In common-interest spatial-dispersion games the agents common goal is to choose distinct locations.We experimentally investigate the role of cognition in such games and compare it with the role of cognition in spatial matching games. In our setup cognition matters because agents may be differentially aware of the dispersion opportunities that are created by the history of the game.We ask whether cognitive constraints limit the agents ability to achieve dispersion and, if there is dispersion, whether these constraints affect the mode by which agents achieve dispersion.Our main finding is that strategic interaction magnifies the role of cognitive constraints.Specifically, with cognitive constraints, pairs of agents fail to solve a dispersion problem that poses little or no problem for individual agents playing against themselves.When we remove the cognitive constraints in our design, pairs of agents solve the same problem just as well as individuals do.In addition, we find that when playing against themselves agents do not change the mode by which they solve the dispersion problem when our design removes the cognitive constraints.
|Place of Publication||Tilburg|
|Number of pages||20|
|Publication status||Published - 2005|
|Name||CentER Discussion Paper|
- noncooperative games
- laboratory group behavior