Today’s Opening Day — at least, from my Mets fan perspective — so baseball is front and center in my mind.  My e-migo Sky Kalkman tweeted this out earlier today:

I replied with the below and we got to talking.

So here’s my case for the six man rotation.

The caveats:

  1. Starting pitchers typically throw bullpen/side sessions at some point during the week when they are not scheduled to pitch. I’m assuming that the point of these sessions is to keep loose during the season, as a five day layoff between starts seems long. If there are things that need to get done during these sessions — e.g. work on mechanics, test out new pitches — my plan probably won’t work.  (Sky’s investigating that right now on Twitter. It doesn’t look promising, but oh well.)
  2. The team needs the right personnel, which, of course, can be designed for. What you ideally want is six SPs with a relatively small talent gap between the best and the worst. The 1998 Mets are a good example here (once Armando Reynoso came back from injury) and, in fact, are the team I was thinking about when I came up with this years ago. If your sixth starter is Jeremy Hefner or Chris Schwindin, this is a bad idea, as you don’t want to give them Major League innings unless you have to. 

How it works:

First, add a sixth pitcher to your rotation. Then, on each SP’s “bullpen day,” instead of throwing that session in the pen, he throws it in live action, taking up 1 IP from the bullpen.

In a 162 game schedule, you’re reducing each SP’s workload from 32.4 starts per season to 27, a net loss of 5.4 starts. At 7 IP per game — which is way too high — each SP loses about 38 IP. All of this, of course, goes to the sixth starter.  That’s the typical case against a six-man rotation — why take away 38 IP from guys like Johan Santana, R.A. Dickey, and Jonathon Niese… and then give them to Jeremy Hefner? It’s nonsense.

The bullpen sessions, in part, change the math. Each of the six starters would pitch 27 relief innings a year; together, they’d account for 162 relief innings pitched. And I think it’s safe to assume that each of the five current SPs are better than the bottom of the bullpen. Heck, even Hefner is probably better than the 2012 equivalent of D.J. Carassco and/or Pedro Beato.

But that’s not why this system may work. The real savings come in (a) roster spots and (b) player development.

Roster Spots

In 2011, Miguel Batista (30 IP) and Chris Schwindin (21 IP) combined for about 50 IP. Let’s assume that those two guys didn’t overlap (even though they did) and use them as a proxy for Jeremy Hefner. Basically, in order to get the sixth SP on the roster, you need to lose your swingman RP. Or whatever — you’ll see it really doesn’t matter.

Switching to the six man rotation buys you 162 IP — your SPs still make 162 starts, each of which at their average IP/game (and perhaps more due to the extra rest?), but they also give you 1 IP in relief per game, or 162 extra innings pitched. Subtracting Batista and Schwindin’s 50 IP leaves about another 110-115 innings pitched. 

In 2011, Pedro Beato (67 IP) and D.J. Carrasco (49.3 IP) combined for just over that amount innings pitched. And they sucked. Beato was worth -0.3 WAR per FanGraphs and Carrasco was worth -0.5 WAR. And with this system, we can remove them from the team entirely. Their innings are already accounted for.  So we gain 0.8 wins *and* clear two roster spots. 

While the Beato/Carrasco Mets are an extreme case, perhaps, the fact remains that many RPs are basically fungible, and this is doubly true for bottom of the pen ones. They’re on the roster to eat up innings more than anything else, so getting rid of them feels like a win. What can you do with those roster spots is beyond the scope of what I’m writing, but there’s room for experimentation

Player Development

Keeping with my Mets-as-example theme, the Mets have three guys SP prospects who, a Tommy John surgery notwithstanding, have a very good chance at seeing time at Citi Field this year: Matt Harvey, Jeurys Familia, and Jennry Mejia.

In general, young guys like these are better than the Hefners of the world and often better than the Dillon Gee and Mike Pelfreys, too. It doesn’t make sense to rush these players into a relief role in order to get some small marginal gain (ahem), and in many cases, young starters are on strict innings/pitch counts which make it very hard to dedicate a full season of starts to them. You know, the “Joba Rules” or Stephen Straburg’s 160 innings limit.

The sixth man SP spot with the bullpen boost may meet this unique spot on the prospect/MLer see-saw balancing act. Drop the bullpen duties, and the sixth guy makes 27 starts for a total of 162 IP (at 6 IP/start). You can easily find a swingman type of RP to take his bullpen role and keep the prospect in a light-use, full-time SP position. And because of the additional roster spaces (plus the ability to expand to 40 guys in September), there’s a ton of flexibility here. 

If Mejia were healthy and if the Mets weren’t awful, this would make a ton of sense for them. It would probably mean that the team would forgo signing Jon Rauch (saving $3.5 million) and keeping Miguel Batista in AAA, all while actually improving the team’s pitching, opening up a roster spot, and furthering Mejia’s advancement. And this would allow the team to retain useful-but-tertiary parts like Nick Evans, Josh Satin, Adam Loewen, etc., or perhaps add a second LOOGY.

Open questions:

1) Can this be tuned in a way where top starters are exempt from it — that is, they go every 5th day, regardless, and perhaps skip the bullpen duties all together?

Last year, Justin Verlander averaged 7.38 innings per start in 34 GS — which, in a five man rotation, means that he accounted for his share of a 170 game season. And, of course, he put up a Cy Young Award-winning performance (with a 170 ERA+).  His teammate, Brad Penny, averaged 5.85 IP/GS and sported a 77 ERA+.  Let’s assume that it makes sense to get an Andrew Oliver or Jacob Turner into the rotation, in part because you’re reducing the reliance on guys like Penny.  Is there a way to do it so we do not reduce Verlander’s use?

2) Is there a way to build an organization around this model?

Building a team around it seems hard. But what about a whole organization? It seems necessary — do you really want your SPs to be doing the traditional five-man rotation one year, this the next, and then… who knows? 

This is a hard question to answer because it’s unclear what the right makeup of personnel is. Do you want a bunch of replacement-level, C/C+ prospects to round out the back of the rotation, hoping they can turn into 1.5 WAR guys? Do you have to maintain an annually-replenishing pool of Mejia-types? Who knows.

3) What do you do with the extra roster spots?

The Mets could add Adam Loewen and Vinny Rottino. Grrrrrreeeeeeeeeeeeat.