Weaver's Week 2007-10-21
We regret to report the death of Toše Proeski, the Macedonian pop singer, and entrant to the 2004 Eurovision Song Contest. His was the one with the chairs and the bandages coming out of his coat.
We also regret to report the death of Alan Coren; a fuller obituary will follow next week.
Thursday was a good day to bury bad telephone lines. While the BBC was announcing how it would implement the cuts imposed upon it by the government, and yielded its independence by accepting commercials for its website, ITV quietly let out its long-awaited report into 0898gate. The review, by accountants Deloitte, made three general points.
1. Programme producers didn't always recognise that using premium-rate call-in numbers was ceding editorial control to the Grate British Public. If the public votes for A but the producers put out B, that completely compromises the integrity of the phone-in.
2. There was no consistent control or way of working between programmes. What happened on one show was ignored by another.
3. Supporting technology failed too often. Votes were delayed or not counted, and calls were charged but ignored.
Three shows had particular failings. Soapstar Superstar planned to begin this year's series by eliminating the bottom-placed contestant, and putting the next two up for the public vote. The least popular singer amongst eleven did go, but the overnight vote was between the singers in seventh and eighth, not ninth and tenth. The show also asked the viewers to choose between two songs for each performer – this vote was ignored around 20% of the time.
Antan Dec's Saturday Night Takeaway suffered from a confusing mixed-entry system for Grab the Ads and Win the Ads. There was some geographical bias in the selection of callers for Grab the Ads. Selection of players for the Jiggy Bank was claimed to be random, but actually had to fit in around a pre-booked national tour.
A similar fix took place on the 2005 series of Gameshow Marathon, where the prize pile was awarded to someone selected by a researcher, not by random draw. Commentators found it difficult not to pin some blame on the executive producer of both shows, Antan Dec. It's since been claimed that the executive producer credit didn't actually mean Antan Dec took an active role in the detailed planning of the programmes, and was included for contractual reasons. Television training body Skillset says that an executive producer is "responsible for the overall quality control of productions, and for ensuring that final products conform to commissioners' specifications... They should have current knowledge of the relevant legislation, regulations, and associated procedures, including copyright, data protection, public liability and how to comply with regulatory requirements."
Deloitte also criticised the flakiness of red button technology, and pointed out that almost one in seven votes for the 2005 X Factor final had been discarded. They also found that overnight call-and-lose programmes on the timeshift channel ITV2+1 regularly had the number on display, even though the game had long closed.
ITV chairman Michael Greed said that his channel had zero tolerance for programmes that misled the viewers. Why, in that case, hasn't he pulled the plug on The X Factor, which let viewers think that hired mansions were the real homes of the celebrity judges. Anyway, Mr. Greed has announced a large programme of refunds, as follows:
- Soapstar Superstar (5 Jan 2007 only) – All votes – refund 60p
- Ant and Dec's Saturday Night Take Away
- Series 2 (early 2003) – refund 99p
- Series 3 (spring 2004) – refund £1.14
- Series 4 (autumn 2004) – refund £1.00
- Series 5 (spring 2005) – refund 98p
- Christmas Special 2005 – refund £1.10
- Series 6 (autumn 2006) – calls £1.10, text £1.32
- Ant and Dec's Gameshow Marathon (autumn 2005) – refund calls £1.10 text £1.32
- I'm a Celebrity (2006, whole series) – text refund 60p
- X Factor (17 December 2005, 25 November, 2 and 16 December 2006 only) – text refund 45p
- ITV Play on ITV2+1 (All competitions between 12 December 2006 and 16 August 2007) – 85p per call or text.
Details of how to claim a refund are available at http://www.itv.com/competitions/
Channel Four News reported that the police would not get involved unless they received a complaint from OFCOM, ICSTIS, or another official regulator; it is believed that no complaints from the public had been received before this week. Mr. Greed said that the aim was not to directly fleece the viewing public, but to make a better show, and charge higher rates to advertisers, and those advertising budgets would ultimately be paid by (er) all customers, not just those who enjoyed Antan Dec. John Whittingdale (the Conservative's culture spokey) said that it was fraud, and he wouldn't be surprised if the police did get involved. Peter Hain stuck his oar in; this week's secretary of state for employment said on Question Time that the deception was "almost daylight robbery" and asked for the responsible people to be "nailed". Adam Price of Plaid Cymru called for the police to launch their own investigation, and Charles Kennedy worried that OFCOM was too close to the companies it was supposed to be regulating. By Friday, the Serious Fraud Office had announced was talking with OFCOM about ... GMTV. That's so last month!
Other news that was buried on Thursday included ITV's announcement that it would start working with RDF Media again – the two had parted ways over some piddling little matter in the summer. And Dermot Murnaghan renounced his title as TV's Mr. Six, and resigned from the BBC.
We mentioned last week that John Humphrys had been quoted as saying that men do better in the specialist subject round on Mastermind than their general knowledge, while women's scores are more equally distributed. Is this a fact, or a misperception?
With our faithful records of every score since the television revival began in 2003, we can put this to the test. For each contender, we have the number of correct answers in the specialist subject round (S), and the number of correct answers in the general knowledge round (G). From this, we can work out the total of the two scores, and their difference. For convenience, we define the difference as S-G.
Given the large sample size, we can invoke the Law of Large Numbers and expect that the difference will be Normally distributed around a mean. If Mr. Humphrys' proposition is correct, the mean for female contenders will be close to 0; the mean for male contenders will be significantly higher. It may be necessary to normalise the difference, as a difference of +1 in 11 correct answers feels more significant than a difference of +2 in 32 correct answers. To this effect, we will also compute a metric of divergence (difference / total). Again, we would expect to find female contenders clustered around a divergence of 0, and male contenders clustered slightly but significantly higher.
We discard data relating to passed questions: this is only of interest in the context of an individual programme. We do include the same contestant if they have appeared more than once, either in the same or different series. In this analysis, we've only included data relating to the main adult series: celebrity editions, junior editions, and the Dr. Who special have all been ignored.
Over the shows aired between 2003 and last week, we have seen 118 female and 376 male contenders. Differences range from +12 to -10, and are illustrated by a graph here.
We can see that the most common differences for women are 0, +2, +3, and +6; for men, +4, +1, and +2. The averages come out as women +2.58, men +2.42; the standard deviation for women is 4.18, for men 4.14. That is clearly an insignificant difference – the lookup tables show something like a 98% confidence that it's arisen by chance, and we could only call it a significant result if that confidence was down at 5%.
We can draw up a similar table (though not a pretty graph) for the Divergence metric – this ranges from -7/9, for someone whose specialist subject round scored precisely one point, to +8/10, for the contestants whose general knowledge round fell into pass hell. The average for women is +0.126; for men, it's +0.116; standard deviations of 0.225 confirm this is another insignificant difference.
We don't even need to bring out the heavy theories of chi-squared tests. It's clear that there is nothing between the sexes, and that Mr. Humphrys' hypothesis is false. An average contestant, regardless of sex, can expect to score about 2.5 more correct answers on their specialist subject than on their general knowledge round.
Repechage 1: Lancaster v Birmingham
As a point of honour, we must remind readers that Weaver obtained his degree from Birmingham University. Though he tries to be impartial, this doesn't always happen.
It's a very long time since we saw these sides. Lancaster lost to Nottingham in the opening game of the season, back on 9 July, when the day's maximum temperature was 17°C, and the heat was from the contest – Lancaster's 185 is the highest losing score of the series. Birmingham fell to St Andrews three weeks later, and we thought their score of 145 wouldn't get them through. It was, we're surprised at that, and more surprised to find the highest temperature recorded on 15 October was 16°C.
Birmingham gets the first starter, on a region of Paris. It's answered by their French student. Birmingham gets the run of the green on the second question, Thumper accepts "Daffodils" for "I wandered as lonely as a cloud". Neither side was particularly remarkable on the bonuses in their heat, coming in around 1/2 right, but Lancaster was faster on the buzzers, surely. After the picture starter – supersonic flight – is dropped, Birmingham's lead is 70-0. We'll take Word of the Week.
- Q: Of uncertain etymology, but thought to date to the 1920s, what word of four letters indicates a simple musical phrase repeated over and over...
- Birmingham, Mark Goodwin: Riff.
It's beginning to look awfully one-sided now. Every member of the Birmingham side has answered at least one starter correctly, Lancaster has a grand non-total of one missignal, and the lead is up to 130 points. Though the side from the northern end of the M6 gets a starter right, Birmingham has already equalled its score from the heat before we reach the audio round, on Russian musicians. The lead: 145-25.
We did not know that the Kennel Club published a book on thoroughbred horses. Birmingham manage to arrange things so that they each have three starters correct; at this time, Lancaster has combined for precisely two. The visual round is on paintings by Canaletto, and Birmingham's lead is 220-25.
If there's one thing we know from bitter history, it's that Birmingham can conspire to lose any match from a seemingly impregnable position. Lancaster will have to start buzzing in on starters (which they do), getting them correct (not always), and getting the bonuses correct (generally not.) Birmingham, meanwhile, is continuing to get questions correct, leaving Lancaster requiring 225 points in four minutes.
Almost inevitably, the collapse doesn't happen. Not even the Buffalo Bills could lose from that sort of position. Birmingham wins the game – and that's only happened twice before in the entire Paxman Era – but does so with a stunningly large margin. It's not just the highest score of the year, but the highest score since Trinity Hall Cambridge crushed Magdalen Oxford at the start of October 2005. At the gong, it's 315-50. Three hundred and fifteen!
"You got off to a bad start," commiserates Thumper very politely, not mentioning the poor middle or the not much better end. Jake Goodman got two starters for Lancaster, the side made 4/12 bonuses with two missignals. Mark Goodwin was hot for Birmingham, getting six starters, but every team member had at least four. 25/55 is not the most impressive bonus conversion rate, though there were no missignals.
Next match: Magdalen Oxford v Liverpool
This Week And Next
At some point, we may actually get to publish the review of Don't Call Me Stupid that has twice been squeezed out by events this month, and will make way next week for an Alan Coren tribute.
Simon Cowell said that he would cease judging television talent shows after Pop Idle US in 2010.
Ratings for the first week of the month are in. The first skirmish in the Saturday night war went to ITV, as X Factor (8.1m) beat Come Dancing (7.7m); the Sunday results shows also went to ITV, 7.9m to 7m. In It to Win It had 6.35m, Antan Dec suffered from a move to teatime, making just 4.85m viewers. University Challenge had 3m viewers on BBC2, just 50,000 ahead of QI. Link, Eggheads, and Deal all had slightly more than 2m, as did The Restaurant. Mastermind's 1.95m is another best for this year.
Xtra Factor (985,000) remained ahead of QI (680,000) for the lead in the cable channels. Hell's Kitchen USA came in third on 510,000. Slips for Come Dine With Me and Deal or No Deal on More4, and Dancing With the Stars on UK Gold. Challenge has a new most-watched show, Ninja Warrior debuted with 120,000 viewers.
A quiet week for new shows: Raven The Secret Temple comes to BBC1 (4pm weekdays), and the unexpected and not entirely welcome return of All Star Family Fortunes (ITV, 7.45 Saturday). In the meantime, 2004 Mastermind champion Shaun Wallace explores and explains the Dewey Decimal System (Radio 4, 11.30am Tuesday).
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.