bunts.R

###############################
# getting 2015 retrosheet data
# and computing runs expectancies
# assuming chadwick files are installed

library(devtools)
setwd("~/Desktop/retrosheet")
source_gist("https://gist.github.com/bayesball/8892981")
source_gist("https://gist.github.com/bayesball/8892999")
parse.retrosheet2.pbp(2015)
setwd("download.folder/unzipped")
d2015 <- compute.runs.expectancy(2015)

####################################
# saved in d2015.Rdata
#####################################

library(dplyr)
library(ggplot2)
d2015 <- mutate(d2015,
                BAT_TEAM_ID=ifelse(BAT_HOME_ID == 0,
                  as.character(AWAY_TEAM_ID), 
                  substr(as.character(GAME_ID), 1, 3)),
                League=ifelse(BAT_TEAM_ID %in% 
                              c("OAK", "BOS", "TBA", "NYA", "BAL",
                                "HOU", "CHA", "KCA", "SEA", "DET",
                                "ANA", "MIN", "TOR", "CLE", "TEX"),
                              "AL", "NL"),
                BAT_ORDER=factor(BAT_LINEUP_ID))
bunts <- filter(d2015, BUNT_FL==TRUE)

# what teams are bunting?

teams <- sort(table(bunts$BAT_TEAM_ID))
bunts$BAT_TEAM <- factor(bunts$BAT_TEAM_ID, 
  levels = names(teams))

ggplot(bunts, aes(BAT_TEAM, fill=factor(League))) + 
  geom_bar() + coord_flip() +
  ggtitle("Number of Team Bunts, 2015 Season")

# who is bunting? Look at batting order position

ggplot(bunts, aes(BAT_ORDER)) +
  geom_bar() + coord_flip() +
  ggtitle("Team Bunts by Bat Order Position")

# are there specific players who like to bunt?

S <- filter(summarize(group_by(bunts, BAT_ID),
               N=n(),
               SH=sum(SH_FL),
               H=sum(H_FL),
               Success_Rate = round(100 * (H + SH) / N)), N > 19)
master <- read_csv("~/Desktop/STUFF/Lahman2015/Master.csv")

Top <- inner_join(S, select(master, retroID, nameFirst, nameLast),
           c("BAT_ID" = "retroID"))

arrange(select(Top, nameFirst, nameLast, N, SH, H, Success_Rate), 
        desc(N))

# when are they bunting?

# Inning?

ggplot(filter(bunts, INN_CT < 10), aes(factor(INN_CT))) +
  geom_bar() + coord_flip() +
  ggtitle("Team Bunts by Inning")

# game state (bases occupied and outs)?

S <- summarize(group_by(bunts, STATE), 
               N=n(), SH=sum(SH_FL==TRUE),
               H=sum(H_FL),
               Success_Rate = round(100 * (SH + H) / N),
               Type=ifelse(SH > H, "Sacrifice", "Hit"))
S50 <- filter(S, N >= 50)
ggplot(S50, aes(STATE, Success_Rate, color=Type)) + geom_point(size=3) + 
  coord_flip() +
  ggtitle("Success Rates for Different Base/Out Situations")