Atomic Nerds

Out on the theoretical frontiers of biology operate the game theorists, whose major occupation in science is constructing mathematical models to demonstrate that X or Y given scenario in sociobiology- that area of biology that covers how certain social behaviors increase or decrease fitness and therefore might be subject to natural selection in some fashion- is mathematically possible or impossible. Because they have to describe massively complex systems with ridiculous numbers of variables and lots of emergent behaviors, the models have to be pretty stripped down and simple- spherical race horses aren’t just accepted, they’re the rule. Hardy-Weinberg equilibrium accurately represents no population of organisms that lives or has ever lived, but instead exists to prove mathematically that genetic polymorphisms are stable over time absent any other pressures on a given allele no matter how rare the allele. Like Hardy-Weinberg, the models of the biological game theorists exist to demonstrate that certain strategies have demonstrable effects and certain choices are objectively more fitness-enhancing depending upon the organism’s starting conditions. The models don’t have to be direct analogues of reality, they just have to prove that the strategic choices HAVE a significant effect. If someone later comes along and demonstrates that a variable not included in the spherical-horse assumptions destroys the model altogether, then so much the worse for the model, but it least provides a starting point for more detailed and useful research.

So, let’s consider a hypothetical* journal article by a pair of such theorists that, unlike 99.5% of what they do, somehow contained something that attracted attention as being newsworthy to the general public. Let’s further assume that this article was about mate choice, and that it constructed a model meant to apply to any species that reproduced sexually, featured female mate choice as the predominant determinator of what matings do and don’t happen, and didn’t have big “event” breeding where mate choice must be settled on within a very short period of time. Let us further assume that the model was then used to demonstrate mathematically that a female’s fitness interests were maximized if she waited a bit before mating to test the male, because if the male was “high-quality” his best interests were to mate with the best female he could find even if that meant waiting for awhile, but if he was “low-quality”, his best interests were to quickly break off before she figured out he wasn’t worth it and find a female that would allow mating sooner. Can you guess, after listening to the theorists explain their research, what a journalist might then turn around and report?

If you guessed that it would be headlined something like “Refusing to have sex on the first date increases the chance of finding a good man”, come on up and take your kewpie doll.

Journals have learned that it’s good practice for a number of reasons to relax their normal “subscriber only” access to online editions when an article in a particular issue gets wide attention, inside the community of that specific kind of research or out of it, and the Journal of Theoretical Biology proves to be no different. Since we have access to the original publication, we can actually clap a pair of eyeballs on it and see what the article ACTUALLY said.

We consider a male and a female in a courtship encounter over continuous time. Both parties pay participation costs per unit time. The game ends when either one or other of the parties quits or the female accepts the male as a mate.

Translation: Wasted time and energy matter, as both are finite for the male and female alike. Time and energy is what’s lost for both if either quits and what’s invested by the male if it ends in a mating and for the female if it ends in a mating that was worth the pregnancy.

We assume that there is a binary variable which determines whether the male is a “good” or “bad” type from the female’s point of view, according to either his condition or his willingness to care for the young after mating.

For ease of modeling and so that the model applies across many species we have assumed that goodness and badness are literally as simple as that.

This variable is not directly observable by the female, but has fitness consequences for her: she gets a positive fitness payoff from mating with a “good” male but a negative fitness payoff from mating with a “bad” male. We assume also that a “good” male has a higher ratio of fitness benefit from mating to fitness cost per unit time of courtship than a “bad” male.

The male gets a fitness payoff either way, but since the bad male has a lower chance of mating with any given female, his opportunity costs are greater if he waits around too long, so he’ll quit sooner.

There’s more in the abstract, but it essentially goes on to say “the math is sound within evolutionary guidelines and this is how this fits with other theories and research into mate choice”. So, to sum, the assumptions made are:

1. The male always wants to mate with the female. No other quality of hers is relevant.
2. The female’s stake in it is always her future fitness interests. (I.e. sex is never a desirable end in itself to the female, pregnancy always is.)
3. The goodness or badness of a male is a single binary trait that can be accurately evaluated solely by length of courtship.

Surely there could be no possible conditions here that aren’t thoroughly broken by the human dating experience! The journalist doesn’t think so, as his lede before he goes on to quote the two researchers (who are careful to keep it within the boundaries of what their model actually does and could do) demonstrates:

A new study shows that refusing to sleep with a partner on the first date could be one of the keys to making a successful match.

Researchers used a mathematical model to show that more reliable men were willing to wait longer before having sex for the first time.

By contrast, less suitable men were not as likely to continue dating.

Males turns into men, good turns into “reliable”, the period of courtship gets contracted to “first date”- the better to start working out which date exactly is the correct one to surrender the “prize” for his investment, presumably. “The model is a starting point for future research” turns into “it has now been scientifically proven that sleeping with someone on the first date is bad”.

Which, if a guy wouldn’t keep dating you if you didn’t put out at the earliest opportunity, then yeah, that’s probably a bit of a bad sign right there. And it’s not a good idea to try and make men more interested by sleeping with them. But if you really need to be told that, then you have far further to go along the learning curve of “making a successful match” than anything short of therapy can help with. Humans are the most advanced social species on the planet; we’ve radically diverse in temperament, ability, goals, culture- and perhaps most importantly, all the cheater/cooperator arms races that may have begun as early as colonies of single-celled organisms and continue all the way up to Survivor. If you try to sort out your love life with game theory, you’re almost certainly going to get gamed by someone whose tactics are sophisticated enough to be unmodelable. (Which only requires being more advanced than a moth.)

Also, journalists reporting on science are approximately as reliable as neolithic hunter-gatherers reporting on a space shuttle launch, so when and if you can, find the full text of the original article.

*Not really.

Posted Ranting, Science: we'll fuck you up, bullshit |