Late last night, the Eurovision.tv website posted the ‘split vote’ results of the 2013 Song Contest. The average jury and televote rankings for each song in the two semi-finals and the grand final was announced. Take a few minutes to look over the message boards, forums, and other community sites and you’ll see discussion on who the juries helped and hindered, the televote results that were over-ruled, and other distractions from these numbers.
Yes, distractions, because ultimately the data on the jury and televote rankings released by the EBU tells us only a little more than nothing. It is a panacea for those who want to see how the Contest is scored, how the songs are judged, and who want to dig deeper into the cultural data set that the Eurovision voting represents.
A job half done is a job not done. The split vote ranking data, as released by the EBU, is less than half the job. Frankly, I think it would have been better if the EBU had not released any more data from the televote or the jury vote beyond what we saw on Saturday 18th May if this is all they feel they can release.
Nobody Really Understands Averages
Let’s illustrate this with a limited example of two songs and two juries, that are both listed as having an average jury ranking of ‘7’. When the combined rankings of these songs are translated into a placement, they could each receive a different score.
It could be that song one was first with one jury, and thirteenth with another. Translate that to the classic Eurovision Bourda points (12, 10, 8, 7, 6, etc) and it would score ten points. The second song could have finished sixth and eighth. Again in the old scoring system, that scores eight points (five and three). The same jury ranking, the same place in a table constructed as the 2013 split votes have been calculated, yet the scores awarded are different.
Tonight Matthew, I’m going to be Kate Winslett
The ranking data provided by the EBU rightly shows a victory for Denmark. But it’s entirely possible to construct a set of televotes and jury votes that would hand the victory to Azerbaijan. With a bit of work I reckon you could take the published data and fashion a voting set that puts the United Kingdom in the Top Ten, and perhaps even top the scoreboard.
Why would you release data that allows for so much flexibility in its interpretation?
It Sends The Wrong Message
The chatter online has already started. ‘Montenegro were fourth in the televote, they should have qualified’, ‘The jury voted Austria fifth and they were sent home’,’ Romania won the televote in the second semi final’, ‘the jury had San Marino as qualifiers’, and so on. All of these statements will rapidly crystallise into the ‘facts’ of the 2013 Eurovision Song Contest, when they are in actuality little more than an intermediate step towards the final result. Any tiny crumb of information is being treated as a feast.
All of these discussion points come down to fairness. The split vote might put Denmark at the top of the televote and jury vote’s ‘average rankings’, but when you start looking at other results, the implication is that this system is not fair. That’s not the case, but the explanation of the new scoring procedures has not been clear. Everyone was comfortable with what to expect with jury and televote results in previous years, but those rules of thumb no longer apply.
Unfortunately they’re still being used to try and analyse the split vote.
A Eurovision Grand Final is the final frontier for Montenegro
Anyone entering a Contest wants to believe that they have a chance of winning and that it will be conducted fairly. I’m sure that performers and delegations are aware of the changes and how the scores are calculated, but the Eurovision Song Contest is one of the few annual moments where it feels like a nation is entering a Contest – and the people of each nation must believe in the inherent fairness of that Contest.
Why would you release data that destroys the sense of fair play that should be present in any contest, rather than highlighting it?
It Makes It Easier To Hide Things
Okay, I may be veering slightly into ‘tin foil hat‘ territory here, but just because these seem outlandish, it doesn’t mean that the average Eurovision Song Contest fan won’t be thinking them – after all the latest theory behind Cascada’s loss this year is because ‘everyone hates the imposition of economic restrictions by Germany’ rather than anything to do with the performance on the night. No matter how crazy that might sound, conspiracy theories like that can damage ‘Brand Eurovsion’ for a long period of time.
This year was a 50/50 combination of jury votes and televotes, just as last year was a 50/50 combination of jury votes and televotes. That is a fact. But it’s the exact method of taking the two 50% sides and synthesizing them that’s in question here. How the combination process worked during 2013 is vital to understanding how the Contest unfolded, and it is this exact process that is being hidden by the EBU.
The EBU’s answer here is that to reveal the country splits would show where the televote did not reach the threshold limit (set targets in a number of areas to show that a televote is representative). Where a threshold is not met, the 50% jury vote becomes the vote from that broadcaster. It’s fair to say that San Marino would be a 100% jury vote given the issues in identifying enough votes from inside the Serene Republic, compared to votes from the surrounding Italian countryside.
Third time’s the charm!
The argument goes that if the countries not meeting the threshold are known, it would be easier to influence the televote in the following year. This ignores the fact that in previous years the countries not meeting the threshold were announced on the night by the then-Executive Supervisor Svante Stockselius, and that televote organizers Digame have procedures in place to negate the potential ‘power’ voting of people looking to influence the final result.
Of course with the average rankings and some smart maths, you can work out a few things. EscXtra’s Ervin Juhász has looked at the semi-finals and can say with a high level of confidence that one country used 100% jury vote in the first semi final, and two countries used a 100% jury vote in the second semi final. But that’s about all we can be sure about.
Putting the tin foil hat back on, therefore it must be a conspiracy!!! What would the result have been under 2012? Would it have been an Azeri victory, and all the political baggage that would have brought? Would we see that a significant number of countries did not reach the televoting threshold in the grand final, implying that tens of thousands of viewers might have wasted money casting votes that would never have an impact on the result? Or perhaps the cultural voting blocks are so pervasive that there is no easy way for the EBU to diminish their effects?
Why would you release data that creates the illusion of a hidden agenda?
It Would Have Been Better To Remain Silent
Looking at the issues around it, the ‘split vote’ announcement is a weak compromise between preserving the competitive nature of the Song Contest and being open with the data. It has failed on both counts. If meaningful split voting results are too sensitive to be released, then don’t try and hide behind a smokescreen of ambiguous numbers… make a decision to not release them.
These numbers given out by the EBU are incomplete at best, and it would be difficult to rely on them for any rigorous study. It feels like the official Eurovision press team have thrown us ‘something shiny‘ in the hope that the discussions will be distracted away from the numbers that actually matter.
Perhaps the best solution here would have been for the EBU to not release any split figures at all.
That is exactly my opinion….! Thank you!
But aren’t we all the wiser because they have? This is a total shambles. They should bring back the old system, pronto.
An interesting read, Ewan! A particularly relevant statement was your, “Any tiny crumb of information is being treated as a feast.”
I am however, one of those suggesting a return to form for Eastern-Bloc dominance, especially in light of recent reports into voting irregularities. The betting market for Ukraine was particularly strange, bearing in mind Azerbaijan’s odds lengthened during the show.
Under these new rules, juries can exercise far more influence over the result. Buy several jurors and a series of 20+ rankings can kill a song’s win potential.
One of the crumbs I was able to feed off suggested a new outlook and marking criteria for the juries. Ballads are not routinely marked up. Instead, their mantra seems to be focused on rewarding commercial/current songs over the more traditional olde-Esc entries. Effectively, ESC appears to be channeling Melodifestivalen.
Betting markets definitely your area, the patterns there would make an interesting read. As to criteria, the top of the Jury forms suggested that jurors look for were:
Vocal capacity, The performance on stage, The composition and originality of the song, and The overall impression by the act.
Anything else and the jurors added that themselves
“The argument goes that if the countries not meeting the threshold are known, it would be easier to influence the televote in the following year.”
This reason for the EBU to not release the data is a load of bollocks. Anyone wanting to influence the televote will know which countries to target. It’s not terribly hard to figure out.
Also, the EBU releasing such a statement implies that tampering with the televote is something that might have happened at sometime in the past, and might happen again in the future.
I agree, either release the full split results per country, or not release them at all.
FInlands jury-
Sarin-record producer.
Salokorpi-musician.
Mustonen-lyricist.
Saari-singer.
Laine-DJ.
I have a couple of problems with this jury.Laine-the “relevant DJ” only works with songs in finnish-her station (radio Suomi) only plays songs in the finnish language.I dont know how she is qualified to judge on an international song festival.Mustonen is best known for her teenage lyrics in finnish-“frontside ollie” Again-to my knowledge she has never written anything in English.
Saari-the singer-finalist in finlands final.He is a 24 yr old student and does not have a lot of musical experience.Good singer though.
The other 2 -fair enough.Overall I do not feel this is a jury qualified to judge in an international song competition.Finland scraped the barrel with the judges somewhat.I know the world does not revolve around english-but Finnish is such an unusual language-and there were LOADS of people in Finland better qualified..
https://svenska.yle.fi/artikel/2013/05/18/har-ar-finlands-jury
I totally agree with this article, nice writing.
Going through some of the blogs and forums this morning was really frustrating after reading for the 100th time how the results were obviously fake because there’s no way Spain could be last in both Jury and Televote results, but still beat Ireland in the final! Or how the splits prove that the juries prevented Greece from winning, or prove that EBU were responsible for crop circles, or whatever crackpot theory the blog wanted to spout to generate themselves more clicks this time.
On a different note, I still don’t understand how the EBU can have complete faith in Digame and their anti-fraud / power vote system, but still want to hide exact details in case people learn how to game the system better. I’m quite sure that countries interested in “enhancing their scores” already have a pretty good idea of where to focus their attentions.
All the EBU is actually achieving instead with their policy of “security through obscurity” is hiding their mistakes and preventing *us* from having any trust in their system.
It does seem strange that they would release these average rankings instead of the split points like usually, which are slightly more helpful. I wonder if there was an increase in fishy televoting this year and they wanted to hide the fact that more than one or two countries (say, tiny San Marino, or perenniel televote-issue having nations Albania and Norway) had to rely exclusively on the juries.
I’m with Peter in regards to the seeming contradiction between the EBU’s proclaimed trust in Digame but fear that someone may learn how to game the system.
Something to note, the points given via televoting for years past has always added up to the correct number except when there was something specifically indicated (Albania and Norway have had to use only jury voting in the past for technical reasons). So this makes me think that something happened this year that caused in unusually high number of countries to rely exclusively on their jury. And, given the ESC Extra article that Ewan linked to, there’s a possibility that someone was able to get away with not using their jury and relying on only televoting for the second semi-final.
Well said. This is very much smoke and mirrors.
I this case Ewan I’m going to disagree with you on releasing split results. I think that it is important to see roughly what inputs the juries had and how the public vote went, from a point of view of understanding how a song does well in the competition.
I don’t think the EBU meant for these votes to be extrapolated in the way that you suggest, and to the extent so fans are doing so.
It’s face it – no one could reverse engineer the old style of split results. So why is there uproar about these results?
These results have done what the EBU wanted them to do. Raise awareness of which way the vote went.
Why couldn’t the EBU have given us points instead of average numbers? Just translate each country’s jury votes into points – you know, if the jury in country A has given country G the lowest score, that would be 12 points, and country M with the second lowest score would be 10, and so on until country C with the tenth lowest score which gets 1 point. Add all the jury points up and there you go. Do the same with the televotes and we have what we want: Who would have won if only the juries had counted, and who would have won without them. No need to publish individual countries’ results for that if the EBU really think that’s such a terrible thing to do.
But these numbers do nothing but confuse people. It’s true, nobody understands averages. I know I can’t take these numbers to mean what they seem to mean at first glance, but I don’t have a clue as to what I can take from them at all. But all over the Internet, people are going “Look, the juries put Sweden on 3” (even though Sweden’s average rank is 8 or something) and declare who has been “robbed” or how this has been “rigged”, when there is no basis at all for any of this in these numbers. What good does that do anyone, really?
More tin-foil hattery here but just a question.
Why averages?
The fans are aware of the new system. What could the need possibly be to show us an average rank instead of a raw overall European ranking in the split results? It wouldn’t change anything.
Unless Denmark wouldn’t have won under the old 50/50 system.
Dun dun dun.
I just did a rough reverse-engineering job and was able to distribute average ranks across countries to give possible score ranges of
__ Ireland 0 … 168
__ Denmark 68 … 360
Might not be exact, but it illustrates the point.
The EBU abandoned the built-in shredder and now have no strategy for sanitising the info people want.
I suspect they believed these numbers would somehow validate the system purely because ‘vertical’ combination will tend to converge on actual placings.
Of course people will slot these numbers into their habitual framework because it looks like it might fit, and they’ll take any shortcut available to avoid hard math.
By next year we can expect a better toolkit of concepts and methods.
Sean,
My issue is that we have not seen, even roughly, the jury inputs. Take the Spain/Ireland final positions… two or three juries must have given Spain relatively high marks for them to leapfrog Ireland in the final scores. Which countries? Public or Jury input? Two 12’s? 4 sixes? Neighbours or distributed all over Europe? AN average mark flattens information and makes it much harder to get anything useful.
@ewan, for me. It’s time for the ebu and digame to put up or shut up, in regards to the results and voting data.
Yep, I agree.
The split rankings are better than nothing, but they reveal barely anything about the voting split. In the name of transparency the EBU should release the full split voting figures, like they did in 2009. Won’t happen though…
Ewan Spence,
Very good article. I think most eurovision fans are just happy with the results and don’t really care about calculations. There are always countries issues, blockvoting and so on. As you mentionned, the only reason why true results in terms of points were not released is because they didn’t want to pay back televoters where threshold wasn’t reached. It is almost sure that results would be different according to last year calculation rules… This system allows jury to push down strongly countries they dislike since a last position from jury would almost cancel a top choice from televoting… Very frustrating you cannot know split results from your country to see whether jury or other fans follow you… Cheers, Maxime
I take your point Ewan but I maintain that the old results didn’t really tell us much more information than we have at the moment from the averages.
I do think more “in depth” split results are ideal, but I only have ever used split results to understand in broad terms what the public like and what juries like.
Split results should only inform us as we move on to next year, rather than leave us stuck in the past.
Sean, to be fair I argued that the old results weren’t enough back in the day (see here), but they at least gave us some cardinal numbers. Even that little crumb has been taken away.
I’ve given this some thought, and I think it might be possible to extract the full split results per country based on these average rankings and the final score board scores!
It would require coding a program that goes through all possible combinations of jury and televote scores for all countries and matching them with the final scoreboard scores and the average rankings.
I’d guess there are thousands of combinations of jury and televoting scores possible that would lead to the exact final scoreboard scores, but very few that also would lead to the exact same average rankings. Perhaps even just one.
As Ervin Juhasz pointed out in his excellent EscXtra editorial, some countries may have used only jury televoting, or perhaps even just the televote, so the program should take that into account as well, but I’m certain it can be done. The data is out there, it is just waiting to be found.