Minnesota Twins News & Rumors Forum
View RSS Feed

Teflon

Revisiting the 4-Man Rotation

Rate this Entry


Parker Hageman posed his thoughts on how the Twins can improve their rotation through free agency. The players he suggested were either once-good pitchers who are now in decline or once-promising pitchers who have failed to keep their promises. If you are the Twins GM and those are the best pitching options the team can hope for in 2014, you wonder if there might be a simpler (and much cheaper) solution that nobody's thinking of.

The 4-man rotation.

Like nearly everyone, you were ready to dismiss this as foolish or irrelevant in this day and age of delicate arms, guaranteed mega-contracts and pitch counts. But what if you didn't use it to divide more innings among fewer starters. What if you adapted it to address this fundamental premise:

A mediocre starting pitcher's third trip through the lineup is usually crap.

And to that end, you could attempt to optimize those opportunities for disaster by replacing them with something better.

How would it work? First, you'll need to say arrivederci to your 5th starting pitcher and 3rd catcher. Replace them with two relievers and then smile inwardly just a bit because you know this fundamental premise to be true, but not leveraged: Serviceable relief pitchers are easier to find than serviceable starting pitchers. (Cheaper, too.)

With your 25-man roster now aligned as needed, you then set these two rules for utilizing your mediocre 4-man rotation:


  • Have no expectation to EVER pitch a starter more than 5 innings.



  • Plan for regular use of a 6-man combo of middle relievers to share middle innings.


Seems pretty simple. You're exchanging that third trip through an opponent's batting order by a mediocre pitcher for some fresh-armed innings of relief from an expanded relief corps. Hell, you can even do some lefty-righty match-ups you didn't have a chance to do before.

You've looked at the K-rates and opposing batting averages of your pitchers for years acknowledging (but not exploiting) the fact that a player pitching out of the bullpen does better in those categories than when he starts. (You might be looking at Glen Perkins as you muse about this) You know that's because he can throttle-up in a much shorter appearance and he doesn't get exposed to any batter more than once. Why then, haven't you thought it would be advantageous to have more of those kinds of innings as opposed to the tail end of a Mike Pelfrey start?

You shake your head when you realize that your team's old pitching philosophy was to let a mediocre starter pitch until he got into trouble. This meant that you were usually engineering a trouble situation to be your key to make your pitching change.

This is definitely better, you think. It does all of the following:


  • Takes away 25-30 starts from a horrible starting pitcher



  • Doesn't add any innings to the 4 starters. Just shifts their innings from the ends of games when they're tired to the beginnings of games when they're fresh.



  • Limits starting pitchers' exposure. Starters will not get a third time through a batting order.


This really is just evolution, you think. This philosophy is already used for closers. They don't get used to respond to specific game situations but rather to optimize the 9th inning. Same with the setup pitcher. He optimizes the 8th inning. Why shouldn't you optimize the middle innings as much as possible, too?

Is there a downside, you wonder? Your four starters will still pitch comparable innings to other starters so don't have to worry about how their stats compare at contract time - plus they should get a couple more wins because they're still going 5 and getting more starts. Hell, pitching more innings when they're fresh might even improve the quality of their stats. Quite possibly their health and longevity, too.

Inter-league play? Does losing a bench player mean you lose an opportunity to pinch-hit? You still have 12 batters. You could still pinch-hit 3 or 4 times in a National League game. You can live with that.

Workload for the bullpen? You're not getting them up in knee-jerk response as much and you added 120-140 innings worth of arms out there. Pretty close to the innings lost from a 5th starter.

Prep time for the starters? Can they adapt to three off days instead of four? Is that adequate time to recover from a now lessened 80-pitch workload? That's going to be the main question to answer. Hopefully one you can test this season.

Addendum:

The Rockies Interesting Rotation Experiment

Rockies four-man rotation gathering traction as season continues

Updated 09-07-2013 at 02:41 PM by Teflon

Categories
Uncategorized

Comments

  1. jorgenswest's Avatar
    PA K/BB OPS
    1st 1260 2.32 .790
    2nd 1191 1.65 .806
    3rd 858 1.28 .850
    4th 45 1.25 1.139

    The table above shows performance by Twins starter the 1st through 4th time through the line up in 2013.

    At first glance, it appears that the third time through the line up is a significant drop off.

    However, it is not quite a dramatic as it looks. Those 350-400 fewer plate appearances are mostly missing from the bottom of the opponents batting order. That wouldn't account for the entire difference or the near halving of k/bb ratio. The k/bb ratio takes a big hit just going to PA#2.

    Relievers rarely go through a line up twice. In their first time through the line up this year Twin relievers have posted a 2.74 k/bb ratio while allowing an OPS of .655 in 1895 plate appearances.

    It might be worth a shot. While I can't see a team with an ace going this route that doesn't describe the Twins situation.
  2. jorgenswest's Avatar
    A second thought about roster composition. A return to an 11 man pitching staff would enable the bench flexibility to platoon. Is there a way to do this with a 7 man bullpen and retain the effectiveness of the relievers?
  3. Willihammer's Avatar
    Its an interesting strategy. I was a little skeptical of the assumption that a lousy group of starters like ours gets through 5 IP every day, esp. only facing lineups twice. Because as it is, with a 5 man rotation and 100 pitch count limit, the 2013 Twins starting staff is averaging less than 5.2 IP per start.

    I pulled some stats. These include 3rd and 4th time through the order, so they are a little skewed, but as a staff, the Pitches per PA and Whip are this

    PPA WHIP
    3.814 1.522

    That's 4.522 batters per inning, and 17.25 pitches per inning on average. If you take them out after 2 trips through the order, that would only equate to roughly 4 innings.

    If you change the criteria to Pitch count, instead of trips through the lineup, you see this:

    HTML Code:
    <PRE>                                  
    Split            BA  OBP  SLG  OPS
    Pitch  1-25    .258 .314 .395 .708
    Pitch 26-50    .292 .338 .438 .775
    Pitch 51-75    .274 .342 .424 .765
    Pitch 76-100   .318 .374 .490 .864
    Pitch 101+     .328 .418 .517 .935
    </PRE>
    
    Provided by <a href="http://www.sports-reference.com/sharing.shtml?utm_source=direct&utm_medium=Share&utm_campaign=ShareTool">Baseball-Reference.com</a>: <a href="http://www.baseball-reference.com/teams/split.cgi?t=p&team=MIN&year=2013&utm_source=direct&utm_medium=Share&utm_campaign=ShareTool&utm_source=direct&utm_medium=Share&utm_campaign=ShareTool&utm_source=direct&utm_medium=Share&utm_campaign=ShareTool#pitco">View Original Table</a><br>Generated 9/7/2013.
    Not surprisingly, the same problem shows up. it would require over 86 pitches to get through 5, on average. That's 11 pitches or so too many.

    So, realistically, you'd be asking your bullpen to complete innings 5-9. A 6 inning guaranteed workload on 9 guys. Sounds taxing, but maybe it could be done. See the days rest splits:

    HTML Code:
    <PRE>                                   
    Split        PA   BA  OBP  SLG  OPS
    0 DaysGR    247 .235 .313 .304 .617
    1 DayGR     607 .227 .299 .345 .644
    2 DaysGR    437 .222 .282 .365 .647
    3 DaysGR    365 .269 .316 .424 .740
    4 DaysGR    155 .210 .265 .343 .607
    5 DaysGR     39 .237 .256 .421 .677
    6+ DaysGR   128 .231 .278 .274 .551
    </PRE>
    
    Provided by <a href="http://www.sports-reference.com/sharing.shtml?utm_source=direct&utm_medium=Share&utm_campaign=ShareTool">Baseball-Reference.com</a>: <a href="http://www.baseball-reference.com/teams/split.cgi?t=p&team=MIN&year=2013&utm_source=direct&utm_medium=Share&utm_campaign=ShareTool&utm_source=direct&utm_medium=Share&utm_campaign=ShareTool&utm_source=direct&utm_medium=Share&utm_campaign=ShareTool#dr">View Original Table</a><br>Generated 9/7/2013.
    Probably skewed a little because the guys pitching back to back are limited to the back end guys (Fien, Burton, Perkins).

    We really don't know much (or I don't know much) about optimal rest days and pitch counts for relievers. If you can get away with throwing these guys 4-6 days a week, for 15-25 pitch stints, then it would be a very intriguing strategy. I'm a little dubious about the ability to get through that kind of guaranteed workload over the course of a season though.
  4. Teflon's Avatar
    Quote Originally Posted by jorgenswest
    PA K/BB OPS
    1st 1260 2.32 .790
    2nd 1191 1.65 .806
    3rd 858 1.28 .850
    4th 45 1.25 1.139

    The table above shows performance by Twins starter the 1st through 4th time through the line up in 2013.

    At first glance, it appears that the third time through the line up is a significant drop off.

    However, it is not quite a dramatic as it looks. Those 350-400 fewer plate appearances are mostly missing from the bottom of the opponents batting order. That wouldn't account for the entire difference or the near halving of k/bb ratio. The k/bb ratio takes a big hit just going to PA#2.

    Relievers rarely go through a line up twice. In their first time through the line up this year Twin relievers have posted a 2.74 k/bb ratio while allowing an OPS of .655 in 1895 plate appearances.

    It might be worth a shot. While I can't see a team with an ace going this route that doesn't describe the Twins situation.


    And these were the breakdowns for the MLB in 2012 from the Fangraphs study at the time of the Rockies move to the 4-man rotation:

    1st PA vs SP: .247/.310/.393
    2nd PA vs SP: .260/.321/.417
    3rd PA vs SP: .271/.332/.444

    while the first PA versus relievers was:

    .241/.316/.375

    I agree that if you have Justin Verlander and Max Scherzer this isn't the option you should take - but if you have a sorry bunch of starters it seems a no-brainer to jettison as many badly pitched innings from your games as possible for better ones.

    Starting pitchers will undoubtedly hate it since it's calling into question their manhood among their starting pitching brethren.
  5. jorgenswest's Avatar
    Willhammer is correct. The third time through will often come up sometime in the 5th inning. They need to get through that inning for this to work. Too much burden on the bullpen otherwise. It is still a worthy idea.

    I am not as certain on the 11 pitches. That would assume there is a big drop off on pitch 76 and pitch 76 was the same as pitch 100.
  6. Willihammer's Avatar
    I'm much more willing to roll with short benches offensively than conventional wisdom dictates though. We have mounds of data that prove that on average, if you take anyone off the bench cold and ask him to pinch hit, well, the effect is the opposite of relief picthing. They stink. They take a horrible penalty, worse than the platoon penalty for the starter he's PH-ing for, in most cases. So the primary reason to keep bench players is as pinch runners, defensive replacements, and injury backups, IMO.

    If you had couple Eduardo Escobars in your system, guys you could stick anywhere in a pinch, then you could roll with a 14 man pitching staff. That gives you 10 relievers. I think it would be a viable strategy at that point (provided the bullpen guys are, you know, real MLB Pitchers).
  7. jorgenswest's Avatar
    My interest in the bench is platooning. The most likely place to find a platoon piece is on a corner. The typical platoon player will have a good bat on his platoon half and limited ability to defend anywhere up the middle.

    A 3 man bench requires all three players to be solid up the middle. Those guys usually aren't very good hitters. A 5 man bench allows the possibility to platoon at two of the corners.

    Is there a way to reconfigure the current pitching staff use in an 11 man staff? Can this idea make it possible?

    I think some team will figure Out howto teturn to an 11 man staff. There are players out there that could have a Bill Robinson type career if the someone can figure out how to roster him.
©2014 TwinsCentric, LLC. All Rights Reserved.
Interested in advertising with Twins Daily? Click here.