Unless it can be proven that a 6 man rotation keeps your starters healthy, I'm not sure it will gain traction. Until every team goes to a 6 man rotation, wouldn't the 5 man rotations have an advantage, in that the 6th starter would be facing the 1st and 2nd starters quite a bit? Seems like that would be on the case without delving too deep into the details.

Please correct me if I'm wrong on that.

I still don't understand why teams don't just mix up the back end of the rotation in situations like this. What I mean is, if you have 2-3 good, dependable starters, have them pitch every 5 games, then fill in the rest of the starts with your other 2-4 starters as needed. Wasn't that often done in the past?

I really doubt the six-man is a trend that will take hold. Teams scramble to find five guys to fill out the rotation many seasons. On top of that, it seems most pens are filled with actual, permanent relievers, so using a sixth starter would require bringing someone up. It just doesn't seem there are any consistuencies calling for its use.

Are the White Sox even planning to stick with six now that they've dumped Jackson?

I still don't understand why teams don't just mix up the back end of the rotation in situations like this.

The Brewers tried something like this briefly a few years ago. They tried a home/road platoon with 2 pitchers (IIRC Bush and McClung). I think it lasted a few weeks.

The Sox are the only team that tried this. Teams with six "starters" pay lip service to it so that no one feels bad. Then they put someone the DL for sucking (Kyle Davies) or move someone to the bullpen.

Rob Neyer: Is The Six-Man Rotation Here To Stay?

no. next question.


there aren't enough roster spots available for a team to commit to carrying 6 players who won't impact the game 5 days out of every 6.

@AO, the same argument could have been made about going from 4-man rotations to 5-man. Yet the perception of advantage was there.

As Randy notes, the 4-man rotation was a bit of an aberration anyway. The dominant 20th-century doctrine was to run out your better pitchers on 3 days' rest. When there were more doubleheaders, this led to de facto five-man starting staffs with long relievers and swing men getting into the picture. There have never really been a lot of pitchers who could go consistently with three days' rest and be really good at it. So flexibility was the order of the day in the 20s, 30s, 40s, and 50s. Also, during those years, the top starters would frequently pitch in important relief situations on their middle day of "rest."

The four-man rotation was seen in its Platonic-ideal form on the 1966 Dodgers. The regularity of that rotation is somewhat eerie to look at on B-Ref. Four starters went in lockstep order, and had only three relief appearances among them; the fifth starter was a guy named Joe Moeller who threw mostly in relief and had just eight starts, five of them in doubleheaders. There were three factors making this possible: air travel, home games in a place where it doesn't rain, and most important of all, the fact that the four rotation guys were three Hall of Famers plus Claude Osteen, and the next-best starter was Joe Moeller.

If you look at a more normal pitching staff, the difference between the fourth and fifth starter can be a toss-up. It makes little sense to bench the fifth guy when working him in will give the staff more regular work, longer and more predictable rest, and just as good a chance to win. And the larger rotations get, the less difference there is, on average, between the worst starter and the best starting prospect. I daresay there are lots of teams right now with a purely arbitrary difference between their fifth starter and the sixth-best starter in their organization.

Teams don't mix up the work anymore because the value of predictable regimens is great psychologically and probably physically as well. Managers of the Stengel era used to play a guessing game, assigning work erratically, sometimes not even naming a starter till they'd slept on it (the famous "ball's in your glove, you're starting today" tactic). It sounds cool, but it might not really be an optimal strategy.

It's been a long time since I enjoyed one of Rob's articles.

Please correct me if I'm wrong on that.

You are wrong on that, in that rotations rarely "line up" after the first week, anyway.

You know what is good strategy? Take your best starter and give him fewer starts. Then do the same with your #2 and #3 starter. While you are at it, bench your best hitter twice a week....you don't want him to wear down. And replace your centerfielder with your backup catcher once a week....that way he's not offended at only being "just a backup catcher".

If ever there was a place to try a mix and match it would seem to be the Yankees this year.
Sabathia starts every 5th day. Colon, Burnett, Garcia, Hughes and Nova in rotation behind him, with Hughes/Nova getting skipped to keep Sabathia on schedule, and serving as the long man in the bullpen in between...Keeps the best pitcher on the roster getting the most innings, gives extra rest to colon and garcia, and gives you the rest of the season to evaluate whether hughes can get back to where he was/nova can survive long term. If any of them really step up or flame out during that time, you can adjust and go back to a 5 man, with the loser of the battle going to the pen or AAA. Main issues would seem to be that you need to get Hughes some more innings than he will get this way, and that Burnett turns into even more of a head case without a more rigid routine.

The Sox are the only team that tried this.

Last turn through, the Yankees started Sabathia-Hughes-Burnett-Colon-Nova-Garcia. Though that might just be due to the doubleheader.

As for the White Sox, looks like a temporary situation. Their 6 have accounted for all 107 starts by the team, but Edwin Jackson is gone. They aren't planning on bringing a 6th starter up from the minors, are they?

It's a rare situation, as 5 of the 6 are having above average years by ERA+, and the one who isn't, Jake Peavy, has a big contract, a great FIP, and track record to keep getting him chances. Few teams are going to have 6 good starters at one time.

It's been a long time since I enjoyed one of Rob's articles.

Yeah, this is Peter King level of complete over-reaction to one team doing something different and assuming the rest of the league will follow suit.

In addition to taking your best starters and pitching them less as #10 points out, this would increase bullpen usage, since the 6th starter isn't going to pitch as many innings as the guys above him would. Otherwise he wouldn't be the 6th starter. Say the Phillies went to a 6 man rotation, every start that the 6th starter takes away from Halladay or Lee, instead of the bullpen only pitching one, maybe two innings they're pitching 3+.

Unless it can be proven that a 6 man rotation keeps your starters healthy, I'm not sure it will gain traction.

Lack of evidence it helps anything didn't stop a five man rotation from gaining traction.

I can't see the six man rotation gaining traction unless rosters are expanded to 26. Six man rotations coupled with the seven man bullpens that have become the norm leave you with just a backup catcher and two reserves in the AL.

Say the Phillies went to a 6 man rotation, every start that the 6th starter takes away from Halladay or Lee, instead of the bullpen only pitching one, maybe two innings they're pitching 3+.

The effect wouldn't be quite so pronounced, since each appearance by the sixth starter wouldn't be taking a full start away from the front line guys, but rather 1/5th of a start (since your #6 is just going to pitch the day before your #1 and not instead of him).

Still, it's a terrible idea.

Over 30 games with a 5 man rotation, every starter gets 6 starts. Over 30 games with a 6 man rotation, every starter gets 5 starts. Figure that you're adding just an inning or two to the bullpen's workload on the days when your #6 is out there, and over 30 games maybe that's another 10-15 innings total (if you figure it's 2-3 innings per start made by the #6 guy), and 50-75+ over the course of season, conservatively. The flip side is that you reduce your starters' workload by probably 5 or 6 starts over the course of the season, which isn't necessarily an advantage for the team or the pitcher.

So you're making the bullpen work more often while eating a bullpen roster spot by employing a #6 starter who is likely sub-replacement level (do you really have 6 above-average pitchers?). The only marginal advantage is that you rest your starters somewhat so they are arguably fresher for fewer starts, but you've got to figure that any advantage there is more than cancelled out by the increased likelihood of losing games started by the #6 guy.

Not to mention the fact that your #6 is probably just a guy you've promoted from the bullpen, and now you've got to (a) fill his spot with your next least-bad pitching option and thus subtract a hitter from your bench, or (b) keep your hitting roster intact and go with a smaller bullpen, even though they're going to be pitching more innings, meaning that now each guy's got to pitch even more innings.

Of course, a savvy real-life GM is welcome to prove me wrong. I'm just some anonymous guy on the internet.

The Mets might be a candidate for a six-man rotation if/when Santana comes back this season. Pelfrey's pitching the worst out of the current group but they've given no hints about moving him to the bullpen. Gee did not fare well early in the year out of the pen and he's not likely to be moved with a 10-3 record, anyway. It's either a six-man rotation or move Capuano to the pen.

I'm sure that Earl Weaver would point out that it's easier to find five guys who can pitch than find six, :-)

Pitchers and the union will flip out about the six-man, arguing that it artificially keeps the innings and pitching wins of a great starter down.

The 3 top teams in the AL East are using 6 man rotations at the moment, although both the Red Sox and Yankees are probably only doing it once through the rotation while they decide who t demote (Red Sox are choosing between Miller and Wakefield, Yankees between Nova and Hughes). The Rays have been on a 6 man rotation since Niemann came off of the DL though, and haven't really announced any plans to change that.

As Randy notes, the 4-man rotation was a bit of an aberration anyway. ..flexibility was the order of the day in the 20s, 30s, 40s, and 50s

it's very striking if you look at the number of individual pitchers who made 35 or more starts by decade:
20s 9
30s 4
40s 6
50s 8
60s 44
70s 110
80s 84
90s 68
00s 45

of course, this is complicated by expansion and by going from 154 to 162 games, but the point is made

what I discovered that was even stranger, is that, virtually ALL the pitchers in the 90s and 2000s who made 35 or more starts made either exactly 35 or 36 starts--if you change the criterion to pitchers with 37 or more starts per decade, you get:

20s 9
30s 4
40s 6
50s 8
60s 44
70s 110
80s 45
90s 1
00s 0

he not-so-obvious reason is that baseball players are more sensitive, their agents and their union more powerful, than ever before. If you read between the lines a little, you might get the impression (as I have) that some this stuff about six-man rotations is about not wanting to offend the sixth-best starter at hand .

Yeah, I can see the Yankees losing sleep over offending Nova or Hughes.

The Yankees managed their pitching staff for years by running the "Columbus Shuttle", now the Scranton/Wilkes-Barre Shuttle. Fifth starters are fungibly bad, and you keep trying until you find one; and you pick up an emergency starter for double headers/minor injuries-missed starts.

The Yankees will not be using a "six man rotation", once they figure out that Hughes is more like a 7th or 8th man these days.

I suppose we're now expected to guess who the 1 pitcher in the 90s to start 37 games in a season was. My guess is Charlie Hough.

I suppose we're now expected to guess who the 1 pitcher in the 90s to start 37 games in a season was. My guess is Charlie Hough.

Not Charlie, but a future Hall of Famer.

*tries to look inconspicuous*


I think we're fast approaching the day when shortening the rotation is what's called for. Three fresh guys airing it out are going to outperform your starter his fourth time through the order, so I'm completely understanding of the trend toward pulling starters earlier; but if they're going to be asked to do less per start, why do they also need to start less often?

This is the kind of garbage article you get when someone is trying to produce X amount of content in a given time. Neyer presumably understands this will not happen, so it's probably dishonest as well.

I still don't understand why teams don't just mix up the back end of the rotation in situations like this.

I agree with this in theory and it was how teams did it in the past (at least moreso) but I also have to say that the time I tried to do this in DM, it was really hard to make work. I was trying to set up a lefty/right platoon in my 5th starter slot. The problem occurred when, say, my RHP was starting today but, looking ahead, I knew I wanted my LHP to start 5 days from now. So my RHP starts and obviously it's at least 3-4 days before I can use him in relief. Meanwhile, since my LHP is starting in a few days, I can't use him in relief. When the RHP is gonna start the one after, I put myself in the same bucket. Essentially I found myself one pitcher short in the bullpen for 10 games whenever I wanted to swap pitchers. That often meant the rest of my bullpen got tired and at least a few times I ended up having to use my next 5th starter in relief (blowouts, extra innings), meaning I had to keep the other guy in the rotation against a sub-optimal lineup.

I'm sure it could be handled with a more planning and forethought but I've got to say it wasn't as easy as I thought it would be.

As #13 points out, the key question here is number of starter innings. I don't think teams can afford to reduce this much further -- at least not without pushing relievers to more than one inning per appearance. The combination of current roster size, current staff size, current starter usage and current reliever usage is pretty much at its breaking point right now. So as #13 notes, unless by some magic adding a 6th starter allows your #1 and #2 to pitch 8-9 per outing, you end up burning a roster spot on the 6th starter and you've added maybe another 50 innings to your bullpen load ... and now you've got a 14-man staff instead of a 12. Barring roster expansion, you can really only add a 6th starter if it allows you to cut back to 6 relievers but it's hard to see why that would be under current reliever usage patterns.

37+ starts in the 90s I assume was Maddux, possibly Clemens or Johnson. The Braves ran a 4.5 man staff through the mid-90s which allowed them to get 35-36 starts out of their big guns (and only about 20 out of their #5). Checking I am correct on Maddux but he did it with the Cubs in 91.

but if they're going to be asked to do less per start, why do they also need to start less often?

That's the $162 million question. :-)

Not Charlie, but a future Hall of Famer.

with emphasis on the "future".

-- MWE

EDIT: Walt's first guess was correct. 'Twas Maddux.

Are pitcher innings going up with the offense down?

Teams should also expand their 1-8 batting lineups by adding, or designating, a ninth hitter.

it's very striking if you look at the number of individual pitchers who made 35 or more starts by decade:
20s 9
30s 4
40s 6
50s 8
60s 44
70s 110
80s 84
90s 68
00s 45

of course, this is complicated by expansion and by going from 154 to 162 games, but the point is made

what I discovered that was even stranger, is that, virtually ALL the pitchers in the 90s and 2000s who made 35 or more starts made either exactly 35 or 36 starts--if you change the criterion to pitchers with 37 or more starts per decade, you get:

20s 9
30s 4
40s 6
50s 8
60s 44
70s 110
80s 45
90s 1
00s 0

So nobody from the 20's through the 70's made 35 or 36 starts?

edit: also, I don't understand your data. In 1930 alone, 7 different pitchers made 35 starts.

Are pitcher innings going up with the offense down?

No. But before long, I predict either rosters will expand, or complete games and long relievers will make a small comeback.

meaning I had to keep the other guy in the rotation against a sub-optimal lineup.

But that's what you'd be doing anyway without the platoon. Since the last man in the bullpen is often a long man anyway, I don't see how such a platoon would really cost you much. You'd basically just be giving some starts to the long man when he was rested and had the platoon advantage over the 5th starter. Makes sense to me.

Unless it can be proven that a 6 man rotation keeps your starters healthy, I'm not sure it will gain traction.

Lack of evidence it helps anything didn't stop a five man rotation from gaining traction.

Pretty sure the evidence was that very few teams could put together and maintain a healthy, productive four man rotation. Teams weren't getting 40, 41 starts each out of their four best pitchers and suddenly deciding they had to get Bobby J. Jones, v2000-2002 into the rotation.

Omar Minaya couldn't do much, but he could dig up number 6 starters at will.

35 starts in the 162 game era is 21.6%. For a 154 game season that is 33 starts. Now I'm pretty sure that Pasta's chart is probably listing unique pitchers and not every single time it happened. Now I don't have the capability of following in his footsteps hear at work but I can do how many times a pitcher had 33 starts or more in a season for each decade.

Now I'm pretty sure that Pasta's chart is probably listing unique pitchers and not every single time it happened.

Perhaps, but as I said, it was done 7 times in 1930, compared to his claim it was done only 4 times in the entire decade. Surely 3 pitchers didn't do it twice in 1930, and no one else did it ever for the rest of the decade? But, to partially answer:

1930 - 7
1931 - 4
1932 - 4

That's 15 in the first 3 years of the 30's. No repeats. 15 different pitchers started 35 or more games in MLB from 1930-1932.

edit: I'm not playing gotcha here, just trying to understand what he is saying. I think he messed up his charts, especially when he shows how the 90's and 00's starters with 37 or more starts almost disappear vs the 35 or more starts, but the 70's and previous guys are unaffected, as if every one from the 70's prior who started 35 games or more also started 37 or more.

edit edit: And just to pile on, his chart claims there were only 6 in the entire 1940's, yet there were 7 unique pitchers in 1940 alone, and 2 more in 1941.

something other: as far as I understand it, it was that plus the desire to have everyone on a schedule (everyone needs a specific role and all that) that caused the shift from a kind of helter-skelter rotation where the good pitcher(s) got 40 starts and everyone else just kind of slotted in wherever, to the rigid 5-man rotation we've had the past 20-25 years.

The thing is, I'm a long way from convinced it's worth taking six or eight starts a year away from your good pitcher(s) and giving them to bad pitchers, just to keep everything nice and scheduled, and I'm a long way from convinced it does anything at all to prevent injuries to the good pitcher(s).

Just using PI his numbers don't remotely make sense.

Number of individual starters to make 35 starts or more in a season by decade

20's 48 (9 pitchers made 35 or more starts 3 or more times in the decade)
30's 33 (4 pitchers made 35 or more starts 3 or more times in the decade)
40's 30 (4 pitchers made 35 or more starts 3 or more times)
50's 30 (8 pitchers made 35 or more starts 3 or more times)
60's 97 (24 pitchers made 35 or more starts 3 or more times)
70's 111 (42 pitchers made 35 or more starts 3 or more times)
80's 81 (23 pitchers made 35 or more starts 3 or more times)
90's 48 (7 pitchers made 35 or more starts 3 or more times)
00's 29 (4 pitchers made 35 or more starts 3 or more times)

I was thinking maybe he was going by pitchers doing it 3 or more times, but that didn't track remotely with his info. I have no clue what he was trying to say.

number of individual seasons with 35 or more starts(37 or more in parentheses)

20's 77 (43)
30's 53 (19)
40's 46 (15)
50's 62 (27)
60's 196 (94)
70's 290 (167)
80's 153 (45)
90's 76 (1)
00's 45 (0)

Unless it can be proven that a 6 man rotation keeps your starters healthy,

FWIW, college pitchers often lose 2-4 MPH on their fastballs when they go pro. The reason usually cited for this is that in college they get 6 days of rest between starts and in the pros they get just 4. Of course, there are other reasons a college pitchers fastball might slow down, like the difference between throwing 100 innings a year and 200 innings a year as well as the need to stragetically slow the fastball down to gain command and/or movement, but it might be worth exploring if a pitcher in a 6 man rotation shows better stuff and gets better results than he did in a 5 man rotation. I'm skeptical that it would be a significant improvement, but it is possible.

In addition to taking your best starters and pitching them less as #10 points out, this would increase bullpen usage, since the 6th starter isn't going to pitch as many innings as the guys above him would. Otherwise he wouldn't be the 6th starter.

It is possible the bullpen usage wouldn't increase as a in a 6 man rotation your starters would be getting an extra 20% more rest between starts and might be able to go deeper into each start.

Teams don't mix up the work anymore because the value of predictable regimens is great psychologically and probably physically as well.

Bobby Cox and Mazzone were often rigid with their bullpen assignments - pitcher A is my closer, pitcher B is my 8th inning guy, pitcher C is my 7th inning guy - and Braves announcers often used to cite the fact that the guys knew what their role was and when they were going to pitch as an explanation for Cox and Mazzone always getting monster years out of the Darren Holmes, Antonio Alfonsecas and Chris Hammonds of the world.

EDIT: And Kevin Gryboski! Jesus, Kevin freakin Gryboski. Getting 3.5 years of 130ish ERA+ out of THAT guy is on par with walking on water

EDIT: And Kevin Gryboski! Jesus, Kevin freakin Gryboski. Getting 3.5 years of 130ish ERA+ out of THAT guy is on par with walking on water

Well to be fair his component numbers do suck, and he had a tendency to allow a much higher than average inherited runners scored while pitching for the Braves.(he still got better performance in Atlanta than he did in the rest of his career)

EDIT: And Kevin Gryboski! Jesus, Kevin freakin Gryboski. Getting 3.5 years of 130ish ERA+ out of THAT guy is on par with walking on water
Quoted for emphasis.

Just because the worst manager in baseball is going with a 6 man rotation does not mean it is a good idea. Your #1 and # 2 SPers get 5 fewer starts over a season. Heck, this is the guy who bats Adam Dunn cleanup against CC, a guy who was 3-77 against LHP'ers.

Most teams do not have even 4 competent starting pitchers, let alone 6. Makes more sense to piggy back your #5 and # 6 starters in the same game. One starter starts the 1st, the other starts the 4th or 5th. Both guys available to pitch the 3rd day in the pen. This gives the bullpen a blow since games when the #5 SP'er goes are usually long days for the pen since they have trouble going more than 5 IP anyways.

The thing is, I'm a long way from convinced it's worth taking six or eight starts a year away from your good pitcher(s) and giving them to bad pitchers, just to keep everything nice and scheduled, and I'm a long way from convinced it does anything at all to prevent injuries to the good pitcher(s).

There might be an interesting compromise possible. Take a guy like Sabathia and perhaps one more horse, and pitch them both on three days rest. Piece your remaining three best starters in around your horses, such that your third best starter gets the ball in games 3, 8, 15, 20..., and so on.

I didn't follow it out to the 162nd game, but I'm guessing it gets your two best starters 41 starts each, your third best starter around 28 starts, your fourth best 27 starts, your fifth man 25 starts.

I can see why it'll never happen: the third through fifth starters go on irregular rest, and they'll whine to their agents. Or, if you're the GM do you really want to risk CCs arm and your job by running him out there so often? OTOH, if you luck into a guy making the minimum who looks extremely durable (or you're a small market club with an arb guy whose value you and he want to build), and the rest of your staff is a good, 35 year old knuckleballer followed by dreck, I can see giving it a shot.