Lisbon 2018 Retrospective: Semi-Final Two #eurovision #allaboard

The first 2018 semi-final offered an embarrassment of riches: there being a sense of disappointment with what would follow on Thursday seemed inevitable. But even with a lowered set of expectations, what was delivered on Thursday felt, with a few exceptions underwhelming.

There was another element to the luck of the 2018 draw that seemed cruel. The second semi-final had one less entry (18 versus 19). In effect, the better songs faced somewhat lesser chances of qualifying regardless of song quality because of that extra entry: 10 from each show would advance, rather than the top 20 from across both shows.. There is no way to avoid this, short of moving the producer-led performance order to a wholly producer-led allocation of semi-final spaces. We think we should be moving in the other direction of travel—back towards random draws for all the shows—rather than creating greater scope for delegations and fans to allege bias and favouritism.

Let’s drill down into the results from the second semi-final.


There were 21 voting nations for this semi-final: the France, Germany and Italy all voted on Thursday night, along with 18 participants. The final top 10 was:


Country Total Score Total Jury Total Public
Norway 266 133 133
Sweden 254 171 83
Moldova 235 82 153
Australia 212 130 82
Denmark 204 40 164
Ukraine 179 65 114
Netherlands 174 127 47
Slovenia 132 67 65
Serbia 117 45 72
Hungary 111 23 88

Norway won the semi-final in a couple of remarkable ways. First, it was the show opener—the first of 18 songs performed—and managed to be second with juries and third with the public. Second, their score was identical in terms of televoters and juries: 133 points from each. Alas Norway’s Grand Final result would prove to be quite different.

Denmark won the televote with 164 points, but Moldova was not far behind with 153 points. Sweden handily won the jury votes again: their 171 points were almost 40 points ahead of Norway. Norway, Sweden, Denmark, Moldova and Ukraine all scored at least one point from every single delegation. Australia was blanked only by Georgia.

Unlike the first semi-final, Norway was the only consensus entry. In the first semi-final, the threshold was having broken the 100 point threshold in both scoring components. Even with lowering the threshold to 90 points (one less entry compared to Tuesday means the maximum score in each component is reduced by 12 points), no one else achieved this.

In other words, there was a significant skew between the component scores for the remaining nine qualifiers.

The jury favourite was Sweden. Then we had three.  entries across six points: Norway, Australia, and the Netherlands. Malta was fifth with juries (93) but last with the public (8 points). Latvia was just behind Malta (92 points), but their somewhat higher televote score (14 points) still wasn’t enough. Romania were 8th with juries (67 points), but only 13th with the public (40 points): they ended up 4 points behind 10th place qualifier Hungary.

The televote pumped up the scores of qualifiers Denmark (164 points versus 40 from juries), Moldova (153 points, though they were 7th with juries on 82 points), Ukraine (114 points, though they were 10th with juries on 65 points) and Hungary (5th place for 88 televote points versus 13th place and 23 jury points).  Australia were third with juries and 7th with the public—an improvement over last year’s semi-final results—putting them comfortably in 4th overall.

Just as in the first semi-final, the overall scores from this semi-final break into three bands: the leaders, those with variable support, and those that limped across into qualification.

The leaders

Norway’s mean public and jury scores were identical: 6.65 points. That means That’s How You Write a Song earned an average public ranking of 5th overall across all voting delegations—a remarkably low “winning” score. Sweden’s fortunes were more mixed: an impressive average jury score of 8.55 points (3rd overall), with a lower televote average of 4.15 points (7th overall). Consistency is what gave Rybak the win on Thursday night.

Variable support

Moldova’s big televote score (153 points), an average of 7.65 points or 4th place: their 82 jury points is an average of 4.1 points (7th overall). Australia’s story was the opposite story: a high jury vote (130 points; average 6.5 points, 6th overall) and a lower public vote (82 points, 4.1 average, 7th overall). Denmark rocked the televote (164 points; average 8.2 points, 3rd overall) and limped through with the juries (40 points; average 2 points, 9th overall).

Ukraine (65 jury; 114 televote) nearly had enough points from the televote to qualify without any jury support.  The Netherlands (127 jury; 47 televote) would have qualified on jury support alone.

Ukraine’s Mélovin did, however, break the frock block: until Under the Ladder qualified, every Ukrainian entry that qualified from a semi-final was performed by a lead artist wearing a dress. Doubtless this will inspire an entire generation of piano-burning, contact-lens sporting singers.

Limping across

Slovenia nearly matched Norway with her near identical score components (67 jury; 65 televote). We could say that Hvala Né was liked by some but loved by none. Hungary 5th place in the televote would not have been enough: the 23 jury points were nearly not enough. Serbia was a bit more popular with the public than jurors (72 versus 45 points).

Looking up

Poor Romania: a mere 5 points were between scraping into the Grand Final and losing their perfect qualification record. Bearing in mind that their most reliable ally—Moldova—hooked them up with a double douze points (from the Moldovan jury and the public; the Romanians returned the favour, by the way) and it’s clear that Romania didn’t favourably land on the radar of any particular constituency. Rock? Hungary was rockier. Anthem? Denmark or Australia were better options. Jury fodder? Nothing about the composition or staging to make the pros say “yeah, awesome”—and having one of the best vocalists wasn’t enough in its own right.

At the other end of the table, Georgia and San Marino were well back of the rest. If Georgia were to have any chance it would have been thanks to jury support: 11 points, including 8 from Ukraine, tells the whole story. Their 13 televote points were also paltry. San Marino’s 14 televote points included 2 from Australia and topping the Maltese televote. Maltese lead artist; it makes sense. You know, except for the whole “quality” aspect.

And while no one got null points. Malta nearly did in the televote (again): 7 points from Australia and one from Denmark was it. But Christabelle was 5th with the juries. 13th overall in a Grand Final wouldn’t be a disaster; 13th in an 18 entry semi-final it’s not a great result.

Bloc’d Out

The current “pot” system was designed to flatten out skew related to “friends and neighbours” inflating the scores of unremarkable or weak entries. There are three persistent voting blocs in the 21st century Eurovision that are the main targets of the pot system:

  • The ex-Soviet states (plus Israel)
  • Scandinavia
  • The ex-Yugoslav states (plus Switzerland)

Let’s take a look at how each bloc’s members did in the Tuesday night show.


There is usually one bloc member who benefits most from the others: for the ex-Yugo bloc this is almost always Serbia. In fact, when Serbia does not appear in a Grand Final, the next highest ranked entry doesn’t seem to gather that freed-up support. This year’s semi-final had Serbia, Montenegro and Slovenia participating: both Serbia and Slovenia qualified.

Once again Serbia topped its table. Montenegro double douzed Serbia. The Slovene public also ranked Serbia first, but their jury “only” gave them 4 points. More importantly, 11 out of 20 delegations awarded Serbia null points in the second semi-final—the weakest distribution among our qualifiers.

Slovenia also qualified this year, but generally they do not benefit much from being in this bloc. There are lots of reasons for this—their language isn’t mutually intelligible with Serbian/Croatian/Bosnian/Montenegrin, their artists are less likely to participate in the yugosphere, their culture is in a lot of ways an outlier from the other southern Slavic nations—but when they sing in English, that should flatten out the intelligibility issue, at least. The Serbians gave Hvala Ne some love: 8 televote and 2 jury points. The Montenegrin televoters gave Lea Sirk 10 points, but their jury blanked her. Malta was the only delegation to blank Slovenia on Thursday.  Slovenia were 25 points clear of  11th place Romania: deduct the 20 bloc points and Sirk still qualifies.

Finally, we have Montenegro, who were 16th of 18. Their 40 points consisted of 17 from Serbia (10 televote; 7 jury) and 8 from Slovenia (7 televote; 1 jury): over 60 per cent from the bloc.

Has the pot system eliminated the bloc effect among ex-Yugos? Not entirely, but we no longer see most of the bloc qualifying, year upon year, regardless of quality.


This year was the first in over a decade where Russia’s entry was considered weak. As head of this bloc—and as the focus of the Russophone cultural sphere, including a massive, lucrative music market—their 21st century entries have ranged from good to epic. Before rehearsals started, many of us theorised Russia’s perfect qualification record might be in trouble. As a corollary of that theory, where would the ex-Soviet bloc points go, in Russia’s semi-final, and in the Grand Final? Would Kirkorov’s involvement—as song writer and in terms of his ubiquitous presence on the ground in Lisbon—facilitate a bloc shift toward the Moldovan entry? Somewhat…

First off, Russia: Julia Samoyleva did not get typical levels of support from the bloc: half of her 14 jury points came from Moldova: the rest of the bloc’s juries blanked Flame is Burning. But bear in mind both Ukraine and Georgia have problematic political relationships with Putin’s Russia: both have been invaded and had part of their territories occupied by Russia in the last decade. The televoters in both countries awarded Russia a total of 9 points (3 from Ukraine; 6 from Georgia), with the high profile of the russosphere in both countries means public support is  often significant. But the politics of each country means their juries are unlikely to reward many points to a Russian entry. Moldova awarded Russia 8 televote and 7 jury points. The Latvian public, however, delivered its douze points to Russia. They often do. Their jury blanked Russia, however.

Russia, meanwhile, totally blanked Latvia; otherwise the bloc treated Funny Girl OK. The Georgians awarded Laura Rizzotto 15 points (8 jury; 7 public), the Ukraine 11 points (10 jury, 1 public), and Moldova 3 points (all jury). Latvia had targeted the juries for support and did well with them: 92 points (out of a total of 106 points) and 6th with the juries. In the end they were 5 points out of a qualification spot. Latvia’s ex-Soviet bloc traffic is often one way…towards Russia. When the Latvian artists has been a Russophone—regardless of the language performed on the Eurovision stage—the rest of the bloc hasn’t sent many points unless the entry in question is one with massive appeal across the Eurosphere. Yeah I’m lookin’ at you, Aminata.

Moldova worked the bloc to great effect. As in last year, Doredos are from eastern Moldova and two of the three singers are Russophone. Combined with their high profile Russian songwriting team and they hoped to top the bloc tables. The Russian public and jury obliged with a double douze points (24 points). Georgia provided 15 more points (12 public; 3 jury), Ukraine 12 points (all from the televote), and Latvia 14 points (8 public; 6 jury). Those 65 points certainly helped!

But Ukraine also did well with the bloc. Georgia provided 17 points (10 public; 7 jury), Russia 16 points (10 public; 6 jury), Moldova 15 points (8 jury; 7 public), and Latvia 11 points (10 televote; 1 jury).  So Ukraine grabbed 59 points from the bloc. Moldova was number one, but only but a small margin of 6 points.

Alas, poor Georgia were last in this semi-final. 23 of its 24 points came from the bloc (Hungary provided that last one via its jury), 13 of which came from Ukraine (8 jury; 5 public).


All three heavy hitters in this pot ended up in the second semi-final. All three qualified; these entries crowded the top 5 (first, second and fifth).

Norway won the semi-final having placed second in the jury scoring and third with the public. Sweden won the jury vote and was 6th with the public. Denmark won the televote…but was only 12th with the juries. Here’s how they supported one another, point-wise, on Thursday:

  Norway  (Public) Norway (Jury) Sweden (Public) Sweden (Jury) Denmark (Public) Denmark (Jury) Total
Norway X x 10 12 12 6 40
Sweden 10 12 x x 10 10 42
Denmark 12 0 12 0 x 24
Total 22 12 22 12 22 16


A “perfect” score from within the Scandi-bloc for this semi-final would be 48 points: douze points in both the jury and public scores from the other two bloc members. Sweden scored 42/48 possible points; Norway was a bit behind on 40 points—very nearly the max for both. Where the bloc failed to deliver was with the jury votes for Denmark: both Norway and Sweden blanked Higher Ground, though, to be fair, so did 11 other juries (Rasmussen only scored 40 of his 204 points with jurors). Otherwise the Scandis locked one other into first or second place in every scoring dimension.

Whilst people seem to moan about “the Russians” and “the Yugoslavs”, the Scandinavian bloc remains the longest, most persistent, and most efficient voting bloc with respect to semi-final results.


A couple of nights later and these results proved inconsistently predictive. Norway fell from first on Thursday night to 15th on Saturday, despite being given a prime slot in the first half of the draw. Norway arguably took the greatest tumble between a semi-final and a Grand Final.

Sweden tanked thanks to a drop in public points despite the larger Grand Final public vote being on offer, though the juries kept them at number one. Denmark’s performance from Thursday was, arguably, replicated: very popular with the public and not that well regarded by jurors. Sweden (7th), Denmark (9th) and Moldova (10th) were the only top 10 results from this semi-final’s qualifiers.

At the other extreme, the other six qualifiers occupied places 17th to 22nd in the Grand Final rankings. That too is unprecedented.

What if

The inequity of this year’s semi-final allocations—which we accept are totally random—got us thinking about ways to manage this. So here’s an experiment, based on a query we received: What if we aggregate the total scores from across both semi-finals and see if we get a different list of 20 qualifiers? We already knew the answer, but thought it was worth illustrating:

Take a look:

Rank Country Total Rank Country Total
1 Israel 283 21 Romania 107
2 Norway 266 22 Latvia 106
3 Cyprus 262 23 Malta 101
4 Sweden 254 24 Azerbaijan 94
5 Moldova 235 25 Belgium 91
6 Czech Republic 232 26 Switzerland 86
7 Austria 231 27 Greece 81
8 Australia 212 28 Poland 81
9 Denmark 204 29 Armenia 79
10 Estonia 201 30 Belarus 65
11 Ireland 179 31 Russia 65
12 Ukraine 179 32 Croatia 63
13 Bulgaria 177 33 Montenegro 40
14 Netherlands 174 34 San Marino 28
15 Albania 162 35 FYR Macedonia 24
16 Slovenia 132 36 Georgia 24
17 Lithuania 119 37 Iceland 15
18 Serbia 117
19 Hungary 111
20 Finland 108

In other words, nothing changes. Because these scores are ordinal in nature: they are based on rankings of scores generated by other rankings (public votes and jury rankings). It is possible one or two of the 20 qualifiers might change…if the scores were positively skewed towards the top entries in one semi-final and they were less skewed in the other. That didn’t happen in 2018. Finland almost would have been swapped out for Romania. Almost does not count, alas.

A weaker semi-final doesn’t automatically have a skewed lower point distribution. Each semi-final is a distinct event, in terms of scoring. Points are allocated on an ordinal basis. Whether then 4th most popular entry is about as good as the third or much weaker, is not indicated in the douze points scoring system.

Leave a Reply

Your email address will not be published.