# principal_agent_reputation-t3
three-periods repeated principal agent game with varying stakes.
good agent with probability 0.1
bad agent with probability 0.9.
state of world 0 or 1
action 0 or 1
good agent shares the same preferences with the principal --> matching the true state of world
bad agent always prefers action 1.
after agent takes action principal observes both the action and the state.

This is treatment 3 where interaction starts small and stakes increase by multiplying with 3.
This is the equilibrium treatment which provides the highest expected payoff to the principal.