Sometimes the best inventions and ideas come at us unexpectedly. The Slinky. Post-It notes. Penicillin. The same could be true of art, as the Tumblr accidental aRt proves.
The site features submissions of data visualizations that went wrong and turned out to be beautiful. As the blog Flowing Data put it, "There comes a time late at night when your screen grows fuzzy and the code runs together. Mistakes happen, and with visualization, the bugs often manifest themselves into abstract images that sort of resemble data."
The submissions come from data geeks who describe their fortuitous blunders. From one submission: "I tried to plot ellipses around groups of points from the same survey site, but ended up with this beautiful stained glass-esque plot." The capital "R" in the name, accidental aRt, is a reference to the software, called "R," that many statisticians use to run models and visualize their data.
Accidental aRt was started by two data analysts Kara Woo, an information manager at the National Center for Ecological Analysis & Synthesis, and Erika Mudrak, a statistical consultant at Cornell University. One day Woo tweeted a graph "gone awry," says Woo, and Mudrak tweeted back that she also had saved many such mistakes and had been meaning to post them.
Their site is an example of how researchers might do a better job celebrating failures or ideas that don't work out as planned. From what I understand, that’s an important part of staying both sane and innovative in many research fields.