Timing matters in life. And in the Eurovision.
In a live, competitive competition, what you do and when you do it both are important. But how important is the “when” bit, when the immediate goal is qualifying for the Grand Final?
While the producers (coughChristercough) have largely set the performance order for several years, Eurovision fans nonetheless believe that some performance slots are better than others. In this article John Egan looks at the question of Grand Final qualification in the two semi-final era of the Eurovision.
Once the final delegation list is confirmed—usually in December—the EBU assigns all semi-finalists to one of six pots. Each pot represents one or more voting blocs, whose members have a history of supporting one another. This year’s pots are as follows:
- Balkan: Serbia, Croatia, Montenegro, North Macedonia, Albania
- Scandi: Sweden, Denmark, Norway, Iceland, Finland, Estonia
- Russophone: Armenia, Azerbaijan, Belarus, Russia, Georgia, Ukraine
- Pairs: Romania & Moldova, Cyprus & Greece, Malta, San Marino
- Miscellaneous 1: Austria, Hungary, Czechia, plus Netherlands, Belgium, Switzerland
- Miscellaneous 2: Ireland, Lithuania, Latvia, Poland, Portugal, Australia
Half of each pot will go into each semi-final—but there’s no purposeful division within any pot. As are a result we still sometimes see both Cyprus and Greece, or both Moldova and Romania, in the same semi-final. The pots are designed to prevent most of the Balkan, Scandi or Russophone pot ending up in a single semi-final, making it nearly impossible for unaligned delegations to qualify for Saturday night.
The pot names are ours, rather than the EBU’s.
At the January Heads of Delegation meeting each country performing in a semi-final draws one of the following:
- Semi-final one, first half
- Semi-final one, second half
- Semi-final two, first half
- Semi-final two, second half
The Big 5 (United Kingdom, Spain, Italy, Germany and France) and host (Israel in 2019) are also assigned semi-finals for voting purposes. The EBU does entertain requests for a particular semi-final day from some delegations: Israel will sometimes have its Independence Day during Eurovision week, for example.
The draw Performance Order
After the March HoD meeting, all the final versions of entries have been submitted. These are the ones that appear on the official CD and the ones that will provide the backing track for the shows themselves in May. It is now up to the producers to work within each semi-final half’s list of entries to create a performance order.
There is always some controversy about what the order looks like. Some countries seem to avoid ostensibly terrible slots; others rarely get the ostensibly awesome slots. This sort of discomfort is unavoidable. Unless we were to return to 100% random draws. Which, on balance, we support. The argument for the producer-led draw is to ensure the live broadcasts are maximally entertaining for the audience. I can’t, however, recall any stories of a gaggle of Eurovision viewers turning off a semi-final or Grand Final because of the song sequencing. It has always struck me as a self-serving argument.
Let’s crack on with the analysis itself.
We have gone back to 2008, the first year with two semi-finals. We have compared the performance order and assigned each entry a binary score: one (1) for having qualified for that year’s Grand Final and zero (0) for not having qualified. We have not looked at actual points.
We have used the actual top 10 from each semi-final: in the years where 10th ranked entries from Macedonia were bumped for a jury favourite, we have gone with the actual top 10 rather than the 10 qualifiers. This ensure consistency of data and comparing like for like. We also happen to quite like Let Me Love You and felt they was robbed.
Tamara, Vrčak & Adrian – Let Me Love You (Source: YouTube/Musicoftheworld2018)
The smallest semi-final in this period had 15 entries; the largest had 19. We initially tried to look at the reverse order in the second half of each semi-final (last, second last, third last to perform), but this got messy very quickly. So we stuck with ordinal sequence. That means that performing 15th could be the “pimp” slot or performing with several songs yet to follow. Too hard!
We then aggregated the number of times an entry qualified from that slot in the order. There were 22 semi-finals in the period (two each 2008-2018, over 11 years), but slots 16-19 occurred less frequently. This is important because we wanted to see what the mean (average) rate of qualification was for each slot. Nineteenth slot occurred 7 times rather than 22 times; 16th occurred 21 times. This year we have one 17 entry semi-final and one 18 entry semi-final.
Once tabulated, we clustered the results into five bands:
- Great: qualified 80% or more of the time
- Good: qualified 60-79% of the time
- Average: qualified 50-59% of the time
- Fair: qualified 40-49% of the time
- Poor: qualified less than 40% of the time
Here’s how the distribution ended up:
It is worth noting that the only slot in the second half that is considered poor is 11th. Conversely, only two slots in the first half are good: 6th and 10th (none are great). But in some semi-finals 10th ended up in the second half of the order: any of the eight semi-finals with 17 or fewer entries.
Let’s look at this year’s performance orders and see how things cluster for semi-final one:
|3||Finland||Darude feat. Sebastian Rejman||“Look Away”||English||Poor|
|4||Poland||Tulia||“Fire of Love (Pali się)”||Polish, English||Fair|
|5||Slovenia||Zala Kralj & Gašper Šantl||“Sebi”||Slovene||Average|
|6||Czech Republic||Lake Malawi||“Friend of a Friend”||English||Good|
|7||Hungary||Joci Pápai||“Az én apám”||Hungarian||Average|
|9||Serbia||Nevena Božović||“Kruna” (Круна)||Serbian[c]||Average|
|11||Georgia||Oto Nemsadze||“Keep on Going”||Georgian[d]||Poor|
|12||Australia||Kate Miller-Heidke||“Zero Gravity”||English||Good|
|13||Iceland||Hatari||“Hatrið mun sigra”||Icelandic||Good|
|16||Greece||Katerine Duska||“Better Love”||English||Fair|
|17||San Marino||Serhat||“Say Na Na Na”||English[e]||Average|
In this semi-final, Finland, and Georgia are in the worrisome slots. Czech Republic, Australia, Iceland, Estonia and Portugal are all in good slots. But the two slots with the highest rate of qualification (18th and 19th) are missing this year.
For semi-final two:
|4||Switzerland||Luca Hänni||“She Got Me”||English||Fair|
|6||Romania||Ester Peony||“On a Sunday”||English||Good|
|7||Denmark||Leonora||“Love Is Forever”||English, French[f]||Average|
|8||Sweden||John Lundvik||“Too Late for Love”||English||Average|
|10||Croatia||Roko||“The Dream”||English, Croatian||Good|
|12||Lithuania||Jurij Veklenko||“Run with the Lions”||English||Good|
|14||Albania||Jonida Maliqi||“Ktheju tokës”||Albanian||Good|
|15||Norway||KEiiNO||“Spirit in the Sky”||English[g]||Good|
|17||North Macedonia||Tamara Todevska||“Proud”||English||Average|
In this semi-final, Moldova, and Malta are in the worrisome slots. Azerbaijan has one of the slots with a near 100% qualification record, 18th. Romania, Lithuania, Russia, Albania and Norway are all in good slots.
Still a Song Contest
This analysis doesn’t take into account the song, the artist, the genre, the bloc membership—all of which can contribute to an entry’s score. All of those things can matter. The current bookie’s favourite, the Netherlands Duncan Laurence’s Arcade has only a “fair” slot.
As well, every single semi-final performance slot has produced several Grand Final qualifiers. In fact, both the poor slots (3rd and 11th) have still produced 8 qualifiers. There’s no reason for delegation (or fan) despair.
Very shortly rehearsals start. And so much more will become clearer.