Weaver's Week 2008-03-30
The Grand Final
We're gathered in what looks like the Election '97 studio, a mass of orange and purple lighting with a man in a suit behind a desk. We're actually in the Saltire Centre at Glasgow Caledonian university, a building personally endorsed by the original Mastermind host Magnus Magnusson. Lest we forget, this is the first opportunity the show's had to take the final on the road in the year-and-a-bit since his death. Indeed, Mr. Magnusson presented the prize in the last final.
Stewart Cross has previously taken European Female Painters and Steeleye Span, and will tonight discuss the History of Venice to 1797. As is traditional, the finalists have been allowed to do some research, but they've had to lug a BBC film crew around with them. Venice is, of course, the city in the lagoon, and the significance of 1797 is that it was conquered by Bonaparte of Waterloo, the effective end of the city's independence. The round ends on 12 (0).
Anna Torpey has had The Sopranos and Joanne Harris; tonight, the Life and Films of Steve McQueen. "There's nothing like getting your brain into something," says the contender, who is filmed with her full-time children. McQueen was an actor from the 1950s until his death in 1980. This round finishes on 13 (1).
David Down went through Claude Monet and Bobby Charlton, tonight it's the D-day Landings. Mr. Down only took up the challenge of Mastermind after his retirement, and his filmed insert is made in the well-known Normandy annexe of, er, Bushey Park. The landings were an invasion by French, British, and other troops on the north coast of France in June 1944. The round proceeds to plan, ending on 16 (0).
Sandra Piddock has had Coronation Street and Poldark, and now transfers to It Ain't Half Hot Mum. Looks like she's already won the prize, meeting the series' creator Jimmy Perry at his home in London. The subject is a comedy, based on the exploits of British soldiers in India during the second world war. Final score here: 12 (1).
David Clark offered Henry Ford and George IV, this time he's telling us about the History of London Bridge. He shows us to his old school and round the Museum of London, but doesn't get to fly out to Arizona where the previous incarnation now stands. London Bridge is - oh, you work it out. The round is much more free-flowing than the cars on the structure, ending with 15 (0).
Derek Moody's had The Survivors and The Edge of Darkness, and now offers The Godfather trilogy. The film shows Mr. Moody watching the films and eating in an Italian restaurant. Why an Italian restaurant? It's important in the motion pictures. We have a new leader, the round ends on 17 (0).
Before the general knowledge round, we pause for a commercial from the Glasgow Tourist Board, voiced by Muriel Grey. Our host suggests that we could end up with a five-question tie-break to decide the show, but we can't have two in a row, surely!
Mr. Cross suggests that if the Doge hadn't handed the keys of the city to M. Bonaparte, there would just be a rubble-strewn swamp where the city now stands. One of the islands has already disappeared, and a general rise in sea-levels would claim the entire city. He has a very good round, ending on 26 (1).
Sandra Piddock likes the fact that It Ain't Half Hot Mum is a very clean show, one for Aunt Edna and the children to enjoy without blanching. Her round ends, appropriately, with a question on Calcutta, and a score of 27 (2).
Anna Torpey reminds us that Mr. McQueen has become an iconic figure, known more for his t-shirts than his movies. Like Michael Crawford, Mr. McQueen did his own stunts, often because he was better than the stuntmen. The round progresses to 21 (6).
Mr. Clark says that there were a lot of London Bridges, many of which were destroyed by fire or flood before anyone bothered to write about them. People lived on the bridge, there were battles on the bridge, it was a place where the heads of executed criminals were stuck on display. In spite of confusing the Morning Star and Sun newspapers, the round ends on 30 (0).
Mr. Down's father knew Montgomery, and talks about how the largest sea-landing in recorded history was kept a complete secret from the enemy. He gets the Hidden Climate Indicator of the Week, about winter ending in July and recommencing in August. The final score is 28 (3).
Mr. Moody requires 14 to win, 13 to force a tie. The directors of The Godfather were quietly contacted by the local Mafia chapter, who insisted that the M-word, and its Italian translation, were struck from the script, and everything would be fine. Though he gets the final answer - Morrissey - it's not enough, he ends on 29 (0).
So David Clark, a school teacher from London, is our champion, and receives the trophy from Sally Magnusson. He describes the honour as "an indescribable honour".
Congratulations to him, and to all the other competitors in the series.
How Europe Will Vote
(If Terry Wogan is right)
This week's report is one that's been floating around the in-tray for a couple of years now, and one that deserves its moment in the sun. It's this: if there are patterns to voting at the Eurovision Song Contest, then the outcome of the popular votes will be determined regardless of the quality of the performance. Readers may wish to consider this research alongside the trends we spotted after last year's contest regarding which countries had a propensity to vote for which others.
To put this another way: if Terry Wogan's theory that voting is predicated by politics rather than performance is correct, we are able to calculate the scores, work out who will progress to the final, and who will ultimately win the contest. All without singing a note. In fact, we can do it right now, and save at least 41 national delegations the bother of having British Airways send their bags to a dingy hall in Belgrade while leaving the passengers in a collapsing cowshed somewhere near Staines.
This year, there are two semi-finals, where only the countries competing (and some of the direct qualifiers to the final) can vote. Ten countries will progress from each semi to the final, and all 43 countries vote in the final, a process that will probably take longer than the presentation of the songs. Just to pacify those who have said that this year's contest lacks baffling incomprehensibility, only the top nine in the public vote are assured of a place in the final. The tenth qualifier will be picked from the national juries who are standing by in the event that Buzby falls off the top of the Post Office Tower during the voting period, or everyone in Andorra votes for Spain again, or Irish Telecom are involved. For the purposes of this exercise, we'll assume that the public vote and jury vote will be identical.
To establish the estimated vote from country A to country B, here's what we're doing.
1) Working out the average vote from A to B over the past four years. We use the 2004-2007 contests as this period had the semi-final and a reasonably consistent cast of countries voting for each other. This factor is a number between 0 and 12, generally far nearer 0 than 12.
2) Adding a fixed amount to account for the quality of the song. Our theory is that this perception will be the same across the continent. We've arranged this factor so that it's Normally distributed with mean 0.4, sd 0.2. This element will be between 0.2 and 0.6 for about two-thirds of all songs, and almost always between 0 and 0.8. This factor is exactly the same across all countries for any one song, it relates to a shared pan-European perception of the song.
3) Adding a random amount to account for the local view of the song, an unpredictable amount for any given country. This element is Uniformly distributed between 0 and 1. This factor differs between countries for the same song, it's the local reaction to the song, it's how the viewers reacted to the commentary, it's how they fared with the national televote.
4) For the semi-final only, there's evidence that draw position has a small but noticeable effect. Both semi-finals will see 19 songs presented; for the song appearing in position n, we add (n/100 - 0.1) points. For the first song, this factor works out to -0.09; for the last song, +0.09.
Add together these four elements to work out the estimated score from country A to country B. Then do the same for the score from A to C, and so on for all the remaining countries. Rank these in the traditional Eurovision way to determine points from 12 down to 1. Then repeat the whole process points from country B, and so on. Add up these final scores, and work out who is going through. Then repeat the whole process twenty times, with different (randomly-chosen) values for elements 2) and 3), to give a rough-and-ready probability of each country qualifying. The top-rated countries will always progress, the bottom-rated countries might scrape through occasionally.
For the final, we choose not to use element 4), the position in the draw. In part, there is less evidence that early songs are disadvantaged, but we're discarding the element primarily because we do not know the performance order for the vast majority of songs.
In order to reduce the bias created by one year's results, we scale down the average voting tendency from the relative newcomers - last year's debutants Czechia, Georgia, and Montenegro have their scores halved, 2006 entrants Armenia have their scores reduced by a quarter, as do '05 and '07 competitors Hungary. If we didn't do this, we'd be giving loads of points to last year's highly-placed countries on the basis of one popular song, and if we're going to give lots of points to a country, we need more than one year's data. San Marino and Azerbaijan enter for the first time this year; the history for their voting patterns is uniformly zero, making the random factors far more important in their voting patterns.
The running order for the first semi-final, to air on 20 May, is as follows:
Azerbaijan, Russia, and Greece selected their place in the running order. Also voting in this contest will be Spain and Germany.
The first thing we note is that most countries have voted for about 12 of the competing countries. Obviously, it's impossible for anyone to have voted for San Marino or Azerbaijan, and these countries are at something of a disadvantage in our model. Finland (winners in 2006), Greece (winners in 2005), and Russia (perpetual nearly-men) have a particularly strong recent heritage, while Andorra, Estonia, Ireland, and Montenegro haven't done well from this draw, or indeed from any draw.
Our model predicts that Armenia, Bosnia-Herzegovina, Finland, Greece, Romania, and Russia are certain to qualify. Poland and Moldova are very strongly favoured to get through. Norway and the Netherlands are set to be in the frame for the final spot, with Slovenia and Andorra somewhat more long-shots to progress. Only in one of our 20 trials did Ireland make the top ten, and then only in tenth place.
The second semi-final will air on 22 May, and the performance order is:
Denmark, Macedonia, and Portugal selected their place in the order; Serbia, France, and UK will also be voting.
In this semi-final, we note that Czechia and Georgia will have to break new ground, both countries have given points to just eight of their fellow competitors. Albania's only awarded marks to ten of them, while Latvia has spread points around like no-one's business - only Czechia and Macedonia have failed to receive something from Reynars and his chums. Ukraine is always there or thereabouts, Hungary's appearances have been well rewarded from this set of countries, while Portugal, Switzerland, and Iceland have been mostly ignored.
Our model predicts that Bulgaria, Georgia, Hungary, Macedonia, Turkey, and Ukraine are certain to qualify, Sweden and Belarus very nearly so. Two of Latvia, Malta, and Cyprus will take the last two places, with Albania and Croatia outsiders to progress.
The five direct qualifiers know their place in the running order; Serbia chose her own spot.
For our friendly voting model, we assume that the ten most likely countries to qualify from each semi-final will do so; those are the first ten names in each list, down to Netherlands and Malta.
And, if the models are correct, and all voting is pre-determined anyway, the winner will be ... Serbia! In twenty runs of our putative final, they win 18, the other two go to Greece, with Ukraine consistently in the top three. Armenia, Hungary, and Russia are assured of places in the top ten, Bosnia-Herzegovina, Finland, Georgia, Romania, Sweden, and Turkey are competing to join them. The UK is unlikely to finish dead last - that's been dominated by Germany, France, Malta, and Norway - but equally the country is not likely to rise above 20th.
Here are the extremes of the tests we've run - each of the finalists has a Lowest and Highest Score (in the second and third columns), and a Worst and Best Place (in the fourth and fifth columns).
It is worth our while repeating this exercise for the UK, to determine how that nation might vote. This is of more than academic interest, because the UK will be the first nation to give her vote on the night, and it is useful to find pointers as early as possible. If political voting is a reality, Greece is a racing certainty to pick up the twelve points from the UK, with Turkey and Sweden also seeing reason to cheer. The rest of the points from our hypothetical final go to Malta, Ukraine, Finland, Latvia, Germany, Serbia, and Romania. Small points have gone to Bulgaria, Bosnia-Herzegovina, Hungary, Norway, and Russia; large points have traditionally been awarded to Ireland, a country the UK cannot vote for before the final; and Cyprus, present in the UK's semi. We'll be updating this prediction once the finalists are known.
The draw order for the voting is as follows:
And, of course, we should re-iterate that we don't actually believe a word of this. This column's thesis is that political voting is not enough of a reality to decide the result, and the only collusion is by the people of Europe, whose ears on the night detect great music where the BBC's commentators hear only grated cheese, and vice versa.
This Week And Next
ITV has confirmed that it will be making a prime-time tribute to Jeremy Beadle. An Audience Without Jeremy Beadle will be hosted by Chris Tarrant, and will be co-produced by the inimitable Clive Doig.
Nominations for the BAFTA Craft Awards have been released. The only game show nominated is: Editing (Factual) - Tris Harris – The Apprentice (BBC One/Talkback Thames)
Ratings to 16 March saw Dancing on Ice bow out with a record 12.1m viewers. Saturday Night Takeaway also had a series-best, 7.75m, and I'd Do Anything debuted with 5.65m. Sport Relief's Apprentice take had 5.15m, Millionaire had 4m, and Duel slipped back to 3.85m. Deal was the biggest on the minor channels, securing 2.7m; Mastermind had just 1.7m. Come Dine With Me topped the digital channels for the first time this year, 855,000 tuned in to the More4 repeat. Pop Idle US' (ITV2, 530,000) just beat America's Next Top Model (Living, 525,000).
Highlights of the coming week include the return of University Challenge: The Professionals (BBC2, 8pm Monday and Tuesday), the new series of Banter (Radio 4, 6.30 Thursday), and a million more programmes in The Apprentice and its myriad spin-offs.
To have Weaver's Week emailed to you on publication day, receive our exclusive TV roundup of the game shows in the week ahead, and chat to other ukgameshows.com readers sign up to our Yahoo! Group.