Punishment and Gossip: Sustaining Cooperation in a Public Goods Game - Advances in Social Simulation - page 110

behaviors in response to the percentage of detected free-riders in their group:
when the number of known defectors in a group is too high, i.e., it exceeds a
certain threshold, agents become active and their strategy changes (as described
in Algorithm 1). Each and every agent is endowed with an initial amount of
50 points that can be used to cooperate or to punish others; regardless of the
strategy, agents are culled from the game when their payoff goes to zero and
they are not replaced. The other strategies are:
- Tit-for-Tat (TFT): They start as cooperators but if active, they start free-
riding until exiting from the active state.
- Nice Punisher s (NP): They contribute in the passive state; once active, they
punish free-riders at a cost to themselves and making FR s pay a cost, but
they continue to cooperate in the PGG. This behavior will continue until
the agent exits the active state (by being assigned in a group with a number
of cheaters below the threshold).
- Mean Punisher s (MP): They contribute in the passive state; once active,
they punish free-riders and free ride themselves in the PGG until they exit
the active state.
To the above types, developed by Carpenter [4], we add two more types, in order
to compare punishment and reputation. Active Gossipers transmit and receive
information and they integrate their personal experience and the reputational
information received from other Gossipers in order to react against Free-Riders.
Gossipers' behavior is described in Algorithm 2 (2). Gossipers can be:
- Nice Gossiper s (NG): Agents contribute in the passive state; once active,
they start spreading information about free-riders, and cooperate in the
PGG.
- Mean Gossiper s (MG): Agents contribute in the passive state; once active,
they start spreading information about free-riders, but they always defect in
the PGG when active.
Algorithm 2. Description of gossipers' behaviors
for Number of Timesteps do
Random group formation of the population;
if Group has bad reputation then
Apply reputation strategy
else
Take First stage decision according to active/passive status.
end if
Gather and Distribution of the Public good in each group;
First Stage Decisions are made public within the group;
Agents make Second Stage decision;
Reputation diffusion;
end for