Making the Move to Triple-A (aka the NL)

Anyone following this season in baseball, especially the Red Sox, has seen the remarkable difference in pitching in the National League as opposed to the American League. Everyone has taken note of John Smoltz moving from the Red Sox to the Cardinals and then Brad Penny doing the same weeks later (to the Giants).

I don't really want to get into the reasons why pitchers pitch much better in the NL. We all know about the designated hitter and the disparity between payrolls (specifically the 1st, 3rd, 4th, 5th and 6th-highest payrolls in baseball in the AL in 2008) and any number of other theories. But the real question is - is it true? Is there really a wide disparity between NL earned run averages and AL earned run averages? Most people have theorized the number is about .75 runs fewer in the NL than it is in the AL. No one ever quotes a specific study to support this though and I'm guessing it's just one of many theories in MLB that go assumed and unchecked.

So what if we checked every pitcher in baseball over the past five years that has switched leagues. What would the numbers show? I compared both pitchers who transitioned during the season (by trade or release) and those that signed or were traded in the offseason. First I looked at their season in the AL and then their season in the NL. For pitchers who transitioned midseason (say John Smoltz this year) I compared their statistics during the one season before and after switching teams. For those that transitioned in the offseason, I compared them from one season to the next. I thought it was most important to look at their years back-to-back because that is when they were pitching most similarly. Obviously there are a thousand different factors that could mess with the numbers - everything from injuries to changed pitching coaches to heartbreak caused by losing your trophy wife. You'll just have to excuse those issues because they are impossible to account for. I also looked solely at starting pitchers, though I imagine relief pitchers would follow a similar trajectory.

It turns out the theories are fairly accurate, though they actually low-ball the number by a bit. Over the past five years, the average is actually -0.92 earned runs average when a pitcher transitions from the American League to the National League. There is, in fact, only one season in which the numbers actually went up (2007). That number is largely skewed by Mike Maroth who went from Detroit to St. Louis in midseason and went from a 5.06 ERA to a 10.66 ERA. Why they pitched Maroth 38 innings with a 10.66 ERA and 2.32 WHIP I have no idea, though seven appearances came in relief. And even with that number factored in the overall ERA for the five starters who transitioned during the 2007 calendar year the AL ERA was 4.35 in 600.2 IP and the NL ERA was 4.39 in 606.1 IP - just a miniscule (and statistically insignificant) difference of .04.

The most stark difference came last year when the ERA for the six pitchers in the AL was 4.20 in 891 IP and then plummeted to 2.93 in the NL - a difference of 1.27. Those numbers can largely be attributed to three pitchers - C.C. Sabathia going to Milwaukee (1.65 ERA), Rich Harden going to Chicago (1.77 ERA), and Johan Santana joining the Mets before the season (2.53 ERA). But even pitchers like Odalis Perez and Joe Blanton saw significant declines once they joined the NL.

You can take a look at the entire statistical analysis. I didn't want to post it, because it's so dry (it's a list of names and their statistics). And while it's not necessarily surprising, it is quite interesting to read (if you're crazy like me).