Prize Structure and Information in Tournaments: Experimental Evidence

This paper examines behavior in a tournament in which we vary the tournament prize structure and the information available about participants' skill at the task of solving mazes. The number of solved mazes is lowest when payments are independent of performance; higher when a single, large prize is given; and highest when multiple, differentiated prizes are given. This result is strongest when we inform participants about the number of mazes they and others solved in a pre-tournament round. Some participants reported that they solved more mazes than they actually solved, and this misreporting also peaked with multiple differentiated prizes.