Interesting -

Wow DK dropped min salary on NBA to 2k. Interesting.

— Michael Hofeld (@MichaelHofeld) April 24, 2015

So I went to DraftKings and used their handy CSV download to grab the NBA salary table for Friday, April 24, 2015.

DKSalaries.20150424 <- read.csv("/data/Projects/draftkings/nba-playoffs/DKSalaries-20150424.csv",
  stringsAsFactors=FALSE)

# scale Salary by 1000 and scatterplot
DKSalaries.20150424$Salary <- DKSalaries.20150424$Salary / 1000
plot(
  DKSalaries.20150424$Salary, 
  DKSalaries.20150424$AvgPointsPerGame,
  main = 'DraftKings NBA 20150424',
  xlab = 'Salary (Thousands)',
  ylab = 'Average Points per Game')

Wow! Except for the bunch-up at Salary = 2000, this looks like a pretty straight line. So … let’s do a robust line fit.

robust.line <- line(
  DKSalaries.20150424$Salary,
  DKSalaries.20150424$AvgPointsPerGame)
plot(DKSalaries.20150424$Salary,
  DKSalaries.20150424$AvgPointsPerGame,
  main = 'DraftKings NBA 20150424',
  xlab = 'Salary (Thousands)',
  ylab = 'Average Points per Game')
abline(coefficients(robust.line))

So the “value line” is

## [1] "AvgPointsPerGame = -1.9 + 4.8 * Salary / 1000"

Interesting indeed.