Pitchers frequently have significant splits within a particular year. He's faced weaker opposition, for one thing . . . random variation with respect to balls in play is another likely factor.

If he was blown out every single start he wouldn't still be in the rotation - even a well below-average MLB starter can string together some solid performances. That's nothing new, it happens all the time.

If you want to look at his performance by half-season, OK, since the beginning of 2015 he will have 6 half seasons, and he was bad in 4 out of the 6.

So for any given half-season, there's a 1/3 chance he will be pretty good and a 2/3 chance he will be bad. Maybe that's a good ratio to you, in which case we just have an honest disagreement over SP expectations.

Don't disagree with you in principle, but we need to look at stabilization rates for particular statistics. For example, a pitcher's BIP doesn't really stabilize until about 2000 balls in play. That's why I included the range of statistics I did.

For reference, Kyle faced 376 batters in the first half, 248 in the second half. I can't find splits for balls in play, but this year he's had a total of 461 balls in play.

Stabilization rates are as follows:

K rate: 70 batters faced

BB rate: 170 BF

GB rate: 70 balls in play

FB rate: 70 balls in play

So, while we'd expect to see streaks in ERA, wins, and even HR rate (1320 batters faced), a significant change in strikeouts, walks, ground balls and fly balls become predictive much more quickly.