Darwin, Erdős, Mystery

Here it is argued that Darwin was wrong about dating cause women can be as promiscuous as men. Then it brings examples of revisions in studies like in the number of sexual partners that were previously reported quite higher for men than woman and further research does not show this gap.

Just a quick mathematical thought on that. The average number of lifetime sexual partners can not be a good measure for sexual attitude as the sexual attitude may be different in genders but this number has to be equal between them (Graph theory: total out-degree equals total in-degree in a bipartite graph, aka high school prom theorem and in this case works also for average rather than total as both sets are about the same size, read more):

Beyond the abstract world of graphs, men may like casual sex more than women, or not. Whatever the case is, the average number of lifetime sexual partners in straight couples must be exactly equal. There are dozens of studies reporting this number for men versus women as 7 vs. 4, or 12 vs. 7 or … They are all wrong. The error could be due to sampling. Either improper statistical sampling, like leaving out sex hubs such as pick up artists and prostitutes, or local sampling, i.e. in the community asked the numbers were 7 to 4 at a cost of another community having it as 4 to 7. The error could also come from the fact that that genders lie about it: Women report it lower and men higher. And let’s say it’s not even a lie. Men and women estimate this number differently.

Anyhow such ratio not a good measure for a sexual attitude comparison, simply because sex happens between the two genders. So they didn’t really need to put so much effort to perform new research to prove this number is equal as Boromir says:

One does not simply walk into the male side of a heterosexual bigraph to count the outgoing edges to the female side. And if does he doesn’t go to the other side to count the incoming edges, wrong, to come up with two different numbers! And even if he does, he wouldn’t report this error as some shocking result!

And about replacing mean with median: It is only introduced to fix the tautological problem. That can be even more problematic: the median is reported lower for women, which could arguably be even higher, due to the different skewness of the degree distribution between the two genders. That depends on if there are more sexual hubs in men or in women. Which basically asks if we have more womanizers and sex machine in men, than more prostitutes in women? A median is neither a measure of sexual attitude.

So such results doesn’t necessarily prove ultra-Darwinians wrong: In an extreme yet consistent scenario a male may evolutionarily have the desire to mate a thousand females lifetime and a typical female may be hardwired to pick the one and the best. Both fail! It’s simply not possible. Men don’t have more sexual partners than women but want it. Women don’t want as many but they get it.

Remember religion suggested heaven when it firstly encountered death. Trying to please everyone with such limited resources and mathematical barriers, solutions like afterlife 72-virgins are the only way of explaining such statistical difference.

Author: admin

Paranoid Data Scientist Based in Oslo

1 thought on “Darwin, Erdős, Mystery”

  1. Interesting. However, your comparison to a bipartite graph could be more detailed. Firstly, in total 51% of the world population is male, secondly I don’t know the proportion of virgin males or females. Both factors means that there could be a difference (probably not as big as referenced by the article), but still significant. Imagine you have 10 guys and 20 girls sexually active in a population. And they all have heterosexual sex with whatever is available. Then the guys have 20 partners on average and the girls 10. So what we need to find out is what proportion of the population has had sex, or find the proportion of people who have not. That would be very difficult to find out (for similar reasons to the ones you mention) but is essential in order to say that the survey results truly are logically impossible.


    Should we exclude virgins? I assume the average should be calculated over the whole population (including 0-values) and not only over the active part. There’s one tricky part though, as you mentioned about sex ratio which can be even more tricky if we include genders’ lifespan differences:

    The current world wide sex ratio at birth is 107 boys to 100 girls but for every 96 men there are 100 women due to the high death rate in males [wikipedia]. So should the survey include the number of sex partners even if some of them are not alive any more? On the other hand should it dig dead people’s graves (which are more guys!) to ask their number too? Time can make it quite complicated I would say.

Leave a Reply

Your email address will not be published.