Magic on the Internet: Evidence from Field Experiments on Reserve Prices in Auctions

The minimax argument represents game theory in its most elegant form: simple but with stark predictions. Although some of these predictions have been confirmed in field data, tests of minimax have generally failed in laboratory experiments that make use of student subjects. In a seminal study, Palacios-Huerta and Volij (2007) present evidence that potentially resolves this puzzle: soccer players (both amateur and professional) play minimax strategies in laboratory experiments. They attribute this result to an ability on the part of the soccer players to transfer expertise they have learned shooting penalty kicks to card games played in the lab. These results induced us to carry out similar experiments using subject pools with unrivaled experience applying analytical thought in card games: world class bridge and poker players. In contrast to Palacios-Huerta and Volij’s soccer players, we find little evidence that real-world experience transfers to the lab for either subject pool. These poker players, who have vast experience with high-stakes randomization, play the mixed strategy equilibrium in these games roughly as often as our student subjects. We proceed to test whether a sample of professional soccer players from the U.S. play according to theory. Again, we observe little evidence in line with the predicted mixed strategy equilibrium. Coupling complementary experimental treatments that pit professionals against preprogrammed computers with answers to post-survey questionnaires permits us to explore why professionals do not perform well in the lab.


First version: February 2007

This version: May 2007

Download the manuscript.