[yorwba]

> the expectation for offspring is equal between male and female so the selection pressure generally keeps it close to 50

It's the other way around: the selection pressure drives the expected offspring to become equal. The expected offspring for an individual of a certain gender is (total offspring)/(number of individuals of that gender). The fewer individuals of one gender there are, the higher the expected offspring and the greater the selective pressure to produce more offspring of that gender. So the gender ratio is self-stabilizing.

[FabHK]

This is really a quite surprising and profound fact, worth pondering.

Naively, one would think that if you have (in some antediluvian setting) village A with 50 men and 50 women, and village B with 10 men and 90 women (and children born with that gender ratio, respectively), that village B could "produce" way more offspring and grow faster.

A naive "group selection" view of evolution might even predict something like that to happen.

But, in village B, there is an incentive, so to speak, to cheat and have boys at a higher ratio, because that would increase expected total number of offspring. So, the gender ratio stabilises around 50/50 - a Nash equilibrium, if you will.

[hrolfgar]

Sort of like the prisoners' dilemma; at the end of the day, the selfish actors will choose to locally optimize.

Btw, this argument about the sex ratio stabilizing at 1:1 is known as Fisher's principle. Fisher made another really interesting observation about how extremely exaggerated male ornamentation, like in peacocks, reaches its equilibrium when the aesthetic advantages from a large, attractive tail becomes offset by the practical costs of being slower, needing more food to produce said tail, being highly visible, etc. This one's called "Fisherian Runaway". These purely aesthetic traits will "run away" all the way up until they becoming detrimental to survival.

[mpalczewski]

Thanks for the correction, yes that's what I meant.