Logit Models:
Logit models postulate some relation between the logit of observed probabilities (not the probabilities themselves), and unknown parameters of the model. For example, logit models used in logistic regression postulate a linear relation between the logit and parameters of the model.
The major reason for using logits, as opposed to probabilities themselves, is that in many cases where a linear model using probabilities does not fit the data, a linear model using logits does.