Maybe in early April and May when he was doing so well you should take him out after 7 innings with 5-0 lead and save those innings for later in the year.
In terms of physics, there is no such thing as a ball breaking early or late. It is going to have the same amount of "break", or movement the entire time to the plate (ignoring air resistance, which will obviously reduce the amount of break on the way towards the plate). A ball cannot magically start out straight, then all of a sudden start to move half way towards the plate. The break is based on the spin of the ball out of the pitchers hand.
Therefore what you are seeing with the ball breaking too early out of his hand is your eyes playing tricks on you. Doug's data shows that he is getting less break overall on the pitch. That could explain why you are seeing a difference, although I'm not sure how a pitch getting less break would look like it is breaking earlier. I would have to think about that one.
Another thing to consider is each ballpark has a different camera angle from the CF camera. And some get quite bad (I think the Reds camera is supposed to be one of the worse in the league, but I can't remember). So you might have been watching a game with a tighter angle, and noticed the ball breaking earlier because of this.
My impression was that late in the year he lost the late, hard break on the slider. It got slurvy. Not sure what the data shows, but it seemed clear to me at the time.
"Baseball is a very, very complex business. It's more of a people business than most businesses." - Bob Castellini