It's that time of year again...
By Ruth Alexander
BBC Radio 4, More or Less
Britain's Eurovision hopes will no doubt be dashed as bloc voting sweeps a Balkan or Baltic act to victory on Saturday. What patterns do mathematicians spot in the voting?
Eurovision has long been known as a festival of political skulduggery.
No congratulations: Cliff lost out to Spain in 1968
Claims that Sir Cliff Richard was robbed of first place in 1968 because of General Franco's scheming is just the latest story in a long history of grumbling about questionable voting practices.
So much so that allegations of vote-rigging have become the subject of intensive academic inquiry. Sociologists, engineers, mathematicians and even a molecular geneticist have been trying to determine whether suspicions of neighbourly back-scratching are well-founded.
In last year's contest, when the UK came second - from bottom - even commentator Terry Wogan's sense of humour failed him. "Over the last few years, the scoring has undoubtedly become ridiculous," he spluttered. "The voting is so influenced by Baltic groups, and Russian groups, it's become unfunny really."
Serbia was the Eurovision 2007 winner - a result predicted by Dr Derek Gatherer, a scientist who's been studying the song contest for the past five years.
"I decided that if I can write computer programmes to study patterns in biological sequence data, then presumably I can also write something that will study patterns in Eurovision," he says.
FIND OUT MORE...
More or Less is on Radio 4 on Mondays, 1630 BST
Dr Gatherer uses a computer to generate thousands of random simulations of Eurovision song contest results, using data on all the votes cast since 1975. These simulations are then compared with the real results. Any unusual voting patterns can be spotted.
"What comes out of this analysis is that there are essentially three large voting blocs within the contest. One of them is centred around the Balkan countries and the former Yugoslavia," he says.
"The second one is slightly to the east and is centred on Russia and the Ukraine, and includes countries like Belarus and Poland, Georgia and Moldova. And then there's one further north which is around Scandinavia. As well as the Scandinavian countries, it consists of the Baltic countries of Estonia, Latvia and Lithuania."
Derek Gatherer's statistical analysis also shows several smaller trading partnerships, including the UK and Ireland.
So Wogan's right. There's obviously something suspicious going on. Actually, that's not necessarily true.
France hopes to broaden its appeal with English lyrics
Another academic has been looking at some possible innocent explanations. Dr Michel Vellekoop specialises in mathematical finance at the University of Twente in the Netherlands, and ponders the statistical significance of Eurovision voting in his spare time.
"It may be that some countries will give many points to songs which are in the same language, or in a language which is similar," he says. "Or it may be that there are certain cultural characteristics, which makes a song more interesting to other countries."
Together with a colleague Dr Laura Spierdijk, he has taken a closer look at how individual countries have voted between 1975 and 2003.
After correcting for language, cultural and religious preferences, they found strong evidence of political voting only among Estonia, Latvia and Lithuania. Other voting patterns, Dr Vellekoop says, can largely be explained by language preferences and shared cultural tastes.
Cyprus and Greece, for example, are commonly accused of favouring each other and of all the countries, statistics suggest they are the most likely to vote for each other. Wogan seemed to sum it up when Cyprus awarded Greece 12 points in last year's contest. "Over the years people say this is ludicrous, this is ridiculous," he laughed. "But still they do it. They just don't care."
But Dr Vellekopp says his statistical analysis shows the reason the two countries give high marks to each other so often is because their people speak the same language and probably like the same kind of music.
Dr Gatherer also dismisses the idea of straightforward and widespread political bias. But neither can cultural preferences be the key, as voting blocs have grown so rapidly in recent years.
"In the 1980s, there were only two or three countries that were involved in observable vote trading partnerships. But from the 1990s onwards it increased dramatically. In 1993, there were six countries involved. In 1998, there were 12 countries, and now we have 31 countries involved."
"People observe it happening when they watch the contest and then they're motivated to go out and do it themselves in subsequent years. So it seems to be some kind of social epidemic that's spread through Europe and has infected almost everybody."
A Eurovirus. Whether you're in the camp of the conspiracy or the cultural theorists, the statistics explore in detail what the naked eye can already see - voting patterns.
Might the UK's Andy Abraham win, even if there's bloc voting?
Based on his analysis, Dr Gatherer will make a prediction for this year's winning entry after the semi-finals later this week. But he says the likely winners and losers are already clear.
"There are seven countries which are not involved in any kind of vote trading at all: Malta, Monaco, France, Israel, Switzerland, Portugal and Germany. They're at a serious disadvantage and are quite unlikely to win the contest, whereas countries at the centre of the larger blocs - like Serbia, Russia, Sweden and Iceland - have a much higher chance."
So what hope for the UK's Andy Abraham with his catchy song Even If?
"We simply won't get a lot of countries voting for us on the basis of bloc voting patterns," Dr Gatherer says. "So although I obviously wish the British entry the best of luck, I think we're labouring under a very difficult handicap."