Jackson Pollock’s signature paint splatters aren’t as random as they appear. At least, according to Richard Taylor, a physics, art, and psychology professor at the University of Oregon.  In 1999, Taylor discovered that the 20th-century abstract paintings are composed of geometric patterns, called fractals, that are recognizable by computers. 

“A lot people think that you can just pour paint on a canvas, and you’ll naturally get a Pollock painting, when, in fact, it’s far from the truth,” says Taylor. After studying fractal composition and human balance, he found that “In order to replicate, you have to have his type of physiology, his same muscular build-up and height and things like that, and then you have to be willing to spend 10 years learning to refine the patterns.”  

Despite the challenges of imitating Pollock’s work, there are currently more than 400 fake Pollock paintings in circulation. This poses a problem, according to Taylor, to buyers and sellers because masterpieces like the Blue Poles, can be worth up to $600 million. 

To curb the trade of fraudulent works, Taylor developed an algorithm that can be used to verify a painting’s authenticity. “We do a very rough, cold, quantitative comparison between what the person has found in their attic and what an established Pollock is,” he says. “I’m always giving bad news, unfortunately.” According to Taylor, it’s statistically more likely to discover a replica than an original masterpiece.

When Taylor first began his research, he says he received pushback from the artistic community. “An awful lot of people said ‘You know, you just cannot do this. You can’t use a computer to quantify the magnificent emotional state of the human being,’” says Taylor. “But we’re not trying to strip away the mystique…Our computers can use fantastic computing power to actually look at the patterns on a Pollock painting, but what it can’t do, for example, is tell you why you personally should prefer a Monet over a Pollock.”