Summary: Marshall et al. (2013)

So called "irrational biases" have become increasingly popular in the scientific literature, being studied across the broad disciplines of biology, neuroscience and psychology. Irrational biases can describe those behaviours that may not necessarily maximize fitness, but that individuals behave in a manner suggesting that the probability of a positive, or negative, outcome is greater than it actually is. Overconfidence, optimism and pessimism are considered irrational biases. Marshall et al. (2013) suggest that functional explanations underlying irrational biases is necessary, but an understanding of the benefits of irrational biases (e.g. overestimation of the probability of success during a conflict, or better deception of others through increased self-deception) is lacking. They also suggest that close attention must be paid to the difference between outcome and cognitive biases, even though this can impose challenges (as cognitive biases are subjectively experienced by individuals). Marshall et al. (2013) suggest that "an evolutionary theory of cognitive bias might require a closer integration of function and mechanism".