From this morning's article summarizing the first half of the season:
"See you in September: The Mariners, with the better lineup, comparable bullpen and improving rotation, edge out the Angels on the final weekend to win the division title; the Angels settle for the wild-card spot."
This prediction comes in the last paragraph, which is pretty cowardly. Seems to me to be a perfect example of predicting based on the small recent sample size rather than the larger sample size of the first half of the season, and even the second half of last season.
The Mariners have played well, but they are going to cool off, and the Angels are going to get back to .600 ball.