An Alternative Explanation for the Logit Form Probabilistic Choice Model from the Equal Likelihood Hypothesis

The logit function is frequently used in the probabilistic choice problem. The justification is usually provided in a manner that actual utility consists of observed utility and Gumbel random term; the probability then becomes logit. In this paper, we present an alternative explanation of the logit form probabilistic choice model from the "Equal Likelihood Hypothesis'' without the Gumbel distribution. This hypothesis is that if the total utility U from combinations of actions are the same, then all such combinations of actions are equally likely