Besides its applications to human behavior, game theory can also aid in explaining biological phenomenon, such as the evolution of viruses.

A research paper from American Scientist Online argues that cheating is a dominant strategy in viruses evolving in a laboratory setting.  The results from the experiment are consistent with the evolutionary version of the prisoner’s dilemma, which suggests that “cheaters should take over the population – selfishness turns out to be the evolutionary stable strategy.”

 To test the prisoner’s dilemma, scientists recruited bacteria-infecting viruses as players. The viruses came in two types, an ordinary virus that could replicate on its own and a defective form, which replicated much faster because of its shortened genome but needed to usurp enzymes from the ordinary virus to do so. The scientists divided the viruses into two groups and allowed them to evolve for 250 generations. In one group, viruses infected bacteria individually (single-infection). In the other group, two or three viruses infected a bacterium simultaneously (co-infection). 

The results showed that under co-infection, the viruses grew much faster than under single-infections, pointing to the possibility that “evolution under co-infection had selected for cheater viruses.” 

 The scientists then mixed these “cheater” viruses with an ancestral population (thawed from the freezer) that exhibited a cooperation strategy. As the evolutionary prisoner’s dilemma predicted, the cheaters had a huge fitness advantage in the beginning but began to lose their fitness as their frequency increased. The cheaters also were well on their way to displacing the cooperators after five generations.

Of course, the prisoner’s dilemma presented here differs in many aspects from the one discussed in class. In this case, the two players do not have the same strategies available to them – the ordinary viruses cannot cheat. The game here is much more complex, involving numerous iterations and with payoffs that depend on the relative frequency of the two players. However, the similarity between the two versions – and the interesting observation – is that in both cases, cheating prevails, leading to an overall decline in payoffs.

Finally, the authors make a very interesting point, which is that game theory and Darwin reach very different conclusions about the evolved population. Under game theory, the overall fitness of the population decreases, while Darwin argues that evolution should make the overall population more fit.