Communication à un séminaire :
(Universite de Maastricht), « Extrapolation in Games of Coordination and Dominance Solvable Games. »
, BEE Seminar
, TSE, Toulouse, 27 mai 2010, 15:30-17:00, salle MF 323.
We study extrapolation between games in a laboratory experiment. Participants in our experiment first play either the dominance solvable guessing game or a Coordination version of the guessing game for five rounds. Afterwards they play a 3x3 normal form game for ten rounds with random matching which is either a game solvable through iterated elimination of dominated strategies (IEDS) or a pure Coordination game. We find strong evidence that participants do extrapolate between games. Playing a strategically different game hurts compared to the control treatment where no guessing game is played before and in fact impedes convergence to Nash equilibrium in both the 3x3 IEDS and the Coordination game. Playing a strategically similar game before leads to better (faster) learning in the second game.
We also find evidence for 'naive extrapolation'. Participants tend to choose actions which are labeled similarly to succesful actions in the guessing game with higher probability in the first rounds of the 3x3 game. This effect is much stronger in the Coordination games. On balance extrapolation helps play in games solvable through deliberative reasoning (such as IEDS), but the effect is ambigous in games where some intuition is needed to reach a Nash equilibrium (such as pure Coordination games).
C71 : Cooperative Games
C92 : Laboratory, Group Behavior