clock menu more-arrow no yes mobile

Filed under:

A Modest Proposal for the Fifth Starter Spot

Somewhere, John Maine is still trying to make his comeback despite the aches and pains he reports after every start. Color the average Mets fan unimpressed. Of course, Hisanori Takahashi can make a decent starter, as outlined here in a pitch f/x piece I did for Bloomberg Sports, but I also think that the ideal solution is beyond the imagination of the people in charge.

So let's make a modest proposal that shouldn't be of Swiftian proportions, but probably will be given the state of innovation in baseball. (Really, what were the innovations over the last fifty years in pitching? Lower pitch counts, one-inning relievers, and 'closers?' Do we really think that all arms come in either the 'six innings every five days' or 'one inning every day and a half' variety? What about three innings every three days? It does not seem far-fetched to question if pitchers really come in two different shapes and sizes.)

Given this preamble, let's make the proposal. Once Maine is "healthy," why don't the Mets try a six-man rotation. No, not one where they feature six different starters, but more one where the team has four starters, and then platoons Maine and Takahashi in the fifth-starter role. Platooning starting pitchers has never been done, but let us lay out the reasons that it might not be a bad idea in this particular* case. (*(c) Jerry Manuel)

First, why these here arms may well be suited for the job. Richard Gross over at my old stomping grounds,, points out the following about Takahashi:

Opposing hitters are already warming up to Takahashi as he moves through the lineup two or three times. The first time around batters are hitting just .250. The average goes up a pinch the second time around to .268 but by the third time, it jumps significantly to .346. To further bring the point home, the Florida Marlins have seen Takahashi on four separate occasions this season: three times in relief and once as a starter. The third and fourth go-arounds against the Marlins have resulted in a cumulative 7 earned runs through 8.1 innings.

Those numbers for Takahashi are currently .224/.269/.286, then .245/.288/.405 and then a whopping .355/.394/.484 the third time around. The same trend holds true for Maine this year. Batters hit .305/.414/.424 (!) the first time through, .293/.414/.483 the second time around, and then turn into Barry Bonds the third time around with a .293/.348/.659 line. If we go for a larger sample size, the same trend holds true but the numbers look better: .217/.294/.348, then .229/.325/.402 the second time, and .275/.358/.480 the third time around.

Okay, so this isn't rocket science. Bullpen arms almost always benefit from shorter stints, and Maine in particular struggles to maintain his velocity and should gain from seeing fewer batters. How about the effect on the staff? First, a note on how this would work. Starters have to be set up to pitch every fifth day, but they also have a throw day, so each guy (Takahashi more likely) would be available an extra day of the week. By carefully working the rotation, both guys could be available for an inning, say, on Jonathon Niese's days in case he has a poor outing. They wouldn't necessarily be limited to pitching three innings each every other start.

But let's say they only pitch three innings every other start. Over a year, that would mean you are devoting two pitchers to about 95 innings instead of aiming to have one starter with 180+ innings. How badly would that effect your pen? Let's say the current rotation would have a guy with 140 innings like Maine last had in 2008 - could it survive with a bullpen pitcher pitching 'only' 40 innings in the pen? Um, yes. Here are the pitchers that were in the pen last year and didn't amass 40 innings: Fernando Nieve, Ken Takahashi, Elmer Dessens, J.J. Putz and Lance Broadway.

Or, to put it another way, the Mets' pen put in 501 innings last year, and their starters 924.1 innings. By shifting 50 innings (the 50 innings you 'lose' by targeting only 90 innings from Maine instead of 140) to the pen, they'd be asking six pitchers to put in 550 innings. That sounds bad - the other guys in the pen would have to come up with 90 innings. However, those 924.1 innings by the starters last year was only 5.71 innings per start. If you have Takahashi and Maine platoon, could they average seven innings per start? That doesn't seem unreasonable. Take those 1.3 innings, multiply them by 35 starts, and you save the pen 45 innings. Well, heck, that's easy math. You save the innings you lose if the two of them can average seven innings per start together.

There's yet another way to look at this. The five rotation spots on the Mets averaged 184.82 innings last year, and the six bullpen spots 83.5 innings. Losing those 84 innings might hurt, but that's assuming you can't use either guy on throw days. Getting 40 innings from each over the course of the season would be asking them to throw an inning and a third between each 'start.'

Would this murder Jerry Manuel's precious flexibility? The six remaining pitchers in the pen might be asked to handle a little more, or maybe they wouldn't. By using Takahashi (and maybe Maine later in the season) occasionally on their throw day, you should remove that caveat. Yes, this approach would require more work and more effort with the schedule. Yes, it would require an in-depth look at the mechanics of pitching and the possibility of injury - which means it won't happen this year.

Instead, one of the two will go to the pen to be abused by Jerry Manuel. And that pitcher will probably rescue the other pitcher for multiple innings very often. So, in effect, this may actually happen, via the 'long reliever we will save for the days that our most-likely-to-explode pitcher is on the hill' method. But they won't call it a platoon, and they won't take up this modest proposal in name. But the whole situation is worth daydreaming about on a light-schedule day.