Changeup! (AP Photo/Kathy Willens)

Last year was really a tale of two seasons for Phil Hughes. He was brilliant early on, striking out close to a batter per inning (8.1 K/9) and earning a trip to the All-Star Game. Things started to come apart down the stretch, as his strikeout rate fell (6.6 K/9) and he suddenly become homer prone. Opponents had just a .295 wOBA (.138 ISO) off Hughes in the first half (which exactly matches Jeff Francoeur’s 2010 mark, for perspective) compared to a .330 wOBA (.181 ISO) in the second. Although most of our focus was on Phil’s changeup, perhaps we should have been paying attention to one of his other pitches.

For the most part, Hughes is a three-pitch guy. He throws a regular old four-seam fastball that sat right in the 92-94 mph range all season, a sneaky little cutter in the high 80′s, and a big breaking over-the-top curveball. That last one is the offering we’re going to focus on. The table on the right shows how often Hughes threw his curve, plus how often the batter swung at it and how often they swung and missed, broken down between the two halves of the season.

The first thing that (should have) jumped out at you was the whiff rate. Hitters swung and missed at Hughes’ curve 8.6% of the time in the first half, but that fell all the way down to 3.4% in the second half. That 5.2% drop is drastic, and it’s compounded by the fact that he started throwing the pitch a whole lot more often down the stretch. Turning to PitchFX, we can see that the vertical break of the curve was fairly consistent throughout the season, but the pitch was drifting all over the place horizontally…

Click the image for a larger view or better yet, look at this gif of the two graphs overlaid onto each other. It’s easier to compare them that way. The curve is the splotch of blue in the lower right quadrant.

The majority of Phil’s second half curves ended up about three or four pitches from the center of the plate to his glove side, which for all intents and purposes is right down the middle. In the first half it was more like five to seven inches off center, a pretty big difference. Hughes’ bender was was just far enough away from righties and too far inside on lefties for them to do any major damage in the first half, but they had a little easier time getting to it after the break.

Not only was Hughes’ curve finding the heart of the plate with more regularity in the second half, but he also lost about two miles an hour off the pitch. Hitters had that much more time to react to a pitch over the plate, a straight up bad combination. A power curve that generated swings and misses becamee a little more loopy down the stretch and simply wasn’t missing any bats. The decline of the curveball (theoretically) explains the decline in Hughes’ whiff rate, which in turn explains the decline in his overall strikeout rate.

There’s two things I should mention because they seem relevant enough. First of all, the All-Star break is right around when Hughes eclipsed his innings total from the previous year. He threw 105.1 total innings in 2009, and following his first start after the break in 2010, he was already at 106 IP. Could be a coincidence, could be meaningful (fatigue?). The other thing is that his release point changed, raising about six inches from the first half to the second. Here’s a gif comparing before and after. Again, it could mean something, it could mean nothing.

The changeup is undoubtedly going to be priority number one for the Yankees’ young hurler in 2011, but getting the curveball back to where it was in the first half will be key as well. Hopefully it’s nothing more than a fatigue issue and an offseason of rest does the trick. If it’s a mechanical issue, well those can be a pain in the ass.

Big ups to Texas Leaguers for PitchFX data and graphs.