17 Collective intelligence

 

This chapter covers

  • Automated exploratory data analysis
  • Conducting baseline EDA tasks with the tableone package
  • Performing advanced EDA operations with the DataExplorer and SmartEDA packages
  • Applying new functional and aesthetic techniques to ggplot2 bar charts

Our first purpose here is to establish who might be smarter—the small number of Las Vegas oddsmakers with advanced data science degrees who establish opening odds using very sophisticated algorithms where millions of dollars are at stake for the casinos they work for, or the thousands of gamblers, professionals and amateurs with skin in the game, who then wager their hard-earned money and, in the process, influence the closing odds.

For instance, on October 17, 2018, the Memphis Grizzlies played at the Indiana Pacers. The opening total from Las Vegas oddsmakers—that is, the estimated number of points to be scored by the Grizzlies and Pacers combined—was 209. Wagers were then placed on what is otherwise known as the over/under until the betting line closed. Money wagered on the over comes from gamblers who think the Grizzlies and Pacers will score more than 209 points; money wagered on the under comes from gamblers who think the two teams will combine for fewer than 209 points. It doesn’t matter who wins, who loses, or what the final margin is; all that matters is whether or not the combined point total is greater than or less than 209.

17.1 Loading packages

17.2 Importing data

17.3 Wrangling data

17.4 Automated exploratory data analysis

17.4.1 Baseline EDA with tableone

17.4.2 Over/under EDA with DataExplorer

17.4.3 Point spread EDA with SmartEDA

17.5 Results

17.5.1 Over/under

Summary