The natural log of the odds, the Logit function dates back to 19th century. It was given its name by Verhulst, a Belgian mathematician.
The Logit regression is one wherein the dependent variable is not continuous rather a state of existence or non-existence, a mere yes or no, a 0 or 1. It helps establish causal relations i.e. what potentially leads to a result in question and aids in understanding the category behavior. Also, it allows specific economic interpretations instead of empirical ones. At heart it does not allow a grey area as it is altogether a binary. It has no room for the error component as the error is a constitutive component. And the funny thing is that this drives most of our daily lives which nonetheless has a lot of grey areas and error components.
Though a lot of times we enjoy these grey uncertain blips of our lives, the ground reality is that humans by nature need a clear separation among the choices they are offered to take sound and smart decisions. Otherwise, we tend to loose the certainty of choosing the best in interest. And guess where Logit is used? To aid in better decision making. 🙂