Everybody knows about how McCarthy decided to scrap his curveball (which is supposedly a plus pitch) for a slider. Apparently McCarthy wasn't all that comfortable with the curveball and could not control it very well (hat tip: BBTiA). So I briefly looked at some of the pitch f/x numbers from both 2007 and 2008 (beware of the SSS) to see if his discomfort was also reflected in results.
The one thing that stands out about the curveball is that almost nobody swings at the pitch and misses. In 2007, the swinging strike % on the curveball is a bit lower than 5%, and in 2008, it's even worse at 3% - Andruw probably swings and misses more than that in batting practice. The average starting pitcher gets about 7.8% swinging strike rates on the average of all of his pitches, and the swinging strike % are usually higher for breaking pitches than for fastballs - provided that the breaking pitch is any good. By comparison, McCarthy's fastball ended up getting about 7% swinging strikes in 2007 (which isn't bad at all), and further suggests that the curveball doesn't fool anyone (a real man's curveball like that of Ben Sheets gets 11.5% swinging strikes). Nevertheless, it seems mildly effective in avoiding being hit - partly because he threw it for a strike only 54% of the time in 2007, and slightly improved 61% strikes in 2008. As a reference, the strike rate of an average starter (for all his pitches) is about 64%.
On the other hand, his changeup (also supposed to be a plus pitch) got swinging strikes 15% of the time in 2007 and 14.5% in 2008. Unfortunately, his command of the pitch was really bad in 2007 and 50% of it missed the strike zone while being ignored by the hitters. The command improved to 65% strikes in 2008. Interestingly, the average velocity of the changeup increased by a bit more than 1 mph last season, but don't ask me what it means. McCarthy said he is mainly a fastball-changeup guy, and those two pitches do indeed appear to be his better pitches. In 2007, his fastball wasn't bad, but he didn't have much command on either of his breaking pitches. In 2008 on the other hand, his breaking pitches seemed to be better, but his fastball wasn't quite as good.
After seeing the data, I guess I am a bit less upset and flummoxed about his sudden change of heart and repertoire - given how bland his curveball appears to be. I don't know if the slider will turn out any better, but it can't be too much worse than the curveball even if it's just a placebo pitch.
Quickly, somebody teach him a cutter.