Definitions

• Game Score From Basketball-Reference:

“Game Score was created by John Hollinger to give a rough measure of a player’s productivity for a single game. The scale is similar to that of points scored, (40 is an outstanding performance, 10 is an average performance, etc.).”

• Results A basketball game consistents of two teams, fielding 5 players each at a time. At the end of four 12-minute quarters the team with the most points wins. The players are regularly substituted (known as rotation) to rest and prevent injury. Because of the large amount of combinations of players possibly on the floor at any time, many analysts seek to identify the individual impact a player can make all else equal. A typical result could be the team winning by 3 points (+3) or losing by 5 points (-5).

• Marginal Result The predicted effect of 1 more game score (a 25 vs 24 game score) has on the team result (winning by +2 instead of +1). While technically any value is possible, a Marginal Result close to 1 is amazing and 0 or .1 is typical. The impact of 1 player is instrinsically limited.

The Pitch

• Selecting 5 of the top NBA players (1 for each of the top 5 teams in the 30 team National Basketball Association). I’ve plotted their Game Scores vs the Point Differential (Results) of the entire team at the end of each game.
• The app allows you to select which players you want plotted, and a linear model will be put over all the data points (the games) for all the players selected.
• Optionally, you can brush (outline) a rectangular area of the points and a Marginal Result table will form.

Slide with R Output

library(ggplot2)
#Taken directly from the Shiny App 
dataURL <- "https://raw.githubusercontent.com/RafaJones/NBA/master/allNBA.csv"
download.file(dataURL, destfile = "./allNBA.csv")
allNBA <- read.csv("./allNBA.csv",stringsAsFactors = FALSE)
allNBA <- allNBA[,c(10,31,33)]
colnames(allNBA) <- c("Results","GameScore","Player")
allNBA <- as.data.frame(allNBA)

g <- ggplot(aes(x = GameScore, y = Results, color = Player), data = allNBA)
        gg <- g + geom_point() + geom_smooth(method = "lm") + labs(title = "Is Basketball a Team Game?",
                                                      x = "Individual Game Score",
                                                      y = "Point Difference")

Slide with Plot

# See the NBA Compare Shiny APP to get interactive features!
gg