science

McConaughey or Culkin? Algorithm predicts actors' peak years


Macaulay Culkin peaked young, Jane Lynch hit the big time in her late 40s and Matthew McConaughey’s mid-career resurgence is the stuff of legend.

Now researchers say they have developed an algorithm that predicts with 85% accuracy whether an actor is yet to have their most productive year, or whether they have already peaked.

While actors and agents alike might be keen to get their hands on such a crystal ball, the researchers say there is little mystery: an actor’s best year is usually preceded by a steady rise in the amount of work they are getting.

Oliver Williams, the first author of the study from Queen Mary University of London, said: “If I were to give a piece of advice based on my findings, I would say just do more jobs and you’ll get more jobs.”

The research may end up having a star turn itself. Williams said a screenwriter was considering using the work as part of a film about an actor whose career is in the doldrums and seeks help from a scientist to make her comeback.

Writing in the journal Nature Communications, the team reports how they trawled through the International Movie Database (IMDb), looking at the careers of actors from 1888 to 2016 – including more than 1.5 million men and almost 900,000 women. For each actor, the number of annual credits was recorded and the data analysed.

The team found that more than two-thirds of actors had credits for only one year, perhaps suggesting that having dipped a toe in thespian waters, many had a change of heart. That might not be surprising – the authors said previous work had suggested just 2% of actors could make a living from the job.

The results show that few actors had either a long or active career, and that women had shorter careers than men.

“There is a sign of gender bias,” Williams said, adding that, while the data shows actors tend to go through phases – with clusters of credits and career droughts – “we see that men tend to recover from cold streaks better than women do”.

Williams said the finding that few actors had high productivity and many were “one-hit-wonders” could be explained by a sort of “rich get richer” effect. In other words, the more credits actors have, the more jobs they get. “It is not purely meritocratic,” said Williams.

They then looked at actors with careers of at least 20 years and whose peak year had at least five credits, to explore patterns in when an actor had their most productive year. For both men and women, their peak year was more likely to occur in the first part of their career than later – with the effect being stronger in women than men.

The team found that by building a model based just on the number of acting credits per year in this subset of actors, they could predict with about 69-75% accuracy if an actor had already had their most productive year, or if the best was yet to come. They then improved the model by training it with chunks of data from before and after the actors’ “best year”, allowing them to develop an algorithm that accounted for the ups and downs of actors’ careers.

The resulting model was able to predict with about 85% accuracy whether an actor had already had their most productive year. The team said some of those that were misclassified were “comeback cases”, which were “fundamentally difficult to predict”– as those watching McConaughey’s career might agree.

But the study has limitations, not least that the team defined an actor’s annus mirabilis by the number of acting credits. The team argue that for most actors, the key measure of success is making a living from the job, but many might consider high-profile roles in two major films preferable to a large number of smaller, more obscure parts.

“This work is really focused on productivity being what you want to achieve,” said Williams, adding that it was not clear if the same results would hold for winning awards. “This is something that maybe should be looked at in the future.”



READ SOURCE

Leave a Reply

This website uses cookies. By continuing to use this site, you accept our use of cookies.