I'm wanting to create a probabilistic model based on several simulations, where I have a binary outcome (go or no-go) and a variable input. The old way of formulating the proability distribution was a straight-forward, curve-fitting of the input variable versus the "go or no-go" outcomes...