Bayesian and frequentist approaches to inference or prediction are very different. How different? This simple example highlights the difference and the argument in favor of using Bayesian posterior probabilities.