I worked on a project with a guy who did network analysis like this. It was pretty cool... the first three times. After that it seemed like every problem had to

I worked on a project with a guy who did network analysis like this. It was pretty cool... the first three times. After that it seemed like every problem had to involve a whizbang graphic - then I realized he didn't actually know how to program at all. His whole job was 'analysis', which was using Tableau and a graph partitioning program to produce visualizations (pretty much using the defaults), and then try to explain what they meant. I was kind of in awe that this was a whole job by itself - I've never really seen an analysis of one of these graph diagrams that wasn't really superficial, like the examples in this presentation. It kind of makes me wonder if the "complexity" of the graph visualizations doesn't cow the viewer a bit, and make them more impressed by subsequent, not-super-interesting insights. I'd be interested in a comparison with and without the graphics.