While countries are again preparing themselves for another year of the Eurovision Song Contest, we've updated our dataset on the scores since 1975 and ran a network analysis on the last 10 years. The results are not surprising.
As with the previous edition, two years ago, it's obvious that common history, region and current active ties have more influence on the voting behaviour than the actual performance of the artists of different countries.
We've used the data from the finales from 2008 till this year from the site of Eurovision. We included both jury and televoting.
The sizing of the nodes is based on the total amount of points received in those 10 years. The thickness of the edges is based on the number of points given from one country to another in that period.
The coloring of the nodes was done by clustering using the modularity statistics in Gephi.
Layout was created using Forced Atlas algorithm and a few small adjustments to keep it all readable.
We have the data available for all (semi-) finales since 1975. It's also on data.world and Kaggle.
