How can you tell if someone is an expert? If X is more talented than Y? Maybe they are just guessing (or just got lucky). In joint work with the wonderful Ron Graham, over a 40 year period, we studied a model version of this problem. In this talk, I'll tell you what we learned. The problem involves an ordinary deck of 52 cards. They are well shuffled and looked at one at a time. Each time, a guesser tries to guess the next card (before it is discarded) and is told if the guess is right or wrong. How should the guesser use this information to get a high score? For all our study, we don't know the exact best thing to do(!). NEXT, how can we judge if the guesser 'has talent'? A very useful notion of 'skill scoring' helps to solve this problem. I'll explain how this has applications to taste testing experiments, to evaluation of clinical trials and to testing if 'extra sensory perception' is a real phenomenon.