Eurovision Top Tip Update: The Albania Code

Is there such a thing as a natural level for countries in Eurovision? I think there is. That is, if every country sent indistinguishable songs, you could predict pretty accurately how the voting would go based purely on each country’s traditional voting allies.

This is a wider view of the argument I developed in my last article, that Eurovision big hitters such as Greece would probably be in with a fighting chance of a top-ten finish no matter how bad a song they sent.

Look at Greece’s form figures: in the last five years, their finishing positions read 9, 7, 3, 7, 8. While none of those songs were especially bad, those placings do suggest to me that Greece has a natural level somewhere around 7th to 9th, rubbing shoulders with the likes of fellow big-hitter Armenia (8, 8, 4, 10, 7).

Does this apply further down the leaderboard? Yes it does, and no country demonstrates it better than Albania.

Albania’s 2004 debut entry, Anjeza Shahini’s ‘The Image of You’, wowed fans and viewers alike. It came fourth in the semi-final and seventh in the final.

Since then, though, Albania’s finishing positions read: 16, 14 (of 23 in semi), 17 (of 28 in semi), 17, 17, 16. That’s a remarkably consistent record. In this time, Albania has provided us with a huge variety of musical styles and staging. No matter. At the end of the day, their result has come down to collecting points from their main allies – who are, according to Eurovision’s premier statistics site, Macedonia, Greece and Switzerland.

In terms of the quality and quantity of their voting allies, I would put Albania in Eurovision’s second tier – behind the big hitters, but ahead of many other countries with fewer and less reliable friends. This helps to explain why Albania have qualified for the final for the last three years, under a system which puts more final places up for grabs from two semis, having failed to qualify in the two previous years under the old one-semi system.

The same could be said about Portugal. They missed out narrowly in 2007 – eleventh in the only semi, three points behind tenth-placed Moldova – but have qualified each year under the new two-semi system, and have gone on to place with an almost-Albanian level of consistency in the final: 13, 15, 18.

Again, a wide variety of songs represented. And again, a final placing that seems to depend largely on their main voting allies (which in Portugal’s case are France, Switzerland and Spain).

We can think of a country’s natural level as being roughly where you would expect it to finish with an average, unremarkable song. And I have argued regarding the big hitters that a song probably has to be quite a lot worse than average to depress a country’s vote significantly below its natural level.

Is the same true with the lower-ranked nations? Some would argue that last year’s 18th reflected it being Portugal’s weakest recent effort. Maybe. But I would argue that the non-participation last year of Andorra, another traditional voting ally of Portugal, was just as important.

That’s why, incidentally, it’s so important for punters to be aware that not every Eurovision has the exactly same line-up of contestants. Some new countries join, some drop out, some return again. That can have a huge impact on the natural level of other countries – consider what Turkey’s three top-four finishes in the last four years suggests about the recent extension of the contest into the Caucasus.

However, while it may be unlikely that a country will sink too far below its natural level, the same doesn’t apply in an upward direction. The natural level represents something of a floor, but it’s absolutely not a ceiling.

We’ve already seen how a strong song can lift a country, with the examples of Albania’s debut 7th, Kalomira’s 3rd place for Greece in 2008 or Sirusho’s 4th the same year for Armenia interrupting those countries’ runs of similar placings.

Much more strikingly, we can see it with Germany’s win and Belgium’s strong showing in 2010. These two examples prove that even countries whose lack of traditional voting allies gives them a very low natural level can still produce great results with the right song.

I find that each year, there are usually about half a dozen songs that tend to stand out on their merit. These are the cases where we might expect a country to improve significantly upon its natural level in the Eurovision pecking order by adding floating voters or strong jury scores to its base support.

As for the rest, the Albania Code shows us that past performance can often be an excellent indicator of future results.

8 comments to Eurovision Top Tip Update: The Albania Code

  • Rob

    Fascinating and informative analysis as always Daniel. Looking at this first semi-final and applying your so-called Albania Code, the one that I believe has a distinctly weaker entry than usual is Armenia. But obviously it’s starting from a strong position as a ‘big hitter’. The question is, by how much its overall vote will be depressed. Turkey, Serbia and Russia have been stripped of some of their strongest voting allies in this semi but all 3 of their songs, I would argue, are by no means weaker than their average. And to my Western European, at least, Serbia and Russia are stronger compared to their entries last year. Factor in the draw and it’s a real head-scratcher!!

  • Rob

    *Western European ears

  • Daniel

    Hi Rob, your response does indeed show what a conundrum semi 1 is at this stage. Armenia has not been blessed with as many friends as usual in this semi either and has an early draw. It’s a song that could do anything. It is ostensibly weak in that it’s silly: poor staging and vocals may see it sink, especially with juries. But, there is also a catchy, infectious quailty to it if done right and if so, it has the potential to do incredibly well in the televote, whatever you may think of its artistic merit.

  • Alex

    Hi Daniel!
    Great site, have been following it since last year!
    Can you please write an article about Estonia’s place in the bookies? I think their odds are the ultimate proof of how lost the market this year is.

    • Daniel

      Hello Alex, and many thanks for the kind words. You’re right that bookmakers are at a loss about who is going to win this year, probably because punters are too at this stage. I did write an article on ‘Rockefeller Street’ when it won the Estonian selection. I do like the song but I completely agree with you that its place at or near the head of the market does not reflect its chances in the competition. I would not put it in single figures personally. As a lay bet at current prices it is tempting, though not as much as 4.9 Norway which I laid on Betfair just over a month ago. Watch this space for more articles on individual countries’ entries and a look at the betting market as it evolves – and don’t forget it is still early days.

  • David

    While I’m sure you’re fully aware of this already, it’s never a bad thing to point it out in statistics discussions: always be careful not to mix correlation with causality.

    The fact that Albania has had a remarkable run of similar placings doesn’t necessarily imply that there is some external factor behind this. It could also just be random chance. In fact, one could argue that with so many countries participating, it would be a statistical anomality if NO country presented a run like that – even if placings were completely random (which of course they’re not).

    Now, for countries like Armenia and Greece it’s obvious that their high finishes are at least in part due to diaspora and/or friendly voting. Albania has this to some effect, but not much compared to those two. So when it comes to “natural levels”, perhaps it’s even more describing to speak of some kind of “floor” – a finishing position the given country doesn’t go under almost regardless of participating song? I think the variance upwards from that level is much less predictable than the variance downwards – hence, “floor”.


  • Daniel

    Hi David, I completely agree with you. I think the fifth-last paragraph of the article makes this point, followed by the examples of Germany and Belgium last year, but it maybe should have been given greater emphasis.

  • David

    It most definitely does, Daniel – I think this is what happens when I write my comment quite a while after actually reading the article 🙂

 Leave a reply...