Plots groundball statistics for all teams in a particular season

groundball.plot.R

groundball.plot <- function(pbp, season){
   require(dplyr)
   require(ggplot2)
   require(car)
   inplay <- filter(pbp, BATTEDBALL_CD == "F" |
                   BATTEDBALL_CD == "G" |
                   BATTEDBALL_CD == "L" |
                   BATTEDBALL_CD == "P")

    inplay <- mutate(inplay,
              HOME_TEAM_ID = substr(GAME_ID, 1, 3),
              FIELD_TEAM = ifelse(BAT_HOME_ID==0,
                            as.character(HOME_TEAM_ID),
                            as.character(AWAY_TEAM_ID)),
              BAT_TEAM = ifelse(BAT_HOME_ID==0,
                          as.character(AWAY_TEAM_ID),
                          as.character(HOME_TEAM_ID)))

    # BB <- with(inplay, table(HOME_TEAM_ID, BATTEDBALL_CD))

    # with(inplay, table(BATTEDBALL_CD, EVENT_CD))

    # focus on EVENT_CD 2, 18, 19 (out), 20 (S), 21 (D), 22 (T), 23 (HR)

    inplay2 <- filter(inplay, EVENT_CD != 13)
    inplay2 <- mutate(inplay2, event=EVENT_CD)
    inplay2$event <- recode(inplay2$event, "c(2, 18, 19)='out';
                      20='single';
                      21='double';
                      22='triple';
                      23='home run'")

    inplay2 <- mutate(inplay2,
                  event=factor(event, c("out", "single",
                              "double", "triple", "home run")),
                  battedball=factor(BATTEDBALL_CD,
                          c("P", "G", "F", "L"),
                    c("Popup", "Groundball", "Fly Ball", "Liner")))

    # T.BE <- with(inplay2, table(battedball, event))
    # round(100 * prop.table(T.BE, 1), 1)

    gb.data <- summarize(group_by(inplay2, FIELD_TEAM),
        p.gb.out = sum(battedball == "Groundball" & event == "out") /
                        sum(battedball == "Groundball"),
        p.gb = mean(battedball == "Groundball"))

    # ggplot(gb.data, aes(p.gb.out, FIELD_TEAM)) + geom_point()

    print(ggplot(gb.data, aes(p.gb, p.gb.out, label=FIELD_TEAM)) + 
          geom_text() + xlab("Proportion Ground Balls") + 
          ylab("Proportion of Ground Balls that are Outs") +
          ggtitle(paste("Season =", season)))
    data.frame(gb.data, Season=season)
}