On mu from the logistic quantal-response equilibrium

This note reflects on the key parameter of the popular logistic quantal-response equilibrium in order to set some common guidelines on its empirical interpretation. It is stressed that the estimated model must be in harmony with the experimental design, because the estimation results on mu prove to be sensitive to changes in the strategy sets of players even if those are unimportant from a game-theoretic point of view. It is also shown that a simple post-estimation correction of mu can help inter-game comparisons, while pre-estimation treatments of the data may introduce unwanted biases.