It was our good friends over at Swedish newspaper Aftonbladet who caught the exclusive news that Melodifestivalen’s voting system was going to change. They describe the changes to the system as more exciting for viewers, but a possible nightmare for the artists. It’s also lifted the complexity for all of us who follow Melodifestivalen to a whole new level.
Let’s start off with how voting worked last year so we all understand that starting point. If you understand this already then jump over this section. In the heats of Melodifestivalen the results were decided by 8 public voting groups. One of those groups was the good old-fashioned televote. The other 7 groups were determined by votes cast via the Melodifestivalen app with it’s immensely popular button-bashing system where up to five free votes were awarded per song. Viewers using the app enter their age when they sign up, and were put into a different voting group based on their age – with the youngest group being for 3 to 10 year olds and the oldest for the over-75s.
In each of these age groups, including the televote, their scores were converted into a points system resembling Eurovision (12-10-8-6-4-2-1). The points accumulate for each song and 5 songs out of 7 advance. Then a second round of voting took place (which still includes the votes from the first round) and each heat ends with the top two songs being announced as direct qualifiers, with those in 3rd and 4th place qualifying to a Second Chance round for a last chance to make the Melodifestivalen final.
Almost everything said above is the same this year – the changes tweak the ever popular system rather than overhaul it.
Viewer Engagement Key
The reasoning for the adaptations is that SVT wants to fully maximise on their viewer engagement. Viewer engagement statistics are huge for Melodifestivalen – with it speculated last year that there is no show in the world with as much viewer participation in the result and outcome as with Sweden’s Eurovision selection.
However, project leader for Melodifestivalen, Anette Brattström, realises that the previous voting system narrowed that engagement down to a very small window during or immediately after all the songs have performed. That is because the majority of users during the show are using the app to vote, and they use their 5 free votes per song during this first voting window. To increase that viewer engagement, one of the changes will mean that viewers will be able to cast up to 5 free votes via the app again in the second round of voting. Expect to see 15, maybe even closer to 20 million total votes cast even during the heats by the 10 million strong Swedish population.
Theoz, with 2 million TikTok subscribers, will be attracting plenty of support in this first heat (Photo: Annika Berglund, SVT)
However there are further ideas that will prevent this second round of voting being a carbon copy of the first round of voting. The first of those ideas is with how the qualifiers are announced. After the first round of voting the act that received the highest number of votes will qualify automatically to the final. Note that this is votes, not points as well. Taking them out of the equation means that in the second round of voting those who voted for the heat winner may support a different song to get them through to the Melodifestivalen final. One can expect more people to have voted for the heat winner than those who voted for those two songs in last place in each heat, so this change will increase volatility of opinions in this second round about who should take the last spot directly in the final.
These final six songs will be subjected to a show climax that will be very much like the Eurovision Song Contest. Each of the 7 voting blocs, plus televoting, will reveal their scores almost exactly like juries do each May. In Friday’s dress rehearsal for Melodifestivalen it was Oscar Zia in charge of announcing these scores. To add to the suspense the 12 points will be announced first – with the five other scores flashed up on screen in the seconds that follow (Aftonbladet have a photo of show this appears on screen).
Note too that the scoring between these six songs is slightly tweaked from the 12-10-8-6-4-2-1 between the seven songs last year to be 12-10-8-5-3-1. And yes, those of you who know SVT will not be surprised that the order of the voting blocs in the voting sequence will completely be at their decision…so the voting reveal will undoubtedly end in a dramatic ending.
All of this mega reveal will reveal means there’s so much voting drama that there will be no time for a winner’s reprise after the song qualifies to the final in this year’s show. As a part of the show it feels very un-Melfest.
This Voting System Historically
With thanks to the great MelloMikeUK on Twitter we have calculated this voting system for the last three years of Melodifestivalen. We go back three years because it was 2019 which was the first time Melodifestivalen split their app voting in the current voting blocs.
Now we have to make one important statistical presumption. That presumption is that there is no movement of votes from the winner of each heat to another similar act in the second round of voting. Yes some votes in these previous years did come through in the second round of voting, but knowledge of those numbers are not available for us to speculate with.
We were somewhat surprised to find the headline fact that only one qualifier to the final would have been any different. Indeed the same 48 acts would have made either the Andra Chansen round or the final in our system. The example of a change is that, in the first heat of 2021 Paul Rey, who finished 4 points below Arvingarna in that system, would have tied with the 1993 Eurovision entrants in that show. By virtue of having more raw votes Paul Rey would have then qualified to the Melodifestivalen final.
Some others are close, Charlotte Perrelli and Alvaro Estrella would have tied in 2021 for that last qualification berth, yet Charlotte wins the tie-break on virtue of having the most votes. Anna Bergendahl the previous year would only have stumbled through to the final with a 1 point margin over Alvaro Estrella (yes, him again, but this time featuring Mendez) – despite Anna finishing in 1st place that night in Gothenburg.
There are though two other shocking changes that this system showcases. Anna Bergendahl win featured above wasn’t announced in that show – it was her and Dotter who were both simply announced as qualifiers – but Dotter actually received more votes than Anna Bergendahl and would have been announced as, alone, as a shock winner that day.
The other shock comes the year earlier in 2019. The final heat was ‘won’ on points by John Lundvik, however he finished 32,095 votes behind Bishara, an inch in Melodifestivalen terms, and John Lundvik, he who won that year’s Melodifestivalen, would have had to journey through that second round of qualifying.
Yes, Anna Bergendahl and John Lundvik would have still qualified to the final regardless. But that they didn’t win their heat would have been hugely shocking at the time. In terms of momentum for both acts, who finished 3rd and won respectively, the momentum that they were in-it-to-win-it would have dropped off massively if they had to accept a 2nd place. It’s hard to imagine John Lundvik being an obvious favourite in the build-up to the 2019 final if he didn’t win his heat.
Melodifestivalen has played with this format earlier. In 2010 and 2011 Melodifestivalen announced a winner of each heat that qualified directly after the first round of voting. A couple of shocks came up in that 2010 edition that demonstrate this momentum shift so well. It felt so obvious in 2010 that Peter Jöback would qualify from the final heat of that show. He did, but it was Anna Bergendahl who shocked with that 1st place finish. Being the winner that night lifted Anna straight up into being favourite tipped and she went the entire way to Oslo that year.
The other example came the week earlier, when an even heavier favourite was Darin, who still is a megastar here in Sweden. However he lost that heat by a whisker to unknown girl group Timoteij – and all momentum and expectation on Darin representing the nation vanished. Darin finished a respectable 4th place in the contest but it’s little surprise that since this episode he hasn’t returned to this song competition, despite his 2013 interval act in Malmö. This is the type of nightmare situation Aftonbladet refers to.
The Tiniest of Competitive Changes
While the statistics show that this system is unlikely to influence huge dramatic changes in who qualifies or not, there are different acts this year who can consider themselves winners and losers in this system. Both John Lundvik in 2019 and Anna Bergendahl in 2020 have appealed to the older voters and their support dropped away historically with the younger demographic. This is important for getting that first qualifying spot and if there is a tight battle at the top in this voting system the push swings towards those that get the teen votes once more.
What potentially could amplify that effect is that voting on the app is reset after the first round of votes. While all votes in the first round remain in the system, the voters of the act that qualified in the first round may switch their allegiance to somebody else also to their taste. The results of the age based app voting has shown how split viewers are between young and old in terms of musical taste, and should the previous paragraph run through that effect could be magnified in this second round of voting. Alternatively, maybe this system encourages those songs that need a second listen to get chance with a recap and another burst of voting.
Regardless, this system skews the results success the tiniest margin towards songs that can score well across all demographics, rather than highly in a few. The points system of 12-10-8-5-3-1 has the biggest gap between 3rd and 4th place. This means that in a close scoring battle for the final or semi final positions, getting in the top 3 in a voting bloc is vital for a good score.
Of acts in the 2021 final, it is little surprise to see the three acts who, Tusse aside, were most adored by older voters (Clara Klingenström, Arvingarna and Charlotte Perrelli) had the highest standard deviation values for their televote points across the age ranges. Should the competition be tight for the last final spot or a semi final spot we would anticipate it to be harder now for songs mainly appealing to the older demographics to qualify to the final.
Combined all these factors are still tiny in a competition of millions of votes each week, but the trend is clear as to in which competitive direction the changes go in.
Voting Sequence Suspense
The show on Saturday will likely feel strange to many traditionalists watching. Never in the history of holding heats in Melodifestivalen have any scores been revealed on the night, and this whole voting rigmarole will take up so much time (and at least at the Friday dress rehearsal it was painfully slow) that neither direct finalist will get the honour of performing to the audience again.
Now the voting sequence is the big finale. The way the numbers will fly across the screen to decide the qualifiers is an exciting element of Eurovision and Melodifestivalen as music entertainment. But is it necessary now? The divisions between young and old voters are usually only noticed by our stats geeks who pour over the numbers once the show is over. For it to be so visible for the nation to see creates division when I’d rather unity and harmony about the results of the heats.
Yet I know that it is just this division and difference that encourages engagement. The changes made will get more people voting, and voting for longer, than they did before. What I question is if we get quality from this type of engagement, and if the risk of sparking further divides between generations is wise for a public broadcaster to stir up. Maybe there’s a mantra that all publicity is good publicity here, but it sits uneasy from this vantage point before the chaos unfolds.
The voting system may have been created to showcase acts that all of Sweden can love and allow them all to shine. However showing it so openly and vividly this feels very much like splitting people apart rather than allowing them to unite and come together.