Essays on bounded rationality

This research is comprised of three essays studying bounded rationality. We propose a model of bounded rationality based on a two outcome strictly competitive extensive form games of perfect information without chance moves.; In the first essay (Chapter I) we estimate bounded rationality and analyze pricing performance uncertainty across various subject pools. In the baseline sessions, the average subject can reason effectively to about rank 6. Most subjects do not have an accurate expectation about their performance against a procedurally rational opponent. They are too optimistic and value the games at much more than they are likely to earn if they actually play them.; The second essay (Chapter II) attempts to determine if the variation in estimated rationality bounds is correlated with the probability of winning when playing against another person in games that exceed both players' estimated rationality bound. We find that the model of substantive rationality has little predictive power. We ask the question, “Does seeing deeper into a game matter when neither player can see to the end of the game?” We design a model of play that accounts for the individual players rationality bounds as well as the structure of the game. The subjects with higher estimated bounds win 63% of the round robin games regardless of the games theoretical value.; Chapter III studies learning. We design a model to account for learning in PI-games and test to see if playing against an error prone opponent influences. We find strong evidence that subjects learn to play according to various rules and algorithms that lead to significant increases in performance. There is weak evidence that playing against a human increases performance more than playing against the procedurally rational algorithm.