Definitions of #1 starter, #2 starter, etc.

[dadofandrew]

Jeff Sackmann does the legwork and shows what a true #1 is, and so on.

 

This will hopefully help eliminate those who think a 4 ERA guy is a #4 SP. Those who feel Sheets is anything but a #1 (when healthy of course) are off base as well. This also explains why a 200 inning, 5.25 ERA guy has a huge amount of value.

 

Excellent piece by Jeff.

 

LINK

[rluzinski]

Very nice. Someone here posted ERA+ averages for each spot in the rotation and it seemed to show the same thing (back end of rotations stink).

 

I don't think he should assume every spot in the rotation get's one-fifth of the starts, however. Also, the #5 spot looks so bad because a significant number of those innings were actually accumulated by a #6 or #7 pitcher. So, comparing a pre-seaon rotation to those averages is a little misleading, especially for the backend of the rotation. So, resist the urge! http://forum.brewerfan.net/images/smilies/smile.gif

[1992casey]

This is a wonderful piece.

 

I wish he would have used a stat that's better than ERA. Then again, even though it's not a great stat, it's probably a fair tradeoff for the sake of simplicity.

 

Jeff promises more on Friday. It should be interesting.

[Pointer1VB]

When you're given a sample of 4860 starts throughout the league, ERA will do fine.

[rluzinski]

When you're given a sample of 4860 starts throughout the league, ERA will do fine.

 

I'd agree with that. Things like, the subjectiveness of errors, BABIP and left-on-base % all kind of balance themselves out in the aggregate.

[brettac1]

Just for reference, if this thread were to get very active:

 

http://tinypic.com/30vd3pl.jpg

 

The article also says that the Brewers were closest to the league norms last year. Hopefully they wind up above the norms this year, they probably should be.

[dadofandrew]

What it tells me is that very few teams have any depth at all. As soon as someone goes down, they are lost, and like the Crew, have to settle for kids that re not ready.

 

That's really been a strength of the A's, they always seem to have a Sarloos type waiting to go, unspectacular, but serviceable.

[brewerfan82]

Very nice article. It really bodes well for the Suppan signing as well.

 

Using his same technique with the current Brewers three-year averages, you get the following:

 

Current Brewer Three-Year Averages

Ben Sheets: 3.13 ERA (24 starts)

Jeff Suppan: 3.95 ERA (32 starts)

Chris Capuano: 4.17 ERA (29 starts)

Dave Bush: 4.28 ERA (24 starts)

Claudio Vargas: 5.08 ERA (22 starts)

Carlos Villanueva: 3.69 ERA (24 starts)*

Zach Jackson: 5.40 ERA (25 starts)*

Ben Hendrickson: 7.41 ERA (28 starts)*

 

* used Major League ERA from 2004-06, but average starts from all professional levels

 

NL Averages (from article)

Starter GS ERA

#1 Total 32 3.51

#2 Total 32 4.04

#3 Total 32 4.57

#4 Total 32 5.11

#5 Total 32 6.26

 

Brewers with Suppan

Starter GS ERA

Sheets 24 3.13

Villanueva 8 3.69

#1 Total 32 3.27

 

Villanueva 16 3.69

Suppan 16 3.95

#2 Total 32 3.82

 

Suppan 16 3.95

Capuano 16 4.17

#3 Total 32 4.06

 

Capuano 13 4.17

Bush 19 4.28

#4 Total 32 4.24

 

Bush 5 4.28

Vargas 22 5.08

Zach Jackson 5 5.40

#5 Total 32 5.00

 

 

Brewers without Suppan

Starter GS ERA

Sheets 24 3.13

Villanueva 8 3.69

#1 Total 32 3.27

 

Villanueva 16 3.69

Capuano 16 4.17

#2 Total 32 3.93

 

Capuano 13 4.17

Bush 19 4.28

#3 Total 32 4.24

 

Bush 5 4.28

Vargas 22 5.08

Jackson 5 5.40

#4 Total 32 5.00

 

Jackson 20 5.40

Hendrickson 12 7.41

#5 Total 32 6.15

 

Of course there are a number of problems with the above analysis (Gallardo will probably be there instead of Hendrickson, the number of starts for the replacement players will be different, there will be more than two or three replacement players needed, park effects, etc.), but it does give you a rough idea of how good our rotation's outlook is and how much value the Suppan acquisition adds to it.

[Ennder]

The problem with that is that ERA doesn't work nearly as well for individual pitchers. I don't expect a 5+ ERA out of Vargas with his move. I do expect a 4+ ERA out of Suppan with his, in fact I expect around a 4.50 for him. I'd be really iffy about projecting anything for Villanueva based off of that limited sample.

 

I do like the chart though, it does really show the extra depth Suppan gives us. Almost every team has at least one starter go down each year and when that happens to us this year we won't take a huge hit in production unless its Sheets.

[rluzinski]

What it tells me is that very few teams have any depth at all. As soon as someone goes down, they are lost, and like the Crew, have to settle for kids that re not ready.

 

Or Lima.

[pogokat]

assuming that melvin has access to stats like this, i think this article goes a long way to demonstrate why the brewers are reting much easier with suppan in the fold...if you can dump those starts from the 6+ ERA guys, then the brewers would have probably won another 5 games last year...

 

shazam!

[SoCalBrewfan]

A quick post, as I'm barely back in town yet and haven't caught up with everything. Here's the link to the ERA+ numbers referenced above, and the results.

 

Quote:

---

#1 slot - 118 ERA+

#2 slot - 104 ERA+

#3 slot - 97 ERA+

#4 slot - 90 ERA+

#5 slot - 78 ERA+

---

 

www.baseballthinkfactory....ion/44951/

 

It's essentially very similar information (as it should be, whenever you normalize), showing that a league-average ERA is somewhere between a #2 and #3 starter.

[endaround]

Keep in mind of course that a league average starter will have an ERA+ below 100

[Brew Man Dan]

I think that with the ERA and ERA + numbers that you can conclude that the Brewers have a above average pitching staff

[EdgarDiazRocks]

Or Lima.

 

Word.

 

Then again Aaron Sele was effective for the Dodgers for about a month and a half last year.

[rluzinski]

I think that with the ERA and ERA + numbers that you can conclude that the Brewers have a above average pitching staff

 

Just keep in mind that, unless you assume zero missed starts by the starting rotation for the entire year, an average starting rotation will likely get below average production for the year. The only way it wouldn't happen is if their injury replacements end up being better than their expected production).

 

I'd be interested to see how many starts are lost to injury in a 162 game season on average. 10%? Greater?

[1992casey]

Quoth Pointer1VB:

When you're given a sample of 4860 starts throughout the league, ERA will do fine.

---

Quoth Russ:

Things like, the subjectiveness of errors, BABIP and left-on-base % all kind of balance themselves out in the aggregate.

---

Quoth Ennder:

The problem with that is that ERA doesn't work nearly as well for individual pitchers.

---

Pointer1VB and Russ make me feel better about using ERA for the aggregate list of all pitchers and setting cutoffs. Ennder's comment still concerns me for using ERA for the individual pitcher part of the equation. And I'm happier with ERA+ to weed out ballparks.

 

This brings me to some questions:

1. Can we use Jeff's baseline ERAs and place individual pitchers' FIPs, xFIPs, or DIPS next to them? Or do we have to do the baselines over in terms of FIP, xFIP, or DIPS?

2. Is there someone out there that can take Jeff's ERAs and turn them into ERA+s? I'd like to compare them to ERA+s that SoCal posted. I can do the formula, but I'm not sure which 'league factor' to use.

[SoCalBrewfan]

You can get league era's from baseball reference:

 

2006 NL: www.baseball-reference.co...2006.shtml

League ERA: 4.49

 

2006 AL: www.baseball-reference.co...2006.shtml

League ERA: 4.56

 

That the AL is only 0.07 runs higher is pretty surprising to me, but I guess three of the best hitters' parks are in the NL (Colo, Hou, Ariz).

[dadofandrew]

Not to mention how weak the DH's were last year. Hardly anyone had a legit starter as a DH, and it showed.

[SoCalBrewfan]

So I get ERA+ based on Jeff's NL numbers as follows:

 

#1: 127

#2: 111

#3: 98

#4: 88

#5: 72

 

That looks to be a slightly bigger spread for 2006 NL than 2005 AL, but otherwise pretty consistent. I think that suggests that the average #5 starter is basically little better than replacement level if I'm remembering the definition correctly (70% of league average?). It was certainly true in Milwaukee last year...

[TooLiveBrew]

Guys, I don't want to be the poop in the punchbowl, and I know the article is well-done, and so is the individualized research here.

 

But I, for one, just don't care about numbering SPs, just as I don't care about ranking players against others at their positions. I understand the arguments in doing so, but I just don't buy into them.

 

At the end of the game, all that matters is that the production was created, not from which position it came.

 

Likewise, labeling a guy 1-5 does nothing but make us feel better about our uses of stats. Defining a guy's performance gives one a good idea of where he ranks, relative to other SPs (just like all other statistics). So, if the goal is a more simplified form of categorization and regognition, fine, sounds good.

 

But if the point is to craft your rotation around these numbers, you're wasting your time, IMO. Just like the notion that a "leadoff" hitter is super important (he is only guaranteed to 'lead off' once more than any other batter), the numerical ranking of a team's SPing is irrelevant once you get into the grind of the MLB season.

 

Guys get hurt, managers use off-days to restructure based on pitching matchups, and general scheduling leads to the notion of #s 1-5 being moot by the time you hit June or July (if not earlier).

 

It's not that I don't think that 1-5 rankings, like any other stat, aren't useful (obviously the ranks are merely new/different statistical interpretations of other, already existing, effective, pitching stats). It's just that the idea that you can then use these numbers to fixedly structure your pitching rotation is asinine. SP #'s aren't any more or less useful than, say ERA+, FIP, or WHIP

[dadofandrew]

Bumping this so folks can read Part 2. It seems to indicate that playoff teams get much better pitching from their bottom 20% of starts than other teams.

 

It's almost like an ad for SP depth.

 

LINK, Part 2

[rluzinski]

It's almost like an ad for SP depth.

 

Or it shows how well a healthy rotation for an entire year can do.

[SoCalBrewfan]

At Primer, another study along the same lines, with similar methodology and conclusions:

 

www.baseballthinkfactory...._analysis/

 

One thing that these results suggest is something that Bill James said long ago, and that others have repeated time and time again: an average starting pitcher who can pitch a lot of innings is a very valuable player. That belief is borne out by the free-agent market, in which someone like Suppan is worth $10M+. That's a result that the Brewers made painfully clear last season...those guys who take the ball 33 times and pitch 6 innings a start are crucial to a successful team, even if they put up a 4.5 ERA. Why? Because if they can't pitch, their replacements are going to give up a lot of runs and wear out the bullpen, with the strong likelihood of a self-enhancing feedback loop. (In other words, shift a reliever into the rotation or move up an ill-prepared guy from AAA, have another game with lots of runs, and thus lots of pitches thrown, the whole staff is worn down, causing further injuries and/or ineffectiveness, causing you to go deeper into your pitching pool, and so on until you have the 2006 Brewers.)

[Fike85]

So, is this guy a Brewer fan?