suboptimality of monogamy

Biological systems are replete with barbell strategies. Take the following mating approach, which we call the 90 percent accountant, 10 percent rock star. Females in the animal kingdom, in some monogamous species (which include humans), tend to marry the equivalent of the accountant, or, even more colorless, the economist, someone stable who can provide, and once in a while they cheat with the aggressive alpha, the rock star, as part of a dual strategy. They limit their downside while using extrapair copulation to get the genetic upside, or some great fun, or both. Even the timing of the cheating seems nonrandom, as it corresponds to periods of high likelihood of pregnancy. We see evidence of such a strategy with the so-called monogamous birds: they enjoy cheating, with more than a tenth of the broods coming from males other than the putative father. The phenomenon is real, but the theories around it vary. Evolutionary theorists claim that females want both economic-social stability and good genes for their children. Both cannot be always obtained from someone in the middle with all these virtues (though good gene providers, those alpha males aren't likely to be stable, and vice versa). Why not have the pie and eat it too? Stable life and good genes.

Antifragile - Nassim Nicholas Taleb (Pages 162-163)

True monogamy is evolutionarily suboptimal. This will be painfully apparent in near future when everyone's genome will be sequenced at birth.

Note that the most optimal strategy highlighted by Taleb is accessible only in an incomplete-information environment. If everyone could see the genetic landscape, the strategy would fail because

• alpha fathers would not be open to the idea of their children being reared by others and
• non-alpha fathers would not want to spend their resources to rear others' children.