This paper aims to measure differences in risk behavior among expert chess players. The study employs a panel data set on international chess with 1.4 million games recorded over a period of 11 years. The structure of the data set allows us to use individual fixed-effect estimations to control...