Baseball for the Thinking Fan

Login | Register | Feedback

btf_logo
You are here > Home > Baseball Newsstand > Baseball Primer Newsblog > Discussion
Baseball Primer Newsblog
— The Best News Links from the Baseball Newsstand

Monday, December 11, 2017

Texas Rangers: Shohei Ohtani or not, Rangers still considering nontraditional 6-man, 4-man rotations next season | SportsDay

If you are asking your starters to pitch fewer innings per start, why not move to the four-man rotation?

“I think we will get through the offseason and look at where our personnel is before we determine usage,” Daniels said Sunday. “I’ve said there are benefits to a six-man rotation and to a four-man. And there are benefits to a five-man rotation, too. But we’ve got to see where we are.”

Jim Furtado Posted: December 11, 2017 at 09:16 AM | 13 comment(s) Login to Bookmark
  Tags: rangers

Reader Comments and Retorts

Go to end of page

Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.

   1. Boxkutter Posted: December 11, 2017 at 11:38 AM (#5590082)
Didn't the Rockies try a four-man rotation with shorter outings a few years ago... with bad results?
   2. 6 - 4 - 3 Posted: December 11, 2017 at 12:41 PM (#5590151)
I've done a 4 man rotation with shorter innings in OOTP (DH league). How I organized it was:
- 4 starters: 4 innings or 60 pitches, whichever was lowest;
- 2 long relievers (if winning or close): 2 innings or 30 pitches, whichever was lowest;
- 2 long reliever (if losing big): 3 innings or 50 pitches, whichever was highest;
- 3 middle relievers (use whoever was rested): 2 innings or 20 pitchers, which was lowest; and
- 2 stoppers, lefty and righty (use 8th inning or later): 1 inning or 30 pitches.

Exceptions happened on occasion, but I stuck mostly to the script. Occasionally I had to bring one of my weaker long relievers in a winning game or a stronger long reliever in a losing game. My two long relievers who I brought in when winning or close both won well over 20 games. I had three players with double digit save totals. No pitcher threw more than 150 innings.

FYI: there's a "market inefficiency" in OOTP for relievers with high stamina who lack a good third pitch and thus cannot be classified as starters (same may be true in real life). I classified all players as relievers because of how the game is configured (you get less of a short rest penalty for a reliever, but he cannot throw as many pitches in a game). Anyway, it was quite cost effective and I'd consider doing it again if/when I play out every game in an OOTP league. The key is that your best pitchers aren't necessarily your starters; rather, it's your winning long relievers and stoppers.
   3. Walt Davis Posted: December 11, 2017 at 03:51 PM (#5590366)
A model like #2 works fine as long as you've got about 7 relievers that can handle 100+ innings in a year ... preferably year after year. Such beasties have been rare at every point in baseball history so finding 7 for one team seems unlikely. The trend for relievers is in the other direction -- lower seasonal IP loads, not higher. In 2017, 4 pure relievers (100%) made it to 80 innings, another 22 to 70 innings and another 51 to 60 innings. That's just 2.5 per team.

In the real world, obviously the two losing long relief slots are a AAA/waiver/DL shuttle churning through several pitchers which takes some of the load off the middle relievers (i.e. those two slots eat up about 300 innings). That makes this kinda viable without massive changes in most reliever usage ... but with a lot of crappy pitchers taking the mound in losses.

4 starters -- 600 innings
2 winning LR -- 160-180 innings
2 losing LR -- 300 innings (using 6-8 different pitchers)
5 relievers -- 360-380 innings (an increase but not radical)

When the starter only makes it 3 innings, you use a losing LR and, if still losing, your second one. They probably both then get optioned/DL'd and two more called up although hopefully you win the next day. When things go as planned, the starter and your Devenski get you through 6 then you turn it over to the pen as usual.

Alas, Devenskis don't grow on trees (or will be one of your 4 starters) and those losing LRs will be turning some potentially close games into blowouts in the wrong direction. It's not at all clear what advantage such a set-up gives you over the current one -- fewer innings out of your 4 best starters, a need for good 5th-6th starter types (your winning LRs), the same need for AAA depth (your losing LRs plus reliever replacement), more durability from your top relievers. It shifts innings from your top 4 starters to (essentially) starters #5-12. Maybe that improves the health of your top 4 starters or maybe pitching on 3 days rest hurts them.

But sure, I can see this as a model that teams might try to build towards, mainly requiring them to develop a lot of Devenskis.
   4. Walt Davis Posted: December 11, 2017 at 04:04 PM (#5590374)
To put names on that, presumably a potential winning LR is a guy like Jason Hammel (or Hammel of the last 5+ years). First problem: Hammel is a top 4 starter on a mediocre or bad team. Second problem: Do you want to bring Hammel into a 2-1 game in the 5th as your top winning LR?

Granted, we don't really know what the answer to the second one is -- maybe limiting Hammel to 2, occasionally 3 innings per appearance will make him a much better pitcher on a rate basis.

Still, it runs counter to current strategy. At the moment you don't (intentionally) bring a league-average (or worse) pitcher into a close game, especially not one that you're winning. If you've got a 3-1 lead after 4, you want 5 good relief innings not two average ones then hope you still want 3 good relief innings. And if Hammel can give you 2 above-average innings 50-60 times a year, then he can probably give you 150 average or better innings and you're back to him as a 4th-5th starter.
   5. PreservedFish Posted: December 11, 2017 at 04:05 PM (#5590375)
FYI: there's a "market inefficiency" in OOTP for relievers with high stamina who lack a good third pitch and thus cannot be classified as starters


I've noticed that but never exploited it as creatively as you have. Lots of relievers with 80 stuff, two crazy pitches, good endurance, but no starter rating. It seems like the game must just fudge the results, as I doubt the game engine is accurate enough to be able to model a 2-pitch pitcher failing as a starter but succeeding as a reliever (does anyone on earth understand it properly?) ... OOTP likely gives that pitcher an automatic quality loss if he starts a game. But if he throws exactly the same number of innings as a reliever, he's terrific.
   6. PreservedFish Posted: December 11, 2017 at 04:41 PM (#5590395)
2 losing LR -- 300 innings (using 6-8 different pitchers)


I am fairly sure that this will be a new frontier in bullpen management. (Maybe it's already happened but we don't notice because the players are cycled so quickly)

Baseball has always had the mop-up man, of course. But managers will become more and more eager to raise the white flag in lop-sided losses, and as a result will be more aggressive in pushing those unimportant innings onto losers and away from valuable relievers.
   7. 6 - 4 - 3 Posted: December 11, 2017 at 05:18 PM (#5590413)
Well some of those long relievers would be converted starters, so more capable of pitching multiple innings than the modern middle reliever. The laptop that I simmed this on got stolen, so I don't have the stats anymore. But my recollection is that it was just the starters and long relievers who got more than 100 innings each. The key is that you don't overwork either of your stoppers. I'd often bring in a middle reliever to finish a game rather than the other stopper (i.e., I didn't use them like Goose Gossage was in his prime).

On a few occasions, I did use minor league options on occasion with one of the losing long relievers or my worst middle reliever (usually because an extra inning game destroyed my model). I didn't do it often or routinely though because OOTP isn't coded to prevent you from recalling an optioned player with less than 10 day (with an exception for injury to another player), so I didn't like having to keep track of it manually or flaunting the rule. Speaking of which, the only season-ending injury to the pitching staff was one of the starters and (late in the season) my worst middle reliever. I was able to keep the other pitchers in their assigned roles.

In this model, I used my best "starting pitchers" as my winning long relievers (it was a fictional league). So if you take a typical pitching staff:
- (5) starters
- (1) swingman/mopup/long reliever
- (6) middle relievers (includes LOOGY[s])
- (1) closer

So I would use:
- 4 Starters: #3 Starter, #4 Starter, #5 Starter, Swingman
- 2 Winning Long Relievers: #1 Starter, #2 Starter
- 2 Losing Long Relievers: Middle Reliever #5, Middle Reliever #6
- 3 Middle Relievers: Middle Relievers #2, #3, #4
- 2 Stoppers: Closer and #1 Middle Reliever

The thing to keep in mind is that there is a correlation between the number of innings/pitches that the starter will give you and the likelihood of utilizing a winning/losing reliever. I also only took out starters before 4 innings when they reached 60 pitches (i.e., let him get hammered until he reached his pitch count), so it was pretty infrequent that I wouldn't have a starter at least start the 3rd inning (and when he didn't, the losing long reliever would compensate by pitching an extra inning). That's why I gave more innings to the losing long relievers than the winning long relievers.

Again, by recollection, distribution of innings by pitching role was something like:
- Starters: 40x3=120 innings (4x120=480 innings)
- Winning Long Relievers: 90x1.5=135 innings (2x135=270 innings)
- Losing Long Relievers: 60x2.5=150 innings (2x150=300 innings)
- Middle Relievers: 60x1=60 innings (3x60=180 innings)
- Stoppers: 70x1=70 innings (2x70=140 innings)
- Others: 50-70 innings (occasionally a position player, but most came from September callups, which basically expanded the number of losing long relievers)

Note: the distribution among the 5 middle relievers and stoppers was about what you would get from a conventional bullpen setup (60-70 innings). The back-end of the staff was used very similarly to how a conventional modern bullpen is used, except that I would be willing to bring in a closer in the 8th inning if necessary.

To recap, the keys are:
- strict adherence to pitch and inning counts (I took them out the batter after the exceeded either);
- use the winning long relievers only when there is a lead to protect or it's close in the middle innings;
- try to use only one winning long reliever per game;
- try to use only one stopper per game;
- use middle relievers for more than one inning if pitch count is below 20;
- be willing to use a position player to pitch in a huge blowout loss (don't think I mentioned this earlier, but did this a few times);
- stoppers need not finish the game (how I wound up with 3 or 4 guys with double-digit saves).

Finally, my recollection is that it totally broke Pythagorean winning percentage (I won something like 90 games, but RA=RS almost exactly). That is, I won a disproportionate number of close games, but had a ton of blowout losses (by design). Also probably had more than my fair share of blowout wins, but the losses were more memorable.

If I were to do it again, then I'd probably use a "shuttle system" for the losing long relievers and the worst middle reliever slot as you described. I'd also probably be a bit more liberal with allowing my stoppers to pitch more than 1 inning, which might get me into pitching guys more innings than closers typically pitch these days.


EDIT: I may be a little off on the innings estimates. But I remember that my long relievers (both winning and losing) had more innings pitched than all but one of my starting pitchers.
   8. RMc's Unenviable Situation Posted: December 11, 2017 at 05:25 PM (#5590420)
Re 2 and 7: The problem is that if you used your system in real life, the minute the team suffered a four-game losing streak (or somebody got hurt), the entire fanbase would freak out and you'd probably get fired.
   9. 6 - 4 - 3 Posted: December 11, 2017 at 05:29 PM (#5590423)
Re 2 and 7: The problem is that if you used your system in real life, the minute the team suffered a four-game losing streak (or somebody got hurt), the entire fanbase would freak out and you'd probably get fired.

Probably. Going into the season, my "owner's expectation" was only that the team play .500. So I exceeded expectations.
   10. PASTE, Now with Extra Pitch and Extra Stamina Posted: December 11, 2017 at 09:36 PM (#5590507)
Every year we all read these articles as this team and that team consider 6-man or 4-man rotations. Every year, sure enough, they roll out the same 4.5 man/5 day rotation everyone else does, because it always turns out to be far too much trouble getting everyone in the organization on board with such a drastic change to bother with.


I've noticed that but never exploited it as creatively as you have. Lots of relievers with 80 stuff, two crazy pitches, good endurance, but no starter rating. It seems like the game must just fudge the results, as I doubt the game engine is accurate enough to be able to model a 2-pitch pitcher failing as a starter but succeeding as a reliever (does anyone on earth understand it properly?) ... OOTP likely gives that pitcher an automatic quality loss if he starts a game. But if he throws exactly the same number of innings as a reliever, he's terrific.


This is correct. OOTP sharply reduces a pitcher's effectiveness if he doesn't have at least three serviceable pitches and starts a game. Start some random flotsam for just the first batter and then bring the same pitcher in in relief, however, and all's well. (It appears the game won't let you start a position player as a pitcher, though; you can try, but it'll just automatically substitute a random reliever.)
   11. PreservedFish Posted: December 11, 2017 at 09:56 PM (#5590512)
To solve that problem you actually have to understand what's happening, which I'm not sure we do. But I guess this is a good theory: because we know that the 2nd/3rd/4th time through the lineup is real - and, moreover, that the more pitches a hitter has seen, the bigger the penalty is - OOTP should apply a tiny incremental (batter-specific) penalty every single time a pitcher reuses a pitch. So if you're facing Daniel Descalso and your fastball is an 80, maybe the next fastball is a 78, and the next a 76. But if you throw your first curveball to him, even if it's in his 3rd time facing the starter, that curveball is still at full strength.

Of course we're talking about a game where they spend more time making sure that there's an accurate database of Sri Lankan surnames than they do trying to make the aging curves or trading AI even slightly realistic...
   12. PreservedFish Posted: December 11, 2017 at 10:04 PM (#5590515)
I guess the other thing you'd do to make it hyper-realistic is designate an effort level for your pitchers. Relievers don't throw harder by magic, they throw harder because they know they won't last more than 25 pitches. It would improve velocity and pitch quality with an immediate known cost to endurance. These things together would likely throw a wrench into 6-4-3's creative plans.
   13. 6 - 4 - 3 Posted: December 12, 2017 at 12:44 AM (#5590563)
So I just simmed 29 games with the 2017 Giants, going 18-11.

I traded for Chris Young and Ian Krol. Then configured the staff as follows (projected IP in parentheses):

SP: Samardzija (180.2), Moore (121.0), Young (188.0), Cain (178.2)
WLR: Bumgarner (134.0), Cueto (117.1)
LLR: Ramirez (102.1), Kontos (110.0)
MR: Gearin (59.2), Krol (48.1), Law (65.1)
ST: Melacon (52.0), Strickland (70.2)

There were no major injuries and I didn't promote/demote anyone during the first month. Obviously, this is a lot more difficult to manage in the NL without the DH and only having a 4 man bench. Consequently, I wound up leaving in Shark and Young (two relatively good hitters) in a bit longer when they had low pitch counts, which is why they wound up on track for higher-than-expected innings totals (while relievers have slightly lower ones). I needed some innings from Cain, which is how he wound up higher (was actually hoping to injure him by overworking him, but no such luck). As in real life, Moore just plain sucked, which is why he's a little low on the innings total. No position players pitched, although I used Bumgarner as a pinch hitter a few times in games where I knew he wouldn't pitch.

Neither of the more fungible guys have options except for Law, so it would not be possible to do a minor league rotation with the LLR or MR with this particular team.

In case you're wondering, Bumgarner is on track for 34 wins. Samardzija is on track for 3.3 WAR, while Bumgarner and Ramirez are both on course for 1.9 each. Moore is on track for -2.9 WAR, so if I were to continue then he'd probably get demoted to LLR and promote Ramirez to SP? I don't know whether I'll continue with this little experiment. Despite my best efforts to prematurely end his career, Cain is on course for 0.7 WAR (that would represent his best total since 2012).

NOTE: OOTP has their own WAR, so it doesn't really map well to either BBRef or FG's.

You must be Registered and Logged In to post comments.

 

 

<< Back to main

News

All News | Prime News

Old-School Newsstand


BBTF Partner

Support BBTF

donate

Thanks to
danielj
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Hot Topics

NewsblogMLB's plan on how they can make baseball games at least 10 minutes shorter
(37 - 7:12am, Jan 22)
Last: bunyon

NewsblogOT - 2017 NFL thread
(1529 - 7:09am, Jan 22)
Last: ERROR---Jolly Old St. Nick

NewsblogInside Baseball | Scott Boras Weighs In On Slow Market
(25 - 6:49am, Jan 22)
Last: You Know Nothing JT Snow (YR)

NewsblogRyan Thibs has his HOF Ballot Tracker Up and Running!
(1578 - 5:48am, Jan 22)
Last: QLE

NewsblogOTP 15 January 2018:Mississippi's anti-gay marriage law is hurting two college baseball teams
(2087 - 3:33am, Jan 22)
Last: David Nieporent (now, with children)

Gonfalon CubsStatus of the Roster mid-December
(102 - 1:33am, Jan 22)
Last: Walt Davis

NewsblogA Newly Discovered Wasp Is Named After the True Hit King Ichiro
(22 - 1:11am, Jan 22)
Last: Rennie's Tenet

NewsblogWhat Is Baseball’s Equivalent of the Vikings’ Miraculous Victory?
(101 - 10:03pm, Jan 21)
Last: Howie Menckel

NewsblogOT - NBA 2017-2018 Tip-off Thread
(2756 - 8:16pm, Jan 21)
Last: Crosseyed and Painless

NewsblogOT: Winter Soccer Thread
(569 - 7:56pm, Jan 21)
Last: Mefisto

NewsblogComparing a Player Outside His Era | Articles | Bill James Online
(37 - 7:48pm, Jan 21)
Last: McCoy

NewsblogWhy is J.D. Martinez's big bat still available
(47 - 5:29pm, Jan 21)
Last: Jose is an Absurd Doubles Machine

NewsblogChristian Yelich can only blame himself for Jeter entanglement | New York Post
(62 - 5:07pm, Jan 21)
Last: Walt Davis

NewsblogPrimer Dugout (and link of the day) 1-18-2018
(13 - 4:57pm, Jan 21)
Last: Walt Davis

NewsblogNBC Sports Philadelphia: Morphine found in Roy Halladay's system before fatal plane crash
(32 - 12:45pm, Jan 21)
Last: Adam Starblind

Page rendered in 0.2309 seconds
47 querie(s) executed