There was an effort to discuss this topic awhile ago, but I thought I would pick it up again now that the rumors and suggestions are increasing about what to do with the existing pitching staff, both starters and relievers.

I'm not someone who lives and breathes Baseball Reference, but I think a strong statistical analysis is the best route to take in considering both 1) what the front office is thinking; and 2) what the needs and holes are with the existing roster.

So let's assume that in a 162-game season, you will have 35-40 games a year that you will lose on the road without having to pitch a 9th inning, but you will have a dozen extra-inning games that will stretch the innings. Last year, we had 1443.1 innings pitched in the season; we know (from painful experience) that a handful of those will be pitched by position players (we had 3.2 last year), but given the numbers, let's ignore that for now. We need to fill 1,443 innings.

My math (also from the Twins page on Baseball Reference) shows that 907 of those innings (thereabouts, as I didn't try to winnow out how many innings that Gabriel Moya threw as a starter rather than reliever, or several others, but spitballed it with those types of players) were pitched by starters. That averages 5.2 innings per start, and whether you use the opener strategy or a "traditional" starting pitcher, (then calling the "starter" the primary pitcher rather than the opener) you are hoping to get something close to that out of your primary/starter.

That leaves room for optimism, as we can hope that the slate of starters picks it up a notch (whether that is 5 guys or 15 over the course of a season), but let's assume they don't, that we need 536 innings out of the non-starters. My quick South Dakota math tells me that is 3 1/3 innings per game.

I am also going to assume that we will roll with a 13-man pitching staff, unless the starters are on track to bump their innings to a level where the remaining innings can be handled by 7 guys (and that doesn't appear to be the case, but if we get to a point where starters are putting in over 1,000 innings combined, we might). How can we best divide innings for our 8-person staff so as to make that possible?

1) Simple but wrong - that is 67 innings per reliever per year (and no, I don't assume that the same 8 guys will man the posts all season, but perhaps like a second unit in basketball, the substitutes will take the innings/minutes not played by the first string). 8 guys getting 67 innings per year means that within a 26-week season, each relief pitcher would pitch roughly 3 innings per week.

2) More specialized - A look at the "typical" workload of a closer shows that they average pretty close to that number - 65-75 innings per year, maybe 30-40 of which are in closing situations. I don't expect the Twins to vary significantly from that role, even if they don't have that player on their roster right now (I think they do in Trevor May). But for purposes of this analysis, let's say we do have a closer (whether May or Allen from the Indians or someone else, that can be debated) and that person pitches 70 innings.

3) Who does what? That leaves us with 466 innings. I think we can pencil in Reed (for all of these guys, let's just say "assuming health" without really assuming that all of them will stay healthy for the year) Hildenberger, and Rogers for 70 innings each. That's not unreasonable, given their traditional workloads, again citing this page: https://www.baseball-reference.com/teams/MIN/2018.shtml#all_team_pitching.

Now we're down to it. That is 280 innings pitched by the core four relievers, leaving 256 remaining innings for 4 pitchers. The names of those now-existing pitchers includes Romero, Stewart, Gonsalves, Moya, Littell, Slegers, Vasquez, Curtiss, DeJong, Duffy, and Magill (I'm assuming for our discussion that Mejia is our 5th starter, but you can change names if you want to).

While only 4 of them can be on the roster at once, many of the remaining names have options and can be shuffled between leagues as need arises.

4) Opener strategy - if the team is indeed - as has been hinted - considering an opener strategy, it seems that Stewart and Gonsalves might be the primary candidates for those roles (lining up with Mejia and Pineda, possibly Odorizzi). If we utilized them for the typical 30 starts that your #3 through 5 starters get, and hoping for 2+ innings - once through the order and averaging a little over a hit and walk per inning means that you face 9 batters while getting roughly 6 2/3 outs or 2 1/6 innings per open). 2.16 times 30 equals 65 innings, so 130 innings with two openers.

5) We have now defined several spots (which can be intermingled or altered depending on performance), but we allot 907 innings to Berrios, Gibson, Odorizzi, Pineda and Mejia; we allot 280 innings to May, Hildenberger, Reed and Rogers; we allot 130 innings to Stewart and Gonsalves.

That leaves - ta-daa! - 126 innings to be covered by two remaining pitchers with Romero and Vasquez and Moya as your primary candidates, the third of them stashed at AAA and the remaining names, plus potentials in Tyler Jay, Jake Reed, et al lined up behind. That strikes me as do-able, especially if Romero emerges as the long man to back up short starts from Berrios, Gibson or Odorizzi, and Vasquez/Moya as the short-stint lefty.

6) Now who do you want? Once whittled down to innings and performers for those innings, any discussion of a reliever or starter should also include the ramifications of what happens to the pecking order and opportunities for the above players. If we added a 5th starter, for instance, a Wade Miley (no offense, but please don't) and he took up his share of those 900 innings, either Mejia is bumped back into the core of 4 relievers, or one of the opener positions or one of the two remaining slots. Who should be bumped? Knock out Gonsalves for Mejia and leave the rest the same? Knock out Stewart? Move Mejia into Hildenberger's innings, Hildenberger into Moya/Vasquez's innings? Do we consider not giving Romero an audition in the regular season?

While I think the strategy of strengthening the bullpen is a sound one, I get far more lost in the thinking process when I follow the logical chain and try to decide what this means for our chances of promoting from within or giving real opportunity to a slew of candidates.

Best guess - no more starters will be signed (other than minor league contracts with an invite in case someone intriguing slips through the wintertime cracks), 1 more dependable reliever, only one opener (either Stewart or Gonsalves) to pair with Mejia, the other to start at AAA, plus Hildenberger, Rogers, May and Reed. Remaining two slots are taken by Romero as more of a long man and either Moya or Vasquez as the 8th arm/lefty. I think you also have to cut bait with some of the remaining names and keep your roster alive with players who could be optioned as the season progresses, even if that player isn't the one causing the gaping void. The more you fill this roster with veterans and guaranteed contracts, the more you make that flexibility impossible and are forced into waiver/release positions on players. I'd rather be in a position to take on someone else's unfortunate waiver casualty than be forced into one of our own.

Happy holidays everyone!