Learning in Tournaments with Inter-Generational Advice

We study learning in a simulated tournament using an inter-generational framework. Here a group of subjects are recruited into the lab and play the stage game for 10 rounds. After his participation is over, each player is replaced by another player, his laboratory descendant, who then plays the game for another 10 rounds as a member of a fresh group of subjects. A particular player in generation t+1 can (1) see the history of choices by his generation t predecessor and (2) receive advice from that predecessor via free-form messages that generation t players leave for their generation t+1 successors. We find that the presence of advice makes a difference in that the experimental groups who get advice perform better – their decisions are closer to the Nash equilibrium – compared to a control group of subjects that plays the game with no recourse to such advice.