This paper examines data from the Norwegian television game show Joker, where contestants make well-specified choices under risk. The game involves very large stakes, randomly drawn contestants, and ample opportunities for learning. Expected utility (EU) theory gives a simple prediction of choice under weak conditions, as one choice is always first-order stochastically dominating. We document frequent, systematic and costly violations of dominance. Most alternative theories fail to add explanatory power beyond the EU benchmark, but many contestants appear to have a systematic expectation bias that can be related to Tversky and Kahneman's (1973) "availability heuristic". In addition, there seems to be a stochastic element in choice that is well captured by the so-called Fechner model.
