Listen up, record executives. A new computer program could put you out of a job. The creators of the software, developed at Britain's University of Bristol, claim it can use algorithms to determine whether a song will reach the Top 40. Is pop music really that predictable? Here, a brief guide to the technology:
How can software spot a hit?
It rates a song based on 23 separate characteristics, such as tempo, time signature, song duration, and loudness. Then, using the Top 40 charts from the past 50 years, it inputs those figures into a "hit potential" equation, which then classifies a song as either a hit or a non-hit. The goal was to find "an equation that distinguishes between a hit and something that dangles at the bottom of the charts," lead researcher Tijl De Bie tells BBC News. "We can expect to get it right in about 60 percent of cases."
Isn't pop music always changing, though?
It is. Which is why, to increase accuracy, the team's software takes shifting musical trends into account, something similar experiments have overlooked. For example, in the '50s and '60s, "harmonically simple songs were more popular," says Sebastian Anthony at Extreme Tech. And before the '80s, "the danceability of a song wasn't very important — but ever since, it has remained a very significant factor." One of the most important factors to a hit song now? "Rather depressingly," loudness.
Okay, but isn't 60 percent kind of low?
De Bie asserts that those odds are favorable because the equation can't account for factors outside the song itself (like the amount of marketing a song receives from a major label). And there will always be select tracks that don't fit into an era's standard hit profile, but end up as major successes anyway, including Guns N' Roses' 1992 "November Rain" and Bloc Party's 2005 sleeper "So Here We Are." (See a comprehensive list here.) "At every moment in time the equation can be different," notes a cautious De Bie. "It's not perfect."