Risk-Taking in Hide-and-Seek and Hide-and-Avoid Games

We used novel two-person, hide-and-seek and hide-and-avoid games to study risk-taking, patterns of hiding, and search and avoidance behaviors. In the hide-and-seek game, hiders-defenders in a loss frame-first decided how to divide 16 coins belonging to them and then chose which of 16 compartments to hide the bundles in. Seekers opened eight compartments and were rewarded with the coins found; hiders kept the unfound coins. In the hide-and-avoid game, the coins initially belonged to the seekers. Hiders-attackers in a gain frame-had 16 damage-inflicting thumbtacks to bundle and hide; thumbtacks encountered by seekers penalized them and rewarded the hiders. In complementary variations of the games, it was the seekers who determined the bundling. We focused on the risk taken in bundling. Creating 16 one-item bundles entailed no risk, placing all items in a single bundle entailed maximum risk, and other bundlings fell in between. We also measured bundlers' optimism and actual gains. In four experiments (N = 320), we found that players in a loss frame took greater risks than players in a gain frame and that seekers took greater risks than hiders. Participants were optimistic, expecting to earn more than eight coins. Hiding was nonrandom: certain locations were over- or underused irrespective of the desired outcome; seekers apparently expected hiders' pattern of hiding, thereby earning significantly more than hiders. Introducing a device that performed one or both roles randomly reduced the effect of role on risk-taking. Optimism remained high irrespective of agency.