 0.2000000 |
| (84,85] | 34 | 6 | 0.1764706 |
| (85,86] | 3 | 1 | 0.3333333 |
| Velocity Range | Total Pitch | Swing | Swing Strike Percentage |
|---|---|---|---|
| (79,80] | 3 | 0 | 0.0000000 |
| (80,81] | 43 | 7 | 0.1627907 |
| (81,82] | 131 | 17 | 0.1297710 |
| (82,83] | 148 | 18 | 0.1216216 |
| (83,84] | 102 | 32 | 0.3137255 |
| (84,85] | 35 | 9 | 0.2571429 |
| (85,86] | 7 | 1 | 0.1428571 |
| Velocity Range | Total Pitch | Swing | Swing Strike Percentage |
|---|---|---|---|
| (76,77] | 2 | 0 | 0.0000000 |
| (77,78] | 6 | 0 | 0.0000000 |
| (78,79] | 4 | 1 | 0.2500000 |
| (79,80] | 13 | 0 | 0.0000000 |
| (80,81] | 9 | 1 | 0.1111111 |
| (81,82] | 2 | 0 | 0.0000000 |
| (82,83] | 3 | 2 | 0.6666667 |
| Velocity Range | Total Pitch | Swing | Swing Strike Percentage |
|---|---|---|---|
| (77,78] | 7 | 0 | 0.0000000 |
| (78,79] | 6 | 1 | 0.1666667 |
| (79,80] | 11 | 2 | 0.1818182 |
| (80,81] | 16 | 2 | 0.1250000 |
| (81,82] | 2 | 0 | 0.0000000 |
| (82,83] | 4 | 0 | 0.0000000 |
| (83,84] | 1 | 0 | 0.0000000 |
Junis’ 2020 slow sliders (which behave like 2019 curveball) do not get strike swings anymore. His sliders showed much better performance when in [82, 86] mph range; however, most of his 2020 sliders fall under [76,81] mph range, which has poor performance.
From the analysis, we noticed that the 2020 slider is not getting as many called-strikes or strike-swings as his 2019 slider and one of the causes is because his 2020 slider behaves like his 2019 curveball (lower mph and spin rate). For this reason, right-hand hitters are not falling for bottom-right slider in 2 strike situations, which lowered his slider’s overall performance.
To solve this issue, Junis should have a new pitch design to fix his slider or add a different pitch type. However, this process takes a long time, which he will have to work on during the spring before the next season start. For now, we will look at if there is a quick fix in his strategy that Junis can use for the rest of the season.
Pitch Before/After Slider 2019
Pitch Before/After Slider 2020
Both in 2019 and 2020, Junis tend to throw slider back-to-back. About 43% of pitches he threw before and after slider was a slider.
Slider After FB vs Not after FB 2019
In 2019, his slider showed better performance when it was paired with non-fastball (higher strike-swing and called_strike percentage). It explains why Junis tends to throw his slider back-to-back. The attached plots show his slider performance per location per batter hand.
Slider After FB vs Not after FB 2020
However, in 2020, his slider showed better performance when paired with the fastball (higher called strike percentage) which is the opposite of what happened in 2019. As a quick strategy fix, he could pair up his slider with a fastball to increase his called strike percentage even though it won’t fix the bigger problem, low swing-strike percentage. The Swing-strike percentage dropped to 10% in 2020, regardless of what he paired his slider with.
Conclusion
Based on the analysis, we learned that Junis’ 2020 slider is not getting as many called-strikes or strike-swings as his 2019 slider. This underperformance seems to be since the 2020 slider is not like the 2019 slider; instead, it is more similar to the 2019 curveball, which he wanted to remove due to poor performance in 2019.
Ideally, Junis should change his slider or add a different pitch as a long-term solution. However, it is not simple to work on pitch design in the middle of the season. For this reason, the quick strategy change is recommended and as a new strategy, data suggest that he should pair up his slider with fastball more often.
This report’s primary focus was to analyze what happened and how to fix it rather than why it happened. It would be beneficial to look at what caused this change in a future analysis since it might give us a better understanding of how to bring a good performance back. Also, most of the report’s approaches are based on numbers and it would be helpful to add more baseball perspective in the future since there are many other factors we need to consider in baseball strategy.