Actually, that's not the argument I am making. I'm saying the job of the GM is to anticipate and predict future performance. Obviously, some times those predictions will be more accurate than others. In making that assessment, statistical analysis of previous years will only be one piece, and sometimes a piece of lesser importance, in a much larger picture - it is one piece of data among several, and in the end it does require intuition and guesswork. That's what GMs are paid to do.