Game Theory Analysis Shows How Evolution Favors Cooperation's Collapse

First time accepted submitter Ugmug (1495847) writes Last year, University of Pennsylvania researchers Alexander J. Stewart and Joshua B. Plotkin published a mathematical explanation for why cooperation and generosity have evolved in nature. Using the classical game theory match-up known as the Prisoner's Dilemma, they found that generous strategies were the only ones that could persist and succeed in a multi-player, iterated version of the game over the long term. But now they've come out with a somewhat less rosy view of evolution. With a new analysis of the Prisoner's Dilemma played in a large, evolving population, they found that adding more flexibility to the game can allow selfish strategies to be more successful. The work paints a dimmer but likely more realistic view of how cooperation and selfishness balance one another in nature."

Comments

TIt-for-tat fallacy

[Baldrson]

The notion that "tit for tat" is relevant to evolution in the iterated prisoner's assumes that defection is detected -- an unrealistic assumption. The only reliable evolutionary system in which cooperation is sustainable is one in which the replicators (genetic and memetic) share a common fate aka vertical transmission. This is why the meiotic lottery works in multicellular sexual species and it is how symbiosis between species can evolve in ecologies where migration is restricted -- migration being the origin of the evolution of virulence via horizontal transmission. However, since restricting migration is not practical in much of nature, there is an "optimal virulence" in which a replicator tests the limits of its ability to, in essence, "take the money and run", and exploits to that limit.

Re:TIt-for-tat fallacy

[hey!]

What's unrealistic is believing one strategy is always favored by evolution. Evolution tries everything, so you get all strategies tried.

The substantive argument here should be over this question: what is it that makes H. sapiens such a successful species? The vast majority of discourse on this, unfortunately, is tainted by ideological bias.

I think what makes us successful can't be boiled down to one strategy without being simplistic. The minimum number of strategies that's interesting, in my opinion, is two, because realistic strategies have to interact. Personally the two I'd go with would be cooperation and behavioral flexibility, noting especially that behavioral flexibility sometimes works *against* cooperation. People cooperate to build a successful village, but during a disaster having a few selfish bastards who grab what they can and run is good for the survival of the species. But just because a *little* bit of something is good, doesn't mean a *lot* of it is good. So much selfishness people can't cooperate efficiently is too much selfishness. So little selfishness that nobody saves themselves when they can't save anyone else is too much selflessness.

One more thing to chew on: nature doesn't owe you a justification for your behavior, and it's certainly not going to provide you a logically complete and non-contradictory ideology. It doesn't even give us that for arithmetic.

Re:The Selfish Gene

[Anonymous Coward]

This is explained in Dawkins' book. It's an evolutionary stable strategy.

*Sigh* ... there is no such thing as evolution, creation is unchanging, the earth was created 6018 year ago this has been conclusively proven by analysis of scripture you insensitive clod.

Re:Justifying

[Anonymous Coward]

Rand

Rand's "philosophy" is so full of holes that only a sociopath
or someone with the mind of a child embraces Rand's ideas
wholesale. The sort of people who truly believe in the "fuck you,
I've got mine" position and who are also incapable of understanding that
society has a duty to help those who truly cannot fend for themselves
are the sort who embrace Rand.

I think it works until it doesn't

[EmperorOfCanada]

Basically cooperation is the best strategy as long as there is also a built in punishment system for the selfish. For instance if a disease wipes out too many hosts then it will fail to spread very quickly. If it wipes them all out then it won't spread anymore.

But evolution often will sacrifice to deal with the selfish. So our immune systems are sitting here primed and ready to have a go against all kinds of invaders; our immune systems are fantastically costly. But in a pristine system evolution might eliminate our immune system and then we would be wiped out by the first disease to come along.

The same with having the police. Police are expensive but we keep them around to deal with those who won't cooperate in ways that we find so egregious that we make laws.

But just as we have seen with our bankers there are those diseases that will subvert our punishment systems to not only ignore them but to actively abuse the us. AIDS would be an example of this (and yes I am saying bankers are as bad as AIDS).

So I would think that if you look carefully I think that what you will find is that what evolution will do is to evolve systems that punish the non-cooperative(bad diseases), reward the cooperative (things like digestive bacteria) and then continue living just fine.

Even within animals that group together there are often many systems for punishing animals that don't play by the rules.

But there is one huge problem with evolution from the standpoint of the individual. It might take a 95% die off for evolution to develop a way to fight off a disease, or the disease might end up being just deadly enough to continuously hurt individuals while not killing enough to drive evolution.

But this is where we might have just jumped some kind of hurdle. We demolished smallpox, we have polio on the ropes, malaria might have a bullet heading its way, and other diseases are lined up in the crosshairs. But taking out diseases to the point of extinction takes global cooperation. In Pakistan they recently killed 4 polio workers which will now probably dissuade polio workers from going back into that area and I suspect that if they were there then polio was there as well.

The key is that when gaming any relationship like evolution there are a huge number of rows and columns to work with. But quite simply we have way too many animals that cooperate in pretty magical ways for it not to be a key evolution friendly solution.

Re:Academic Beclowining

[Anonymous Coward]

This is mainly because people forget that the person who 'invented' game theory ended up being committed to an institution due to actually being a psychopath.

Neither John von Neumann nor Oskar Morgenstern was ever institutionalized. They are the two people credited with "inventing" game theory. You may be thinking of the subject of the movie A Beautiful Mind, John Nash. Nash was a prominent early theorist in non-cooperative games. The Nash equilibrium is named after him. Nash was a paranoid schizophrenic, not a psychopath.

Testing game theory through experimentation is much newer. Most of the early work was done purely on a theoretical basis and founded in pure logic. I.e. it didn't try to explain why people did things; it tried to determine how a perfectly logical person should react.