I have a friend who is a statistician and has published a couple of books and has a well-regarded blog, teaches at an Ivy League, etc. His general thesis is that raw numbers/data can be made to support just about anything.
https://junkcharts.typepad.com/