Your consulting company was appointed by sports regulatory organization to build the model which related attendance over the season for sports teams to all the different factors that affect attendance. They wanted to employ this model to measure expansion opportunities, to see whether the cities being proposed were likely to support good attendance. You run regarding a gazillion regression models with 60 or 70 independent variables, and find out that only one independent variable is significant -- the number of games out of first place team finished last year. The R-squared isn't great -- just under .45.
Was this study just the failure? Why or why not?