Game Theory Metrics of Human Choice Behavior Show That Suboptimal Strategic Performance is Due to High, not Low, Cognitive Ability

Game theory behavioral paradigms pit subjects against one another in competitivedistillations of real world social exchange scenarios. In the “inspection game,” two competing players make binary choices for rewards by matching or mismatching an opponent’s move. Martin et al. (2014) found that chimpanzee pairs outperformed human pairs in the inspection game, as defined by proximity to the Nash Equilibrium, and hypothesized that evolution had endowed chimpanzees with the ability to excel in strategic, competitive social contexts while hindering humans from doing the same. I hypothesize that counterintuitively, suboptimal human performances could be attributable to superior cognition as indicated by game theory metrics known to indicate complex abstract reasoning. In my own version of Martin et al.’s (2014) human-subjects experiment, I replicate the finding that humans fail to equilibrate in the inspection game. It is additionally observed that subjects often fail to randomize their choices and exhibit a win-shift lose-stay tendency consistent with the gambler’s fallacy. Based on ideas advanced by Tversky and Kahneman (1971), I contend that this finding in fact indicates higher, not lower, cognitive ability to strategize for rewards in humans, thus questioning Martin et al.’s (2014) account of cognitive superiority evidenced in chimpanzee performances.