(Ln(x))3

The everyday blog of Richard Bartle.

RSS feeds: v0.91; v1.0 (RDF); v2.0; Atom.

Previous entry. Next entry.


3:00pm on Tuesday, 13th May, 2014:

Eurovision Voting

Comment

The Eurovision Song Contest votes were put online after the competition for "transparency". I've had a look at how the UK voted to see how widely the audience and the jury diverged.

For the sake of non-devotees of Eurovision, here's how it works: there are 26 countries and each one gets to vote for their favourite song of the other 25. The votes come in two parts: half from a panel of industry professionals and half from a popular vote. The reason for the split is that if it all went on the popular vote then the biases of the general public would get in the way (not that the industry panels of Greece and Cyprus ever really differed with the public over which country had the best act).

That's what we're told, anyway. However, how do you apportion half the votes from a panel of 5 people and a millions-strong popular vote? If it were just the popular vote it would be easy: order the acts by how many votes they got and the one with the most gets the most points. For a panel of 5, that doesn't work as there are fewer panellists than countries.

What seems to happen is that the panellists individually order the countries from best to worst and the sum for each country is totalled. These totals are then put in order to give an "average" ranking. This ranking is the one that counted as the panel's vote. For example, the five panellists ranked Malta 3, 2, 2, 4 and 9; this comes to a total of 20; 20 is the lowest such total, therefore Malta was ranked first by the panel. Finland was second on 2, 13, 3, 1, 3 = 22; Poland was last on 25, 25, 24, 24, 25 = 123.

I've run this through a spreadsheet (you can download the marks in Excel format) and can confirm it works.

Thereis a problem with this approach, though: what happens if two countries get the same total? This only happens once with the UK's panel votes: 7, 22, 19, 11, 5 = 64 for Norway and 15, 15, 13, 10, 11 = 64 for Hungary. The Eurovision web site for the UK votes puts Norway at rank 11 and Hungary at rank 12. How did they decide that, I wonder? I'm guessing that if there's a draw, the chair of the panel (presumably judge A) has the casting vote. Norway was ranked 7 by our judge A, Hungary was ranked 15, so Norway comes before Hungary.

Right now we have an ordering for the panel we can do a similar thing to combine the panel ordering and the popular vote ordering to get the final ordering. So, the panel vote ranked Austria third and the popular vote ranked Austria third: 3+3 = 6; 6 is the lowest total so Austria won. Montenegro was 21, 25 = 46, the highest total so Montenegro came last. Sorted!

Again, I put this in the spreadsheet and the ordering is the same as on the web page. This does definitely look to be how they do it.

However, Malta got 1, 5 = 6. That's the same as Austria. Why did Austria get the UK's 12 points rather than Malta? Similarly, Sweden was 3, 8 = 13 and Finland was 2, 11 = 13. Why is Sweden above Finland? Iceland and Denmark are both on 19, Poland and Hungary are both on 26, Norway and Russia are both on 28, Belarus and France are both on 38. How do you decide which one has precedence? Well, it looks as if the popular vote wins in this case. France was ranked 15 by the popular vote and Belarus was ranked 20, so France comes 20 overall and Belarus 21.

It's nice to see the popular vote being important here, but I'm still a little wary of how this all works. See, Poland came last with the panel and first with the popular vote. The panel's decision meant that the lowest score Poland could get would be 26. For Poland to get any points from the UK in the contest, voters would have had to vote in large numbers for other songs that the panellists didn't like. The panel vote was already known before the public vote, so the public could have been told in advance that they were pretty well wasting their money voting for Poland...

I'm pretty sure there must be a better way of sorting this than totalling rankings and using the resulting total to establish another rank, though.

Oh, and as for why the UK public voted Poland as having the best song, I suspect it was a combination of its only-in-Eurovision bonkerness and the large number of Polish immigrants we have in the UK at the moment. Nothing to do with buxom milkmaids at all.




Latest entries.

Archived entries.

About this blog.

Copyright © 2014 Richard Bartle (richard@mud.co.uk).