I've been waiting to put together this look at the Nationals rotation for a little while now. What I'm doing here is splitting up our pitching staff into a theoretical rotation, allowing us to see the rough performance of our pitchers by slot in our rotation. I am working off of the methodology used in this article by Jeff Sackmann of The Hardball Times. Let me start by introducing our 2010 Nationals starting pitching staff:
Creating a Rotation
In a perfect world, each of five starters in the rotation would pitch about 32 games each. What I am going to do is create a #1 starter out of our best performing pitchers, then move down the list until we have a theoretical pitcher for each slot. The core of our team's #1 starter is naturally going to be Stephen Strasburg. He only accounts for 12 of the 32 games we need, so we will mix in Matt Chico's start and 19 of Livan's starts for a theoretical #1 starter with a 3.36 ERA in 32 games. The rest of Livan's starts form the core of our theoretical #2 starter, and so forth. (note - I am discounting Batista's start, awesome as it was, since he was not a regular starter)
As you can see, the Nationals staff was anchored in 2010 by a #1 slot that put together an ERA of 3.36, which is equal to the ERA of the entire pitching staff for the World Champion San Francisco Giants. Unfortunately, it was all downhill from there for us. In his article, Zuckermann calculated league averages for each slot in 2006, and our pitching staff ended up with numbers very close to what he found to be league average four years ago.
Ups and Downs
The bright spots in the rotation appear to be young arms that struggled to stay healthy, like Strasburg, Detwiler and Zimmermann. They will all have chances this year to work their way back into the rotation. J.D. Martin is stuck in AAA with a glut of pitchers ahead of him despite his role in holding down the #2 spot.
Rizzo has weeded some weaker pitchers out of the rotation. Both our #4 starters are out of the picture at this point, with Stammen sent to the bullpen and Atilano back in AAA. Of the three #5 starters, however, only Scott Olsen has lost his job.
The 2011 rotation
Let's take a quick look at the 2011 starting rotation and see how it resembles what we used last year. At the top, we have our #1/#2 pitcher Livan Hernandez back. The pair of #3 pitchers, John Lannan and Jordan Zimmermann, are also a lock. #5 starter Jason Marquis is slated to pitch regularly as well, while the last slot is a battle between #5 Maya and #2 Detwiler to start the season.
If we use the most likely scenario via conventional wisdom, the 2011 rotation appears to be made up of #1/#3/#3/#5/#5. The Nationals have lost a pitcher's worth of #1 or #2 production, and appear to be relying on #3 and #5 pitchers and one aging veteran in Livan Hernandez.
Any one of you may now be pointing out that last years ERA is not reflective of what many of these pitchers can contribute, given the plethora of injuries to our staff in 2010. For example, Zimmermann is expected to be stronger this year. It is also unlikely that Jason Marquis will put up a 6.60 ERA over a full, healthy season and probably projects to be more of a #3 or #4 starter. On the flip-side, nobody expected Livan to put up #1 starter numbers last year, and it is doubtful that he will repeat that performance this year. What can we do to predict our rotation's performance in 2011 more accurately than by looking at last year's ERA?
In Part 2 we will go through this same drill again, but I will be crunching the numbers by xFIP rather than ERA. xFIP is an ERA related stat that tries to weed out things like fielding and that dreaded factor called luck. It is accepted to be far better for predicting future performance. I can already promise that this will turn some of what we discovered today on its head.