Statistics and population numbers don't prove that. They're simply data points to interpret.

Statistics need context. For example, if a population drops off, where did they go? I don't recall ever seeing a good analysis on correlation/causation, just a lot of speculation. That the population left is one thing. Where they went helps tell a story. Where they came from to begin with adds even more context because we're looking at consumer behavior, not just individual data points. Did you lose a customer because they stopped playing FPS games or because they switched to a different FPS? Did you lose a customer because they didn't like your product... or because they preferred another product more because it had certain features? Satisfaction is often assumed to be negative when it was actually neutral.

The Xbox platform gives MS a chance to do see a wide array of data beyond the statistics and population data that you have. I'd love to see, if given access to the same data, how your conclusions might change.