I was saying this offseason that I want us to elaborate a new relief role: guys who face exactly 9 hitters. These guys should be easy to collect, because while they need stamina to go multiple innings and some sort of approach to lefties and righties, they don't necessarily need the full repertoire that a starter would need. It could also be a good role for breaking in young candidates for the rotation.To the extent this is a leaguewide issue, one thing you would think more teams would be developing, especially smart teams like the Red Sox, since the change in starters going deep are the sort of classic "swingman" role. Guys who can go once through the order, 2-3 IP, 2-3 times a week and fill the gap between the 5th and the 8th or 9th without using everyone in your pen. Michael King did this last year for the Yankees, but by the end of the year he was more of a starter. Maybe there is evidence that these guys are going to blow their arms out or teams just can't find them but it would seem like that kind of player would be insanely valuable especially if you can get more than one on your roster. For a long time we tried to fit players into either starter or short relief buckets, and long relievers would always just be failed starters -- what if there are guys who are best somewhere in the middle? But it seems like we're still just developing one or the other, except now starters are only going 2X through the order, and like SJH says we have to use 4 or 5 relievers every night to get through the game, greatly increasing the chance that one will not be good that night.
But instead of having one, they'd be fully half of your bullpen. You'd have a 13-man pitching staff with a 5-man starting rotation, and then four of these once-through-the-lineup guys, and four short relievers: two firemen, specialists in getting hot fast and coming in with men on, and two late-inning guys, set up and closer. The idea would be that your starters will often only be getting the lineup twice, so everyday you'd have at least one (or ideally two) of these guys in the pen adequately rested to take one turn through the lineup and aim to bridge you from your starters to your late-inning guys.
With our current staff, this would break down something like this. Obviously the innings allotments are per roster spot, not per pitcher; there will be injuries and depth called up to soak some of that up:
SP – ~32 games/rotation slot; 160 IP/slot
Bello
Crawford
Pivetta
Houck
Whitlock
Firemen – ~80 appearances averaging < 1 IP; 60 IP/slot
Weissert
Rodriguez
Once-Through-the-Lineup – ~40 appearances, averaging > 2 IP; 100 IP/slot
Winckowski
Slaten
Bernardino
Anderson
CL/SU Single-Inning Guys – 70 appearances x 1 IP; 70 IP/slot
Jansen
Martin
That sums to 1460 IP, or an average of almost exactly 9 IP per game — hopefully the ninth innings not pitched in road losses counteract the extra innings pitched in ties.
So if Crawford goes 4 1/3 or something, then one of the firemen (say, Weissert) comes in to wrap up the fifth, before Winckowski or Slaten (or whoever is up in the rotation) comes in and faces 9 guys, which hopefully gets us through the 6th and 7th. Then it's Martin in the 8th and Kenley in the 9th if we're ahead by a few runs, and whichever of Winck and Slaten didn't already pitch if we aren't.