Edinburgh Fringe fame for a one woman show is proof of the randomness of celebrity
The world of the arts is one of inequalities: very few do extremely well, and most do badly. It’s on show now at the Edinburgh Fringe, and it proves that celebrity is almost accidental, writes Paul Ormerod
The Edinburgh Festival is in full swing and the number of performers is almost incredible: the overall total this year is estimated to be around 50,000.
A small fraction are already established names; an even smaller fraction may become famous as the festival unfolds. But most will return home just as unknown as when they arrived.
There was an outpouring of support – and subsequent publicity – for performer Georgie Grier after she tweeted that only one person had turned up to watch her show. She now seems to have become something of a celebrity and wound up performing to a packed house.
This is very much the exception rather than the rule. Those responding to the tweet had not seen Georgie’s show; they would have been completely unaware of its inherent quality. So following basic rational choice theory, in which people evaluate the attributes of a product or service and see how it matches their own preferences before forming a view, endorsing the show makes little sense.
This small vignette tells us a great deal about the artistic sector as a whole. The arts are inherently massively inegalitarian. A select few do extremely well and most do badly, very badly indeed.
The work of Duncan Watts helps us understand why this is the case. Watts, who was originally at Columbia University, went on to be paid huge salaries by Yahoo and Microsoft and is now ensconced back in academia.
Twenty years ago he published a paper in the prestigious Proceedings of the National Academy of Sciences with the title “A simple model of global cascades on random networks”.
Watts set up a computer model in which the agents could make one of two choices. For example, you could either go or not go to a particular show at the Edinburgh Festival. Your decision to go wouldn’t be based in any way on the quality of the show; it would be determined by the proportion of people who have seen the show and could influence your choice. Given Grier’s experience, this seems a perfectly reasonable assumption to make.
Watt’s model started with no one having been to the show. Then a few are selected at random to go. Watts then measured, across a large number of individual solutions of his model, how many eventually went. In any context in which the opinions of others have a strong impact on the choices a person makes, the eventual outcome will be very unequal.
The results were intriguing. Most of the time, very few people would go. But very occasionally, huge numbers do. This is exactly like the real life outcomes in the arts.
Watts then followed it up a few years later with some experiments. He got a group of students to listen, one at a time, to 48 music tracks and to rate them and select the ones they really liked. In one set, they just listened to bits of the music. Overall, some were more popular than others, but none stood out dramatically. In another set, students were told what others had selected before them. In this case, a few were hugely popular and most were not. The choices of others had a dramatic impact on the outcome.
This inequality is a deep seated feature of any part of the economy in which the choice which you make is significantly affected by what others do – especially the arts.