"But the challenge was to combine the data in such a way to make not only good looking maps, but ones that made sense. Mr. Tait said, “We always started with the data and often had two interesting variables or two variations of a single variable (density and number of or per capita and number of) that we wanted to show. They often get two different aspects of a phenomenon. For example, quarterbacks by state is shown in the cartograms as size equals per capita but it is also interesting to know number of so we colored them that way. For the bivariate choropleth maps (Golf and Olympic athletes) the desire was to get at economic factors helping explain golf course distribution in the one case and physical geographic factors helping explain athlete distribution in the other.""