Tuesday, April 18, 2017

An all-middle-reliever pitching staff

About the time Kyle Gibson was getting pulled from yet another of his "is that all there is" starts Monday -- 5.2 innings, eight hits, three runs -- somebody in my Twitter feed opined that he'd make a fine middle reliever.

A few innings later, Tyler Duffey wrapped up his day's work: 2.2 scoreless innings. Duffey, of course, was a starter last year for the Twins and now is in their bullpen. He has now worked 8.2 innings this season without allowing a run.

Would Gibson be better out of the bullpen than as a starter? Probably, if only because almost everybody's stuff is better in short bursts.

All of which led me once again to one of my favorite outside-the-box ideas: A pitching staff without traditional starters.

Imagine a nine-man staff, divided into three groups of three pitchers. Each pitcher in each group works three innings at time, and the groups form a three-day rotation. In theory, you'd have nine pitchers making 54 appearances and throwing 162 innings apiece.

In theory. This outline assumes no extra-inning games, no scheduling headaches such as doubleheaders, and -- perhaps most unlikely -- that every one of those 486 appearances are successful enough that nobody needs to get pulled before completing his three-inning assignment. And history suggests that 160-plus innings is a career-sapping workload for a reliever.

So nine pitchers probably aren't enough to make this work. Maybe 11 or 12 is. (Even 12 would be fewer pitchers than the Twins are carrying now.) But there are other problems.

Pretend that the Twins decided to do this. Are veteran starters like Ervin Santana going to be happy working three innings at a time? Do you want to pay Santana and Phil Hughes $26 million combined for 162 innings?

And if you trade your established starters so you can commit to this radical idea, what happens if it fails? Answer: You get fired. And your organization is probably set back for years.

Back in his final years in Oakland, Tony LaRussa experimented with something like this idea, and he abandoned it about two weeks in. It probably takes a manager of LaRussa's stature to try it and not be immediately crucified by the media or abandoned by his players.

Or an expansion team, if we ever see one of them again. The roster compiled by an expansion team should be filled with pitchers willing to take any role to be in the majors, even that of a three-inning starter (who can get a loss but not a win).


  1. I think a hybrid approach would be worth considering. One could have three bona fide starters (each taking their turn every 5th day), five or six long relief/failed starter types working 3 or 4 inning shifts, and then three more traditional relievers for one-inning & high-leverage situations. This would provide a team with the flexibility to experiment a little, while keeping its best pitchers happy in their more traditional roles.

  2. I have been advocating for this for 3 years. I think there is no question it would reduce your team ERA, probably by half a run. My version would use 6 starters in three teams of two. Each starter would aim to pitch 3 innings, 4 if their pitch count is really low, every 3rd day. So one starter would go innings 1-3 and the second innings 4-6. Innings 7-9 would be for the bullpen to share. Alternatively, you could have two groups of 3 starters, where each pitcher goes 2 innings every 2nd day. Again that gets you innings 1-6, and then you use the rest of the pitching staff as a bullpen to handle innings 7-9. My plan differs from Edward's in having a bullpen handle innings 7-9 rather than planning on starting pitchers doing that.