Why Eurovision Song Contest Never Fails To Entertain

The Eurovision Song Contest (ESC) has been attracting millions of fans every year since it was launched back in 1956. At the same time, it represents a unique opportunity for research. Almost no other major cultural event has been so well documented over such an extended period of time. Data on songs, voting and rule changes is freely accessible, making Eurovision an ideal example for computer-assisted social sciences of a data-based culture system.

A research team around Dirk Helbing, Professor of Computational Social Science at ETH Zurich, investigated how participating nations and organisers have learnt from one another over the decades and how this collective learning has manifested in the competition itself. The project started when complexity scientist Luis Amaral, a professor from Northwestern University, visited Helbing's team during one of the ESC competitions.

Together with Arthur Capozzi, a member of Helbing's team, the researchers analysed nearly 1,800 songs from Eurovision's 70-year history, combining classic data collection with Spotify analyses, evaluations of lyrics, AI models and language data. Each song was measured against more than 35 metrics, including danceability, acoustics, emotionality, language, genre and subject matter. "These days, anything that produces data can be investigated scientifically - that includes culture," says Helbing. The researchers published their external page study in the Journal of the Royal Society Open Science.

Everyone learns and differences fade

The researchers identified three stages of development for the Eurovision Song Contest across more than seven decades. During its "formation phase", between 1958 and 1974, the acts were incredibly diverse. Nations sang almost exclusively in their national languages, music styles differed enormously and there were no real strategies for success to speak of. The focus was less on strategising with a view to winning the competition and more about showcasing one's own culture.

In the subsequent "consolidation phase", which lasted until 2003, participating nations began to systematically learn from one another. Certain features, such as catchy melodies or lyrics that an international audience could understand, proved successful and were increasingly adopted by competitors. At the same time, the rules stabilised. The competition became more predictable.

The researchers describe the period from 2004 onwards as "expansion phase", with organisers counteracting the increasing homogeneity among performances by inviting new nations and revising the voting system on multiple occasions, making performers' chances of success less predictable. Their aim was to keep the competition exciting and varied, with the potential for the unexpected.

For Helbing, this makes Eurovision a typical example of co-evolutionary, learning systems, i.e. systems in which actors influence one another. As soon as successful strategies become entrenched, the organisers change the rules to break them up, so that the competition stays interesting and keeps evolving.

From national languages to global pop

The analysis of several decades of song data illustrates clear trends: "Over time, songs have become more mainstream pop and more danceable, but now almost all of them are sung in English," says Capozzi. These features have persisted because they have proven particularly successful over an extended period of time.

Today, however, nearly every participating nation has adopted these strategies. Thus, what was once a competitive advantage is now a baseline expectation.

The researchers call this the "Red Queen effect" after the character in the children's book Alice in Wonderland. "What was once a competitive edge is now the standard," says Capozzi. "English-language pop songs with a danceable beat become a basic requirement." So, if you want to win Eurovision, you need to have an extra, something special. Something that breaks with the norm.

Countries like France, Italy, Portugal and Spain are conspicuous in their explicit rejection of the dominant trend. "They're outliers, in that they keep on singing in their own language, even though that's not one of the established success factors," says Capozzi. The researchers' explanation for this is that these nations are specifically leveraging their cultural identity as a strategy for standing out from the crowd.

When rules learn

It's not just the participating nations that adjust their strategies - Eurovision organisers are also learning along the way. "There's no one formula for success that will always work, neither for the participating nations nor for the organisers," says Helbing. That's why the institution is undergoing targeted transformations to maintain the great interest in the competition.

One example is the introduction of the semi-finals, which were first held in 2004 in light of the growing number of participating nations. Since 2008, there have been two semi-finals - a structural adjustment to account for the expansion of the competition.

The voting system has also been modified several times. Following the introduction of televoting at the end of the 1990s and increasing criticism of tactical voting, the ESC Organisation reacted by reintroducing juries and making other adjustments. The aim was to rebalance popularity and musical scoring and to reduce the predictability of results.

For the researchers, this is also part of the co-evolutionary process: as soon as rules create undesired effects, they are adjusted. The competition learns - also on an institutional level.

What Eurovision can tell us about other systems

The relevance of these results goes beyond pop culture. The dynamics observed - adjustment, convergence, and subsequent loss of the competitive edge - can be found in lots of complex systems. Even fields of academic research have a tendency to homogenise once successful approaches are widely adopted. To get fresh impetus on a field, established trends must be questioned.

Similar patterns are found in businesses and organisations where problems cannot always be solved with common solution approaches. Sometimes, one needs to deliberately deviate from established strategies. Helbing has real-world experience of this from his time consulting a business: "The company had highly qualified engineers but they were unable to solve the problem they faced from the perspective they had learned. The company needed someone who could look at the situation from a new perspective."

For the researchers, this interplay between adaptation and mutation points to a key principle of co-evolution: actors learn from one another, reacting to changes, but at the same time this changes the system in which they are acting.

Eurovision still has a surprise factor

Although the vote share achieved by winners has remained largely stable since 1974, voting behaviour in the competition has shifted dramatically. In previous years, certain nations systematically achieved a particularly high level of success over a longer period of time. Eventually, however, the pool of winners has become more diluted.

With institutional changes, especially during the expansion phase, victories have become increasingly evenly distributed. Previously dominant nations saw their lead eroded and the winners changed more frequently. The analysis suggests that rule changes helped to level the playing field and reduce the predictability of the results.

The competition appears to continue evolving. "It has to - in order to stay interesting," says Helbing. He knows that rule changes for future editions are already in the works.

Despite his team's extensive data analysis and results, Eurovision will remain unpredictable, unless tastes or votes become manipulated. "We haven't found a formula that guarantees success," says Helbing. The Eurovision Song Contest will continue to surprise - and that's exactly where the competition excels.