All empirical evidences point at the inevitability of disappointment even if realistic expectations are placed on individuals. This problem emerges due to an inherent difficulty in obtaining sufficient behavioral samples to construct an accurate model, as well as, our tendency to ignore, misinterpret, and forget empirical evidences as a consequence of emotional involvement with the subject.
So, the solution is not to forgive and forget, but to forgive and update the model!