Eurovision Song Contest research

Regular readers will know that I have been live-blogging the Eurovision Song Contest final since 2011. Every year I sit down in front of the TV, metronome and guitar in hand, and write live musical analysis of the songs in real time, then attempt to predict the winner before the voting starts. Here’s the full archive 2011-2019 (spoiler – I only managed a home run in 2015, with all the top 3 correct, in the right order). My approach (in all musicology) has always been to try to analyse and understand the underlying songs/tracks on a musical level, and move past all the camp theatricality and geopolitics. I take the view that all popular music is ‘good’ by someone’s standard, and that it’s as interesting to analyse songs from a mainstream TV event like the ESC as it is to look at more obscure/highbrow material. Those who carry the scales of musicology should be blindfolded.

In 2020, the planned contest in Rotterdam was canceled due to worldwide restrictions, but in a strange turn of events, Netflix asked me to undertake some academic research into the musical characteristics of the songs over the past decade. The research was commissioned in part as a celebration of the contest, and also to promote the release of the 2020 movie Eurovision Song Contest: The Story of Fire Saga starring Will Ferrell and Rachel McAdams. Working with my friend and colleague Simon Troup (Digital Music Art UK), we analysed the 259 finalists from 2010-2019, immersing ourselves in the corpus over an intense 15-day listening period, and extracting high-level data (key, BPM, style, lyric theme etc) to see what we could learn about voter preferences and song evolution.

We’ll be publishing a pre-publication version of the research paper very soon, and the full academic paper sometime in 2021 after peer review, so this blog post is a preview of some of the early findings, with some song excerpts, and commentary on the songs from the 2020 contest that never was, and also some analysis of the songs from the movie.

Songs for Europe: a music and lyric analysis of 259 finalists from the Eurovision Song Contest 2010-2019

A research project in progress 2020-2022

Project Summary:

Two musicologists[1] have published the early findings of academic research analysing the 259 songs that have appeared in the Eurovision Song Contest finals 2010-2019. They used a combination of computer analysis, music transcription, immersive listening, and lyric interpretation to identify every song’s lyric theme, style, and musical attributes, and then analysed the data to explore the characteristics of the songs that attract the most votes in the contest.

Headlines:

• There are six broad ‘archetypes’: Euro-pop, Ethno-pop, Ballad, Anthem, Schlager and Chanson.
• There are six broad lyric subjects: Love, Unity, Self-Assertion, Partying, History and Music, with love songs accounting for 69% of the whole, and 83% of the top 3.
• The most popular styles are Euro-pop, Ballad, Ethno-pop and Anthem, accounting for 79% of the total. The older styles (Schlager and Chanson) account for less than 5% of finalists.
• The classic Eurovision cliché of a key-change in the final choruses is alive and well, appearing in almost 20% of the finalists, but not among any of the winners.
• Eurovision appears to enjoy sadder songs in recent years – “winter” love songs have been more popular since 2018.
• The most successful finalists are Azerbaijan and Sweden, who qualified 9 times out of 10 (UK, Germany, Spain, Italy and France aka the ‘Big Five’ are guaranteed a place in the final).
• 65% of all songs were in a minor key; of these, more than half used the Aeolian mode aka natural minor scale.
• The mean average tempo is around 104 BPM, although the actual tempos tend to group around 70BPM (Ballads) and 125 BPM (Euro-pop).
• Eurovision winners are slowing down; the average tempo of the top 3 scoring songs dropped from 148 BPM (2010) to 76BPM (2019), helped by some successful Ballads in 2017 and 2019.
• Many popular styles of music outside Eurovision (metal, hip-hop, rap, punk, trap, country, techno) are almost completely absent from the contest, although they have an influence on the production of the archetype styles. Eurovision song styles appear to have ‘evolved’, in the cultural Darwinism sense, independently from mainstream global pop music. The researchers speculate that this is due to the particular cultural and structural factors at play in the Contest, in contrast to the more unregulated market forces that decide the popularity of mainstream hits.

Eurovision archetypes:

• Euro-pop: a danceable song in 4/4 time, faster than 120 beats per minute (BPM).
• Ethno-pop: A fusion of traditional music from the home country and contemporary pop.
• Ballad: a slow, deeply emotional song (40-85BPM) featuring a quiet introduction and finishing with a loud double chorus.
• Anthem: A mid-tempo song (85-119 BPM) with a big chorus.
• Schlager: An ‘easy listening’ singalong, with a sentimental lyric theme and short, cutesy melodic lines.
• Chanson: A poignant, delicate slow Ballad without the crescendo.

• [Niche]: A minority of the songs opt for a particular real-world style – punk, metal, rock, opera, folk-pop etc. We classified these as ‘niche’.

Lyric themes:

• LOVE
  • (Spring): Looking for love: flirting with someone, or searching for the One. Examples include Love Me Back (Turkey, 7th place, 2012) and Popular (Sweden, 3rd place, 2011).
  • (Summer): Being in love: celebrating the relationship, or celebrating love itself. Examples include Beautiful Mess (Bulgaria, 2nd place, 2017) and Grande Amore (Italy, 3rd place, 2015).
  • (Autumn): Love on the rocks. Eurovision examples include Satellite (Germany, winner, 2010) and Too Late For Love (Sweden, 5th place, 2019).
  • (Winter): After love: loneliness, regret or heartache. Examples include Arcade (Netherlands, winner 2019) and Nobody But You (Austria, 3rd place, 2018).
• DANCE PARTY:
  • Dance and partying themes. Examples include Alcohol is Free (Greece, 6th place, 2013) and Party For Everybody (Russia, 2nd place, 2012).
• UNITY:
  • Concepts relating to groups of people working together for a greater good. Eurovision examples include Heroes (Sweden, winner, 2015) and A Million Voices (Russia, 2nd place, 2015).
• MUSIC:
  • Songs about listening to, creating or performing music. Examples include That’s How You Write a Song (Norway, 15th place, 2018) and Me and My Guitar (Belgium, 6th place, 2010).  
• SELF-ASSERTION:
  • The protagonist asserts their confidence and spirit in the face of life’s tribulations. Can be first-person or second-person. Examples include Rise Like a Phoenix (Austria, winner 2014) and Yodel It! (Romania, 7th place, 2017).
• HISTORY:
  • An historical event in the entrant’s home country, or across Europe. Examples include 1944 (Ukraine, winner, 2016), Apricot Stone (Armenia, 7th place, 2010) and Non Mi Avete Fatto Niente (Italy, 5th place, 2018).

Commentary: “Eurovision Song Contest: the story of Fire Saga” (Netflix, 2020)

The movie, which tells the story of Icelandic Euro-pop due Fire Saga’s journey to Eurovision, features original songs whose styles would be very recognisable to ESC fans. Here are our best-guess predictions of how each song might fare in real-world voting.

Ja Ja Ding Dong (old-school Eurovision crowd-pleaser)

This is the song that Fire Saga are forced to play over and over in their covers set, because the Icelandic locals demand it. It’s a classic example of old-school Eurovision Schlager – an oom-pah singalong tune with nonsense words and a happy sentiment. Ja Ja Ding Dong might have done well in the 1970s, but in modern Eurovision it would almost certainly bomb. Lars and Sigrit want to move on from these old clichés and concentrate on their own new material.

Looking In The Mirror (Icelandic contender)

Early in the movie, there’s a meeting at Icelandic Public Television centre in Reykjavik, where we hear the management team’s favoured candidate for the Eurovision entry: Looking In The Mirror, performed by Katiana Lindsdóttir (played by Demi Lovato). It’s a slick slice of Scandi-pop with Swedish influences (rarely a bad thing at Eurovision). At 106BPM it’s a little slow to raise the roof of the stadium, but in our opinion the TV execs have the right idea – it would surely be good for Iceland to enter something like this.

Double Trouble (Icelandic entry)

This is the song that Fire Saga play in the run-up to the Eurovision final (set in Edinburgh in the movie). It opens deceptively as a pastoral minor-key Ballad, but soon kicks into Euro-pop territory with a four-on-the-floor electro beat at 125BPM. Iceland have never won the Contest, and often finish near the bottom of the voting; their best placing in the last ten years was 10^th (2019). Double Trouble is good enough to be a finalist, but we wouldn’t see it winning (slight spoiler: we think Lars and Sigrit might have a better song inside them).

Lion Of Love (Russian entry)

This up-tempo flamenco/disco mashup is sung by the super-slick Russian tenor Alexander Lemtov (played by Dan Stevens), and it’s one of the favourites to win. It’s the perfect showcase for Alex’s vocal prowess and love-god image. The chorus opens with “let’s get together, I’m a lion lover and I hunt for love…”. The Gm-D chords give us a harmonic minor tonality, which has a good track record in the final. At 140BPM it’s at the upper end of the successful tempo spectrum, and the lyric metaphor might be a bit too much for some voters, but we agree it would do well.

Running With The Wolves (Belarus entry – setlist opener)

Don’t be fooled by the horned helmet, skull costume and rock guitars – this is pure Euro-pop in wolf’s clothing. The tempo is a highly danceable 128BPM: Graham Norton (as himself) comments “a very fast start to this year’s competition”. Finnish band Lordi famously pulled off a pop-metal coup, winning in 2006 with Hard Rock Hallelujah. Much as we love to see any hint of rock at Eurovision, we suspect that tastes have probably moved on.

Other (fictional) finalists

Finland’s Fool Moon is solid feelgood high-energy disco, but too predictable to really stand out. Hit My Itch (San Marino) is stronger, with some refreshing soul harmony vocals and a wonderful flat-9 chord supporting the high vocal note before the chorus (which Graham calls “crying for no apparent reason”). Mita Xenakis’s Masquerade is a great 12/8 minor-key fast dance number that we think would score well, with its sci-fi vamp staging and cool chromatic chorus melody, but it might miss out on the top spot due to its inter-chorus key change – we’ve not seen one of these in a winner for more than a decade. Sweden’s Johnny John John performs the possibly problematic Coolin’ With Da Homies; in the movie it functions as a brilliant satire on Eurovision’s brave but usually unsuccessful attempts to incorporate hip-hop into the contest.

My Home Town (Sigrit’s song)

This is the beautiful Ballad that Sigrit is writing through most of the movie. It’s partly a ‘Spring’ love song, where the protagonist is asking for love “all I need is you and me”, plus some ‘Unity’ themes with a hint of environmentalism “where the whales can live ‘cause they’re gentle people”. At around 66 BPM, with big timpani rolls and a dramatic build towards the end, it would be a crowd-pleaser on the night. We don’t think it’s close enough to a single archetype to win, but we’d put it in the top half of the voting.

Commentary: the 2020 contest (that never was)

Sadly, the 2020 contest in Rotterdam did not go ahead due to the global pandemic, so to make up for the lack of real-world voters, we listened to these 41 entrants, and added some comments and scores of our own.

UK: My Last Breath (James Newman)

The chorus opens with the unusual and creative (but perhaps unrelatable) hypothetical metaphor “If we were deep sea divers”. It’s an anthem in the sense that it’s mid-tempo (89.7 BPM) and also due to the nicely accessible “whoa…” chant at the end of each chorus. But we worry that the one-note melody hook isn’t striking enough to lift this higher than the middle of the scoring. They shot the video in a snowy forest, which is a nice nod to the Scandinavians.

Score: 5 points

Ireland: Story Of My Life (Lesley Roy)

We love this one, and we think it would score well. It’s classic all-synth Scandi-pop, with echoes of real-world bubblegum hits (it reminds us of Carly Rae Jepsen’s 2011 hit Call Me Maybe). At 132 BPM it’s a tiny bit faster than the average for this type of song, but the upbeat major-key sentiment and simple scalic chorus melody make it really accessible. Our best guess for the voting would be in the top 5.

Score: 8 points

Croatia: Divlji Vjetre (Damir Kedžo)

Divlji Vjetre seems to be an attempt by the Croatians to have their cake and eat it – it’s a beat that can be felt at both Ballad and Euro-pop tempo, but in our opinion doesn’t succeed fully in either. The song deploys the well-loved chord loop vi-IV-I-V (6-4-1-5) for most of the chorus, with a semitone key change at the end. Damir delivers a great soaring vocal, but the song doesn’t easily stand out from similar material elsewhere in the pack.

Score: 1 point

Denmark: Yes (Ben and Tan)

This is a really nice major-key anthem, and the verses are bravely in 7/4 time. The tempo works well for the verses, but in the chorus the words come far too fast, so the performers Ben and Tan fall over all their consonants, despite super-clear Danish diction. We like the hooks a lot – both the title vocal and the major-pentatonic instrumental theme at the end of each chorus. But there is no recent track record of a song like this doing well.

Score: 4 points

Georgia: Take Me As I Am (Tornike Kipiani)

An all-too-rare rock Ballad, which statistically harms its chances. The performance tempo is incredibly slow and ponderous, and we suspect this sinks it. Great vocals though. The lyric covers a lot of European ground “How do you want me to talk, like an Englishman? Where do you want me to dress like an Italian? Now do you want me to dance like a Spanish guy? / do you want me to smell like a French homme? / do you want me to play like a German?”, and is quintilingual “ti amo, te quiero / Je t’aime, ich liebe dich […] I love you”. So a good score from the geography and languages departments, but not so much from the voters.

Score: 3 points

Latvia: Still Breathing (Samantha Tīna)

An incredibly strange mashup of trip-hop, rap, drum ‘n’ bass, and Euro-pop, played at medium Ballad tempo, with one-note melody lines in the chorus. Includes not one but three inexplicable synth breakdown sections. What were they thinking?

Score: 0 points

Our winner: Iceland!

Iceland: Think About Things (Daði og Gagnamagnið)

Iceland’s Daði Freyr is a real-life Lars type, having grown up in a small Icelandic town and tried for several years to achieve Eurovision success, eventually winning the country’s nomination with a band made up of local friends and family. Think About Things is a cheerful minor-key electro-pop 127BPM dance number with an 80s synth production, augmented with snappy brass riffs and a charming dance routine, under Daði’s deadpan understated vocal. It’s slightly tongue-in-cheek, which helps it to stand out from all the straight-face dance music we hear in the finals. There’s a classic-era whole-tone modulation in the final chorus, which we haven’t seen in a winner in more than 10 years. Perhaps the time has come.

Score: 12 points

Appendix 1: ACADEMIC ABSTRACT

This corpus study provides an overview of the common musical and lyric characteristics of songs from the Eurovision Song Contest (ESC) during the past decade. It uses as its dataset 259 songs, comprising the 25-27 finalists in each year’s Grand Final 2010-2019. Hard data includes high-level musical characteristics such as tempo, key, tonality, dynamic range, and melodic pitch range. Soft data includes lyric themes and musical style. Song styles were categorized according to six ‘archetypes’, adapted from existing categories suggested by previous ESC journalism and academic writing.^[2] Lyric themes were categorized according to seven core themes identified by the researchers through detailed listening and lyric analysis. The full corpus was analysed, to discuss common musical and lyric characteristics of each archetype, and high and low-scoring subgroups were compared, to investigate whether there are common music or lyric characteristics that correlate with success (measured in numbers of aggregated points per song). Methodology included computer-aided audio analysis (BPM detection, dynamic range), subjective listening (archetype and lyric classification), music analysis (key centre, tonality) and statistical analysis of the dataset.

The research, which is an ongoing project 2020-2021, aims to identify the high-level characteristics (of tonality, archetype, lyric theme, dynamic range, tempo etc) exhibited by finalists, and to investigate any characteristics that differentiate winners and other high-scoring songs.

Appendix 2: Data visualisations (selected)

Data visualisations may be used freely under Creative Commons ‘Attribution CC BY’ terms. Credit: Digital Music Art / Joe Bennett

[1]   Professor Joe Bennett, Berklee College of Music USA, and Simon Troup, Digital Music Art UK. Thanks to Milton Mermikides for additional work and advice. This document, and the images, may be distributed under Creative Commons CC BY, and may be used freely with attribution. The actual research paper and full dataset is © Joe Bennett and Simon Troup 2020, and is pre-publication at the time of writing, pending journal submission and peer review. Contact via joebennett.net.

^[2]   Jess Carniel, “‘Schlager’, Scandi-Pop and Sparkles: Your Guide to the Musical Styles of Eurovision,” The Conversation, May 9, 2018, http://theconversation.com/schlager-scandi-pop-and-sparkles-your-guide-to-the-musical-styles-of-eurovision-96268.