My lineups are determined by ratings for the first half of the season, and by splits for the second half (after the all-star break). It's counter-intuitive because, as Mike says, the game is driven off of ratings and not stats. But sometimes you do have players who tend to over-perform or under-perform their ratings, and that sometimes carries through the entire season.
By game 120, it's difficult to keep sending your .235 hitter out there because of his better ratings while benching the .290 guy.