Quote:
Originally Posted by OutS|der
via player ratings.
League totals have too much control. They should not dictate 18 no hitters/Perfect Games, vs a slight adjustment down to 6.
Player ratings should generate all statistics
|
Player ratings are relative to statistics. A .345 hitter in one league will have a different contact rating than another .345 hitter playing in a different league. I don't think it's even possible to generate statistics purely off of ratings when it's relative to itself. Even when generating off of the absolute ratings in the player editor, you'd have to assign statistical values to ratings and their variance.
The problem is that baseball is a zero-sum game. For every home run hit, a pitcher just gave up a home run. Because these ratings are on the level of the individual player, and because the generation of statistics operates on that basis, it would have to deviate from the assigned statistical value when the engine is faced with the fact that the ratings created for a batter have to have their statistics generated by facing pitchers who also have ratings that need to generate their assigned statistics.
That problem is erased with league totals/modifiers because these stats are generated at the league level, instead of the individual player.
So, essentially, computers cannot simply work with player ratings to determine statistics. They need hard numbers, and as a result, need certain statistical outputs assigned to an absolute rating (for example, an absolute Power rating of "100" equals a HR total over 500 plate appearances of "20"). Because of the variation in pitcher ratings over those 500 plate appearances, that total most likely could not be reached without screwing with the HR allowed total of those pitchers according to their own ratings.
TL;DR: Absolute ratings cannot be used as the determinant for statistical generation because absolute ratings exist in a vacuum, independent of the skill of the other party, while statistics depends on the absolute ratings of both parties. The statistical output of one set of absolute ratings would differ from the necessary input in most cases as a result, and would be unable to be computed. Meanwhile, with league totals/modifiers system, stats can be assigned in a way which allows specific stats to be aligned with a league total or average, rather than with absolute ratings, which would solve the problem mentioned previously.
I could be wrong though.