A1392
Title: A network analysis of success at the Eurovision song contest
Authors: Alessia Paccagnini - University College Dublin (Ireland) [presenting]
Alessia Morrone - University of Calabria (Italy)
Barbara Bedowska-Sojka - Poznan University of Economics (Poland)
Sabrina Giordano - University of Calabria (Italy)
Claudia Tarantola - University of Pavia (Italy)
Abstract: What determines the success of a song at the Eurovision Song Contest? Is it driven by individual taste and geopolitical alliances, or do shared musical structures shape audience preferences? The purpose is to use the Eurovision Song Contest as a natural laboratory to explore the complex interdependencies between music, culture, and politics through the lens of network analysis. Using a purpose-built dataset, two types of networks are constructed: undirected similarity networks connecting songs based on musical, lyrical, and performative features, and bipartite networks linking voting countries to the songs they voted for. Centrality measures and clustering algorithms are applied to uncover structural patterns within these networks. Findings reveal that winning entries are often not the most central in the similarity networks. Instead, they tend to occupy peripheral positions, distinguishing themselves through unconventional features. At the same time, recurring characteristics emerge among highly voted songs: the use of English lyrics, minor keys, medium-to-high tempo (BPM), and strong visual performances. Furthermore, unexpected voting affinities are detected between countries with limited historical or cultural ties. It is illustrated how network-based approaches can uncover latent structures in cultural phenomena, offering an original perspective on collective behavior, aesthetic preferences, and strategic alliances in an international contest.
