With this in mind, I took total attendance percentage for each team and total population for each teams' cities and created a regression model that would predict what kind of attendance percentages each team should have.
The graph below looks confusing but is actually quite simple once you figure it out.
Along the x-axis are the "fitted values." These values represent the attendance percentages that should be expected based on the size of the city and win percentage.
For example, based on the regression model created, one would predict that Orlando would have attendance percentages of about 81%.
However, the attendance percentages that Orlando fans have been recording have averaged nearly 95% this season.
This is where the residuals (the y-axis) comes into play. Orlando's residual is +14, meaning that we exceed the expected attendance rates by 14%. This makes us only second to Dallas in the entire nation (and they just won a championship two years ago!)
The loyalty of the fans in Orlando is quite impressive.
P.S. Funny to see Detroit's residual as -22... I wouldn't go there after what Ron Artest did either.