Business Analytics Blog

Business Analytics Blog

Opinions expressed on this blog reflect the writer’s views and not the position of the Capgemini Group

fiX Factor? by Andrew Browell, Shivam Desai and Russell Hodge

The viewing figures for this year’s X factor must be making worrying reading for ITV’s executives.  As it plods unrelentingly on into the third week of its eleventh series, it is being consistently trumped by Strictly Come Dancing in the ratings.  To put it simply, they need every viewer they can get.

Unfortunately for X factor, they lost one of their most loyal viewers a few years ago, he is a member of our blogging team and was a long time fan of the show.  When I asked him why he had given up on the show, the answer he gave was quite disturbing.  He believed that the show was rigged, or to be more precise, the producers were attempting to influence the outcome by carefully shaping the order in which acts performed.

This keen-eyed viewer believed that the later in the show an act performed, the less likely they were to be eliminated.  For this week’s Figure It Out I decided to investigate this; to either validate this theory, or confine it to the ever-expanding vault of disproved conspiracies.

The Numbers

An initial look at the numbers showed that maybe he was onto something.  Out of a total of 93 live shows, only 7 acts were eliminated having performed very last.  If the process of elimination was purely random and talent was taken out of the equation (I’ll let you make your own mind up whether the X factor already does this for us), on average there should be a 15% chance of the act being eliminated, compared to the 8% we saw in reality.  This was a significant result.

Diving slightly deeper into the issues I had a look at how this effect was seen over the course of all 10 of the series

figure 1

In figure 1 we can see this phenomenon in action over the whole ten series.  It is plain to see that the closer to the end of the show an act performs, the less likely it is that they will be eliminated.  The first 40% of acts are 3 times more likely than the last 40% to get eliminated.

figure 2

This is very exaggerated when we look at the first five seasons in figure 2 (which our blogger watched), with a very big bias in acts being eliminated performing early on in the show.  In fact acts performing in the first 40% of the show were four times more likely to be eliminated than those performing in the final 40% of the show.

figure 3

Finally if we take a look at the most recent 5 series in figure 3 of the X factor, we can see again that the first 40% of acts do very badly in comparison to the final 40%, here is 3 times worse. Still very biased.
From this analysis it is very obvious that across all of the years the X factor has been running, performing early on in a show, at any stage of the competition is very bad for your future in the competition.  The question is, are the producers conscious of this, and are they using this with malicious intent.  What is Simon Cowell up to? Two options:
  • Making a better show, in which the quality of the acts improves over the course of the evening.  A perfectly reasonable piece of show planning.
  • Controlling the likely candidates for elimination, by deliberately placing them at the start of the show, week by week. Scheming and controlling of the outcome.
I believe that there is a case to be made that it is the second, more troubling option.  There is a phenomenon in human memory known as the primacy effect, in which the more recent event a human experiences the more easily it can be recalled.  We are all aware of this effect, as are indeed the producers of the X factor.  It could be hypothesised that the effect seen in the figures is produced by this primacy bias.

It seems logical; the acts that performed first would be forgotten by the time it comes to vote.  Voters would call in and vote for the acts that most recently made an impression on them. If we say that the producers are aware of this effect even on a trivial level, then we are left with only two options.
  • The producers are mitigating the unfairness of this effect by randomising the order; i.e. you are simply unlucky if you are made to perform early on.
  • They are taking advantage of the effect. This could mean that preferred acts are getting put in the final 40% of performances and acts to be discarded are being put on in the first 40%.
It is up to us to decide how and if this effect is being used in anger.  One thing we can say with a degree of certainty now; is that if you see your favourite act on this week’s X factor lined up to perform first, second or even third, you’d better cross your fingers...

Source: 4.bp.blogspot

About the author

Russell Hodge
Russell Hodge
2 Comments Leave a comment
Interesting article Russell, you might be interested to read a related analytical study earlier this year into the Eurovision Song Contest on ESC Insight, in which it was concluded with statistical significance that running order did indeed affect outcome.
Thanks a lot Ben. That is really interesting and I will be definitely read that study.

Leave a comment

Your email address will not be published. Required fields are marked *.