R study of called balls and strikes -- uses CalledStrike package

calledstrikework.R

# load some packages

library(baseballr)
library(tidyverse)
library(CalledStrike)
library(gridExtra)

# scrape data for four pitchers

Aaron <- scrape_statcast_savant(start_date = "2018-03-15",
                                     end_date = "2018-10-15",
                                     player_type = "pitcher",
                                     playerid = 605400)

Lucas <- scrape_statcast_savant(start_date = "2018-03-15",
                       end_date = "2018-10-15",
                       player_type = "pitcher",
                       playerid = 608337)

Dylan <- scrape_statcast_savant(start_date = "2018-03-15",
                             end_date = "2018-10-15",
                             player_type = "pitcher",
                             playerid = 605164)

Clayton <- scrape_statcast_savant(start_date = "2018-03-15",
                             end_date = "2018-10-15",
                             player_type = "pitcher",
                             playerid = 477132)

# construct four called strike graphs

Aaron %>%
  setup_called() %>%
  gam_fit() %>%
  cplot(P = c(.5, .9)) +
  ggtitle("2018 Aaron Nola") +
  centertitle() -> plot1

Clayton %>%
  setup_called() %>%
  gam_fit() %>%
  cplot(P = c(.5, .9)) +
  ggtitle("2018 Clayton Kershaw") +
  centertitle() -> plot2

Lucas %>%
  setup_called() %>%
  gam_fit() %>%
  cplot(P = c(.5, .9)) +
  ggtitle("2018 Lucas Giolito") +
  centertitle() -> plot3

Dylan %>%
  setup_called() %>%
  gam_fit() %>%
  cplot(P = c(.5, .9)) +
  ggtitle("2018 Dylan Bundy") +
  centertitle() -> plot4

grid.arrange(plot1, plot2, plot3, plot4)

## Extensions

# break down by side of batter -- stand == "R", or stand == "L"
# break down by the count -- use Count variable
# break down by the pitch type -- for example, pitch_type == "CU"

Clayton %>%
  setup_called() %>%
  filter(pitch_type == "CU") %>%
  gam_fit() %>%
  cplot(P = c(.5, .9)) +
  ggtitle("2018 Clayton Kershaw - CU") +
  centertitle() +
  increasefont() -> plot5

Aaron %>%
  setup_called() %>%
  filter(pitch_type == "CU") %>%
  gam_fit() %>%
  cplot(P = c(.5, .9)) +
  ggtitle("2018 Aaron Nola - CU") +
  centertitle() +
  increasefont() -> plot6

grid.arrange(plot5, plot6, nrow = 1)

###############

# locations of all Aaron Nola's curveballs

Aaron %>%
  filter(pitch_type == "CU") %>%
  setup_swing() %>%
  ggplot(aes(plate_x, plate_z,
             color = Miss)) +
  geom_point() +
  add_zone() +
  increasefont() +
  ggtitle("Swings on Aaron Nola Curveballs") +
  centertitle()

Aaron %>%
  filter(pitch_type == "CU") %>%
  setup_swing() %>%
  gam_fit2() %>%
  cplot(P = c(.1, .3, .5))  +
  ggtitle("Miss Rates on Aaron Nola Curveballs") +
  centertitle() -> plot7

# use all 2018 statcast data (not available here)

statcast2018new %>%
  filter(p_throws == "R", pitch_type == "CU") %>%
  setup_swing() %>%
  gam_fit2() %>%
  cplot(P = c(.1, .3, .5))  +
  ggtitle("Miss Rates on All Right-Handed Curveballs") +
  centertitle() -> plot8

grid.arrange(plot7, plot8, nrow = 1)