So you would have faith that a player who produced stats for one year would have the same odds of doing it again as someone who did it for six years?? Not meaningless as you say.
No. That's not what I mean. What I mean is that there is actually an ocean of difference between projections drawn off a 200-game (or, for ease of baseball use, 162-game) sample coming from across the seven years than there is of using one complete season.
Put otherwise: If you want to know how good a player actually is, would you rather know his stats in (1) one complete season out of seven or (2) 162 games chosen entirely at random out of those seven seasons? (Note that this is not an opinion question: there's a right answer and a wrong one.)
