James Dow, a professor of evolutionary anthropology at Oakland University in Rochester, Michigan, has published a result from his program, called Evogod, that proports to study the evolution of religious feeling. His model of how religious ideas can spread as reported in NewScientist, is a rather unsophisticated application of game theory. (The source code is available from the link).
The "utility" of expressing a religious/supernatural belief was merely assumed. Any discrete value can be entered, and the higher the utility factor the sooner the trait spreads. What is absent is far more important than what was present.
The first question that was sidestepped is "Why would a group of non-believers attend to fantastic stories?" The answer is approximated from game theory as an attempt to minimize risk by minimized uncertainty. The second would be "What advantage would there be to the person promoting their "visions." This is also in the utility factor assumed by James Dow. From ethnographic and historical sources, we are aware that the role of shaman is very dangerous- witchcraft accusations are frequently lethal. So, this must be offset by an increased reproductive advantage. Dow has seemingly lumped these together as "attractiveness."
There is actual biological evidence for the evolution of religious experience, the responce that humans have to dissociative drugs is different from any other mammal. Most mammals when challenged by a dissociative, for example phencyclidine (PCP), will lose consciousness. Humans instead report euphoria, and hallucination as well as a dissociative state. These are hallmarks of mystic/religious experience.
What could the evolutionary advantage be to dissociation- the feeling of not being in one's body? This needs to be looked at as either a direct advantage, or as a disadvantage offset by some other advantage. Considering this as a direct advantage we recall that PCP was developed as a surgical anesthetic. It had the advantage of not being a strong CNS suppressant, and so respiration and cardiac function were not as depressed as with opiate anesthetics, or ether. Anesthesia was so effective that patients didn't even need to be unconscious during surgery. Dissociative states can be entered in ways other than drugs, and members of a group protected by warriors (temporarily) unaffected by pain would have a great advantage in intergroup conflicts. I think this direct advantage alone would be adequate to promote a dissociative trait.The secondary benefits of "mystic" dissociation would be related to improved group cohesion, etc. which have long been considered in the spread of religious practices.