Created
November 27, 2020 15:35
-
-
Save bayesball/247da8335883130e22c471d6f981b6ee to your computer and use it in GitHub Desktop.
R code for illustration of Called Strike package
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
# here are the packages I'm using | |
# CalledStrike depends on other packages | |
library(CalledStrike) | |
library(readr) | |
library(dplyr) | |
library(ggplot2) | |
# read in Statcast data for the 2019 season | |
statcast2019 <- read_csv("~/Dropbox/2016 WORK/BLOG Baseball R/OTHER/StatcastData/statcast2019.csv") | |
# DJ LeMahieu example | |
dj <- filter(statcast2019, | |
player_name == "DJ LeMahieu") | |
# scatterplot of in-play locations with hits | |
# indicated by color variable | |
dj %>% | |
setup_inplay() %>% | |
ggplot(aes(plate_x, plate_z, | |
color = H)) + | |
geom_point() + | |
add_zone() + | |
increasefont() + | |
coord_fixed() | |
# smoothed hit probabilities over zone | |
hit_contour(dj, | |
L = seq(0, 1, by = 0.02), | |
title = "2019 LeMahieu Probability of Hit") | |
# Swing and contact part | |
statcast2019 <- read_csv("~/Dropbox/2016 WORK/BLOG Baseball R/OTHER/StatcastData/statcast2019.csv") | |
# pick up FanGraph batting leaders for 2019 season | |
fg <- fg_bat_leaders(2019, 2019) | |
# identify low and high swingers | |
ggplot(fg, aes(Swing_pct_pi, Contact_pct_pi, | |
label = Name)) + | |
geom_point(size = 3, color = "blue") + | |
geom_text(data = filter(fg, | |
Swing_pct_pi < 37 | | |
Swing_pct_pi > 57), | |
nudge_y = -0.6, color = "red") + | |
geom_point(data = filter(fg, | |
Swing_pct_pi < 37 | | |
Swing_pct_pi > 57), | |
color = "red", size = 3) + | |
xlim(30, 62) + | |
ggtitle("2019 Swing and Contact Fractions") + | |
centertitle() + | |
increasefont() | |
low_swingers <- filter(fg, Swing_pct_pi < 37) %>% | |
pull(Name) | |
high_swingers <- filter(fg, Swing_pct_pi > 57) %>% | |
pull(Name) | |
swing <- c(low_swingers, high_swingers) | |
# compare nine players on probability of swing | |
sc2 <- filter(statcast2019, | |
player_name %in% swing) | |
swing_contour(split(sc2, | |
sc2$player_name), | |
L = seq(0, 1, by = 0.02), | |
NCOL = 3) | |
# look at Mike Trout | |
# break down by side of pitcher | |
statcast2019 %>% | |
filter(player_name == "Mike Trout") %>% | |
split_LR() %>% | |
swing_contour(L = seq(0, 1, by = 0.05), | |
title="Mike Trout Swing Rate") | |
# break down by type of pitch | |
statcast2019 %>% | |
filter(player_name == "Mike Trout") %>% | |
split_pitchtype() %>% | |
swing_contour(L = seq(0, 1, by = 0.05), | |
title="Mike Trout Swing Rate") | |
# break down by pitch count | |
statcast2019 %>% | |
filter(player_name == "Mike Trout") %>% | |
split_count() %>% | |
swing_contour(L = seq(0, 1, by = 0.05), | |
title="Mike Trout Swing Rate") | |
# compare with Jeff McNeil | |
statcast2019 %>% | |
filter(player_name == "Jeff McNeil") %>% | |
split_count() %>% | |
swing_contour(L = seq(0, 1, by = 0.05), | |
title="Jeff McNeil Swing Rate") | |
# look at contact rates | |
ggplot(fg, aes(Swing_pct_pi, Contact_pct_pi, | |
label = Name)) + | |
geom_point(size = 3, color = "blue") + | |
geom_text(data = filter(fg, | |
Contact_pct_pi < 67.5 | | |
Contact_pct_pi > 86.9), | |
nudge_y = -0.6, color = "red") + | |
geom_point(data = filter(fg, | |
Contact_pct_pi < 67.5 | | |
Contact_pct_pi > 86.9), | |
color = "red", size = 3) + | |
xlim(30, 62) + | |
ggtitle("2019 Swing and Contact Fractions") + | |
centertitle() + | |
increasefont() | |
low_connect <- filter(fg, Contact_pct_pi < 67.5) %>% | |
pull(Name) | |
high_connect <- filter(fg, Contact_pct_pi > 86.9) %>% pull(Name) | |
connect <- c(low_connect, high_connect) | |
# plots of smoothed contact probabilities | |
sc1 <- filter(statcast2019, | |
player_name %in% connect) | |
contact_swing_contour(split(sc1, | |
sc1$player_name), | |
L = seq(0, 1, by = 0.02), | |
NCOL = 3) | |
# Luke Voit -- how does eontact prob depend on pitch | |
# type? | |
filter(statcast2019, | |
player_name == "Luke Voit") %>% | |
split_pitchtype() %>% | |
contact_swing_contour(L = seq(0, 1, by = 0.05), | |
title = | |
"Luke Voit - Prob(Contact)") | |
# depend on pitcher side? | |
filter(statcast2019, | |
player_name == "Luke Voit") %>% | |
split_LR() %>% | |
contact_swing_contour(L = seq(0, 1, by = 0.05), | |
title = "Luke Voit - Prob(Contact)") | |
# depend on both pitcher side and pitch type? | |
filter(statcast2019, | |
player_name == "Luke Voit") %>% | |
split_LR_pitchtype() %>% | |
contact_swing_contour(L = seq(0, 1, by = 0.05), | |
title = | |
"Luke Voit - Prob(Contact)") | |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment