Skip Navigation

Physicists uncover Eurovision biases

May 20, 2005 By Philip Ball This article courtesy of Nature News.

Voting analysis reveals European cliques and feuds.

Britain is in harmony with Europe, Nordic countries fancy each others' stars, and France is out on a limb. That's the verdict of a group of physicists who have studied the much loved and loathed Eurovision Song Contest.

This annual event pits musical groups from each of 20 or so European countries against one another in an orgy of kitsch costumes, unique dance routines and songs with titles such as Diggi-Loo Diggi-Ley. To non-Europeans it might seem a puzzling and even vaguely disturbing cultural phenomenon. But to the contestants, it is serious stuff.

In 1974 Eurovision shot the Swedish group Abba to international fame when their song Waterloo triumphed. This year's contest, held in Kiev on 21 May and marking the event's 50th anniversary, will be broadcast to an audience approaching 1 billion people.

But Neil Johnson and his co-workers at the University of Oxford, UK, think that the significance of the Eurovision Song Contest goes deeper than music. They regard it as a barometer of European nations' feelings about their neighbours, and thus as a testing ground for compatibility as the creation of a European constitution looms.

Null hypothesis

In the contest, each country performs a song, which is awarded points by all the others; the one that gathers the most points overall is the winner. The system sparks bitter accusations of reciprocal voting: for example, some accuse the Scandinavian states of favouring each other. Others think that the votes reflect traditional antipathies between adjoining countries, meaning the Dutch won't vote for the Belgians.

"This is arguably the only major international forum in which a given country can express its opinion about another, free of economic or governmental bias," the researchers say. They argue that a systematic pattern in voting can be revealing and have analysed vote networks between 1992 and 2003. Each country represents a point in the network, and links between them are forged when one country awards points to another.

Statistically, these networks turn out to be far from random: some countries form cliques that tend to vote in the same way and for each other. This is most evident for Greece and Cyprus, but the Nordic countries also tend to operate a block vote. Geographical proximity doesn't guarantee harmony, however: Spain generally doesn't vote the same way as Portugal.

Fair exchange

Johnson and colleagues argue that the Eurovision Song Contest is an example of an international marketplace in which players repeatedly exchange goods. In this case the goods are votes, but "they could equally well be ideas, opinions, money or food". They claim to have found a measure of how close each nation feels to the rest of Europe.

And to their surprise, it seems that Britain is the most 'European', despite the fact that it is an island. France, which is often thought of as a 'core' European state, shows the least compatibility.

The researchers admit that their analysis is based on one contentious assumption: that all the songs presented are equally good, so that votes are a reflection of national taste rather than the absolute quality of the entries. However, for many aficionados, assessing the awfulness of the entries is the whole point.


  1. Fenn D., Shleman O., Efstathiou J. & Johnson N. F. Arxiv, (2005).


Need Assistance?

If you need help or have a question please use the links below to help resolve your problem.