2017 forecast: final
Image source: Andres Putting (EBU)

2017 forecast: final

The yodellers have qualified, proving that there is some justice in this world.

Sadly, the same can’t be said for Serbia, who were the top pick from the model. Still, they were only rated at 76%, which is a long way off certainty. If we look at all of the entrants who were rated between seventy and eighty percent probability of qualifying (Greece, Armenia, Serbia, Norway, Denmark), we find that four out of five qualified, which is roughly in line with expectations.

The other two misses from the model were Lithuania and Malta, both of which were near 50-50 choices. Instead we have Belarus and Bulgaria. Belarus may be benefiting from some displaced Russian diaspora votes, and Bulgaria have quite a strong song (they’re currently third favourite on Betfair).

Final predictions

Winning probabilities for 2017 final

Again, the qualifications were unsurprising, so there hasn’t been much change to the win probabilities. Despite Serbia being a strong qualification contender, their win probability was less than 2%, so they’ve not made much of a difference.

The three countries which are doing well at the bookies are Italy, Portugal and Bulgaria. Italy were strongly favoured going into the contest - they’ve declined, but they’re still rating about 33% win probability. Portugal have moved up to join them - the two have been trading places as favourites for the last day, and at the time of writing they’re completely tied at 33%. Bulgaria are holding up the rear at 21%, and the closest other country is Belgium at 4%.

Of the three, the model rates Italy the strongest (6%), then Bulgaria (2%), and finally Portugal (1%). Based on last year’s analysis of what does well in the televote versus the jury, I would guess that Italy will do best in the televote, Portugal with the juries, and Bulgaria somewhere in between.

Italy, Portugal, Bulgaria

It’s interesting to ask what a contest looks like where each of these countries wins. I’ve plotted below the likelihood of getting twelve points from each voter, assuming a win for each of these three countries.

Douze points likelihood for 2017 final

Portugal has extremely strong support from Spain, Switzerland and France (no Andorra this year), while Italy has Malta and Albania (but weirdly not San Marino). Bulgaria has just Macedonia to rely on, and frankly they’re not that reliable.

It’s much easier to see how the votes come together for Italy - they have a good base of western support, and can probably get by without any of their individual supporters. The other two are much patchier - losing out on one or two countries could spell disaster.

Finally, remember that the voting procedure has been engineered for maximum drama. The jury votes from each country will be announced one by one, and then the televotes will be added all at once, in reverse order. It’s quite likely, as happened last year, that the country which is leading at the end of the jury vote stage will not be the eventual winner.