I tend to view Actual vs Pythag differences as more random than anything, particularly when the number isn't all that high. So two things I found looking at the league's Pythag numbers overall: 1) MLB teams as a whole had eight fewer Pythagorean wins than actual wins. That seems odd in that one would expect the positives and negatives to more or less cancel each other out. Not sure what would cause that other than an accumulation of rounding errors, although you'd think that those errors wouldn't collectively tilt in one direction enough to skew things about 1/4 of a win per team. 2) A difference of three wins (in either direction) isn't that much. Eight teams won four or more games more they were 'supposed to' [8, 7, 7, 6, 5, 5, 4, 4] based on Pythagorean projections, and five showed a negative Pythag of more than three wins [-4, -4, -7, -9, -11]. So if the Mets Win total is different from where it 'should' have been, it's a variation of less than what nearly half of all other teams (13 of 29) saw this year.