Market theorists assume that expected utility predicts preferences at the market level even as evidence mounts that it predicts poorly at the individual level. The arguments for better‐performing markets are grounded in the assumption that individuals respond to the competition of the market. The objective of this study is to test empirically the validity of those assumptions using the betweenness property of expected utility. I conclude that expected utility does indeed predict better in markets, but analyses suggest that improved performance may be due to the statistical role played by markets introduced by market price selection rules.