Re: Evolutionarily Stable Strategies (ESSs)

From: Stevan Harnad (
Date: Mon Oct 06 1997 - 19:18:06 BST

> From: "Katherine Lyne" <>
> This is the way that I understand the principle of ESS, as explained in
> "The Selfish Gene"
> The principle of the Evolutionary stable strategy is to do with the
> question of aggression among species.

Not just aggression, actually; any competitive selfish-gene trait. It
need not even be behavioural, although the behavioural traits are the most
interesting ones.

I am inventing now: Suppose that snails could eat more mushrooms if
they (the snails) were bigger, so the snails that have the selfish
genes that happen to programme them to grow bigger manage to eat more
and survive and reproduce more.

Is growing bigger an ESS for these snails? I am continuing to invent
now: Suppose that as they get bigger, birds of prey that normally only
eat rodents (because normal snails are too small to see) now take a
fancy to these big, slow snails: Then what the snails gained with
increased food they lose with increased predation. Growing bigger
seemed promising, but it turned out not to be evolutionarily stable.

Most of the examples of ESSs in The Selfish Gene are about
within-species competition between the survival machines programmed by
their selfish genes, and the traits are mostly behavioural ones. But
the concept of ESS can be applied to any trait; it's not "strategy" in
the sense of being a conscious or deliberate strategy.

Nor does it need to be behavioural. "Stability" just means that the
"strategy" works: It has not been "invaded" by other competing genes
using other strategies. Unstable strategies don't last; they are
evolutionary "false starts" that gain ground for a while and then lose
ground, perhaps even to the point of failing to be passed on at all.

With behavioural traits it's especially interesting to go through the
reasoning about why a trait that looked like a good idea, and was
successful for a while, proved to be unstable, and was invaded by

> All members of any species are in
> fact in competition for the same resources. Whilst it might seem that
> it would be advantageous for one individual to be aggressive to the
> other members to minimise competition, this would not actually be such
> a good idea. By killing one member, that aggressor would have not only
> helped himself, by killing a competitor, he would also have helped the
> other remaining members. This would actually be bad for his own
> situation.

Right; and he would have risked the injury, while others would have
gotten the benefits without having to try.

> The ESS is the principle by which different characteristics are
> balanced among a species. In a hypothetical situation green spotted
> people are aggressive and prone to fighting till the bitter end. Red
> blobbed people are only mildly aggressive and will retreat if attacked.
> Green spotted people will always win if fighting reds. If however
> greens fight each other one will win and one will die. Greens will also
> sustain injuries if they engage in combat. In this situation, Greens
> will always win in a species of reds and greens but will sustain more
> injuries than the reds and so the net gain for the aggressive greens
> will be less than for the peaceloving reds. (Dawkins explains this
> numerically)
> So, at one extreme the population will consist mostly of reds. However,
> in this situation the few greens around would profit by winning all
> combat with the peaceful reds and greens would increase in number. They
> would increase until they started killing each other again and the reds
> began to increase in number Thus a equilibrium of reds and greens is
> established and is "Stable". The population consists of both reds and
> greens. The principle is a strategy because it is a "pre-programmed
> behavioural policy" ie it works within an individual without their
> knowledge.

Yes, it's important to keep reminding yourselves that none of this is
conscious, any more than the zebra's "strategy" of looking black and
white -- so it can blend into the background better -- is conscious:
It's just the blind outcome of the effects of selfish genes and the
outside world on the survival and reproduction of their survival

> This strategy also has room for "Cheats" which Dawkins discusses later
> in the book

A strategy that can be invaded by cheats is not an evolutionarily stable
one (ESS).

This archive was generated by hypermail 2b30 : Tue Feb 13 2001 - 16:23:07 GMT