Learning in a black box: trial-and-error in voluntary contributions games
Many interactive environments can be represented as games, but they are so large and complex that individual players are in the dark about others' actions and the payoff structure. This paper analyzes learning behavior in such 'black box' environments, where players' only source...
| Main Authors: | , , , |
|---|---|
| Format: | Working paper |
| Published: |
University of Oxford
2013
|
| _version_ | 1826270118376636416 |
|---|---|
| author | Young, H Nax, H Burton-Chellew, M Westor, S |
| author_facet | Young, H Nax, H Burton-Chellew, M Westor, S |
| author_sort | Young, H |
| collection | OXFORD |
| description | Many interactive environments can be represented as games, but they are so large and complex that individual players are in the dark about others' actions and the payoff structure. This paper analyzes learning behavior in such 'black box' environments, where players' only source of information is their own history of actions taken and payoffs received. Specifically we study voluntary contributions games. We identify two robust features of the players' learning dynamics: search volatility and trend-following. These features are clearly present when players have no information about the game; but also when players have full informaiton. Convergence to Nash equilibrium occurs at about the same rate in both situations. |
| first_indexed | 2024-03-06T21:35:50Z |
| format | Working paper |
| id | oxford-uuid:463a78c0-fdea-4ddc-813b-cea2740ea4ef |
| institution | University of Oxford |
| last_indexed | 2024-03-06T21:35:50Z |
| publishDate | 2013 |
| publisher | University of Oxford |
| record_format | dspace |
| spelling | oxford-uuid:463a78c0-fdea-4ddc-813b-cea2740ea4ef2022-03-26T15:12:28ZLearning in a black box: trial-and-error in voluntary contributions gamesWorking paperhttp://purl.org/coar/resource_type/c_8042uuid:463a78c0-fdea-4ddc-813b-cea2740ea4efBulk import via SwordSymplectic ElementsUniversity of Oxford2013Young, HNax, HBurton-Chellew, MWestor, SMany interactive environments can be represented as games, but they are so large and complex that individual players are in the dark about others' actions and the payoff structure. This paper analyzes learning behavior in such 'black box' environments, where players' only source of information is their own history of actions taken and payoffs received. Specifically we study voluntary contributions games. We identify two robust features of the players' learning dynamics: search volatility and trend-following. These features are clearly present when players have no information about the game; but also when players have full informaiton. Convergence to Nash equilibrium occurs at about the same rate in both situations. |
| spellingShingle | Young, H Nax, H Burton-Chellew, M Westor, S Learning in a black box: trial-and-error in voluntary contributions games |
| title | Learning in a black box: trial-and-error in voluntary contributions games |
| title_full | Learning in a black box: trial-and-error in voluntary contributions games |
| title_fullStr | Learning in a black box: trial-and-error in voluntary contributions games |
| title_full_unstemmed | Learning in a black box: trial-and-error in voluntary contributions games |
| title_short | Learning in a black box: trial-and-error in voluntary contributions games |
| title_sort | learning in a black box trial and error in voluntary contributions games |
| work_keys_str_mv | AT youngh learninginablackboxtrialanderrorinvoluntarycontributionsgames AT naxh learninginablackboxtrialanderrorinvoluntarycontributionsgames AT burtonchellewm learninginablackboxtrialanderrorinvoluntarycontributionsgames AT westors learninginablackboxtrialanderrorinvoluntarycontributionsgames |