Exploring three seasons of Jake Arrieta's pitches

arrieta.R

# Read in data and a ggplot2 theme for the title

library(tidyverse)
sc15 <- read_csv("statcast2015.csv")
sc16 <- read_csv("statcast2016.csv")
sc17 <- read_csv("statcast2017.csv")

TH <- theme(
  plot.title = element_text(
    colour = "red",
    size = 16,
    hjust = 0.5,
    vjust = 0.8,
    angle = 0
  )
)

# obtain data for Arrieta
# library(baseballr)
# playerid_lookup(last_name = "Arrieta")
# id is 453562

j15 <- filter(sc15, pitcher == "453562")
j16 <- filter(sc16, pitcher == "453562")
j17 <- filter(sc17, pitcher == "453562")

jake <- rbind(data.frame(j15, Season = 2015),
              data.frame(j16, Season = 2016),
              data.frame(j17, Season = 2017))

# what pitches did Jake throw each season?

library(cowplot)
ggplot(filter(jake, pitch_type %in%
        c("CH", "CU", "FF", "SI", "SL")), 
        aes(pitch_type,
        y = (..count..) / sum(..count..))) +
  geom_bar() +
  facet_wrap(~ Season, ncol = 1) +
  ggtitle("Frequencies of Pitch Types") + TH

# boxplots of release speeds

filter(jake, pitch_type %in%
         c("CH", "CU", "FF", "SI", "SL")) %>% 
  ggplot(aes(pitch_type, release_speed, 
             color=factor(Season))) +
  geom_boxplot() + coord_flip() +
  ylim(70, 100) + TH + 
  ggtitle("Release S
          peeds for Each Pitch Type") +
  draw_image("jake.jpg", x = 3, y = 75, 
             scale=15) 

# horizontal breaks

p1 <- filter(jake, pitch_type %in%
         c("CH", "CU", "FF", "SI", "SL")) %>% 
  ggplot(aes(pitch_type, pfx_x, 
             color=factor(Season))) +
  geom_boxplot() + coord_flip() + TH +
  ggtitle("Horizontal Break for Each Pitch Type")

# vertical breaks

p2 <- filter(jake, pitch_type %in%
         c("CH", "CU", "FF", "SI", "SL")) %>% 
  ggplot(aes(pitch_type, pfx_z, 
             color=factor(Season))) +
  geom_boxplot() + coord_flip() + TH +
  ggtitle("Vertical Break for Each Pitch Type")

plot_grid(p1, p2, ncol=1)

# spin rates

filter(jake, pitch_type %in%
         c("CH", "CU", "FF", "SI", "SL")) %>% 
  ggplot(aes(pitch_type, release_spin_rate, 
             color=factor(Season))) +
  geom_boxplot() + coord_flip() 

####### swing and miss rates

jake %>%
  mutate(swing = ifelse(description %in%
                c("foul", "foul_bunt",
                  "foul_tip", "hit_into_play",
                  "hit_into_play_no_out",
                  "hit_into_play_score",
                  "missed_bunt", "swinging_strike",
                  "swinging_strike_blocked"), 
                1, 0),
         miss = ifelse(description %in%
                 c("missed_bunt", "swinging_strike",
                  "swinging_strike_blocked"), 
                       1, 0)) -> jake

# find number of swings and misses for each season

jake %>% group_by(pitch_type) %>%
  summarize(S = sum(swing),
            M = sum(miss))

# plot the miss rates

filter(jake, swing == 1,
       pitch_type %in%
         c("CH", "CU", "FF", "SI", "SL")) %>% 
  group_by(Season, pitch_type) %>%
  summarize(Miss_Rate = 
              mean(miss, na.rm = TRUE)) %>%
  ggplot(aes(pitch_type, Miss_Rate, 
             color=factor(Season))) +
  geom_point(size=3) + coord_flip() + TH +
  ggtitle("Miss Rates for Each Pitch Type") 

# exit velocities

filter(jake, swing == 1, miss == 0,
       pitch_type %in%
         c("CH", "CU", "FF", "SI", "SL")) %>% 
  group_by(Season, pitch_type) %>%
  ggplot(aes(pitch_type, launch_speed, 
             color=factor(Season))) +
  geom_boxplot() + coord_flip() + TH +
  ggtitle("Exit Velocities for Each Pitch Type")

# launch angles

filter(jake, swing == 1, miss == 0,
       pitch_type %in%
         c("CH", "CU", "FF", "SI", "SL")) %>% 
  group_by(Season, pitch_type) %>%
  ggplot(aes(pitch_type, launch_angle, 
             color=factor(Season))) +
  geom_boxplot() + coord_flip() + TH +
  ggtitle("Launch Angles for Each Pitch Type")

# hr rates

jake %>%
  mutate(HR = ifelse(events == "home_run",
                     1, 0)) -> jake

filter(jake, type == "X",
       pitch_type %in%
         c("CH", "CU", "FF", "SI", "SL")) %>% 
  group_by(Season, pitch_type) %>%
  summarize(N = n(),
            HR_count = sum(HR),
            HR_Rate = mean(HR)) %>% 
  ggplot(aes(pitch_type, HR_Rate, 
             color=factor(Season),
             label = HR_count)) +
  ggtitle("HR Rates by Pitch Type and Season") +
  geom_point(size=3) + coord_flip() + TH +
    geom_label()