9 May 2014

Nul points! How will Ukraine tensions hit Eurovision voting?

Will Moscow’s involvement in the Ukrainian crisis see Russia hit where it really hurts – the Eurovision song contest? Two statisticians explore the likely voting patterns for Channel 4 News.

Eurovision finalists Mariya Yaremchuk of Ukraine and the Tolmachevy Sisters of Russia (pictures: Getty)

Above: left – Ukrainian Eurovision finalist Mariya Yaremchuk, right – Russian finalist the Tolmachevy Sisters.

The Eurovision song contest is one of those television events that nobody really likes – in fact many of us have quite vocal and strong negative opinions about it, write Gianluca Baio of University College London.

So can we expect a shock in these dynamics as a result of the latest developments in Ukraine? Maybe.

And yet, come its time of the year, especially in the UK, people and the media start to frantically argue about the chances of our act and how frustrating it is that other countries gang up against us.

One of the myths around the Eurovision is that tele-voting is strongly affected by tactical and political bias. We looked at the available data from the last twenty years to investigate the presence of “negative bias” – is there evidence of systematic propensity to score a given performer higher points?

Who are this year’s finalists? Click on the pictures below to find out.

One interesting result of our model is that in fact no such systematic negative effect seems to show up, substantially. When comparing two “similar” acts (for example two different performances based on a female act singing in their own language in the same year of the contest), none of the voters showed a particularly negative attitude towards one of them, at least not to a degree that could really and consistently skew the final results.

On the other hand, as is quite reasonable to expect, some countries tend to “favour” certain acts; again, one of the myths is that the countries in the eastern European bloc tend to deliberately vote for each other.

It could be that formerly neutrals in Ukraine will now react and deliberately score Russia no points at all.

This is sort of true – but again only in a very few cases the extra advantage given is measured as reasonably high.

And in any case, we did see some sort of clustering across Europe, with geographical proximity and migrations playing, arguably, important roles (for example, the countries in the former Yugoslavia which are close geographically as well as “culturally” do tend to slightly favour each other. The Turkish act is usually scored highly from Germany, possibly due to the large contingent of Turkish migrants).

The Soviet bloc

But what about the infamous former Soviet bloc? Our model shows that the countries that are systematically more likely to score the Russian acts are Armenia, Azerbaijan, Ukraine, Georgia and Belarus, with Macedonia, Greece, Estonia and Lithuania also showing some positive propensity.

However, while on average all these effects are positive, none of these effects can be classified as substantial or large enough to indicate real bias.

Just in case you missed it last year - here is a video of Krishnan Guru-Murthy jamming with some Russian grannies who found fame at the 2013 Eurovision song contest.

Some of these relationships are “symmetrical” – for instance, the Armenian acts tend to be favoured by Russian, Georgian and Greek voters, with Ukrainians also showing some positive attitude towards them (again none of these are so large and precise effects to be deemed substantial, though).

From the past twenty years dynamics, Ukraine tends to receive higher points from Azerbaijan, Russia and Georgia.

Ukraine impact

So can we expect a shock in these dynamics as a result of the latest developments in Ukraine? Maybe.

One of the main provisions is that we should recognise the implicit self-selection present in the data: not only are the votes cast by people who choose to watch the Eurovision contest; they are cast by people who can be bothered to spend their own money to phone in to score their favourite act.

So there is a strong argument against easy generalisations such as “all of Ukraine scores Russia high points”. In other words, it could be that formerly neutrals in Ukraine will now react and deliberately score Russia no points at all, which means we could see a massive decrease in the overall propensity.

But it’s also possible that most Ukrainian voters were pro-Russia anyway and will continue to vote as such.

Gianluca Baio is a lecturer in the Department of Statistical Science at University College London. You can read his blog here. Marta Blangiardo of Imperial College London also contributed to the modeling.