R script for 3000 hitting club exploration

3000hitclub.R

# read in three packages

library(readr)
library(dplyr)
library(ggplot2)

# helper ggplot2 function 

helper_ggplot <- function(){
  theme(text=element_text(size=18)) +
  theme(plot.title = element_text(colour = "blue", 
                                  size = 18,
                                  hjust = 0.5, 
                                  vjust = 0.8, 
                                  angle = 0))
}

# read in Lahman datasets (can also use datasets in Lahman package)

Batting <- read_csv("~/Dropbox/Google Drive/Lahman/core/Batting.csv")
People <- read_csv("~/Dropbox/Google Drive/Lahman/core/People.csv")

# identify players with at least 3000 career hits

Batting %>% 
  group_by(playerID) %>% 
  summarize(H = sum(H),
            .groups = "drop") %>% 
  filter(H >= 3000) %>% 
  pull(playerID) -> players_3000

# add Cabrera to list

People %>% 
  filter(nameLast == "Cabrera",
         nameFirst == "Miguel") %>% 
  pull(playerID) -> mc_id

players_3000 <- c(players_3000, mc_id)

# collect season data including ages

Batting %>% 
  filter(playerID %in% players_3000) %>% 
  group_by(playerID, yearID) %>% 
  summarize(H = sum(H),
            AB = sum(AB),
            BA = H / AB,
            .groups = "drop") -> S3000

# create dataset of players in 3000 hit club

People %>% 
  select(playerID, nameFirst, nameLast,
         birthYear, birthMonth) %>% 
  inner_join(S3000, by = "playerID") -> S3000

# add ages

S3000 %>% 
  mutate(birthyear = ifelse(birthMonth >= 7,
                      birthYear + 1, birthYear),
         Age = yearID - birthyear) -> S3000

# initial graph -- too much plotting!

ggplot(S3000, aes(Age, BA, color = playerID)) +
  geom_point()

# try plotting smooths

ggplot(S3000, aes(Age, BA, color = playerID)) +
  geom_smooth(se = FALSE)

# separate into two data sets by midseason

S3000 %>% 
  group_by(playerID) %>% 
  mutate(MidYear = (max(yearID) + 
                      min(yearID)) / 2) %>% 
  select(playerID, MidYear) -> S_midyear

inner_join(S3000, S_midyear,
           by = "playerID") -> S3000

# smooths over two groups of players

ggplot(filter(S3000, MidYear <= 1976,
              AB >= 300),
       aes(Age, BA, color = playerID)) +
  geom_smooth(se = FALSE,
              method = "loess",
              formula = "y ~ x")

ggplot(filter(S3000, MidYear > 1976,
              AB >= 300),
       aes(Age, BA, color = playerID)) +
  geom_smooth(se = FALSE,
              method = "loess",
              formula = "y ~ x")

# look at each trajectory and categorize
# by the shape of the smooth
# add this information to my main dataset

plot_single_trajectory <- function(j){
  ggplot(filter(S3000, 
                     playerID == players_3000[j] ,
                     AB >= 300),
              aes(Age, BA)) +
    geom_smooth(se = FALSE,
                method = "loess",
                formula = "y ~ x") +
    ggtitle(players_3000[j]) +
    helper_ggplot()
}

plot_single_trajectory(1)

# I estimated the peak age and the type of
# trajectory for each player in this group

peak_ages <- c(25, 27, 33, 30, 27, 27, 33,
               30, 32, 25, 25, 35, 25,
               25, 23, 27,
               26, 35, 27, 27, 27, 25, 
               35, 21, 30, 35, 30, 30, 25,
               29, 25, 30, 30)
desc <- c("E", "E", "L", "M", "E", "E", "L", "M",
          "L", "E", "E", "L", "L", "M", "E", "E",
          "E", "L", "E", "E", "E", "E", "L", "E",
          "M", "L", "E", "M", "E", "M", "E", "M",
          "M")

Players <- data.frame(playerID = players_3000,
                      Peak_Age = peak_ages,
                      Type = desc)

inner_join(S3000, Players,
           by = "playerID") -> S3000

############### six graphs

ggplot(filter(S3000, 
              Type == "E",
              MidYear <= 1976,
              AB >= 300),
       aes(Age, BA, color = nameLast)) +
  geom_smooth(se = FALSE,
              method = "loess",
              formula = "y ~ x") +
  ggtitle("Early Bloomers - MidCareer <= 1976") +
  helper_ggplot()

ggplot(filter(S3000, 
              Type == "E",
              MidYear > 1976,
              AB >= 300),
       aes(Age, BA, color = nameLast)) +
  geom_smooth(se = FALSE,
              method = "loess",
              formula = "y ~ x") +
  ggtitle("Early Bloomers - MidCareer > 1976") +
  helper_ggplot()

ggplot(filter(S3000, 
              Type == "L",
              MidYear <= 1976,
              AB >= 300),
       aes(Age, BA, color = nameLast)) +
  geom_smooth(se = FALSE,
              method = "loess",
              formula = "y ~ x") +
  ggtitle("Late Bloomers - MidCareer <= 1976") +
  helper_ggplot()

ggplot(filter(S3000, 
              Type == "L",
              MidYear > 1976,
              AB >= 300),
       aes(Age, BA, color = nameLast)) +
  geom_smooth(se = FALSE,
              method = "loess",
              formula = "y ~ x") +
  ggtitle("Late Bloomers - MidCareer > 1976") +
  helper_ggplot()

ggplot(filter(S3000, 
              Type == "M",
              MidYear <= 1976,
              AB >= 300),
       aes(Age, BA, color = nameLast)) +
  geom_smooth(se = FALSE,
              method = "loess",
              formula = "y ~ x") +
  ggtitle("Common Trajectory - MidCareer <= 1976") +
  helper_ggplot()

ggplot(filter(S3000, 
              Type == "M",
              MidYear > 1976,
              AB >= 300),
       aes(Age, BA, color = nameLast)) +
  geom_smooth(se = FALSE,
              method = "loess",
              formula = "y ~ x") +
  ggtitle("Common Trajectory - MidCareer > 1976") +
  helper_ggplot()