Analyzing Axelrod's Prisoner’s Dilemma Tournaments

Nicole Christian

Evolutionary Psychology

Dr. Corby

July 25th, 2015

Final 

Robert Axelrod organized two computer tournaments based on the Prisoner’s dilemma.  Why did he do this?  What were the results of the tournaments?  What were the factors that made the computer programs successful or not?   How is this relevant to evolutionary psychology?

Prisoners dilemma, is a theoretical game played between two individuals who can choose to either cooperate with one another or defect. Either choices give completely different payoffs. Both individuals cooperating, gives a greater pay off allowing for both individuals to benefit whereas one individual defecting while the other cooperates allows for one individual (defecting individual) to receive

a greater payoff than their opponent (cooperating individual). This theoretical game helps to understand why rational individuals might choose to defect even if it means that cooperation would be in the best interest of both individuals. 
 Robert Axelrod organized a computer tournament based on the theory of the prisoners dilemma, the reason for doing so was to further understand the evolution of cooperation and whether it has the ability to prosper in an environment or whether it would be overthrown by individuals who choose to exploit others. In organizing this computer tournament, Axelrod had come to understand that among all the program strategies played, Tit for Tat was most successful in beating all other program strategies. This program strategy would cooperate on the first round and for all other rounds thereafter, would copy cat whatever their opponent did. If the opponent defected the second round then Tit for Tat would defect, if their opponent cooperated in the following round then Tit for Tat would also cooperate. Basically, this strategy would cooperate until given a reason to otherwise defect. 
The overall results of this tournament was that Tit for Tat won.

In the completion of this computer tournament, Tit for Tat achieved the highest score against all other strategies and was proven to be the better strategy in the prisoners dilemma. According to Axelrod, there were four properties that will make a strategy successful. The first being the ability to cooperate as long as the opponent was willing to cooperate and this is turn would avoid unnecessary conflicts. The second being provocation by defecting once the other opponent has defected. Thirdly, forgiveness, whereas the player was able to revert back to cooperation after being provoking to their opponent. Lastly, allowing for the players strategies to be clearly understood to allow for the other player to recognize their plans and course of action as to adapt to this pattern. Other factors making Tit for Tat so successful was it was robust, thus having strength to beat all strategies that it came up against. Tit for Tat also had stability whereas it could not be invaded by any other strategies. Also Tit for Tat was viable in that it worked successfully amongst all other strategies. All other program strategies that did not possess these properties were unsuccessful.

This is relevant to evolutionary psychology because it shows some strategies are evolutionarily stable, whereas others are not. Evolutionary Stable Strategies (ESS) meaning if a strategy is adopted by a population in a particular environment, this strategy

would not and cannot be invaded or abolished by any other strategies. Tit for Tat, therefore being an ESS. From a real life perspective, when two animals meet, in the case of this example, two elephants, they interact with one another. In the event of the prisoners dilemma, if the two elephants
 were to interact on the premise that they either cooperated with one another, neither of them cooperate with one another or one elephant cooperates and the other does not. After making this decision these elephants will then move on with their lives. In the event that these elephants were able to cross paths again and identify who cooperated with them and who did not, they will choose whether they would like to help or avoid helping one another based on the decision previously made. If one had not cooperated in the past and the other did, the cooperating elephant may choose to only stick with those who have cooperated in the past and avoid interacting with those who defected them. Eventually, those who are defecting other elephants will continue to do so as long as there are fully cooperative elephants around. Once the cooperative elephants disappear, because they have chosen to stay clear of those who have exploited them, there then will not be any easy prey for those defecting elephants to exploit, therefore leading to their extinction. This will then cause for those who have cooperated with one another to increase in population. This is the basic dynamic of what Axelrod tried to understand when creating this computer tournament and trying to understand how cooperating or defecting would benefit the species in the long run.

Nicole Christian
Evolutionary Psychology
Dr. Corby
July 25th, 2015

Final
Both genes and memes are entities that replicate themselves.  Compare and contrast evolution by natural selection (genes) and the evolution of memes.  What makes a gene endure?  What makes a meme endure?

When comparing and contrasting evolution between natural selection (genes) and memes it is important to understand that while both are replicators, they are both completely different.

Genes transfer biological traits which are transferred from parent to offspring and typically confer some benefit to its holder. It may not be optimal benefit, however does confer some benefit. Memes on the other hand transfer psychologically appealing information. Genes typically need a generation or so to be transmitted whereas memes spread quicker than would genes and can be transmitted in a matter of seconds. While there are some substantial differences the two are also very similar. Both memes and genes are very good at getting themselves copied. These genes that are good at copying themselves will become more prevalent in the environment and comparably so will those memes that are most psychologically appareling. A meme, according to Dawkins, can be considered a cultural counterpart to a gene. For example, although we don’t know the specific origins of where God came from, the idea of a divine being can often be seen in writings, art and music and even word of mouth. These tools have essentially been responsible for carrying on this idea of a divine being, thus making this a

meme. 
Genes endure by copying all of itself into the next generation through sexual reproduction and Memes endure as expressed behaviors every time they are replicated. What makes a meme and gene successful, according to Dawkins, and therefore endure, is both genes and memes must posses the following characteristics; longevity, fecundity and copying fidelity. Longevity in a sense that they must be capable of living long enough to be passed on from the brain to other individuals and for a gene to be passed on through reproduction. Both meme's and genes must have fecundity or better known as fertility to hold the ability for a meme to transmit the psychologically appealing information and for the gene to pass on the trait. Lastly, Copying fidelity concerns the different ways in which memes are transmitted, for example, music, billboards, etc and for the gene, it means having the ability to successfully copy itself. A meme-idea is the way in which the idea is transmitted from brain to brain, the differences in how this idea actually gets interpreted is not apart of the meme. An example of this would be, if someone asked if anyone knows what the Oedipus complex is, everyone will know and have an idea of the Oedipus complex in their minds and have a general idea of what the concept is, whether the information the individuals have interpreted is correct or not, this does not matter, the concept does.

Nicole Christian
Evolutionary Psychology
Dr. Corby
July 25th, 2015

Final
Robert Trivers introduced the concept of reciprocal altruism both between and within species.  Discuss the implications of reciprocal altruism for our human psychology.

Robert Trivers explores reciprocal altruism amongst different species as well as within species. For example, cleaning symbiosis where the cleaner fish cleans the host fish and the host fish, after being cleaned, has the opportunity to eat the cleaner fish however chooses not to in hopes to receive further cleaning in the future. Similarly, where the Grouper fish lives in a tank for six years, by itself, and when other small fish are dropped in its tank, Grouper ate the other fish. On the contrary, when a cleaner fish was dropped into the tank, Grouper, extended its gills to allow the cleaner fish to clean itself. All the while, the Grouper had never seen a cleaner fish. This behavior shows that the cleaner is worth more to the Grouper alive than dead.  Even in the case of birds. If there is a predator nearby, the first bird to see the predator will make a warning call even though this means placing itself at risk. It is believed that the bird who makes the call is in the best position because the call will potentially discourage the success of the predator, preventing predators from specializing on a certain species. On the other hand, if the bird who spots the predator flies away without making a warning call, the other birds will be fed on. It is believed that the warning call is motivated by selfish behavior. 
Although altruism within species and between species is rather interesting, most interesting is human reciprocal altruism. Trivers explains that there has been no direct evidence linking the degree of reciprocal altruism and its evolution with genetics. However, from much studying of altruistic behavior, there has been a link between the underlying emotional dispositions that have an effect on altruistic behavior.He explores that reciprocal altruism amongst the human species would be best understood by comprehending 10 theories of psychological systems underlying human reciprocal altruism. 
The first theory being "A complex regulating system", basically meaning natural selection will favor people's ability to regulate their own altruism and cheating tendencies and what their responses will be towards others with either altruistic or cheating behaviors. The second theory "Friendship and the emotions of liking and disliking." is where altruistic behavior is shown among friendships rather than towards someone who is a neutral individual. Those who like an individual are more altruistic towards that individual and will be more altruistic towards other altruists. Although this has not been proven to be the sole cause for altruism, there is also evidence that those who show mutual altruism will most likely continue to be altruistic towards one another. ”Moralistic Aggression" which is a way for an altruist to protect themselves from cheaters. Those who are moralistic aggressors will want the cheater to know what they have done wrong and possibly even have the cheater punished for their wrong doing. A fourth theory is "Gratitude, sympathy and the cost/benefit ratio of an altruistic act" This concept explains that those who have someone who is altruistic towards them will go through what is the cost/benefit of the altruistic behavior and whether the individual should reciprocate altruistic behavior and to what extent the altruistic behavior should be returned.  It has been noted that psychologically, when understanding the cost/benefit of altruistic acts, if the need for help is great then the tendency for the recipient to reciprocate altruism will also be great. Fifth, "Guilt and Reparative Altruism" explaining that if the individual cheats in what was once a reciprocal altruistic relationship, the cheater therefore pays for what he has done and makes a reparative gesture. This creates guilt and allows for the cheater to compensate and make right what he has done wrong. Sixth, "Subtle cheating: the evolution of mimics" When the emotions that favor altruistic behaviors get developed, certain mimicking behaviors arise causing the individual to use others to benefit themselves. In other words the mimic is a deceptive individual allowing others to feel that they will be altruistic when in actuality they will not be. "Detection of the subtle cheater: trust - worthiness, trust and suspicion" meaning the ability to detect subtle cheaters will eventually be selected for as well as the ability to detect when altruistic acts are insincerely done. "Setting up altruistic partnerships.” Eventually a mechanism for seeking and establishing a reciprocal relationship will be selected for. In short this is the simple notion of "Do onto others as you would have them do onto you." Meaning that an individual will perform altruistic acts towards strangers in hopes of gaining a new friend and eventually reciprocity will take place. "Multiparty Interactions" yet another theory is favoring and forming reciprocal conduct allowing for the selection of effectively dealing with cheaters and ultimately rules of exchange. "Developmental plasticity.” being the final theory, explains eventually it will be selected for that the ability for reciprocal altruism will grow, individuals will learn and become adaptive.
Conclusively, Robert Trivers, thrived to make it known that Altruistic behaviors have been selected for in humans. For example, a post-menopausal female, will typically direct all acts of altruism towards her kin. We engage in reciprocally altruistic behavior because it ultimately benefits oneself. There may be a benefit to other individuals, however we mostly gain from this altruistic behavior. Overall, Altruism is random whereas cheating may win if there is no record keeping of who was the cheater and who was altruistic. In this case, cheating will be selected for. On the other hand keeping record of who is and who isn’t altruistic will result in a reciprocally altruistic system to emerge and in this case altruism will win and be selected for.