- Expected payoff to a deceiver meeting a verifying partner: $(1-\phi)T_p - \phi k$.
- The population **verification intensity** $q$ is itself best-response: agents verify more when they expect more deception. This is the **screening response** (Stiglitz).