I would like to know if there is any way of quantifying the performance of the predictive power of a model in comparison to say a random number generator... for example, let's say with have two models:
Model A is able to predict the correct outcome 73% of the times as opposed to a random number generator that can do that only 19% of the times. Model B is able to predict the correct outcome 78% of the times as opposed to a random number generator that can do that 38% of the times.
Is there anyway of quantifying the performance of both of these models relative to random number generator in a simple number?