When in doubt, pull A.J. out

There aren’t many positives to take from A.J. Burnett’s 2011 season. If we’re grasping at straws we can say that it’s an improvement on his 2010 season, but that consolation goes only so far. Of the 112 qualified starters, he ranks 95th in FIP and 74th in ERA, and is striking out about a batter less than his career average. In fact, his strikeout rate is barely better than last year’s, while his walk and home run rates are just about the same. On Monday he incited a collective groan among the Yankee faithful when he blew the 2-0 lead that the Yankees had handed him in the top of the seventh. It felt like more of the same for Burnett.

One thing Burnett has actually done this year is avoid early exits. In all 18 of his starts he’s gone at least five innings. That might not seem like a big deal, but when the expected No. 2 has a year like Burnett’s 2010, we look in odd places to find positives. In those first five innings he has actually pitched fairly well. THere are a few innings that are worse than others — the 4th seems to be a problem for him, at least for the 18 innings this year — but overall he’s not all bad to start off a game. It’s when he gets into the later innings that things fall apart.

To be clear, most pitchers follow a similar trend as the innings progress. It’s the simple effect of the pitchers having seen them more often. But with Burnett it’s quite a standout. Five of the 15 homers he’s allowed this season have come in the sixth and seventh innings, and many of them have ruined otherwise winnable games. It truly starts to get bad as he faces hitters for the third time, when they’re hitting .292/.379/.508 against him. Such are the perils of being a two-pitch pitcher.

This isn’t to say that Burnett always needs to exit after five innings. There are some games where he’s going strong. For instance, he’s pitched into the eighth three times this year, going 1.2 innings and allowing just one hit and a walk. But it’s pretty clear, both in the stats and in our eyes, that he sometimes just falls apart after five or six. Given what we know about Burnett’s volatility and what we know about his failures once the lineup turns over a second time, there shouldn’t be any hesitation to pull him when he starts to show signs of vulnerability. In other words, the Kearns homer on Monday never should have happened, because he should have been out of the game after Duncan’s single. What followed was a movie we’ve seen too often.

It does seem that Burnett has made some improvements from his 2011 season, particularly at the beginning of games. He’s not getting shelled and handing it to the bullpen in the third, which is a dull but real positive. But it’s clear that he’s prone to lose it once hitters see him for a third time. There are times when he can get through those late innings, when his curveball has plenty of bite and he can locate his fastball. But when signs of trouble appear, Joe Girardi should waste no time in removing him from the game. We’ve seen it, and the stats bear it out. The Yankees are better served with a reliever than watching another late-inning Burnett implosion.