Evolution favours positively biased reasoning in sequential interactions with high future gains.

Area of Science:

• Evolutionary game theory
• Cognitive science
• Behavioral economics

Background:

• Empirical evidence indicates human behavior frequently deviates from game-theoretical rationality.
• Humans may exhibit unrealistic expectations about future outcomes, but the evolutionary basis of these biases is not well understood.

Purpose of the Study:

• To investigate the coevolution of reasoning abilities and cognitive biases.
• To model how different reasoning strategies, including biased ones, fare in evolutionary dynamics.

Main Methods:

• Utilized evolutionary game theory to model interactions in a population.
• Implemented level-k reasoning strategies, encompassing both unbiased and biased (positive and negative) approaches.
• Applied the incremental centipede game to represent sequential interactions.

Main Results:

• Selection consistently favors positively biased reasoning strategies.
• Rational behavior can become extinct under these dynamics.
• Bias coevolves with bounded rationality, limiting reasoning depth.
• Positively biased agents can coexist with non-reasoning agents, forming a novel equilibrium.

Conclusions:

• Cognitive biases, such as wishful thinking, may be adaptive features for navigating social dilemmas.
• Deviations from rational judgment can be evolutionarily advantageous.
• The study provides insights into the evolutionary roots of cognitive biases and their role in social interactions.