Despite all that I’ve found, all that I’ve read, and all the optimistic data I’ve calculated, there is nothing that will make me like A.J. Burnett. Simply put, I strive for consistency, every baseball fan, manager, player, and general manager does; Burnett is as erratic as a starting pitcher gets. You can see it in his hit by pitch totals, you can see it in his wild pitch totals, you can see it in his walks, and worst of all you can see it in his ERA. Despite the trade rumors, Burnett is likely the Yankees’ die to roll for the next two years, so it’s time we discuss how unlucky his 2011 season was.
First, let’s look at two starting pitchers.
Pitcher A, A.J. Burnett, has one less strikeout and half a walk more per nine innings than Pitcher B, but that pitcher also plays in the National League, increasing his strikeout rate and decreasing his walk rate on the pitcher hitting ninth. The big difference between the two is the HR/9 rate; Burnett gives up more than double the amount of homeruns than Pitcher B. From there the ERAs are startlingly different. AJ finished with a 5.15 ERA in 2011, while Pitcher B, Giant’s ace Tim Lincecum, finished with a 2.74 ERA.
One number above predicts that the two should have had closer performances, that was xFIP. In case you’re unfamiliar, xFIP assumes that a player will put up a league-average HR rate (10.6% HR/FB) and applies it to their FIP. It’s one of my least favorite statistics because some pitchers can be more prone to homeruns based on their type of pitching or stadium. While AJ does pitch in a homerun prone stadium, he is not a fly ball pitcher, and his career average 11.3 HR/FB% is very close to the league average. For that reason, the 3.86 xFIP is significant in showing how unlucky his homerun rate was. 17% of his fly balls were homeruns in 2011, which is a number that you’d assume is bound to stabilize in 2012.
If you don’t have full faith in the xFIP numbers, there is another statistic that might convince you. Pitch Type Linear Weight was created with the introduction of Pitch f/x, and its purpose was to calculate the run value of a specific players pitch. If you read this Fangraphs article on how to use it, you’ll see that these numbers can help in determining how certain pitch types faired, however you will not be able to use the numbers to predict how good or bad a pitch was. The main reason for this is sample size, and as we’ll see with Burnett’s Run Values (RVs), BABIP can skew the statistic. To fix the problem, Expected Run Value (RVe) was created, which records the contact of a pitch as ground ball, fly ball, or line drive, and runs it through league average outcomes, unlike the actual outcomes used in Run Value.
With RV and RVe, we get two numbers, one that shows the effectiveness a pitch had and one the effectiveness it should have produced. Back in September, Beyond the Boxscore took a look at the “Unluckiest” pitches in 2011 and found that AJ’s four-seam fastball was one of the worst fairing pitches in the game. The article is a must read if you’re a statistics nerd like me. The difference between Burnett’s total run value and expected run value for 2011 was +20.556, which was second behind only Max Scherzer. This means that his four-seam fastball was hit for 20 extra runs than it should have been. Giving up 20 fewer runs would put Burnett’s ERA at 4.21, remarkably closer to his xFIP.
As much as I dislike A.J. Burnett, his numbers indicate that he ran into some extremely bad luck last year. With the revelations of his xFIP and RVe, there’s a case to be that he has upside to sit in the middle of the rotation. Predicting him has been a losing battle his entire career, but when you consider the money owed and possible production, AJ should be the fifth starter.