Confidence in Interactive Decisions

When the conditions of a competitive market are simulated in a game, the number of people deciding to enter each round approaches the game-theoretic equilibrium after a few rounds. This occurs even when there is no communication between players, and the only feedback on each round relates to the number of entrants. However, in real markets entry is generally too high, resulting in too many start-ups failing and suboptimal profits arising from excessive competition. Previous research has suggested that this over-entry is the consequence of entrants believing themselves to be more likely to succeed than other potential entrants. This hypothesis has not so far been adequately tested because beliefs regarding chances of success relative to others have not been fully manipulated or measured. We report the results of two experiments designed to test this hypothesis and also an alternative hypothesis that it is a different positive bias – entrants‟ belief in their skill per se rather than relative to competitors – that is responsible for excess entry. The results support the general hypothesis that positive biases are largely responsible for excess entry, with the balance of evidence being slightly in favour of absolute overconfidence in one‟s own abilities – rather than overconfidence in one‟s chances relative to those of others – being the better predictor of entry into the market. Asymmetric Dominance Effects in Games Andrew M. Colman and Briony D. Pulford (University of Leicester) and Fergus Bolger (University of Durham)