W35 Wagga Wagga stats & predictions

Clicca qui per vedere altri articoli correlati!

• Come analizzare le performance dei giocatori nel tennis moderno?
• Le migliori strategie per vincere nelle scommesse sportive online: consigli pratici da esperti!gslab-econ/retrosheet<|file_sep|>/README.md # Retrosheet Data This repository contains the code and data to reproduce the analyses and figures from the paper "An Econometric Analysis of Baseball's Steroid Era," available at [http://www.gsl.pepperdine.edu/~seigler/research/](http://www.gsl.pepperdine.edu/~seigler/research/). The data used in the paper is available for download at [https://drive.google.com/open?id=0B8zVUuBkGfCJYmZUNlJqRkZIYjQ](https://drive.google.com/open?id=0B8zVUuBkGfCJYmZUNlJqRkZIYjQ). The link will expire on June **30**, **2020**. ## Acknowledgements This work uses data provided by [Retrosheet](http://www.retrosheet.org/), which is maintained by the following people: * David Smith * Paul Hirsch * David Allen * Sean Forman ## Contact Information If you have questions about this project or would like to request access to the data before June **30**, **2020**, please email me at [email protected]. ## References 1. Seigler M (2019) An Econometric Analysis of Baseball's Steroid Era. Economics & Human Biology (forthcoming). <|repo_name|>gslab-econ/retrosheet<|file_sep|>/code/07_summary_stats.R rm(list = ls()) library(data.table) library(ggplot2) # Read in data -------------------------------------------------------------- data <- fread("data/data.csv", header = TRUE) summary(data) # Summary statistics -------------------------------------------------------- table(data$season) # Average number of at bats mean(data$ab) sd(data$ab) # Average number of singles mean(data$singles) sd(data$singles) # Average number of doubles mean(data$doubles) sd(data$doubles) # Average number of triples mean(data$triples) sd(data$triples) # Average number of home runs mean(data$homeruns) sd(data$homeruns) # Average number of walks mean(data$walks) sd(data$walks) # Average number of strikeouts mean(data$strikeouts) sd(data$strikeouts) # Average slugging percentage mean(data$slg) sd(data$slg) # Average on base percentage mean(data$obp) sd(data$obp) # Average batting average on balls in play (BABIP) mean(data$babip) sd(data$babip) # Proportion of players with at least one year with positive steroid test results sum(table(table(unique(data$id[data$steroids == "Y"])))) / nrow(unique(subset(data[, .(id)], steroids == "Y"))) <|repo_name|>gslab-econ/retrosheet<|file_sep|>/code/05_preliminary_analysis.Rmd --- title: "Preliminary Analysis" author: "Michael Seigler" date: "April `r format(Sys.Date(), '%Y')`" output: html_document: toc: true toc_float: collapsed: false toc_depth: '2' --- {r setup, include=FALSE} knitr::opts_chunk$set(echo = TRUE) ## Introduction This document contains preliminary analysis to explore the relationships between steroid use and various baseball performance measures. {r} rm(list = ls()) library(dplyr) library(ggplot2) library(scales) library(stargazer) data <- read.csv("data/data.csv") ## Overall batting average First we estimate a simple linear model relating a player's overall batting average (`avg`) to his steroid use (`steroids`) controlling for several other player characteristics. {r} model1 <- lm(avg ~ steroids + hgt + wt + batted_handedness + age + season + team + league + division + ballpark_factor + ab + hr_season + hr_season_squared + ob_season + ob_season_squared + babip_season + babip_season_squared + strikeouts_season + strikeouts_season_squared + walks_season + walks_season_squared, data = data) stargazer(model1, type = "text", omit.stat = c("f", "ser"), title = "Linear Model Predicting Batting Average", covariate.labels = c("Steroids", "Height (inches)", "Weight (lbs.)", "Batted Right-handed", "Age", "Season Fixed Effects", "Team Fixed Effects", "League Fixed Effects", "Division Fixed Effects", "Ballpark Factor", "# At Bats", "# Home Runs Season Squared", "# Home Runs Season Squared", "# On Base Season Squared", "# On Base Season Squared", "BABIP Season Squared", "BABIP Season Squared", "# Strikeouts Season Squared", "# Strikeouts Season Squared", "# Walks Season Squared", "# Walks Season Squared")) According to this model there is a statistically significant positive relationship between steroid use and overall batting average; players who tested positive for steroids had a .013 higher overall batting average than those who did not test positive controlling for other factors. Next we estimate a linear model predicting batting average as a function of steroid use but we split the sample by whether or not the player played during the Steroid Era (`season >=1998`). We do this because players who played before the Steroid Era could not have been using steroids during their careers so their steroid use is actually measuring something else entirely. {r} data_pre <- subset(data, season <=1997 & !is.na(steroids)) model2_pre <- lm(avg ~ steroids + hgt + wt + batted_handedness + age + season + team + league + division + ballpark_factor + ab + hr_season + hr_season_squared + ob_season + ob_season_squared + babip_season + babip_season_squared + strikeouts_season + strikeouts_season_squared + walks_season + walks_season_squared, data = data_pre) data_post <- subset(data, season >=1998 & !is.na(steroids)) model2_post <- lm(avg ~ steroids + hgt + wt + batted_handedness + age + season + team + league + division + ballpark_factor + ab + hr_season + hr_season_squared + ob_season +