Jim Albert bayesball

## home_run_train_test.R
# load several packages

library(tidyverse)
library(CalledStrike)
library(mgcv)

# read in Statcast data for five seasons

sc <- read_csv("five_seasons_data.csv")

## home_run_prediction_2.R
################################################
## NEW WORK - August 2, 2019
################################################

#########################################
# data for 1631 games through August 1
#########################################

library(tidyverse)
library(CalledStrike)

## home_run_prediction.R
# 2019 statcast data is in data frame sc2019

# collect number of home runs in each game

library(tidyverse)

sc2019 %>%
  group_by(game_pk) %>%
  summarize(HR = sum(events == "home_run",
                     na.rm = TRUE)) -> S

## clean_statcast.R
# read in SC data from  2019 seasons
# through games of June 23, 2019

library(tidyverse)
library(CalledStrike)

sc2019 <- read_csv("~/Dropbox/2016 WORK/BLOG Baseball R/OTHER/StatcastData/statcast2019.csv")
sc2019_ip <- filter(sc2019, type == "X")

# focus on Statcast 2019 in-play data

## calledstrikework.R
# load some packages

library(baseballr)
library(tidyverse)
library(CalledStrike)
library(gridExtra)

# scrape data for four pitchers

Aaron <- scrape_statcast_savant(start_date = "2018-03-15",

## homerunvalue.R
# load in two packages

library(tidyverse)
library(WinProbability)

# assume dataset all2018.csv is in current working
# directory -- these functions compute the runs
# expectancies and WPA values

d2018 <- compute.runs.expectancy(2018)

## compute.win.probs.R
compute.win.probs <- function(d, S){
  # adds variables P.OLD, P.NEW, and WPA
  # to retrosheet data with run expectancies
  invlogit <- function(x) exp(x) / (1 + exp(x))

  d %>%
    mutate(half.inning.row = 2 * INN_CT + BAT_HOME_ID,
          runs0 = ifelse(BAT_HOME_ID == 1,
      HOME_SCORE_CT - AWAY_SCORE_CT + RUNS.STATE,
      HOME_SCORE_CT - AWAY_SCORE_CT - RUNS.STATE)) -> d

## pitch_length.R
# Load tidyverse packages and read in the Retrosheet data

library(tidyverse)
load("~/Dropbox/Google Drive/Retrosheet/pbp.2018.Rdata")

# if you have trouble getting the 2018 Retrosheet data, you can
# use several complete Retrosheet datasets from previous seasons
# see http://www-math.bgsu.edu/~albert/retrosheet/

# Create new variables:  pseq containing only the pitch

## erode_work.R
# read in Statcast data for two seasons

library(tidyverse)
sc <- read_csv("../StatcastData/statcast2018new.csv")
sc17 <- read_csv("../StatcastData/statcast2017.csv")

# erode function will complete individual regression estimates for all
# players who have seen 1000 called pitches

erode <- function(sc){

## distances2.R
# read in packages and data

library(tidyverse)
library(gridExtra)

sc2018 <- read_csv("../StatcastData/statcast2018new.csv")
sc2017 <- read_csv("../StatcastData/statcast2017.csv")

sc2017ip <- filter(sc2017, type == "X") # 127555 rows
sc2018ip <- filter(sc2018, type == "X") # 126282 rows