Now that we know it won't be Phil Hughes behind C.C. Sabathia and Cliff Lee, what do we make of A.J. Burnett, who had a disastrous 2010, which on first blush screams for a regression to the mean (and will probably do so, to some extent). The man had been a fairly consistent 113 ERA+ pitcher for six seasons, then laid an egg of 81 in 2010.
There are some problems, I think, with the idea that Burnett just had an off year, and that he should be primed to return to his 110-115 ERA+ days of yore.
For one, his BABIP was .319, up 17 points from his 2009, yes, but lower than in his 2008, when his ERA was all of .03 higher than it was subsequently in '09. So he didn't get particularly unlucky with hits falling in.
Similarly, his LOB%, the number of baserunners he stranded at the end of an inning, was 68.8 percent, lower than the 75.9 he posted in 2009 — but basically the same as the 70.5 he had in 2009. Which means the sequencing of his hits and outs was only slightly unlucky, too.
Yet his K/9 has cratered, from 9.56 to 9.39 to 8.48 to 6.99. His walk rate, meanwhile, is basically unchanged, showing only a slight increase from 3.59 to 3.50 to 4.22 to 3.76. So while his FIP indicates he was indeed a little unlucky, he wasn't inordinately so: 4.83 FIP versus 5.26 ERA. He had a big jump in home runs per fly ball, which tends to be mostly luck, and that, combined with the slightly elevated BABIP and slightly reduced LOB%, probably makes up the difference.
But that means, all things being equal, Burnett went from being a 3.80-ERA pitcher to a 4.80-ERA pitcher. What happened?
According to Fangraphs, he used his fastball a whole lot more — 69 percent, most since he was with the Marlins, and well above the 66 percent he had in 2010 and 64.7 percent in 2009 — and his curveball a lot less (27.4 percent, versus 31 percent last season). His changeup, which used to make up a not-inconsiderable portion of his pitches (5-10 percent through 2008), has dropped to 3.5 percent.
What's problematic about that is his fastball in both seasons with the Yankees has been tattooed — 14.3 runs below the average fastball, according to linear weights (which look at the expected runs in a given situation, then, once a pitch is thrown, takes the difference between the former run expectancy and the new one and ascribes it to the type of pitch thrown). In previous seasons, Burnett's fastball was never incredible, but it always stayed within a -6 to +8 range.
Why the change? It's not hard to figure out. Here are Burnett's fastball velocities by year since 2004, when he began posting 120-inning seasons consistently:
2004 – 95.4
2005 – 95.6
2006 – 94.9
2007 – 95.1
2008 – 94.3
2009 – 94.2
2010 - 93.2
So why get away from the curve? Well, because in 2010 it went from being a +16 pitch (and +13 in 2008-09) to a -3.9 pitch.
Stats can only tell you so much. Did Burnett's curve look flat this season? I wouldn't know.
But consider this: Batters didn't swing at a greater percentage of Burnett's pitches than in years past, but they made contact with a greater percentage of his offerings than in any other season. Why? Well, batters against Burnett made contact on an eye-popping 63.7 percent of his pitches that were out of the strike zone (at which they swung), after a previous career high last season of just 51.1 percent. That means when batters swung at Burnett's pitches out of the zone, they were making contact nearly two-thirds of the time.
That seems to indicate a major problem with the curveball not diving far enough away from the strike zone to induce swings and misses — and his swing-and-miss percentage indeed was just 14 percent for the second year in a row, after being at 18 percent in 2007 and 2008.
Is the problem with his curve correctable? I don't know because I don't know what's wrong with it. But even if it is, that velocity drop on the fastball is troubling. Burnett has never been a control artist, so getting by with a slower fastball is going to get tougher and tougher.