Is timing everything? Early artist and song selection and success

Is it indeed November?

Are we really a mere six months away from the 2017 Eurovision Song Contest in Kyiv?

We are!

And while it’s early in the selection cycle, things are slowly picking up. We have 43 ostensive participating countries broadcasters – though there is yet another chance to withdraw without penalty next month. We have our first two (internally selected) artists confirmed…though last year German announced artist was dropped within a couple of days, so anything can change until mid-March.

Cyprus will be sending Hovig, who’s lingered around their national selection in years past and is also an alumnus of the Greek X Factor franchise (he finished seventh). Just like last year, Swede Thomas G:son will contribute to the songwriting for our friends at CYBC. Hovig being a member of the Armenian diaspora cannot hurt Cyprus’ prospects.

Hovig’s unsuccessful 2015 Cypriot National Final Submission

The Netherlands have continued their run of announcing their act rather early. O’G3NE (oh gene) competed in the child labour franchise a few years back (finishing 11th out of 17 entries). They have subsequently resurrected themselves as an adult girl group. O’G3NE won the 2014 edition of The Voice of Holland; since 2014 they’ve had a number one single and album.

We can’t yet ascertain to any extent Hovig or O’G3NE’s prospects for Kyiv: it is a song Contest, after all, and we don’t yet know either act’s songs or staging.

Their Winner’s Single, a #1

But we abacus clackers here at 58points are a bit perplexed at the Dutch once again announcing their act rather early in the Eurovision cycle. Here’s what they’ve done in recent years:

Year Artists Artist Announce Date Song Début Date Result
2013 Anouk 17 October 2012 11 March 2013 9th GF
2014 The Common Linnets 25 November 2013 12 March 2014 2nd GF
2015 Trijntje Oosterhuis 10 November 2014 11 December 2014 14th SF
2016 Douwe Bob 22 September 2015 04 March 2016 11th GF
2017 O’G3NE 29 October 2016 TBA ?

There is a distinct pattern: since 2013 the Dutch broadcasters consistently announce their artist early in the season. In every year save one they have waited until the end of the selection period to début the song. The exception is 2015, where both Trijntje and her entry were released well before any other entries—and look how that turned out for the Dutch.

There were a number of issues with Walk Along – the awkwardness of the English language title, Trijntje’s deer in the headlights panic onstage in Vienna (though she was vocally perfect) and the frocks-ups—but certainly Walk Along was an entry that at least initially was well received. That it ended up doing badly, particularly for an established, highly regarded act, is disappointing.

The frock (Courtesty Wiwiblogs)

The frock (Courtesty Wiwiblogs)

So the question we’ve been mulling over is:

To what extent does releasing an entry early predict Eurovision success—victory, to be precise?

We have purposely focused on winning rather than merely doing well. We’ll explain why in the next section.

Method

Like most of our analyses, we have focused on the televoting era, which began in Dublin in 1997. From 1997 onwards public voting (the National Audience in EBU-speak) has either selected the winner or constituted around half of an entry’s score in a Grand Final. We have focused on the Grand Final results only because we have had a few winners that finished second in their semi-finals who went on to win on Saturday night.

Why haven’t we looked at strong (top 5 or top 10) results rather than outright victory? First, winning matters: that is the nature of a contest, to win it. As well, looking at 5 times 20 (or 10 times 20) years of higher ranked entries would require triangulating a lot more data around artist announcement dates versus entry début dates—and very quickly we realised such data are difficult to find and verify. Finally, with scores in a Grand Final tending to clump together (no more than 3 songs in contention to win, followed by a few more clusters of entries with increasingly lower aggregate scores), the longer your list of “successful” entries, the lower scores get. And the less compelling “success” seems. So the winner takes it all.

61 years, 64 winners

Therefore we focused on the laureate from each year, of which we have 20. We looked at when the full entry was débuted, but we also looked at the selection method—in case selection was either predictive or confounding (turns out it is neither). Here’s what we found.

Timing does matter

Whilst an artist might be announced relatively early in the season, débuting your song later is clearly better to contend for victor. In 2014 Conchita Wurst was announced as Austria’s entry on 10 September 2013—before, in fact, her song had been selected. Rise Like a Phoenix was débuted on 18 March, mere days before the Heads of Delegations meeting in Copenhagen where 2014 entries were officially submitted. Similarly, 2011 winners Ell & Nikki won the Azeri national final on 11 February: Running Scared was launched a month later.

Conchita werks the preview party circuit in Riga

If we look at our 20 winning entries, the earliest any was débuted was late January (Denmark 2013 on the 26th). Six others were released in February, but THIRTEEN were released in March. Drilling down further into the March releases, two were released in the first week, eight in the second week, and three in the third or fourth week. That makes the second week of March the sweet spot.

Here are the data:

Year Artist Country Song Song Début Selection
1997 Katrina & the Waves UK Love Shine a Light 08 March Televote
1998 Dana International Israel Diva 14 February Internal
1999 Charlotte Nilsson Sweden Take Me to Your Heaven 27 February Mixed
2000 The Olsen Brothers Denmark Fly on the Wings of Love 19 February Mixed?
2001 Tanel Padar & Dave Benton & 2XL Estonia Everybody 03 February Jury
2002 Marie N Latvia I Wanna 02 March Televote
2003 Sertab Erener Turkey Every Way That I Can Sometime in March* Internal
2004 Ruslana Ukraine Wild Dances 25 March Internal
2005 Helena Paparizou Greece My Number One 02 March Televote
2006 Lordi Finland Hard Rock Hallelujah 10 March Televote
2007 Marija Šerifoviç Serbia Molitva 08 March Mixed
2008 Dima Bilan Russia Believe 09 March Mixed
2009 Alexander Rybak Norway Fairytale 21 February Mixed
2010 Lena Germany Satellite 12 March Televote
2011 Ell & Nikki Azerbaijan Running Scared 13 March Internal
2012 Loreen Sweden Euphoria 10 March Mixed
2013 Emmelie de Forest Denmark Only Teardrops 26 January Mixed
2014 Conchita Wurst Austria Rise Like a Phoenix 18 March Internal
2015 Måns Zemerlõw Sweden Heroes 14 March Mixed
2016 Jamala Ukraine 1944 21 February Mixed

*Unable to confirm precise date

Sertab’s EPIC preview video

Strategy

It is unsurprisingly, then, that several national selections are timetabled to be wrap up in the first fortnight of March. This makes sense for several reasons:

  • For a multi-semi-final series, entries can build momentum
  • Entries can be workshopped and tweaked
  • Late one-off national finals generate buzz in the absence of earlier releases (a paradox)
  • Broadcasters can synchronise their campaign with the accelerating interest among hard-core Contest fans
  • Clever broadcasts with an internal selection get to hear (and see) other entries, which can create bandwidth to tweak yours. Also known as the Guy Sebastian principle.

When so many went ballad, Australia went uptempo—and top 5

In Sweden, Melodifestivalen entries are submitted and allocated to semi-finals months in advance. But rather than schedule their final in February, Melodifestivalen is timed to climax two or three days before the Heads of Delegations meeting in March. While there are usually perhaps only two or three entries considered contenders for the Swedish golden ticket going into their final, much of the Eurosphere is transfixed by the time the national final hits Stockholm.

Loreen’s Melodifestivalen semi-final performance – diamond in the rough Loreen’s Eurovision Grand Final performance – tweaked to perfection

20 years of televoting

Of course it isn’t only a question of timing or strategy. Around one third of any year’s entries are débuted in March: only one of them wins in May. Entries with purposeful, nuanced campaigns will not win if there’s another entry preferred, on balance, with the juries and voting public.

The 2017 Eurovision Song Contest will be the 21st edition with the public helping to select a winner. That’s two decades! In 2016 a new equal-aggregate method was used to select a winner, which brought an exciting climax to an already well produced Contest. Perhaps the best organized Eurovision ever.

The rules for 2017, however, allow for the proportion between public and jury votes scores to be unequal:

The points of the National Juries and the National Audiences shall be combined according to a ratio which is determined by the EBU subject to the Reference Group approval. For example, if the ratio is 50-50, the points of the National Jury carry the same weight as the points of the National Audiences (sic).

Is this in response to have two consecutive winners that were not the public’s first choice in 2015 and 2016? Perhaps. We look forward to seeing the 2017 winning entry once it’s débuted in March… 🙂

2 Responses

  1. Martin says:

    Great article – probably something most fans suspected but now have the hard data to back it up!

  2. John says:

    Thanks Martin! I was surprised at how skewed it was!

Leave a Reply