Maybe I'm just slow...it's probably that I'm slow, but when I started in R I was dumbfounded by the way people would gush about R's visualization abilities. It never made sense to me. The visualizations, to be frank, looked like they came off the Tandy x286 that my mom could never bear to throw away. I certainly would never show them to others.
Then again, maybe I'm not the only one who is slow when it comes to understanding the purpose of our tools. On the internet I've read blogposts decrying the data loss that occurs when information is presented via bar chart. "NEVER USE A BAR CHART", they scream. I've read long threads on message boards with angry Tableau users who can't understand why their boss doesn't want anything to do with interactive dashboards.
But slowly it occurred to me: Many of the visualizations that we use to explore data are intending for us, the data explorers.
It's okay for R visualizations to be ugly. They're for me.
It's okay to use a bar chart, if it gets a valid point across.
It's okay to show overly simplified visualizations to others, because in terms of communication, these visualizations exist to make your point easier to convey.
Figures that are used in documents (or in live presentations) exist to facilitate discussion. These forms of visualization should make your point understandable at a fundamental, almost emotional, level. They should ease cognitive strain, not induce it.
All figures are visualizations, but not all visualizations should be used figures.
Comments