As regular readers will know, we like to analyse the week-by-week voting statistics by converting them into a percentage of the mean vote available for each week. For example, in week 1, there were 16 acts so the mean vote is 6.25%. Andrea polled 19.2%, which equates to an impressive 307% of the mean. In week 9, there were four acts, so the mean vote is 25%. Andrea polled practically the same, 19.1%, but now this equates to only 76% of the mean.
Here are the full calculations. Unfortunately, the show decided not to release the freeze figures for the Saturday eliminations in weeks 4 and 7, so the figures for Jack and Only The Young respectively are skewed and we’re unable to work out what percentage of the vote came before and after the Saturday freeze. It’s probably not dissimilar to the final, when by our calculations, about 63% of the votes were cast after Andrea’s departure. Interestingly, the Ben:Fleur split was 57:43 before Andrea left, and 64:36 after.
What’s most remarkable is how steady Ben’s vote held. Pretty much every single week he was picking up roughly 130% of the mean, give or take a few percentage points. In other words, he was consistently gaining new fans each week, while Andrea never really expanded the base of support he took into the live shows – a testament to how firmly producers kept a lid on his appeal. We can show this in a graph, with Ben’s line looking impressively consistent, while Andrea’s is a nosedive interrupted only by his sympathy bounce in week 8:
Here are the main things that struck us about the statistics, many of them echoing thoughts which have also been expressed in the comments to the last post.
1. The Star got it wrong – but why?
On the morning of the final, the usually reliable Daily Star leaked that Ben got 50% of the vote, Fleur 22% and Andrea 16%, which would have left Lauren on 12%. As it happens, the figures were 36.5% for Ben, 24.3% for Fleur, 20.1% for Lauren and 19.1% for Andrea.
As was perceptively noted by a couple of Sofabet commenters, that means Ben’s semi final vote was 50% more than Fleur’s, so it is possible that the Star simply misunderstood what they’d been told.
An alternative conceivable possibility, floated by Rob at entertainmentodds, is that the Star saw the figures only for phone votes, and not for apps. If that’s the case, we could deduce that app voting was much more favourable than phone voting for Lauren, somewhat more favourable for both Andrea and Fleur, and much less favourable for Ben.
At any rate, the Star’s record is now sullied and we will need to be more cautious about future leaks from this previously reliable source.
2. It was remarkably close in the early weeks
It wasn’t until week 8 that the winner polled more than 20% of the vote. That’s unprecedented. Week 4 in particular was ridiculously close. Ben topped the vote with just 12.2%, while Stereo Kicks were in the singoff with 8.3%.
3. Ben slipped the field in week 8
Although Ben topped the vote from weeks 4-7, it wasn’t by much. It was only in week 8 that he put daylight between himself and the field, establishing a commanding lead which he never gave up. Which makes it all the more mysterious that producers took their feet off the gas with Fleur that week, something we puzzled over at the time.
Could it be that, after the pimping of weeks 5-7 failed to get Fleur above Ben, producers were sanguine about the possibility of her dropping into the bottom two – as she very nearly did – for a James Arthur-esque bounce to the final? Or was it just a miscalculation? Here are the trendlines producers would have been seeing going into week 8:
4. They really didn’t want Lauren
Lauren was polling very solidly in the early weeks, as can be seen from the graphs above, but was gradually dragged down by a series of poor slots in the running order beginning in week 5.
Why did they not want Lauren? One possibility is that they gave up on her in week 4 when she struggled to cope with some very simple dance moves. Another is that producers perceived at this point that there was room for only one of her and Fleur in the final, and acted accordingly.
5. The takedown of Paul Akister was brutally effective
3rd of 16, 3rd of 14, 4th of 12, 5th of 11… 9th of 9. That’s some going. Paul must wonder what might have been had he said “you know what, ‘Bat Out Of Hell’ isn’t really my thing but sure, it’s Hallowe’en, I’ll get into the spirit of it”.
6. They could have ditched Andrea on deadlock against Stevi
Many of us assumed that if Stevi had finished above Andrea in week 7, producers would have taken the opportunity to achieve a headline-grabbing elimination by taking it to deadlock. It turns out Stevi was above Andrea, but they saved Andrea anyway.
It’s a interesting decision, and not just because it passed up the chance of some easy controversy – getting rid of Andrea at this point would also have eased the path to the final for Stereo Kicks, something most of us assumed at the time that producers would be keen on.
If the decision to save Andrea over Stevi indicated they always envisaged him in the final, does this mean they weren’t seeing Stereo Kicks there at this stage, given that Ben and Fleur will have had their berths booked? Or did they perhaps fear that ditching Andrea for Stevi might have risked an unpredictable response from Mel?
7. A mixed picture on the sympathy bounce
After Steph (weeks 1-2), Jay (weeks 5-6), Stevi (weeks 6-7) and Lauren (weeks 8-9) appeared in consecutive singoffs, there has been much comment about the death of the sympathy bounce. The stats show a mixed picture.
Week 1 singoff survivor Steph didn’t get a bounce at all, going from 46% of the mean in week 1 to 36% in week 2. However, week 2 survivors Stereo Kicks went up from 62% that week to 107% the week after. Week 3 survivors Only The Young went up from 63% to 99%. After their second save in week 4, Stereo Kicks got a second, weaker bounce – 91% to 112%.
Week 5 survivor Jay got no bounce in week 6, his vote declining from 73% of the mean to 68%. Stevi, saved in week 6, did get a bounce – from 86% to 103% – but it wasn’t enough to escape the singoff in week 7. That week’s survivor, Andrea, got a similarly-sized bounce in week 8 – 92% to 105%. However, Lauren didn’t bounce at all, dropping from 89% in week 8 to 80% in week 9.
The conclusion is that producers were very effective at suppressing bounces when they wanted to – as was certainly the case with Steph, Jay and Lauren.
8. Who was the original Plan A?
It’s always interesting to see the opening of the envelopes on Xtra Factor revealing who judges and Dermot predicted to win before the judges’ houses stage. Cheryl, Louis and Simon all mentioned Stereo Kicks, while Dermot chose Lola. Our own 1-16 pre-lives prediction slated Stereo Kicks to win from Lola in second, but their abilities in the live shows put paid to that.
On which subject, hearty congratulations to Sofabet commenter Donald for topping the pre-lives 1-16 prediction contest (you can read everyone’s predictions in the comments under our own), and many thanks to Mech for doing the calculations.
Very many thanks also from us at Sofabet to all commenters for your kind words in the comments to the last post about our coverage this season. As always, it’s the comments that make the site what it is. Do join Daniel for Eurovision coverage in the coming months, and do also keep letting us know your thoughts on the X Factor voting statistics below.