Prisoner’s dilemma finally probed
Game theory has previously told us that, regrettably, the best strategy in the classic ‘prisoner’s dilemma’ problem is betrayal (‘prisoner’s dilemma’ offers two individuals the same choice: cooperate for an equal punishment, or betray. If you betray a cooperating partner, they are punished more severely and you less, but if you both betray you are both punished more harshly). Mathematically, then, our ‘selfish and aggressive instincts’ maximize our chances of survival or victory – unsurprising given the exquisite rigour of evolution in sorting variation for superiority.
For the dismayed rationalist wondering how humans can ever live sustainably and compassionately there is solace in the fact that our evolution moved into a crucial second phase where group selection ruled. Humans lived in packs, and the most successful packs dominated resources. This positively selected for packs that functioned well. Knowing this, it is easy to begin to dissect human behaviour all over again. We conform religiously to the accepted ‘truths’ of the group, to group ideals and values. We have amongst us a brotherhood; in certain circumstances we will transcend survival instincts or solipsistic practice to protect shared interests. Groups with these individuals were far superior to ones without. It is worth noting, of course, that we will, without hesitation, send a competitor group up the river (rival businesses or sports teams etc).
With this in mind, these surprising results from the University of Hamburg should be a little less surprising. Testing the ‘prisoner’s dilemma’ concept on actual prisoners, they found that there was more cooperation than a purely mathematical strategy would dictate. Fifty six per cent of prisoners opted to cooperate, yielding an end result of 30% of total pairs cooperating.
(http://www.sciencedirect.com/science/article/pii/S0167268113001522#)