Friday, 25 September 2009

Six sex sux

In the news this week.
Brits have had 'indirect sex' with 2.8 mln people.
This is the latest marketing gimmick from Lloydspharmacy UK; the Sex Degrees of Separation Calculator. It purports to calculate the total number of direct and indirect sexual partners that you have had in your lifetime. This is a good way to raise awareness about sexual health issues, but as a statistical exercise it bites.

Here is their description of the calculator:
In July 2009 Lloydspharmacy commissioned the polling firm YouGov to ask 6,000 people over 16 about their sexual behaviour, specifically how many sexual partners they have had. We then created 17 age ranges and calculated the average number of sexual partners within each of these ranges.

When you enter the age range of each person with whom you have had sexual intercourse, the calculator raids its database to work out how many previous sexual partners people within that age range have had on average. It then repeats this process for their partners, their partners, their partners, their partners and their partners.

This is added together to give a Sex Degrees of Separation total.
Give it a try. It all sounds very scientific, but all they have really done is take some legitimate survey results and concocted a meaningless calculator so they can then put out a fancy press release so that the media can put the words "sex" and "2.8 million people" in their headlines, so that people click on the inevitable link and increase traffic to lloydspharmacy’s website. Why don't they just take the easy route and put out a press release saying "Please visit our website. It has a lot of information about sex".

I think the calculator is a little bit more sophisticated than I present here (although not much), but from what I can work out, what it does is take the number of people you have had sex with and multiply that by the average number of sexual partners for each of those (adjusted for age) and then multiply this through again 6 times (to complete the 6, sorry sex, generations). Therefore, if you have had sex 5 times and the average sexual partners of those in the groups you identify is 7, then your total indirect sexual partners is 5x7x7x7x7x7x7 = 588,245. That is huge, scary and almost certainly wrong.

While the calculator makes the correct assumption that the person you had sex with as a 16 year old has not, on average, had as many sexual partners as the one you bonked as a 30 year old, it does not then extrapolate this through to the next generation – e.g. it assumes that the first person that your 30 year old wife had sex with as a 16 year old had, at the time, the same number of sexual partners as your wife has had in total, hence the constant use of 7 in my example above. This is completely, ridiculously wrong and hugely over-inflates the real value. Also, why 6 (sorry sex) generations? Why not 8, or 10, or 1000? This arbitrary value seems to have been chosen for three reasons: 1. Because 6 sounds a bit like sex (wink wink nudge nudge); 2. Because of the term six degrees of separation, which is meaningless in this context; and 3. Because for most people the calculator will produce a number in the hundreds of thousands or millions, which is mighty impressive, but not completely unrealistic. Take the generations up to 10 (and why not?), and suddenly you have a problem, because most people will have had more indirect sexual partners than there are people in the world. And on this same point, the calculator makes no correction for the fact that once you are a few generations in, most of the people will be having sex with other people already counted in previous generations.

Here is my simple and I believe far more realistic idea for a calculator. It doesn’t require big surveys or complex age calculations. Simply make one assumption, that each person you have had sex with had had sex with the same number of people as you, at the time that you last had sex with them. This seems reasonable to me, as people in similar social circles are likely to have similar sexual proclivities. The first person you had sex with therefore counts as one, because we assume that you were their first too, the second person counts as 1+1,because they would have had sex with 1 other person, who would have been doing it for the first time. The third person counts as 1+2+1 (they’ve had sex with two people, one of whom had sex with one other person and the other with none). And so on. Doing the maths on this gives the formula x = 2n-1 where n is the number of sexual partners you have had an x is the number of direct and indirect sexual partners you have had. Recalculating my example above then, your five sexual partners would end up being 25-1 = 31 not 588,245. That is quite a big difference. To get to a million indirect sexual partners would require 20 sexual partners, but even that isn’t true, because, just like the Lloydspharmacy calculator, I have not taken into account people having sex with those already counted.

Feel free to tear my formula to bits or point out my error in analysing the Lloyds calculator, but in the end does it really matter whether you have had indirect sex with 30 or 600,000 or 2.8 million people? Always take appropriate precautions, in sex and marketing. They are both dirty little exercises, yet strangely alluring.

No comments: