thoughts on the Mariners, MLB draft, and more homelinksdraftabout me

Rethinking Pitching Staffs

I ran across this post on FanGraphs this morning talking about fifth starters, and how they are mostly mythical. Not as in teams don't use them, but as in it's highly odd for anyone to man the position for the majority of the year. Heck, with the post's cursory look at 2009 rotations, consistent fourth starters are hard to find too.

One thing I've wondered the past few years is why teams are so adamant on five-man rotations. Four-man rotations worked for a long time in baseball, and if it is that hard to find a fifth starter, why not consider them again?

My idea is more of a four and a half man rotation. Every effort is made to keep the top three or four starters in rhythm, meaning they pitch every five days. Thanks to off days, it should be theoretically possible to skip starters from time to time and keep everyone else in regular rhythm. The drawback is that the starter skipped doesn't appear for over a week.

There is a pretty easy solution to that skipped starter predicament though. They become an extra bullpen arm for a few days. That comes in handy from time to time too.

What kind of difference would this approach make?

In a 162-game schedule, if a team never breaks their five-man rotation, the starts will break down as follows:

  1. 33 starts
  2. 33 starts
  3. 32 starts
  4. 32 starts
  5. 32 starts

Taking the Mariners 2010 schedule, and keeping the starters in order, but skipping the fifth starter when possible, yields the following breakdown:

  1. 34 starts
  2. 34 starts
  3. 34 starts
  4. 33 starts
  5. 27 starts
Taking the Mariners 2010 schedule again, and then going starter-by-starter, starting each of them every fifth day as frequently as possible, yields this breakdown:

  1. 36 starts
  2. 35 starts
  3. 34 starts
  4. 32 starts
  5. 25 starts
The drawback of this last one is that the rotation gets out of order for most of the middle of the year, but interestingly enough it gets back on track for September.

Of course, this is all in an ideal fantasyland too, where starters never get hurt. A vast majority of rotations face injuries, and when those happen, things get messed up. Usually the hiccups in the road force re-shuffling, and often great starters pick up a few extra starts thanks to the process.

Situational relievers have revolutionized the bullpen, and if something is going to revolutionize it again, it will be a re-conceptualization of starting pitcher use. Clearly, it isn't that difficult to concentrate starts within four pitchers without ever making any of them start on short rest.

I would like to see a team try something radical: create a "swingman" role on the pitching staff. What if a team featured four starters, three swingmen, and four relievers? The starters would be traditional starters, the relievers traditional back-end relievers. The swingmen would provide long relief and spot starts in the fifth spot. On any given time the fifth spot came up, it would be the duty of the all the swingmen to get through the first six to seven innings.

On this theoretical pitching staff, it would be possible/expected for a swingman to only go up to three or four innings at a time. That would allow him to return to the bullpen quicker, probably after only a few days of rest, meaning the bullpen shouldn't get overexerted too badly. Furthermore, from a competitive standpoint, it asks a fringe starter to only go through a lineup one or two times. Guys often end up in the bullpen, or fifth rotation spot, because they aren't that effective multiple times through a lineup. This model could possibly squeeze a better performance out of the fifth starter than any one of the individuals could accomplish on their own.

Injuries to a pitching staff could be easier to deal with too. If a starter gets hurt, one of the swingmen takes the slot, and it won't take them long to stretch out. It would be an adjustment, but not a radical one, especially because they have been facing major-league hitters already. Then, a player called up from AAA could take a swingman role, meaning the manager could keep them stretched out to a certain extent while also picking places to give them a (hopefully) soft landing. A similar process could work for short reliever injuries too, because it wouldn't be a quantum leap for a swingman to scale down from three to four innings of work to just one.

The swingman pitching staff model is just a thought experiment of mine, but I think it could work well. At its best, it would concentrate innings pitched among the best pitchers on the staff, while also clumping innings among the fringe guys on the staff in such a way that should improve their effectiveness. At its worst, it may overexert everyone. However, the problems largely would come down to pitcher use, which is a problem any bad manager faces already.

If the swingman model was used by the 2010 Mariners, I would make the four starters Felix Hernandez, Cliff Lee, Ryan Rowland-Smith, and Ian Snell. The swingmen would be Jason Vargas, Sean White, and Kanekoa Teixeira. The relievers would be Shawn Kelley, Brandon League, Mark Lowe, and David Aardsma. When Erik Bedard comes back, I would likely put him in the swingman role, and consider bumping him up to the regular rotation if he finds his command quickly.