If you're not familiar with tRA, get familiar. Here's the full explanation and there's also a no-numbers version. Basically, it's the next generation of DIPS. Pitchers receive park-adjusted credit or blame based on eight categories of outcomes (strikeouts, homeruns, walks, ground balls, etc) but aren't held responsible for whether the balls put in play are turned into outs or not. And instead of actual innings pitched, there's an expected IP total based on the same principles.
Some of you may wonder why I'm using tRA and not tRA*. The first is better at measuring past value (because it uses actual batted ball percentages) while the latter is regressed and a better representation of the true talent demonstrated by a pitcher (and thus a better sign of what's to come). As an example, a pitcher with a 0% line drive rate might not be able to sustain it going forward (so says tRA*), but that 0% sure was valuable in 2008 (so says tRA).
Anyway, by using tRA, xIP, and our trusty 5.75 estimate of replacement level ERA, we can compute each pitcher's contribution in runs above replacement. I've only tackled the starters, because I didn't want to deal with reliever leverage. And I've translated tRA onto the ERA scale by multiplying by .92. The ERA scale is just more intuitive to most people. (Thanks to Colin for suggesting the multiplication method instead of simply subtracting .40 runs.) And I've added in .30 runs for NL pitchers because of the lack of DH.
On to the numbers. Here's what the American League leader board looks like for 2008:
- No surprise who the top two finishers are. Halladay probably deserves a bit more respect than he got, though. And when you consider he pitched against tougher hitters...
- Ervin Santana's a great example of a pitcher who figured something out. His career ERAs haven't been that great, and his peripherals always supported that fact. But then this year he goes and sets career bests for K-rate (8.8 per game), BB-rate (1.9 per game) and GB-rate (39%). The question headed into 2009 isn't whether or not his 2008 ERA was a fluke, it's which skillset will show up.
- Ah yes, then you've got Javier Vazquez and AJ Burnett, two pitchers with histories of posting ERAs worse than their peripherals would predict. These two pitchers might very well be examples of guys who give up batted balls that are harder to catch than the average pitcher's batted balls. One season of data certainly isn't enough to make conclusions about individual pitchers, but five seasons should be. Someone should tackle that study. Or we can just wait and hope for Hit f/x.
- Here's the next ten guys in the AL: Andy Sonnanstine, Gil Meche, John Danks, Scott Baker, Mike Mussina, Kevin Slowey, Andy Pettitte, Jered Weaver, Felix Hernandez, Matt Garza. That's three Rays in the top twenty, and none of them Scott Kazmimr. Three for the White Sox, too.
Let's switch over to the National League:
- Once again we see the "obvious" Cy Young choice leading the pack, but only by a nose over the number two pitcher. Brandon Webb deserves more attention, mostly because his home ballpark doesn't get enough recognition as an extreme hitters' park.
- Looking at Dan Haren's 2008 numbers, one has to wonder if Billy Beane regrets that trade at all. Sure would be nice to have him for the 2009 playoff push, no?
- Derek Lowe: stud. If he'll take a three year deal for $15MM per season, it's a steal for whoever signs him. And if Randy Johnson will take anything under that for a single season, that's another steal, even for a 45 year-old pitcher.
- If you combine CC Sabathia's stints in the AL and NL, he comes out at 81 RAR, number one in the majors.
- The next twenty in the NL: Chad Billingsley, Hiroki Kuroda, Roy Oswalt, Edinson Volquez, Ubaldo Jimenez, Ricky Nolasco, Ted Lilly, Aaron Cook, Jair Jurrjens, and Carlos Zambrano.