When Maps Shouldn’t Be Maps

The non-map method they show for correlating income and flooding (and a few other) is a great example. It's nothing but text yest still far more effective at explaining this particular correlation.