Al Central Rotations Ranked by Projected Wins Above Replacement

Hello all. You probably don't remember the enjoyable interactions I had with some (many?) of you during the glorious few hours (for me, anyway) when Gutierrez-for-Teahen seemed like a possibility... Anyhoo, since then, I started writing for Driveilne Mechanics, and I'm currenlty in the midst of a series where I go through each division, take a look at the starting pitching by averaging CHONE, PECOTA< and ZiPS, and project a WAR... there's some other sabermetric pixie dust in there, but you can read about it in the post if you're interested. I know that starting pitching if of particular concern for the Indians, but the results are at least interesting...

almost 3 years ago Matt Klaassen 26 comments 0 recs |

Story-email Email Printer Print

Comments

I don’t think it makes much sense at this point to project the Indians starting rotation value across an entire season, because I think it’s basically a shot in the dark as to how the innings get distributed among the various pitching options. I think it is possible to think about the projected rate stats of the guys and how those might translate over the course of a season. If you look at RAR/IP, here are the rankings from best to worst:

C Lee 0.190
F Carmona 0.141
J Westbrook 0.140
S Lewis 0.130
A Laffey 0.122
D Huff 0.122
J Sowers 0.089
A Reyes 0.081
C Pavano 0.069

by APV on Mar 23, 2009 6:05 PM EDT reply actions

So the key obviously is:

1. Maximize the starts/innings we get out of the top 6.

2. Minimize the number of starts/innings given to anyone outside who isn’t demonstrably a different pitcher than what the projections are expecting.

Well, that's just, like, your opinion, man.

by Jay on Mar 23, 2009 6:22 PM EDT up reply actions

Bingo. Although I think there is reason to be optimistic someone like Reyes can substantially overperform his projection. And although I’m pessimistic about Pavano, I don’t think there’s any reason to think a projection of him is at all meaningful given his lack of pitching over the past 4 years or so.

by APV on Mar 23, 2009 6:31 PM EDT up reply actions

Yeah, that’s a sharp cutoff.

The tricky part of your scheme, Jay, is that it argues for a quick move away from guys number 8 and 9, which would result in cutting our depth. Although Reyes might be the best bet for the guy to move way above his projections….and I swear I wrote this right as APV’s popped up.

by dgcambridge on Mar 23, 2009 6:33 PM EDT up reply actions

great minds, blah blah blah

by APV on Mar 23, 2009 6:34 PM EDT up reply actions

Yes, putting Pavano and Reyes a very short leash might well cut into our depth. I think we have to take that risk. It’s also a big risk to allow them to continue pitching if they’ve done badly, so at that point, it’s really just one risk vs. the other. Gets back to the point made at the end of our “What Went Wrong” article … Plan B can’t work if the club insists on riding Plan A into the ground.

if we cut one of those guys, we still have eight starters (Zach isn’t noted above) and nine once Jake returns. I don’t think you keep running a bad starter out there just to be 9-10 deep instead of 8-9 deep. This is not to minimize the importance of depth, but rather just to emphasize that getting more Wins and fewer Losses is the whole point of having depth or anything else.

Well, that's just, like, your opinion, man.

by Jay on Mar 23, 2009 6:42 PM EDT up reply actions 1 recs

agreed. Plus, if we did cut Reyes/Pavano loose and subsequently lost 3 or 4 of our current starters, we could find another replacement, be it a minor leaguer or a free agent or a small trade that wouldn’t be too far off from what a bad Pavano could have given us in the first place. E.g., Red Sox with Paul Byrd last year. This is why I believe depth is important over the long term, but don’t think its all that important over the course of one year. Ultimately why I would have a very short leash for Pavano and Dellucci.

by Ryan Kelsey on Mar 23, 2009 7:47 PM EDT up reply actions

Just to reiterate, the author stated he was using BPs projections for innings, not what he thought would happen for each team.

The projected 2009 for each pitcher is generated by averagfing the projections of the three best publicly available projection systems: CHONE, ZiPS, and PECOTA. The projected FIP for CHONE and ZiPS are from the player pages at FanGraphs. I generated the PECOTA FIP myself from the big spreadsheet.

Playing time/depth chart projections are based on Baseball Prospectus’s depth charts (not the PECOTA weighted mean) in order to account for spot starters as well. These change every week; the playing times used are those that were up when I was writing the present post.

by talonk on Mar 23, 2009 7:16 PM EDT reply actions

thanks

and I appreciate the comments here, too. I hope people take time to read it if they find this sort of thing interesting. I used BP’s depth charts for playing time predictions because I simply don’t know enough about each team to do it and make adjustments. A “fan survey” at SBNation would the best way, but would take a lot of time and work.

I could have just gone with a “top 5 FiPS” for each team, but that would also depend on figuring out the rotations. I wanted to do a “survey” that also accounted for the relative depth of a teams’ pitching: how many innings they get out of their best starters, how much their fill-in ##6-8 have to contribute, and if so, what they will do.

What was interesting to me about the AL Central is that while no team has what I’d call a “strong” group of starters, all things considered, none of them look terrible, either, not even Detroit (who probably have the best offense). Although I don’t get into this in particular in the post, Cleveland is interesting to me because so many of the guys seem to have a lot of variability — whether it’s Lee coming off of an unbelievable career year, Carmona coming off of a surprisingly bad year, the mystery of how much Westbrook will pitch, the joy of Pavano (seriously though, I think that was a smart deal for the money and potential upside), and the array of #4/#5 type guys like Sowers and Laffey.

Bringing you more-or-less replacement level analysis and commentary to Driveline Mechanics and elsewhere since sometime in 2008.

by Matt Klaassen on Mar 24, 2009 1:48 PM EDT up reply actions

Thanks for stopping by. I didn’t mean my comments to be a criticism of your efforts, more a comment on the challenge the Indians rotation presents when it comes to projection at the moment.

by APV on Mar 24, 2009 2:17 PM EDT up reply actions

No prob, you guys are a lot of fun. I'm just glad I didn't get the "biased agains [insert team]!!!" thing that... sometimes happens some places

if you think the Indians are a problematic, check out Toronto — pretty good rotation, actually. Could be great if Janssen and McGowan can contribute, or could be the Canadian version of the Astros (Halladay/Oswalt + sidekick + garbage).

Every team in the Central seems to have one decent to very good pitcher who can be “trusted” (Lee, Verlander, Danks, KC, curiously, has two in Meche and Greinke, maybe ChiSox, too), one “sidekick” who is coming off a bad year or overly good year who will at least be a bit above average (Carmona, Bonderman, Danks?, Floyd comes back to earth, maybe h im), and then a pu-pu platter of #4/#5 types guys who are young, guys coming off injuries, and then recovery projects, dudes who don’t strike people out, etc.

Like I said, any team in the Central could end up with the best starters at the end of the year, or the worst.

Bringing you more-or-less replacement level analysis and commentary to Driveline Mechanics and elsewhere since sometime in 2008.

by Matt Klaassen on Mar 24, 2009 3:13 PM EDT up reply actions

We all hold out hope that Carmona becomes the best pitcher in the league. From a pure “stuff” standpoint, I think his potential is Cy Young

by Roger Dorn on Mar 24, 2009 3:18 PM EDT up reply actions

His potential is “Hide the women and children, The Fausto is coming!”

by FredOx on Mar 25, 2009 11:49 AM EDT up reply actions

Isn’t this the way most teams are every year?

by Voltaire on Mar 25, 2009 4:34 PM EDT up reply actions

I think the Indians are looking at how to get 1400+ innings out of an aggregate staff this season. Last year, with Sabathia and a terrible bullpen, the Tribe pen had the third-fewest innings of any team in the AL (209.2 IP, just two-thirds of an inning behind the Jays). That’s almost 70 innings fewer than the Rangers, who I believe pitched the most bullpen innings.

Let’s say Lee and Carmona combine for 400 innings (optimistic?) and Pavano and Reyes are good for 200 combined. Maybe you hope for 250 innings from the pen. That leaves 550 innings for Laffey, Jackson, Sowers, Lewis, et al. You have to count on one of those dudes throwing 140 innings. And you have to hope you find a Byrd-like guy midseason. Losing Sabathia gives up a lot of innings and days off for the pen.

by odradek on Mar 25, 2009 12:11 AM EDT reply actions

The bullpen’s innings are determined not by their own quality but by the starters; it’s pretty much a one-way relationship. In fact, a bad bullpen often will throw more innings than a good one, because the fact that they have to throw so many innings is one of the main reasons they’re bad (tiring key guys out and having the worst pitchers on the team throw many more innings).

I’m not sure where you’re getting those numbers, but ESPN is showing 445.1 for the 2008 Indians. Our top four guys had 269 IP by themselves. The last five years, IP and ERA:

08 – 445.1, 5.13
07 – 441.1, 3.75
06 – 422.2, 4.73
05 – 446.0, 2.80
04 – 540.1, 4.90

Bottom line, you’re trying to get 1000 innings out of your starters — averaging a little over 6 IP. That might seem like a lot, but it isn’t — you ought to get as many 8-9 IP starts as 3-4 IP starts. Trying to be conservative and allow for some injury risk, maybe something like:

Lee-Carmona – 375
Laffey-Lewis – 225
Pavano-Reyes – 175
Sowers-Huff – 175
Jackson-Saarloos – 50

Well, that's just, like, your opinion, man.

by Jay on Mar 25, 2009 11:07 AM EDT up reply actions

I found the numbers on a NY Daily News blog from 2008, and I think they were partial stats. But you made my point for me, I think. Shapiro knows his rotation is dodgy. I don’t think he looks in terms of number four starters, etc. I think he’s trying to figure out how to get to 1400 innings. He also knows he’s lost an innings eater, and that there will be more 3-4 IP starts than 8-9 IP starts.

If the pen has to pitch 540 innings, it will suck—both because (as you say) the arms will be worn out and because it indicates trouble with the starting pitching. Look at the IP for the pens last year and in 2006: Wedge was reluctant to go to them.

How well will the Tribe play in its 225 innings of Sowers, Huff, Jackson and Saarloos? That’s maybe 40 starts? A quarter of the season. That strikes me as the critical question in 2009. Though I am concerned again by the offense.

by odradek on Mar 25, 2009 11:24 AM EDT up reply actions

From those four guys, 225 innings is probably more like 50 starts. Leaving Huff aside, there isn’t much mystery about how good the other three will pitch; in terms of a strategic plan for the season, their variability is more about how much we can limit their innings. Lee or Carmona might give us 225, Laffey or Lewis might give us 175, Pavano or Reyes might give us 125. If any two of those guys outperforms in terms of health and IP, that’s 100 extra innings, which eliminates Jackson-Saarloos and suppresses Sowers-Huff.

Well, that's just, like, your opinion, man.

by Jay on Mar 25, 2009 11:35 AM EDT up reply actions

I think that’s the way to look at it. And maybe you get 60 innings from somewhere else, and a midseason trade for someone for another 80 innings, and you don’t have to see Sowers unless you’re really verdreht.

And while I am not among those who think Westbrook will be worth much this season, he might be good for 50 innings, maybe.

by odradek on Mar 26, 2009 12:17 AM EDT up reply actions

You mean, we use more than 10 starters? No. There are only 6-7 games per week, you’re not going to use more than 10 starters.

Funny how I forgot about Westbrook, though. Clearly he’s in the mix for maybe 75 projected innings.

Well, that's just, like, your opinion, man.

by Jay on Mar 26, 2009 12:56 AM EDT up reply actions

If we use the ten starters listed above (a distinct possibility), and Westbrook comes back and the team trades for a midseason veteran, 12 isn’t impossible.

The team used 12 starters last year, admittedly occasioned by departures of C.C. and Byrd.

by odradek on Mar 26, 2009 7:33 PM EDT up reply actions

I like how you didn’t count Mastny. Integrity.

Well, that's just, like, your opinion, man.

by Jay on Mar 26, 2009 8:53 PM EDT up reply actions

I forgot about that start: Just give the team three innings, Tommy Boy.

by odradek on Mar 28, 2009 1:23 AM EDT up reply actions

Interesting how we had a stable rotation for a whole month after Byrd was traded, 28 games:

Lee, Carmona, Sowers, Reyes, Jackson

After that, SLewis replaced Reyes, but otherwise it stayed the same. Bullington’s two starts were both spot starts, where nobody else in the rotation would have been going on full rest. I’m not sure how Ginter got into the picture (late July) or why we dumped him (early August), nor do I care.

Well, that's just, like, your opinion, man.

by Jay on Mar 26, 2009 9:00 PM EDT up reply actions

We literally had to use all our starters on the 40-man by the end of the year, based on a combination of who was on the DL, who had a day-to-day issue, and rest.

by Voltaire on Mar 27, 2009 3:19 PM EDT up reply actions

That 2005 pen was really good. They had 5 guys in the pen with a FIP under 3.25 (Rhodes, Betancourt, Howry, Cabrera and M. Miler), plus David Riske (3.10) and Bob Wickman (2.47) massaging there way into very solid ERAs. Those 7 guys combined for 381 innings.

by APV on Mar 25, 2009 11:36 AM EDT up reply actions