Can we estimate macroforecasters’ mis-behavior?

We answer positively to this question by using Maximum Lq-Likelihood (or Deformed Likelihood) estimator. This is based on a parameter which measures the aggregate quote of judgment in forecasting (game-based) system formed three players—Forecaster, Policy Maker and Reality. For first time econometric literature, we apply estimator dynamic derive robust version Kalman Filter—the Filter (DKF). The evidence from U.S. data suggests that judgmental dynamics exists correlated (but not coincident) with phases Business Cycle. Furthermore its knowledge improves in-sample as well out-of-sample estimation.