The song of the year Grammy winner, predicted using data science

The folks at Musixmatch have gone into the lab to determine which song will be named Song Of The Year at this year’s Grammy Awards. They’ve looked at all 58 winners and 237 non-winning nominees to see what they have in common.

There are a lot of facts jammed into their findings. We learn that nominated songs have gotten wordier and louder over the years, and that they were getting longer until the mid-’90s, when the trend reversed. The songs have also increased in what they call “danceability” over time. (“Danceability,” for those of you who are curious, is a “combination of musical elements that best characterize danceability include tempo, rhythm stability, beat strength, and overall regularity.”)

Interestingly, A-major is the most common key for winners, but C-major is the most common key for nominees. Only one song with profanity has ever won. (That was Adele’s “Rolling In The Deep.”)

We won’t tell you their prediction for the winner—you can go to their site for that—but we will tell you that “I’d Do Anything For Love (But I Won’t Do That)” was both the longest and wordiest song every nominated.

 
Join the discussion...