SONGS 2000-2018
SIZE: SPEECHINESS
OPACITY: LOUDNESS
WINNERS 2000-2018
SONGS 2018
WIDTH: COULD IT WIN?
HEIGHT: COULD IT BE A SUMMER HIT
We analyzed every Eurovision contestant available on Spotify since 2000. By mining nine features of the songs for mere research purposes we used a Random Forest classifier to determine what set of values make a Eurovision’s success. The model only shows a propensity to be a success, since winning in this international contest has many other additional considerations embedded, such as the celebration of the contest itself, the politics of the vote, the popularity of the contest in each country and the language of the lyrics, we don’t assume this model can tell who is most likely to win Eurovision. In order to get a better picture of what is a successful song, we trained a similar classifier with the list of Spanish radio hits, labeling those who where summer hits.
We wanted to show the value of a deeper data sensorization to enrich the way we perceive and understand the world we live in. The ability to have analytic data about songs, not openly and massively available a few years back, open a whole new set of opportunities to create new analytics tools or event failure-proof songs.