Obvious strategyproofness

In mechanism design, obvious strategyproofness (OSP) is a strengthening of strategyproofness that captures a robustness of strategyproofness to cognitively-limited agents.^[1]

Definition

Informally, a strategyproof mechanism is obviously strategyproof if the strategyproofness is robust to agents who don't understand or don't trust that their actions won't affect other players' actions.

The formal definition uses the extensive form formulation of a mechanism. It compares pairs of strategies ${\displaystyle S,S'}$ by considering the possible trajectories from an earliest information set where they diverge. A strategy ${\displaystyle S}$ obviously dominates another strategy ${\displaystyle S'}$ if, starting from each earliest information set where they diverge, the worst possible outcome for ${\displaystyle S}$ is better than the worst possible outcome for ${\displaystyle S'}$.

Examples

The sealed-bid second-price auction is strategyproof but not obviously strategyproof because bidders have to trust that their bids remain sealed.^[2][3] In contrast, the ascending clock auction is obviously strategyproof,^[1] even though for fully rational agents the two auctions are equivalent.

Other examples of strategyproof but not obviously strategyproof mechanisms include:

• Majority vote between two alternatives;
• Deferred acceptance^[4]
• Top trading cycles