Quantifying Artistic Success

March 03 2014

NPR today covers Princeton professor Matthew Salganik's project to create parallel worlds in service of discovering whether good art is popular... or popular art is good.

To see the role of chance you need to see multiple realizations of the same process," Salganik explains. "But we only get to see one outcome. So we see the world where the Mona Lisa is one of the most famous paintings, and it's hard to imagine that something different could have happened."

But Salganik is good at computers, so he came up with a plan.

He would create a series of identical worlds online filled with the same pieces of art, then get thousands of people to choose which they liked best.

If the same art rose to the top of every world, then he would know that success was driven by the inherent qualities of that work. If not, he could conclude, success was essentially random.

"We have the chance of really seeing — as much as we possibly can — parallel versions of history. So rather than trying to argue like that, we just said, 'Let's just create these parallel worlds and see what happens.' "

His experiment shows that once a basic standard of quality is achieved, "what becomes a huge hit and what doesn't is essentially a matter of chance". Read the full story at NPR: Good Art is Popular Because It's Good, Right?

In respect to our algorithm, such research goes far in vindicating the ideological possibility of quantifying the future success of art. The "chance" component is easily quantifiable using vast data sets depicting increasing or decreasing interest across social media, private collections, museums and publications.

Quantifying Artistic Success

Subscribe

Interested in buying works by artist?