Latest Response from CS
Our existing fatigue model is based on a few mostly proven theories.
Pitchers fatigue differently than position players. Pitchers simply cannot throw too many pitches or in too many games without incurring injury. Doing one or the other results in poor performance rather quickly. Leaving a starter in one inning too long always results in manager being questioned. Relief pitcher going an extra inning, same thing.
Position players are very different in that they play through the grind. without occasional days off, they begin showing signs of decay. A position player at 99% is not much different than 100%. At 89%, his decay is much more noticeable. At 79%, he's obviously much worse. And so on.
We debated having the drop off faster beyond 100% but decided not to because of some of the factors at play in SLB.
Opposing owners can strategize to fatigue a team. Start a terrible mopup pitcher can lead to a high scoring game and more at-bats. Playing more games in a division with many offensive parks, same thing. So, we have the 10% padding in place but we also opted to have a slower decay in the 90 - 100% range.
That's the rationale. It is always open for discussion, however.