Long time follower, first time poster. Just hope to share some of my thought on the topic "whether a smaller run differential with a lower variance could lead to a higher winning percentage", assuming that variance can be indeed manipulated:
Imagine:
Team A: run scored = averaging 4 runs with a variance of +/- 0.2, run given up = averaging 3 runs with a variance of +/- 0.2.
Team B: run scored = averaging 5 runs with a variance of +/- 2, run given up = averaging 3 runs with a variance of +/- 2
It is likely that team A would win more often, albeit with a smaller run differential, due to the smaller area of overlap between the "run scored" curve and the "run given up" curve.
