The file S03_65.xlsx lists a lot of data for each NBA team for the seasons 2004–2005 to 2008–2009. The variables are divided into groups:
(1) Overall success,
(2) Offensive, and
(3) Defensive.
The basic question all basketball fans (and coaches) ponder is what causes success or failure.
a. Explore this question by creating a correlation matrix with the variable Wins (the measure of success) and all of the variables in groups (2) and (3). Based on these correlations, which five variables appear to be the best predictors of success?
b. Explore this question in a different way, using the Playoff Team column as a measure of success. Here, it makes sense to proceed as in the Lasagna Triers example in Section 3.5, using the variables in groups (2) and (3) as the predictors. However, these predictors are all basically continuous, so grouping would be required for all of them in the pivot table, and grouping is always somewhat arbitrary. Instead, create a copy of the Data sheet. Then for each variable in groups (2) to (13), create a formula that returns 1, 2, 3, or 4, depending on which quarter of that variable the value falls in. Now use these discrete variables as predictors and proceed as in the Lasagna Triers example. List the five variables that appear to be the best (or at least good) predictors of making the playoffs.