Support ESC Insight on Patreon

The New Ranked Scoring System: Looking Forward By Looking Back Written by on March 11, 2013 | 16 Comments

The Eurovision Song Contest of 2013 will be one marked with many changes – from the performance position being determined by producers, right down to how the staging will feature standing-room-only dancefloor to give the show a more lively atmosphere.  After todays’ unveiling of the rules, the fan spotlight is now on some subtle, yet significant, changes to the way the televote and jury vote will be combined in Malmo.  Naturally, there’s a fair bit of confusion. What has changed? What are the implications? Will it change who wins? How does it work, exactly?  Our special guest writer, and statistical genius, John Egan examines the new voting rules and attempts to answer some of these questions.

2013: a year of many changes

Until Malmo, the televote and jury vote components in each voting country were combined by allocation 12, 10, and then from 8 down to 1 point for the top 10 vote recipients in each component, resulting in a maximum of 24 points (12 points from both the televote and jury). That meant that as many 7 entries from a semi-final with 17 total entries (if the jury and televoters came up with the same top 10) would receive no points from that country. These aggregate scores were added up and the 10 entries that scored the most points from that total were then given the voting scores that we see delivered as the final result – from 1, right up to the douze points. If two songs resulted in the same aggregate score, the one with a higher televote ranking was put ahead.

Arguably this simple system worked reasonably well, so long as the alignment between the jury and televote components was fairly consistent. If both top 10 lists have the same (or reasonably similar) entries, the “televote trumps the jury” makes it simple to implement and nicely populist.  Unfortunately, there’s sometimes a sharp difference between which 10 songs the juries liked versus the viewing public. The changes now being introduced in Malmo are directly addressing this issue, however are nothing new to Eurovision; it is simply revamped.

New rules, new prize winner?

In times past, jury members ranked or scored all the entries before each country’s collective top 10 was calculated and awarded. Of course, this was before the introduction of televoting, but can be applied just as easily in the modern contest.  So to put it simply, for 2013, all countries will now be required to provide rankings of all countries in each of the semis as well as final, in both the jury and televote rather than just their top 10 countries of each show.  These results will then be combined to give an overall total for each, a final ranking given, and then the traditional points applied to the top 10 of that newly created list.

The changes: our analysis

We could drill deeply into the maths at play here, but for those who lose focus when numbers go beyond a simple ‘douze points’ (that would be me – ed) , it would probably be more confusing that clarifying. Instead we’ll look at the results of the 2012 semi-finals and final in detail, and the 2011 final for a quick comparison.  From this you can see the extent to which this change might have made a difference in who qualified from the semi-finals and where everyone ended up in the final rankings for Saturday night’s big show.  Remember it’s the rankings in each of the televote and jury vote components that matters – not the raw score in either.

In the comparisons below we have simply added the jury rankings from across Europe to the televote rankings from across Europe – from first to last – in each semi-final and the Grand Final, and ranked them from lowest to highest total score. In the event of a tie, and there were several ties, the songs with the higher televote rankings were placed ahead.

2012 Semi-finals: One subtle change, one new qualifier

Despite the panic that we are currently seeing amongst the fans and discussions on the chat boards, we can see that 19 out of the qualifiers from last year’s semi-finals would have still made the cut under this new change.

To demonstrate, here is the summary data:

As evidenced, the new change can see positions shift somewhat.  In the first semi-final, Russia moves from 1st to 4th: Albania ends up the winner, Romania and Greece each move up a placing , and there’s a bit of movement between spots 6-10.  However, and most critically, the qualifications remain unchanged.

In the second semi-final, we see Estonia and Lithuania swapping their 3rd and 4th positions, and Turkey dropping behind the likes Bosnia, Malta, FYR Macedonia and Ukraine.  Only the 10th place result would have seen a difference to our qualifiers list for the Saturday finale, with the changes providing Croatias’ Nina Badric with a little piece of ‘Nebo’ (Heaven), and unfortunately for Norway, Tooji no longer ‘Stay’ing in contention.

So do the results change for the Grand Final 2012?

New rules: still a Euphoric result?

Let us take a look at the adjusted scoreboard –

As we can see, Loreen remains our victor for 2012, and despite the dreams of many countries and fans to be elsewhere as our wallets drain, we would still be going to Malmo in 2013.  Admittedly though, after the second most convincing win in the history of the Contest, it’s unsurprising that Loreen still wins under the adjusted rankings.  However under the new rules, Serbias’ Zeljko Joksimovic would have equalled his 2004 Istanbul result and finished runner-up rather than Russia; a result of the juries not being as enamored by the entry as much as the public. Albania would have moved up to the bronze medal position and Russias’ Grannies would have danced their way back to 4th. Azerbaijan also would have found themselves a position lower, whilst Germany – not standing still – moves up two.

Perhaps the most noteworthy other shifts are Turkey falling from 7th to 11th and Moldova climbing into the top 10 by grabbing 9th place. The rest of the top 10 moves round a bit.

2011: a closer result

In a year like 2012, the winner wouldn’t have been anyone besides Sweden. For 2011, where the differences between televotes and juries were sometimes very stark, and the win was much narrower..

We can see that the new aggregate still gives us the same winner before and after the rule change.  Azerbaijans’ Ell and Nikki with their entry ‘Running Scared’ remain the victors at the 2011 contest, though Eric Saade is a bit more ‘Popular’ than before. Ireland moves into the top 5 (from 8th), Serbia moves up five places into the top 10 (14th to 9th) and Georgia is bumped down to 11th. Otherwise, there’s only a bit of shifting around.

Ultimately, what would give a clearer indication of the implications of this change would be to take the complete rankings from a country’s jury and televotes from 2011 or 2012; unfortunately this data does not exist. Jurors last year and the year before only provided their top 10 list within each jury, which then created the overall top 10 for the entire jury.

This change is very much evolutionary rather than revolutionary it seems. Nonetheless, we will find out first hand how this change, as well as the standing room only and the performance order changes will all fare in just over two months time.

 

You Can Support ESC Insight on Patreon

ESC Insight's Patreon page is now live; click here to see what it's all about, and how you can get involved and directly support our coverage of your Eurovision Song Contest.

Share This Post

Have Your Say

16 responses to “The New Ranked Scoring System: Looking Forward By Looking Back”

  1. Ben Cook says:

    I don’t see how we can take these results as any evidence that the new system won’t be a problem, since they aren’t calculated in the same way. We’ve no idea how low the jurors who didn’t vote for Russia 2012 for example, would have put them if they’d had to rank them from 1 to 25. They could have ended up a lot lower than 4th if the jurors that didn’t like the entry all put them 25th on their votes.

  2. Jamie says:

    Can you explain this more clearly please. What data sources have you used for your calculations? I don’t see how it is possible to back-fit historical results accurately without the detailed country by country raw data of 1-26 rankings for both public and jury. If you have used the summary public and jury data published by the EBU then your conclusions are invalid.

  3. Eric Graf says:

    Well Jamie, It’s as close to valid as you’re going to get.

    This all seems like a tempest in a teapot to me. “OMG OMG OMG The evil undemocratic elitist juries will be able to secretly collude and keep my favorite Eurocrap pop idol who can’t carry a tune in a bucket from winning the contest!”

    Uh, no. The idea of ALL 42 juries giving a popular act last place is absolutely absurd. Not going to happen. Ever.

    Now, I can see this as a “Buranovo Grannie Prevention Measure”, in that it has the potential to keep charming but talent-impaired novelty acts from winning the contest. But I fail to see the downside of that.

  4. Zolan says:

    I haven’t found any downsides to new system, only improvements.

    Nice attempt to construct a parallel universe example from available data. It approximates what the new system would do with the numbers we have pretty well, even if it can’t produce what the numbers might have been.

    It might also have been useful to illustrate scoring with an xy-plot, and suggest people imagine what kind of entry could occupy various coordinates for intuition’s sake (relating it to people’s concerns about what seems fair).
    Then show what the 10×10 box does.

    Of course, it’s usually just people’s ambivalence about juries, regardless of the details.

  5. Zolan says:

    People are also overlooking a couple of things that don’t fit the conspiracy narrative. Maybe superfluous, but I thought them worth mentioning.

    Ranking all of the former zeroes promotes as much as it demotes. 11th is actually better than 17th (or 25th), and that counts for something under the new system.

    A top ten can always be produced using both rankings without having to decide between half-missing scores. All preferences are considered, however far down they occur.

    A middling jury rank can help a public favourite surpass a jury favourite, just as already occurs with a top ten.
    Imagine the opposite, in which the jury only supplies a single favourite, to see how more rankings softens influence.

    It goes both ways. The televote can effectively veto deeply unpopular acts even if they are jury favourites.

    More rankings increases transparency, making juries more susceptible to public scrutiny.
    They also supply better data for analysis, hypothesis testing and, if required, further refinements.

  6. John Egan says:

    As it says in the article, there isn’t any data from the previous two years to fully test this new system: we only have the full pan-EBU tele vote and jury rankings. But these are based on televote only versus jury vote only.

    As it also says in the article, we used the data available to explain how the revised calculation would be done for each country. So the example, limited as it is, shows a result would have shifted.

    Imperfect, but concrete. That’s as good as it’s gonna get, for now.

  7. Eric Graf says:

    An observation: The juries’ lives just got way more complicated. Now everyone has to make a top 18 or 26 instead of a top 10, splitting the hairs between whether “I’m a Joker” is really one notch better than “Don’t Close Your Eyes”. Now there’s a thankless task.

  8. Zolan says:

    @Eric. Yes, although it’s not always easy at the top end either, when you might be choosing between different genres, for example.

    But I don’t think there’s any requirement for _individual_ jurors to make exhaustive rankings. There is not much point forcing jurors to break every tie independentlyy since there’s always the possibility of a tie in the aggregate result anyway. It’s better left until the need arises.

  9. Chad Brown says:

    I didn’t understand any of this. What I want to know is… Once they decide the running order will it be announced before the show or do we have to wait and watch to find out? I like to listen to the songs in “running order” to get a feel for the show ahead. Wondering if I will be able to do that this year…

  10. Max Burton says:

    By means of Scoreboard Simulator I analysed the possible results in order to better understand how this new ranked scoring system works. (Don’t care about results of countries. It’s only an example.)

    And these are the results of only JURY voting, but in traditional ESC way – top 10: points from 1- 8 10 and 12 points.

    1 Russia 214
    2 Norway 200
    3 Iceland 194
    4 Malta 149
    5 The Netherlands 147
    6 Denmark 145
    7 Azerbaijan 114
    8 Georgia 101
    9 Belarus 100
    10 Finland 89
    11 Serbia 89
    12 San Marino 84
    13 Germany 82
    14 Denmark 77
    15 Italy 71
    16 Ukraine 69
    17 Austria 52
    18 Armenia 52
    19 France 50
    20 Ireland 50
    21 Switzerland 39
    22 Moldova 39
    23 Estonia 27
    24 United Kingdom 24
    25 Spain 4
    26 Greece 0

    Following Jury voting it is time for the (only) TELEVOTING. Points in traditional ESC way: 1-8, 10 and 12 points.

    1 Greece 247
    2 Norway 188
    3 Russia 168
    4 The Netherlands 167
    5 Sweden 157
    6 Belarus 143
    7 Armenia 142
    8 Serbia 122
    9 Ukraine 108
    10 San Marino 96
    11 Georgia 95
    12 United Kingdom 91
    13 Azerbaijan 85
    14 Switzerland 72
    15 Moldova 67
    16 Finland 64
    17 Italy 51
    18 Germany 42
    19 Ireland 36
    20 Austria 29
    21 Iceland 25
    22 Estonia 24
    23 Denmark 22
    24 France 12
    25 Malta 6
    26 Spain 3

    I combined these points (above) from Jury and Televoting according to a previous method (I added up only top 10 countries in both ranks.) And that’s what I received, as far as the old system is concerned.

    1 Norway 233
    2 The Netherlands 206
    3 Russia 188
    4 Belarus 147
    5 Sweden 124
    6 Georgia 111
    7 Azerbaijan 109
    8 San Marino 108
    9 Serbia 107
    10 Armenia 104
    11 Italy 83
    12 Iceland 80
    13 Ukraine 79
    14 Finland 76
    15 Switzerland 68
    16 Germany 64
    17 Denmark 55
    18 Moldova 54
    19 United Kingdom 53
    20 Malta 43
    21 Austria 43
    22 Greece 39
    23 France 35
    24 Ireland 33
    25 Estonia 20
    26 Spain 0

    Finally, the Final results of the Grand Final in the new scoring system. (The Simulator combined the full rank both of jury and televoting respectively from the last to the first place).

    1 Norway 233
    2 The Netherlands 206
    3 Russia 188
    4 Belarus 147
    5 Sweden 124
    6 Georgia 111
    7 Azerbaijan 109
    8 San Marino 108
    9 Serbia 107
    10 Armenia 104
    11 Italy 83
    12 Iceland 80
    13 Ukraine 79
    14 Finland 76
    15 Switzerland 68
    16 Germany 64
    17 Denmark 55
    18 Moldova 54
    19 United Kingdom 53
    20 Malta 43
    21 Austria 43
    22 Greece 39
    23 France 35
    24 Ireland 33
    25 Estonia 20
    26 Spain 0

    Can you tell the difference between those two scoreboards ? I hope so. As we can see, Greece despite having victory in Televoting, it is last in Jury voting with null points. In the old voting system, it takes 4th place, but in the new scoring system it finishes 22nd. it’s a big gap, isn’t it ? Besides Georgia goes up from 11th to 6th place, Azerbaijan from 12th to 7th, Italy from 17th to 11th but then Iceland goes down from 6th to 12th.

  11. […] von 2012 (soweit die Daten veröffentlicht wurden) bei den Kollegen der englischsprachigen Website ESCInsight weist darauf hin, dass sich am Ausgang in Baku nichts geändert hätte: Loreen war unbesiegbar. Die […]

  12. ReneWiersma says:

    Max, I think you copied a wrong table somewhere, because the last two tables are exactly the same?

  13. Max Burton says:

    That’s right.

    50%/50% system (the old version):

    1 Norway 240
    2 Russia 218
    3 The Netherlands 172
    4 Greece 144
    5 Belarus 133
    6 Iceland 123
    7 Sweden 118
    8 Serbia 112
    9 Armenia 105
    10 San Marino 103
    11 Georgia 102
    12 Azerbaijan 90
    13 Ukraine 80
    14 Finland 71
    15 Denmark 69
    16 Malta 57
    17 Italy 52
    18 Moldova 50
    19 Germany 45
    20 Switzerland 45
    21 United Kingdom 44
    22 Austria 28
    23 Ireland 22
    24 France 20
    25 Estonia 19
    26 Spain 0

    The ranked scoring system:

    1 Norway 233
    2 The Netherlands 206
    3 Russia 188
    4 Belarus 147
    5 Sweden 124
    6 Georgia 111
    7 Azerbaijan 109
    8 San Marino 108
    9 Serbia 107
    10 Armenia 104
    11 Italy 83
    12 Iceland 80
    13 Ukraine 79
    14 Finland 76
    15 Switzerland 68
    16 Germany 64
    17 Denmark 55
    18 Moldova 54
    19 United Kingdom 53
    20 Malta 43
    21 Austria 43
    22 Greece 39
    23 France 35
    24 Ireland 33
    25 Estonia 20
    26 Spain 0

  14. Eric Graf says:

    Well, I don’t think you can dismiss the new voting rules just based on how it affects Greece.

    The big difference now is that if **either** the juries or the televoters really REALLY hate your song, then you will get seriously downgraded in the final standings. And I’m fine with this. Why should Greece come in 4th if the juries hate them that much? That’s unfair too!

    The sad fact is, most of the “general public” knows practically nothing about music. They have no idea if a song is poorly constructed, or if the singers are off-key. They vote for songs based on how cute the lead singer is. They’re easily swayed by pyro, a short skirt, and a stompy backing track. They don’t recognize when a lead vocal is being “buried in the mix” because the singer is lousy. They will vote for an inept performer for the stupidest of reasons. Why shouldn’t the juries be able to vote Greece down to 22nd place if the performance is crummy?

    And please do not forget that it works the other way too. If the juries thought that Greece’s act was fabulous, but the viewers hated it, then it still would land in 22nd place. Any technically accomplished performance that the viewers can’t stand will suffer the same fate. (Poor Cesar!)

    Not one single country is gaining or losing an advantage with this rule change. The only difference is that no act will do well by impressing ONLY the viewers or ONLY the juries – as Greece did in your example. And that’s as it should be.

    The answer to Greece in this case is the same as the answer to most whiny countries who don’t like their result: SEND BETTER SONGS!

  15. Max Burton says:

    First of all, I am not going to dismiss the new voting rules. I am an enthusiast of this. Moreover, I even was thinking about this method in combining points of jury and televoting many months before it was officialy confirmed on 11th March this year. I find it the best possible voting system that has ever been at the ESC so far.

    I decided to make these rank in order to check how the new scoring system works. And as we can see, cliche songs (like Russian entry in last year) aren’t able to achieve the success if the jury doesn’t appreciate them. In previous voting system every big points from the televoting (8, 10 and 12 points) automatically guaranteed points at least in top 5 in the final points of a given country in 50/50 system, even if jury didn’t give any points for these songs. It supported either political or diaspora voting. In current voting system, even if (for instance) Greece will be 1st in the televoting and 20th in the jury voting, it still can finish beyond the top 10, with no points.

  16. […] How might a tweaking of the ranking system for this year’s Contest impact the final results? […]

Leave a Reply