Dunk Analytics 1.0: Which are the most spectacular teams?

The NBA is a great spectacle and the crown jwel of this spectacle is a good dunk.  It is not by chance that the All-star dunk contest generates a lot of discussion every year (even when it is subpar like this year).  So how do teams do in this aspect? Assuming that a measure of the “amount of spectacle” or the “showtime index” for the NBA is the number of dunks, how do teams rank?

I set to explore this by using the play-by-play data for the past two seasons (i.e., 2014-15 and 2015-16). I simply calculated the mean number of dunks per game for each team.



The top figure corresponds to the 2014-15 season, while the bottom to the 2015-16 season.  The amount of overal “dunking” appears to be fairly stable but the actual ranking is different. In fact, the Spearman’s rank correlation coefficient is 0.27 but not significant (p-value = 0.14).  Of course, the ranking represents only the average number of dunks/game, and many times the differences between two teams might not even be significant (e.g., HOU and OKC last year). Consequently, from a statistical point of view, teams could effectively  have the same rank even if the actual means are different. 

However not all dunks provide the same satisfaction and enjoyment to the fans. Fortunately, the play-by-play data also provide information about the type of dunk that it was performed and clearly some of these dunks are more impressive than the rest (e.g., a reverse or an alley-oop dunk are much more impressive than a simple two-handed dunk under the baseket). A weighted average could be generated for a more accurate”showtime index” but then the weights can be different for every person. One interesting technicality is that the play-by-play data for 2014-15 were more detailed and included a larger variety of dunk-types as compared to last season.  Following are the distributions of these dunks:



You can use your own weights and decide how enjoyable is the league in total. Of course it will be hard to visualize the distribution for different teams so I have created an interactive application for you to explore here.  Enjoy and make your own conclusions about the most “showtime” team.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s