A fairer kind of televoting
I'm thinking out loud here, so this probably has a glaring loophole or is completely non-viable for some reason that I haven't spotted yet. At the moment, the points are allocated by counting the absolute number of telephone calls and SMSes made, with all the fun and enjoyment which that leads to when the points get announced. I'm going to propose a system that would - I think - exactly cancel out the effects of friendship and diaspora and focus the results back on the popularity of the songs.

It gets complicated very quickly, but it'd take a matter of seconds for a computer to sort it all out. Country A to Country B</b>
Occurrence 1: 5% (24 options)
Occurrence 2: 2% (27 options)
Occurrence 3: 3% (24 options)
Occurrence 4: 7% (22 options)
Occurrence 5: 1% (25 options)

We'll then normalise that to convert to the number of options available in the next contest - presumed to be 24, in this case.

1: 5% x 24 / 24 = 5%
2: 2% x 27 / 24 = 2.25%
3: 3% x 24 / 24 = 3%
4: 7% x 22 / 24 = 6.42%
5: 1% x 25 / 24 = 1.04%

We then take an average of these five figures, to give us our Target Score (TS) - in this case, 3.54% of the total televotes cast.

We then look at the scores from our new event, in which Country B receives (for example) 3.9% of the televotes cast. This makes their Comparative Score (CS) in this year's contest +0.36 (3.9 - 3.54). If they got 3.4% of the televote, then their CS would be -0.14. and so on.

In order to allocate the points awarded by country A, we rank the CS figures of all their 24 options. Whoever has the best CS figure receives the 12, next best gets the 10 and so on.

Where a televoting pair hasn't occurred 5 times before, the TS calculation is padded by assuming that Country B took part in an event (or events) where all the options received equal numbers of votes - in this case, 100/24 = 4.17% of the votes.

Drawbacks - this will tend to favour those countries which are new to the contest with significant diasporas, or those who have been sending streams of rancid llama poo for the last several years. It also increasingly handicaps those that have been in genuinely good form for several years, as they have to keep outdoing themselves to get good scores. Finally, it *does* mean that the result doesn't truly reflect the votes cast this year. And, it's complicated to explain to the public - assuming that you even try to explain it to the public.

Advantages - the results become truly unpredictable again, and will favour those countries which are in improving form or sending genuinely great songs. It also neatly and precisely negates the diaspora effects on a country-by-country basis.

Thoughts? Other than "My head hurts...", which I think is almost guaranteed with my harebrained schemes!