Go-Jo, a rising pop star from WA, will represent Australia at the 69th Eurovision Song Contest in Switzerland in May.Credit: James Brickwood
Though the pomp and pageantry dominates most discussion of Eurovision, the race for victory in the world’s oldest singing competition is just that: a race. Many come to sing, but only one walks away at the end of the week with a trophy in hand.
Delegations from 37 countries are, as you read this, converging on Basel, Switzerland for the 69th annual Eurovision Song Contest. Montenegro is returning after a two-year absence. Moldova has withdrawn for financial reasons. And Australia, the rest-of-the-world’s wild card, is competing for the 10th time. We’re in it to win it; to suggest otherwise would be naive.
Of course, there is the soft diplomacy of the event, in which feuding nations cast aside real-world conflicts and step into a musical fantasy of equality and unity. But make no mistake, somewhere around the sidelines the delegation heads are pep-talking like cornermen in a boxing ring. Setting aside generosity of spirit for a moment, Eurovision is an exercise in national pride. Nobody minds coming 12th. Nobody wants to come last. And everybody – everybody – wants to win.
So, who will? Well, the answer might be easier to calculate than many realise.
The most solid working theory, which we are going to call the Goldilocks Theory, goes something like this: a Eurovision-winning song should be fun, but not so much fun that it’s annoying; memorable, but not unforgettably awful to the point you’re watching it with wide eyes and an open mouth; and mad, but not so mad that it’s actually too nuts, even by Eurovision’s standards.
And while that explains some Eurovision winners, it does not always explain them all. Monster-masked glam rockers? Dancing grannies? Songs with titles such as Diggi-Loo Diggi-Ley and Ding-a-Dong?
There are general trends in genre: pop does well, so do ballads. Rock and rap, not so much. Cultural connections can be impactful, but they are not required for a winning song. Singing in multiple languages will also not deliver an avalanche of votes, but it’s a nice touch, and Eurovision fans seem to like it.
The 2025 Eurovision presenters, Tony Armstrong and Courtney Act.Credit: SBS
And there is one X-factor neither the country nor artist can control: where they fall in the randomly drawn show order. What is certain, though, is that nobody can deny it is immensely helpful to be the song that just left the stage when the audience start voting.
Unpacking the data is complicated because the history of Eurovision is long, and trends in music make the pop charts of the 1960s and the pop charts of the 2020s uncomfortable statistical bedfellows. So let’s narrow it down to the past decade, excluding 2020 when the contest was cancelled due to the COVID-19 pandemic.
That makes our sample: Rise Like a Phoenix (Conchita Wurst, Austria, 2014), Heroes (Mans Zelmerlow, Sweden, 2015), 1944 (Jamala, Ukraine, 2016), Amar Pelos Dois (Salvador Sobral, Portugal, 2017), Toy (Netta, Israel, 2018), Arcade (Duncan Laurence, Netherlands, 2019), Zitti e Buoni (Maneskin, Italy, 2021), Stefania (Kalush Orchestra, Ukraine, 2022), Tattoo (Loreen, Sweden, 2023) and The Code (Nemo, Switzerland, 2024).
Eight out of the 10 of those artists – Conchita, Mans, Jamala, Salvador, Netta, Duncan, Loreen and Nemo – were soloists. Only two – Maneskin and Kalush Orchestra – were groups. What are the numbers telling you? Groups don’t generally win (sorry, ABBA), which is good news for Australia’s Go-Jo, aka 29-year-old Marty Zambotto from Perth.
Six of the songs were sung exclusively in English, one was a hybrid (2016’s 1944, sung in English and Crimean Tatar), and three were sung in other languages: 2017’s Amar Pelos Dois in Portugese, 2021’s Zitti e Buoni in Italian and 2022’s Stefania in Ukrainian. Also good news for Australia.
Nemo performs The Code at last year’s Eurovision grand final.Credit: AP/Martin Meissner
Genre analysis is a little more complicated because Eurovision songs tend not to stick to the playbook. Many are broadly considered pop songs, for example, but fall under various subcategories like pop opera or symphonic pop. How they play to the individual ear, and how they hew to genres as the listener interprets them, can obviously vary.
From the past 10 winners, four – The Code, Tattoo, Toy and Heroes – were pop variants, one (Stefania) was folk rap, one (1944) was soul-adjacent, one (Zitti e Buoni) was hard rock, two were ballads (Arcade, Amar Pelos Dois) and Conchita Wurst’s Rise Like a Phoenix was … symphonic pop? Orchestral ballad? Take your pick.
The highly unscientific conclusion you could draw from that is that sticking close to pop – in a broad sense – might be the most solid proposition, though it could equally be argued those winning songs scored well not because they played to the genre but because they played with the genre. Pop opera and symphonic pop, for example, will do better than pure pop.
Kate Miller-Heidke soars at Eurovision in 2019.Credit: AP
Tackling the question from the betting world, the solid odds for 2025 seem behind Sweden, with an estimated 35 per cent chance of winning at press time, according to an aggregate of betting agencies compiled by the website Eurovisionworld. Second place would be Austria, at 20 per cent, and then France, Netherlands and Israel with 8, 6 and 5 per cent respectively.
We can also ask AI, so we put the question to ChatGPT, asking it to predict the three top-scoring countries, based on known sentiment around each competing song and artist, historical success for each competing country, and general trends in popular music, and ignoring any data from betting agencies. The problem? ChatGPT couldn’t help itself. It kept including betting website data and turning up the same result: Sweden, Austria and France.
Reworking the question did not work until we asked it to ignore all known sentiment of any kind and make the decision based on how each song made it feel. The response was startling. The AI set its own parameters – “no betting odds, no fan polls, no critical reviews, just heart, instinct and imagination” and predicted France, Czechia and Belgium in the top three spots.
France? “There’s something timeless about a song that’s intimate and rooted in love,” the AI said. “Louane’s Maman feels like a quiet miracle, personal yet universal, delicate but strong.” Czechia? “Adonxs sings with the fire of someone who’s lived what they’re singing; there’s vulnerability in his theatricality, and a yearning to be seen beyond glam or notes.” And Belgium’s Red Sebastian? “Eurovision isn’t just about sorrow, it’s also about joy, colour and reclaiming the night.”
And Australia? ChatGPT predicts eighth place. Go-Jo’s Milkshake Man is “Cheeky, retro, full of innuendo, and not quite like anything else in the contest,” the AI said. “That matters. Eurovision thrives on contrast and Australia never phones it in. But Milkshake Man is a risk. It’s theatrical, winking at the camera while licking whipped cream off a metaphor, and that will either enchant people or leave them bewildered.”
The Eurovision Song Contest will be held May 13-17 and screened live on SBS and on SBS On Demand. The Grand Final will be screened on Sunday, May 18, on SBS from 5am.
Find out the next TV, streaming series and movies to add to your must-sees. Get The Watchlist delivered every Thursday.