I am somehow confused by the term of so called "predictive analytics". It is so often used today that makes me think the old-school statisticians never before tried to create any predictive models.
My thinking is that statisticians can do all that predictive modelers can and more, while predictive modelling focuses only on prediction part of the statistics and disregards descriptive statistics.
What are the areas where predictive analytics shine and statistical analysis will not be sufficient?