AI correctly predicts Childish Gambino Grammy win
AI correctly predicts Childish Gambino Grammy win


US-based DataRobot compares all previous Song of the Year winners with this year's nominees, correctly predicting Childish Gambino's victory for 'This Is America'

MANILA, Philippines – US-based data analytics startup DataRobot successfully predicted Childish Gambino’s “This Is America” Song of the Year Grammy award in the 2019 show.

Three days prior to the actual awards, the company published its findings, analyzing the various characteristics of this year’s nominees via Spotify.

DataRobot data scientist Taylor Larkin said they looked at the “genre of the song, amount of profanity, general sentiment, total word count in the song, and various audio features derived by Spotify.” (LIST: Winners, Grammy Awards 2019)

Using DataRobot’s tools, the company then analyzed all the Song of the Year winners since 1959, and identified the traits of the winning songs. Seeing these traits, DataRobot’s team then identified which ones in this year’s nominees would be closest to the winning formula.

The “various audio features” that Spotify offers to anyone interested are technical details such as time signature, song key, duration, acousticness, danceability, energy, instrumentalness, liveliness, loudness, speechiness, valence, and tempo. 

Using these factors, DataRobot found that these were the most likely to win:

  • “This Is America” by Childish Gambino – 20.38%
  • “Shallow” by Lady Gaga and Bradley Cooper – 19.17%
  • “All the Stars” by Kendrick Lamar and SZA – 17.21%
  • “The Middle” by Zedd and Maren Morris – 16.47%
  • “The Joke” by Brandi Carlile – 15.55%
  • “God’s Plan” by Drake – 15.48%
  • “Boo’d Up” by Ella Mai – 14.54%
  • “In My Blood” by Shawn Mendes – 13.08%

Days later, the prediction would prove accurate as Gambino’s “This Is America” took home the Song of the Year award on Monday, February 11, along with the Record of the Year, Best Music Video, and Best Rap/Sung Performance.

As to why they made the analysis? Larkin says that this demonstrates that “machine learning cannot only be fun but can also have applications well beyond the traditional ones we are used to seeing in fields such as banking or insurance.”

“The music industry could tap into its potential, studying what makes a song successful and understanding why people listen to the songs that they do. With the volume of great music being produced, having quick insights into song popularity could be another tool to help musicians and music producers to refine their expertise,” he finished. –

Add a comment

Sort by

There are no comments yet. Add your comment to start the conversation.