History View of Three Hitting Rates

Historic.R

## Load Lahman Teams data frame

library(tidyverse)
Teams <- read_csv("~/Dropbox/Google Drive/Lahman/core/Teams.csv")

Teams %>% filter(yearID >= 1913) %>% 
  group_by(yearID) %>% 
  summarize(AB = sum(AB),
            HR = sum(HR),
            SO = sum(SO),
            H = sum(H)) -> S
rbind(S, data.frame(yearID = 2018,
 AB = 165432, HR = 5585, SO = 41207, H = 41019 )) -> S

S %>%   mutate(SO_Rate = SO / AB,
         HR_Rate = HR / (AB - SO),
         H_Rate = (H - HR) / (AB - HR - SO)) -> S

S %>% select(yearID, SO_Rate, HR_Rate, H_Rate) %>% 
  gather(Type, Rate, - yearID) -> Sg

TH <- theme(
  plot.title = element_text(
    colour = "blue",
    size = 18,
    hjust = 0.5
  )
)
## ------------------------------------------------------------------------
ggplot(Sg, aes(yearID, Rate, color = Type)) +
  geom_point(size = 1.5) +
  ggtitle("Historical View of Three Rates") + TH

## ------------------------------------------------------------------------
logit <- function(x){ log(x / (1 - x))}
ggplot(Sg, aes(yearID, logit(Rate), color = Type)) +
  geom_point(size = 1.5) + geom_smooth(se = FALSE, 
                    span = 0.2, method = "loess") +
  TH + ggtitle("Logits of Three Rates")

## ------------------------------------------------------------------------
ggplot(filter(Sg, yearID >= 1975), 
       aes(yearID, logit(Rate), color = Type)) +
  geom_point() + geom_smooth(se = FALSE, 
                             method = "lm") + 
  ggtitle("Logits of Three Rates Since 1975") + TH

## ------------------------------------------------------------------------
filter(Sg, yearID == 1975) %>% 
  group_by(Type) %>% summarize(Rate75 = first(Rate)) -> S75
inner_join(filter(Sg, yearID >= 1975), S75, by = "Type") ->
  Sgnew

## ------------------------------------------------------------------------
ggplot(Sgnew, aes(yearID, logit(Rate) -
                          logit(Rate75),
                  color = Type)) +
  geom_point() + geom_smooth(method = "lm", se = FALSE) +
  facet_wrap(~ Type, ncol = 1)

## ------------------------------------------------------------------------
ggplot(filter(Sgnew,
              Type %in% c("HR_Rate","SO_Rate")),
              aes(yearID, logit(Rate) -
                          logit(Rate75),
                  color = Type)) +
  geom_point() + geom_smooth(method = "lm", se = FALSE) +
  facet_wrap(~ Type, ncol = 1) + TH +
  theme(text = element_text(size = 18))

## ------------------------------------------------------------------------
library(broom)
Sgnew %>% group_by(Type) %>%
  do(tidy(lm(I(logit(Rate)) ~ yearID, data=.))) %>% 
  filter(term == "yearID")

#-----------------------------------------------------

library(BApredict)
d <- collect_hitting_data()
d %>%   mutate(SO_Rate = SO / AB,
         HR_Rate = HR / (AB - SO),
         H_Rate = (H - HR) / (AB - HR - SO)) -> S
ggplot(S, aes(SO_Rate, HR_Rate, label = playerID)) +
  geom_point(size = 2, color = "blue") +
  geom_label(data = filter(S, SO_Rate > .35 |
                             HR_Rate > .11),
             color = "red") + 
  TH + ggtitle("SO and HR Rates for 2018 Players") +
  theme(text = element_text(size = 18))