USATSI_10677695_168381442_lowres

The Rotation is Good

What is a good pitching rotation in an MLB environment in which pitching rotations do not exist? Following the ideal model, a pitching rotation is a mechanism that an MLB team can use to line up five “starting pitchers” to rotate on a set schedule, and hopefully produce value by preventing runs during the cumulative turns through the rotation. Trading within the realm of ideals, the contemporary MLB environment exhibits a market failure in terms of delivering consistent pitching rotations.

  • There are some teams that nearly reach this ideal; in recent Brewers lore, the runs prevention and workload combination of Shaun Marcum / Yovani Gallardo / Randy Wolf / Zack Greinke / and Chris Narveson with Marco Estrada as the lone “replacement” is about as good a true pitching rotation as one can find. They prevented 14 runs, which is quite close to the 2017 Brewers rotational performance of 10 runs prevented.
Runs Prevented 2017 Brewers 2011 Brewers
1 Anderson (26) Marcum (11)
2 Nelson (16) Gallardo (6)
3 Davies (9) Wolf (5)
4 Garza (-13) Greinke (-1)
5 - Narveson (-6)
Swing Suter (9) Estrada (-1)
Replace Wilkerson (1) -
Replace Woodruff (-1) -
Replace Espino (-4) -
Replace Milone (-4) -
Replace Guerra (-8) -
Replace Peralta (-21) -
Emergency M. Blazek -
Emergency J. Jeffress -
Total 10 14
  • Other ideals exist with the 2014 Nationals (Roark / Zimmermann / Fister / Strasburg / Gonzalez and two short-term replacements; 95 runs prevented), or the 2016 Cubs (Lester / Hendricks / Arrieta / Lackey / Hammel plus one short-term replacement and five emergency (one GS) starters; they prevented 122 runs).
  • But, let it be known that the ideal fails, too; the 2016 Cardinals rotation of Martinez / Garcia / Wainwright / Leake / Wacha and two short-term replacements was 25 runs below average (worse than the 2016 Brewers rotation, even!);
Runs Prevented 2016 Brewers 2010 Giants
1 Guerra (22) Cain (19)
2 Davies (4) Vogelsong (4)
3 Anderson (-6) Bumgarner (1)
4 Peralta (-8) Zito (-13)
5 Garza (-15) Lincecum (-33)
6 Nelson (-17) -
Replace Cravy (5) -
Replace Suter (3) -
Replace Jungmann (-10) -
Emergency - Petit
Emergency - Hacker
Total -22 -22
  • The famous 2010 Giants won the World Series with about as true a five-man rotation as one can find, but Cain / Vogelsong / Bumgarner / Zito / Lincecum were 22 runs below average for the Champions (exactly as good as the 2016 Brewers rotation. Indeed, what does a Championship Rotation look like?).

Related Reading:
Depth Beats Attrition
83: 2018 PECOTA Projection
Fluctuation Race: Brewers and Cubs Rotations
Aces Don’t Exist: Rotation Spots
Aces Do Not Exist

2017 MLB Rotation (GS) Pitchers Average IP RA/9 Average DRA
30 to 35 41 187.2 4.43 4.31
25 to 29 42 155.5 4.40 4.52
20 to 24 34 127.5 4.77 4.82
10 to 19 75 81.5 5.37 5.45
1 to 9 145 29.6 5.70 5.93

As I wrote on Friday, this MLB environment is one in which teams employ over 300 starting pitching spots to complete the season. There ostensibly is only a pitching rotation for a couple of teams, with the remainder of the league simply finding ways to patch-up injuries or string together a low rotation that does not get too badly shelled on any given turn.

 

The Brewers have gleefully embraced the “replacement” rotation years under both Doug Melvin and David Stearns; Melvin’s 2012 rotation featuring Gallardo / Greinke / Fiers / Estrada / Marcum / Wolf plus four short-term replacements and an emergency starter was twice as good as the “established” 2011 rotation; this ragtag gang prevented 28 runs. Last year, David Stearns improved the rotation significantly, despite starting only Anderson / Nelson / Davies / Garza as regulars and using seven long- or short-term replacements and two emergency starters. This group was just about as good as the 2011 rotation; they prevented 11 runs.


 

In a sense, rotational queries are aesthetic queries. Do you prefer to steadily churn through a “set” group of five pitchers and take the results, come what may? This method can be dangerous as even relatively similar groups of pitchers can vary significantly from year to year. Take the 2015-to-2016 Cardinals, who made minimal rotational changes, and declined by 81 runs from their excellent 2015 core (excluding John Lackey, who left via free agency). One could complain that the Lance Lynn injury skews that, but his total loss for the season was not anything near that of Jaime Garcia or Michael Wacha’s year-to-year variation. The 2017 Cubs rotational decline has been well-documented, but it’s worth emphasizing that they made minimal personnel changes (in terms of starters) and declined by nearly 100 runs as a group compared to 2016. The Nationals, Mets, and Giants (other “consistent personnel rotations” also demonstrate wild variance between seasons).

Preventing runs is difficult even when a team employs five regular starters.

All this¬†frames the lessons of the 2017 Brewers. Last year the club gambled on a set of mechanical and pitching approach adjustments, as well as a group of well-refined replacements, some declining veterans and in-house arms, and an interesting set of n’er-do-wells to effectively match the production of that ideal of Marcum / Gallardo / Wolf / Greinke / Narveson.

If anything, David Stearns effectively applied the lessons of the 2012 Brewers rotation in as effective a way as possible; it’s worth noting that those lessons failed in 2013 when Melvin went back to well, but it’s also worth noting the “solid rotational group” that was reassembled for 2014 also failed to improve. A Lohse / Peralta / Gallardo / Garza main group with swingmen Fiers and Estrada and MLB graduate Jimmy Nelson looks (in theory) like a much more stable group than either the 2012 or 2013 rotations; that 2014 group prevented three runs when all was said and done.


 

Enter the 2018 Brewers rotation, who could conceivably appear as the very best rotation of the last decade in Milwaukee (I suppose that’s not terribly tough when that decade includes 2009, 2010, 2013, or even 2015 and 2016 rotations, but well, improvements have to start somewhere!). The core group, entering the season, likely includes post-adjustment, post-breakout Chase Anderson, steady-as-they-come Zach Davies, underrated rotational depth man Jhoulys Chacin, and MLB graduate Brandon Woodruff. That group gives the rotation considerable merit prior to considering any additional depth options, as these four prevented 40 runs in 556 innings in 2017.

What is especially intriguing about the 2018 rotation is the pre-established depth entering the year. I’ve written about this at length here, so I will not rehash it, but basically, the club features Wade Miley as a potential minor-league contract “walk-on” (Miley has never posted a below replacement WARP); another mechanical & strategic adjustment candidate in Yovani Gallardo; and standard depth and potential injury returns from the 2017 rotation. For scouting, strategic, and performance characteristics, see “Trust the Rotation” and “How the Brewers Beat the Cubs.”

Here’s where I think most of the cognitive dissonance lies with Brewers fans: Brewers fans are idealizing the rotational turns of a club like the 2011 Brewers, where five set starters work the full season, ebb-and-flow, but ultimately produce value by virtue of their regular turns in the rotation. Granted, in terms of Deserved Runs Average (DRA) alone, Milwaukee’s main group of pitching depth entering the 2018 season should be at least average to begin with. Fans often forget that the current MLB could require withstanding a 4.70 RA/G to 4.80 RA/G environment in Miller Park / National League. A group DRA of 4.77 is quite solid if you’re playing 4.70 RA/G to 4.80 RA/G baseball. It’s not even a bad level of production if you’re playing 4.60 RA/G baseball.

What fans are missing, however, is that compared to their roles across the league, the Brewers’ depth pitchers are better than their professional colleagues. The PECOTA replacement depth assessment especially underscores this rotational strength. This is a difficult point to convey to fans and analysts alike, because most people assume the Brewers will be bad because instead of signing Lance Lynn or Yu Darvish to pitch against folks like Kyle Hendricks or Jose Quintana, etc., the Brewers will be running Brent Suter or Miley or Junior Guerra or somebody to the mound. There are two problematic assumptions here:

  • That any of the Brewers depth starters will work regularly enough as starting pitchers to be assessed as a “standard 25-to-35 GS starter.” Instead of this assumption, fans and analysts ought to think of Miley / Guerra / Suter / etc. working in “bursts” across the season; they might reasonably, as a group, average out to a few starts here or a few starts there, but they hardly should be considered “full SP” roles
  • That additional MLB competitors will not be using replacement starters. This is tough to consider, because no one knows the future. So obviously it’s easier to make an argument using a pitcher one expects to be a regular, quality contributor (let’s stick with Quintana and Hendricks here) than trying to project replacements. But, rotational mismatch works both ways, and it’s worth asking how the Brewers’ depth pitchers square against other depth roles across the MLB.

Thankfully, PECOTA keeps a vast database of potential replacements; 653 starting pitchers appear in the Baseball Prospectus March 2, 2018 pitching projection worksheet. What is fascinating about this assembly of projections is that while the projection system features double the arms than those likely used by MLB squads, the proportional alignment of these arms across rotational roles roughly matches the actual distribution of pitching roles in 2017. So, in terms of assessing the quality of pitching roles within one immediate MLB environment (juiced ball, fly ball revolution, etc.), there’s an immediate correspondence or complementary aspect between PECOTA’s projections and what might reasonably be expected to occur based on immediate experience.

I created two PECOTA rotational models. The first “rotates” through PECOTA projections in 30 player sequences, in order to estimate the number of rotational spots that exist in PECOTA. Since the system is used to project system depth as well as established MLB players, there are 10 rotational spots as well as a Replacement and an Emergency group in 2018 PECOTA:

PECOTA 1 Average IP DRA RnsPrv DRA GS
1 179.7 6.9 4.22 29 to 31
2 162.2 1.5 4.49 27 to 29
3 142.5 -5.4 4.91 24 to 27
4 130.8 -6.0 4.99 23 to 24
5 126.0 -8.9 5.20 22 to 23
6 111.6 -17.7 6.00 21 to 22
7 104.2 -14.2 5.79 19 to 21
8 101.6 -12.4 5.67 18 to 19
9 93.5 -15.0 6.01 16 to 18
10 91.6 -15.3 6.08 15 to 16
Replace 53.2 -7.9 5.91 5 to 15
Emergency 30.4 -3.6 5.63 1 to 4
4.24 Average ERA (2016-2017 MLB)
4.57 Average DRA (2016-2017 MLB)

Since this table is quite abstract, I then designed another model in which 2018 PECOTA data are distributed in a manner similar to the 2017 MLB rotation:

PECOTA vs. 2017 PECOTA SP Percentage 2017 SP Percentage
1 30 to 35 13 2.0% 41 12.2%
2 25 to 29 69 10.6% 42 12.5%
3 20 to 24 119 18.2% 34 10.1%
4 10 to 19 197 30.2% 75 22.3%
5 1 to 9 255 39.1% 145 43.0%
Total 653 100.0% 337 100.0%
PECOTA vs. 2017 PECOTA IP 2017 IP PECOTA DRA 2017 DRA
1 30 to 35 182.3 187.2 4.62 4.31
2 25 to 29 160.7 155.5 4.45 4.52
3 20 to 24 120.3 127.5 5.47 4.82
4 10 to 19 84.3 81.5 5.76 5.45
5 1 to 9 39.8 29.6 5.94 5.93
Total 587.4 581.3 5.25 5.01

Here is a model that assesses the Brewers’ primary system depth against the potential MLB roles and replacements. The last two columns judge each pitcher’s runs prevented against both the 2018 PECOTA rotational spot and the comparable 2017 rotation spot (based on Games Started, as a “true” rotational measurement).

Brewers PECOTA GS IP DRA 2017 Spot PECOTA Comp 2017 Comp
Chase Anderson 28 168.0 4.60 2 -2.7 -1.5
Zach Davies 27 162.0 4.44 2 0.2 1.4
Jhoulys Chacin 24 136.7 4.50 2 -0.7 0.3
Brandon Woodruff 23 131.0 4.79 3 9.9 0.4
Brent Suter 13 100.0 5.08 4 7.6 4.1
Junior Guerra 11 66.0 5.05 4 5.2 2.9
Yovani Gallardo 9 45.0 5.04 5 4.5 4.5
Wade Miley 8 42.3 4.60 5 6.3 6.3
Jimmy Nelson 6 34.3 3.83 5 8.1 8.0
Aaron Wilkerson 5 25.0 5.07 5 2.4 2.4
Corbin Burnes 3 16.0 4.66 5 2.3 2.3
Luis Ortiz 3 15.0 5.48 5 0.8 0.7
Jorge Lopez 2 15.3 4.87 5 1.8 1.8
Total 162 956.7 4.77 45.7 33.7

Finally, to assess another range of rotational roles, here is how the Brewers fare against the PECOTA full depth system (1-to-10 spots, plus Replacements and Emergency starters). These sets of columns provide the “Best” and “Lowest” rotational roles assigned by the PECOTA rotation, and then the runs prevented versus each role:

Brewers vs. PECOTA Role 1 RnsPrv Role 2 RnsPrv
Chase Anderson 2 -2.1 2 -2.1
Zach Davies 2 0.8 3 8.5
Jhoulys Chacin 3 6.2 4 7.4
Brandon Woodruff 4 2.8 5 6.0
Brent Suter Replace 9.2 Replace 9.2
Junior Guerra Replace 6.3 Replace 6.3
Yovani Gallardo Replace 4.4 Replace 4.4
Wade Miley Replace 6.2 Replace 6.2
Jimmy Nelson Replace 7.9 Replace 7.9
Aaron Wilkerson Replace 2.3 Replace 2.3
Corbin Burnes Emergency 1.7 Emergency 1.7
Luis Ortiz Emergency 0.2 Emergency 0.2
Jorge Lopez Emergency 1.3 Emergency 1.3
Total 47.4 59.4

These figures tracking runs prevented by role are highly abstract, so I suggest reading them in the following way:

  • While the Brewers’ overall rotation may have an “average” outlook, once replacement rotations begin to appear in the 2018 MLB, the¬†Milwaukee replacements will have a runs prevented advantage of at least 30 runs (assessed against other teams’ replacements).
  • All things being equal, the top four Brewers rotational pitchers are average compared to the major primary rotation roles, meaning that the additional low rotation role mismatches could help the team outperform their Runs Scored / Runs Allowed by approximately three-to-five wins.
  • An alternate implication is that when the Brewers’ top four pitchers face other teams’ replacements, the Brewers will have a larger positive “role mismatch” than their opponents’ primary starters will have against Milwaukee replacements.

Assessed against other depth roles in the MLB, the Brewers 2018 rotation holds up quite nicely. The top of the rotation, even with considerable projected “decline” compared to 2017, remain average across the first four turns of the rotation. That Anderson, Woodruff, Davies, and Chacin indeed might reasonably be expected to work at least 102 starts is quite a solid proportion, and clearly leaves open space to assume injuries during the 162-game slog will occur.

Moving to the replacement and emergency starter candidates, however, the Brewers rotation begins to shine: frankly, over the course of their projected innings pitched, the Brewers depth are notably better than other replacement arms projected throughout the MLB. This should not necessarily be surprising, for where other clubs may have fewer tested options waiting in the wings, the Brewers have two 1,000+ IP veterans (Gallardo, Miley), and previously solid run preventers (2016 Guerra, 2016-2017 Suter). That’s before reaching the more inexperienced prospect depth, as well as potential injury returns (it’s not necessarily worth considering Nelson’s return, if for no other reason than to not establish false hopes or pressure), or even answering any questions about Josh Hader’s eventual MLB role (even he could “start” a game given the right match-up).


 

This type of speculative analysis should begin to show how the rotations of attrition may play out in 2018. It should accompany concrete scouting and strategic analysis of pitching profiles and performance pedigree. Thus far, it appears that the Brewers have indeed assembled rotational depth that is ready to be deployed in flexible roles (Suter as swingman; Guerra as a potential reliever / swingman; Miley as minor league option, “next-man-up”). Moreover, in terms of DRA alone, these pitchers ultimately have production levels that match #3 / #4 starters based on the actual distribution of pitching talent in the MLB. One should not get into exercises of “wishful thinking” about this rotation because this Brewers rotation does not need wishful thinking to work; however, role-by-role, these arms are better than replacement arms to a degree that one wonders which of these arms might be able to step into a more extended role should serious injuries or ineffectiveness take a toll during the season.

At worst, this 2018 Brewers rotation should be a group of arms that can hold games steady in waiting for the strong back-end bullpen to take over. At best, this 2018 Brewers rotation is simply hiding in plain sight, an unassuming depth club that is ready by design to replace starters. Once the wheels of attrition begin churning across the MLB, this is the type of rotation that can shine in fits and starts, around the margins of the team, in assuming ways.


 

Photo Credit: Joe Camporeale, USAToday Sports Images

Related Articles

2 comments on “The Rotation is Good”

Robin's Home Town

Great Analysis! Stearns and Counsell are going to battle with a 40 man roster against the majority of teams that are utilizing a 25 man team.

Early in the offseason, I was in favor of signing one of the free agent pitchers but after the Chacin signing I started to reflect on how to actually win a division on a reduced budget. The formula is to plan on using 10 average to above average starters. This minimizes the attrition risk, as you wrote about.
The Brewers have those starters now and have more in the pipeline.

For those that worry about not having ‘stud’ starters for the post season, be patin

Robin's Home Town

*be patient. Stearns has the flexibility to use our armada of starters and evaluate their performances. With the assets still in the system and payroll room, he has all options open for filling any potential holes. 2B, SP, backup C, DH type, electric bullpen piece….whatever the club need for the playoff run.

IF he had signed one of the 12MM+ guys, he would have been limited in the player pool available to him at the deadline.

Leave a comment

Use your Baseball Prospectus username