1. Weather forecasts are likelihoods.

2. Let’s take a look at different races and the varying likelihoods of Baldness.

3. Begin with a "prior distribution" which may be based on anything, including an assessment of the relative likelihoods of parameters or the results of non-Bayesian observations