Sam Trenholme’s blog
The 80/20 sex myth
Index
The myth explained
An imaginary
world where 20% of men have sex with 80% of women. In this graph, the
further away the blue line representing men is away from the red line
representing women, the more polygyny (men having multiple female
sex partners, where the women are excusive with the man) there is. The
dark grey line highlights the percentage of male and female partners the
“top” 20% have.
Back to top
There’s a famous scene in the Netflix series Adolescence
where one of the characters tells the other character there is a myth
online where 20% of the men sleep with 80% of the women. This hateful
and misogynistic myth is one which gained ground in the 2010s among
far right online communities, including the “incel” community.
Whether this myth originates from a comment by one Baumeister
or from a 2008 OkCupid blog entry,
it has become viral on social media, even
though there is no scientific evidence to support this myth.1
That in mind, I have done some research myself to see whether this
myth is true.
Let’s us imagine a world where 20% of the men have sex with 80% of
the women. In this imaginary world, which I have modeled, we can see
that women have a fairly linear distribution of sex partners; e.g.
26.75% of women have 40.7% of lifetime male sex partners and 99.80% of
women have 100.0% of lifetime male sex partners.
Contrast this with the men in this incel fantasy, where 20.95% of men
have 81.7% of lifetime female sex partners, and there is a nearly 50%
male virginity rate.
The actual data
The actual NSFG
2022 data. This is a “banana curve”; since the curves representing
women and men both bow downwards, this shows a relative minority of people
are promiscuous. Since the lines are close to each other, there is
not much polygyny, but instead promiscuous men by and large have sex with
promiscuous women. The black line highlights the percentage of male and
female partners the “top” 20% have.
Back to top
For me to find the actual figures of reported sex partners, I looked
up the 2022 NSFG data, where they asked about 10,000 men and women how
many lifetime sex partners they have had.
I got the number of female lifetime sex partners from page 448 of the NSFG 2022 female codebook
and male lifetime sex
partners page 254 of the NSFG 2022 male codebook.2
I then took that data and made a graph of the data. Note in
particular that:
- 32.18% of women have 79.7% of lifetime male sex partners
- 27.69% of men have 81.1% of lifetime female sex partners
- There is a 22.3% male virginity rate
Notably, both the men and women have what look to be a polynomial
distribution of partners, instead of the reasonably linear distribution
of partners among women we see when we model the imaginary 80/20 world
incels have made up.
Also, the male and female partner count distributions are very
similar to each other. This indicates that while about 20% of men have
70% of the lifetime female sex partners, about 20% of women also
have about 70% of the lifetime male sex partners. So, while a small
number of men do, in fact, have the majority of female sex partners,
the women they are sleeping with, in turn, have a large number of male
sex partners.
Methodology
Back to top
The files used to process this data and generate these graphs are
available here:
https://github.com/samboy/blog/tree/main/80-20-myth
I will explain some of the methodology below.
The simulated incel fantasy
Back to top
While it’s possible to choose, when a sexual encounter happens, to have
the woman randomly choose one of the “alpha” guys 80% of the time and
one of the “beta” guys 20% of the time, the code actually uses a curve.
Let’s look at the math to make such a curve:
In order to make the integration as simple as possible, let’s have
f(x) be a polynomial:
Let’s integrate that equation:
Then simplify:
For n = 6,
is
about 0.7902848 and
is about 0.8388608; those values
are close enough to 80/20 and keep the math relatively simple.3
That in mind, we do the following when choosing which man will have a
sexual encounter with a new partner:
- We choose a random point (x,y) where both x and y
are random numbers between 0 and 1
- If the point ends up below the curve y=x6, we keep the values of x and
y
- If the point is not below this curve, we choose another random point
until we find one below the curve (maximum tries: 100 times)
- We then multiply x by 10,000 (the number of men in our
simulation), round to the nearest integer, and that man will now have
a sexual encounter
- Now that we have determined which man has a sexual encounter, we
randomly choose which one of the 10,000 women in our simulation he
will have the encounter with
The simulation has 61,000 sexual encounters with a new parter, which is
approximately 10,000 multiplied by the mean number of partners women in
the 2022 NSFG data have.
As an aside, this is how we could calculate the precise value to give
n for a given distribution of imaginary “alpha” men in
the xn function (Note: n has a value of
6 in the actual simulation we ran):
When k has a value of 80% (0.8), this simulates 80/20 (80% of the
women sleep with 20% of the men), and the above equation gives a
precise solution for n in this equation:
By dividing those two logarithms and subtracting one,
we create an exponential degree n such that the curve
y = xn has the property that the area below
the curve between 0 to 0.8
is one fourth of the area below the curve from 0.8 to 1. k,
having a value of 80% (0.8), represents an imaginary world where 80%
of the women sleep with 20% of the men.
If we want to create an curve representing a fantasy world where 90%
of the women sleep with 10% of the men, k would have a value
of 0.9 (90%) with that equation and that curve function. In the
resulting curve, the area below the curve from 0 to 0.9 is one nineth
of the area below the curve from 0.9 to 1. One nineth because 90 in
90/10 is nine times the value of 10.
To calculate this in Lua, here k is a number between 0 and 1:
function chad(k)
return math.log(1-k) /
math.log(k) - 1
end
For 80/20, the value of
n is about 6.212567; for 90/10, the
value is about 20.8543; and for 95/5, the value is about 57.404.
Addressing incel comments
Back to top
I will address some of the comments incels who refuse to believe this have
made:
- 2011 NSFG figures give similar numbers: While the most promiscuous
are less promiscuous, and the number of virgins was lower back then,
the general trend of the most promiscuous women being roughly as
promiscuous as the most promiscuous men was true. Other surveys besides
NSFG give similar figures where notably the most promiscuous women are
about as promiscuous as the most promiscuous men.
- I normalize for over/under reporting by looking at the percentage of
total partners by gender.
- For women and men to lie such that the number of very promiscuous
women is about the same as the number of very promiscuous men would
require a conspiracy.
- Incels frequently change their argument, claiming that the fundamental
claim isn’t that 80% of women sleep with 20% of men, but that 80% of
women are attracted to only 20% of men. This is moving goalposts,
but the claim that women are attracted to only 20% of men, or more
precisely, that women are attracted to only 20% of men while ovulating,
has
been shown to be false.
- Incels claim that this
article shows that 80% of the women sleep with 20% of the men. The
article is exclusively about dating apps, which do not represent real
world dating (dating apps are roughly
70% men and 30% women) and the
article points out that men can get more swipes with better pictures.
I blog elsewhere
about how women can’t really build attraction from just a picture the
way men often times do.
- As a follow-up to the last point, incels claim, based on a 2019 survey,
that everyone meets each other these days on dating apps. These results
have not been replicated; e.g. this
2024 survey shows that only 27% of people getting married met each other on
a dating site or app.
- The claim that there are more single men than women in their 20s
has been addressed in an article
someone else wrote. To summarize: Slightly older men date slightly
younger women.
- Some incels claim it’s different for people 18-29 years old. I have made
a graph for 18-29 year olds, and the numbers are essentially the same.
Further discussion below.
- One incel assertion is that twice as many women reproduce compared to
men (either historically and/or presently). This is not true;
see this
discussion and this peer reviewed paper.
To summarize: “1.4 [women reproduced for every man] in the Yoruba from
West Africa, 1.3 in Europeans, and 1.1 in East Asian samples.”
- Not an incel talking point per se, but it looks like only about 5% of men are actually incels.
What about the 18-29 year olds?
NSFG 2022 data including data for just youth aged 18-29. The dark
grey line highlights the percentage of male and female partners the
“top” 20% have.
Back to top
I have run the stats on just 18-29 year olds and the numbers are not that
different than they are for all ages.
If anything, the trend of there being more virgins of both genders as
well as the most promiscuous people being even more promiscuous continues:
While the “top” 20% of men have more partners compared to all ages, the
“top” 20% of women also have more partners.
In more detail:
- 20.66% of men (18-29) have 75.7% of lifetime female sex partners
- 19.96% of women (18-29) have 68.0% of lifetime male sex partners
- Virginity rates for 18-29 year olds: 30% men, 25% women
- 19.45% of men (all ages) have 71.5% of lifetime female sex partners
- 19.49% of women (all ages) have 64.8% of lifetime male sex partners
- Virginity rates for all ages: 22% men, 18% women
Note that:
- The most promiscuous people have even more sex partners, for both
men and women
- In particular, the difference between all ages and 18-29 is that the
“top” 20% have 3-4% more of the lifetime partner count, again for both men
and women
- Virginity rates are about 8% higher with younger people, once
again for both men and women
Some final thoughts
Back to top
There is a slight skew, where about 5% of men (if even that) are
polygynic, and about 5% more men than women are virgins, but it’s nowhere
near 80/20.
Footnotes
1: The notion that women want to cheat on their male partners when they ovulate has been shown to be false.
2: Note that these are opposite sex intercourse partners.
3: 6.213 is closer to an ideal value for n
to get an 80/20 curve, but 6 is good enough for this simulation.