Acting success is a function of maths, not talent
Machine learning analysis fails to give hope to struggling performers, delivering only gloom and statistical despair. Andrew Masterson reports.
Success in showbusiness, at least according to cynics and more than a few out-of-work actors, does not depend on talent alone. Or possibly at all.
And now there’s some solid mathematics to back the contention up. Researchers Oliver Williams, Lucas Lacasa and Vito Lacora, all of the Queen Mary University of London in the UK, applied machine-learning analysis to the careers of more than two million performers detailed on the International Movie Database (IMDb) and came to several dispiriting conclusions.
In a paper published in the journal Nature Communications, they report that the vast majority of actors don’t get much work, successful actors tend to become more successful (at least for a while), and the chances that the best years of any actor’s career are already in the past are very high indeed.
“We show that the dynamics of job assignment is well described by a ‘rich-get-richer’ mechanism and we find that, while the percentage of a career spent active is unpredictable, such activity is clustered,” they write.
Williams and colleagues make no attempt to constrain their analysis to big name performers only. Rather, and sensibly, they define success in the film and television business as simply getting a gig – any gig, big or small, as long as it gets recorded on IMDb somewhere.
“Sustained productivity (simply making a living) is probably a better proxy for quantifying success than high impact,” they note.
To make their findings, they extracted data concerning 1,512,472 male and 896,029 female actors recorded on the database until January 16, 2016, stretching back to 1888. They then conducted an extensive statistical analysis and combed through the results.
The first outcome they noted was that one-hit wonders – actors who had work during just a single year – “are the norm rather than the exception”. Long careers and lots of gigs are “exponentially rare” and distributed according to a power law – most actors score only a few roles, while a very small number land more than 100.
And although the curse of the one-hit wonder can, and does, hit most actors, women are disproportionately represented in the category, the researchers note, “providing compelling evidence of gender bias”.
In cases where actors enjoyed roles over a number of years – although not necessarily with any consistency and thus, quite possibly, interspersed with long stretches working behind a bar or driving an Uber – Williams and colleagues found there was no way to predict when a role would have a noticeable public impact. Good gigs, it seems, can crop up at any time.
However, running counter to that observation was the finding that work tended to be clustered; that is, a performer could enjoy a hot streak, with several gigs in quick succession. However, she or he could also endure a cold streak, when grease-paint isn’t sniffed for years.
Most performers who transcended one-hit wonder status, the researchers found, enjoyed one particular period in which work was extremely plentiful. This they termed the “annus mirabilis” – which, “for both actors and actresses is located towards the beginning of their career”.
The finding suggests that for actors of all genders above a certain age, the past is always going to look better than the future – mainly because it was.
By running a statistical learning model across the data, the researchers found that it is, indeed, possible to make predictions.
“We can, with up to 85% accuracy, tell whether an actor’s career has reached its most productive year yet or not,” they write.
And talent? The analysis gives insight into career trajectories in an industry where 90% unemployment is the norm, but explicitly finds that talent has little to do with it.
“A producer might offer the job to the actor who had the best audition or to the one who has more followers on Instagram — so productivity is not only, strictly speaking, a performance-driven indicator,” they write.