Maybe if I have trouble sleeping, I'll draw a sample for Santana tonight.

brianS: Feed th' machine.

My point in the first place was that Lohse wasn't really any more inconsistent than other pitchers of his caliber--a point you haven't addressed at all. More data, please!

the hypothesis was that Lohse's results were Poisson-distributed. So with an ERA of 4.75, that translates into the following outcome expectations over 77.2 innings: 46 zero-run innings, 24 one-run, 6.4 two-run, 1 three-run and only a small fraction of any 4+ run innings.

I think we more-or-less expect counts of runs allowed to be pretty close to Poisson-distributed. Pitchers generally are removed before they allow more than 4-5 runs in a single inning, so the upper tail will be censored. And the count of runs allowed in an inning probably isn't independent of runs allowed in the preceding inning, so Poisson isn't quite right. But possibly close enough for government work.

suppose a pitcher has an ERA of 4.5 (expected 0.5 runs allowed per inning pitched). Assume Poisson-distributed outcomes. Over 200 innings, he should produce about 121 innings in which he allows zero runs, 61 in which he allows 1 run, 15 in which he allows 2 runs, 2-3 in which he allows 3 runs.

Hmm. Even I don't believe that. Runs scored data surely is over-disbursed relative to Poisson-distributed expectations. That's got to be too many observations of single-run outcomes and not enough of zero-run and 3+ run events.

Ok. let's look at some game results for Lohse then.

From 2005:
4-08: 4 in 5.1. 4 zeros, 1 one, 0 twos, 1 three run inning
4-13: 4 in 6.1. 3.1 zeros, 2 ones, 1 two, 0 three.
4-18: 5 in 6.0. 3, 1, 2, 0.
4-28: 3 in 4.0. 3, 0, 0, 1.
5-07: 1 in 5.0. 4, 1, 0, 0.
5-14: 1 in 2.0. 1, 1, 0, 0.
5-18: 2 in 6.0. 4, 2, 0, 0.
5-23: 2 in 7.1. 6, 2, 0, 0.
5-28: 1 in 7.0. 6, 1, 0, 0.
6-03: 3 in 6.0. 4, 1, 1, 0.
6-09: 4 in 6.0. 4, 1, 0, 1.
6-15: 5 in 7.0. 5, 1, 0, 0, 1.
6-21: 6 in 4.2. 3, 0, 0, 2.
6-26: 0 in 6.0. 6, 0, 0, 0.

through June 2005, that's 77.2 innings, 41 ER (4.75 ERA) over 14 appearances (13 starts). An average of 2.93 runs per appearance (std dev of 1.8). 56+ zero innings, 14 one-run innings, 4 two-run, 5 three-run and 1 4-run inning. That's WAY too many 3+ run innings for Poisson-distributed data.

Umm, what's my point other than that this sample of Lohse's results isn't consistent with a simple Poisson-distributed DGP? Not much without looking at some other pitchers and larger samples. But I think it IS consistent with our -- other than ubelmann, anyway -- shared impression that Lohse was inconsistent.

That's been my main ongoing frustration with Torii ever since 2002. He seems to have no true self-awareness. It's apparent in his comments about himself, teammates, and the state of the team, and you can see it when he still flails at the same pitches he couldn't hit 5 years ago.

I tend to agree with Mr. Dino, though. Even if Lohse's per-inning run distribution was a bit flakier than average, there's so much inherent statistical variance there that you're not really going to notice it by casually observing.

I think that is probably overblown as well. Without any data (the best way to argue a point) I would think that he wasn't "inconsistent" any more than all pitchers are inconsistent. Generally, a pitcher with a 4.5 ERA isn't giving up a run every other inning, runs are often scored in bunches. Santana's probably the same way, the difference is that he gives up fewer runs, so instead of a Loshe-like game of 7 IP, 3 runs (all in the same inning) you get a game of 8 IP, 2 runs (both in the same inning).

That could be medium-rare crap though.

I don't have that data off-hand. Someday I'm going to learn SQL and get the retrosheet data under control. Someday.

He has an above average curve and maybe a slightly above average fastball. Anyone who says his stuff is great is either wrong or exaggurating. You can be effective with Lohse's arsenal, because he gets enough movement on pitches and generally throws strikes, but his stuff isn't great. Santana has great stuff. Nathan has great stuff. Lohse, not so much.

My point was, how much bandwidth should be wasted talking about Kyle Lohse? Answer: Zero.

My answer is apparently different than yours.

I don't think anybody ever those things about Carlos Silva. Lohse/Anderson friction had as much to do with this departure as anything. I don't think Silva will have the same rope Lohse had - even though it would appear he tried last year to do his best Lohse impersonation. Also, Terry Ryan has a very high opinion of Silva, a factor which cannot be easily dismissed, rather than the notion he is fun to hang out with.

Although with Lohse, my feeling wasn't really about start-to-start variance, but rather the (perceived) within-game variance.

As with Silva last year, I frequently got the shakes whenever Lohse got through three or four scoreless innings because I just knew he was going to fold miserably before Gardy had the sense to yank him.

I wonder what his distribution of by-inning outcomes looks like relative to others on those staffs?

FLAKE - Standard deviation of per-start SNVA for each pitcher. This was previously shown as the variance, and was used to compute the "flakiest" pitchers.

In '06 Lohse was the flakiest on the staff, but he was only given 8 starts, so that could just be a small sample size issue. In '05, Mays, Santana, and Silva all had higher FLAKE scores than Lohse. In '04, Radke, Silva, and Santana all had higher FLAKE scores than Lohse. In '03, Lohse indeed was the flakiest, but that was also arguably his best season as a Twins pitcher. In '02, Mays, Reed, and Santana were flakier than Lohse. In '01, Radke, Reed, and Milton were flakier than Lohse.

In some ways, this problem was similar to Hunter's perception problem. For some reason, people wanted to think that Lohse should be more than what he actually was. He was an innings eater who didn't get great results. He was only flaky to the extent that all pitchers are flaky. Once he got saddled with expectations to be better than you was, though, no one was giving you the benefit of the doubt (like, say, Silva gets because he is a fun guy to hang out with), it's very easy to be perceived as inconsistent.

I've never beem a huge William's fan - regardless of the whole "you have no idea" quote, but the Lohse story is the same as it ever was - lots of talent, very inconsistent results and in the past, a pretty p-poor attitude to boot.

New guy Jeff Cirillo has arrived in camp, leaving Luis Castillo and Ramon Ortiz as the only two key players still not here. E.T.A. on Ortiz remains Thursday or Friday.

I wonder how delayed he would have to be to drop out of the rotation, and how much work it would take to make that happen...

LeCroy, 31, is going to enjoy it whether his career lasts another day or 10 more years. And he has set a goal for his post-playing days. He wants to become a major league manager, and many in the Twins organization view him as a fit for that role someday.

"We all know what kind of a guy he is, how respected he is in this organization," Twins manager Ron Gardenhire said. "He's pretty close to having the body for (a manager) right now. I know he has the speed."

That said, replacing the guy at #23 on the all-time VORP list isn't going to be easy, and with the lack of internal options, it probably won't be cheap either. The good news is that the market for CF "talent" will be considerably better than this past year: Hunter, Ichiro!, Andruw Jones, and Mike Cameron are the headliners, with guys like Eric Byrnes, Corey Patterson, and Aaron Rowand on the lower tier. The bad news is that about half the MLB will probably be looking for a better CF option.

I have no idea where this leaves the Twins in 2008.

I don't agree. He's gotten a tremendous amount of positive coverage in this town. Read what Joe Christenson or Jason Williams wrote about him this week in their blogs. It's only jokers like me who rip him. You want a guy that's been maligned? Try on Kyle Lohse for size. Now, there's a guy who's been maligned. Probably deservedly so, but the treatment afforded Hunter in no way approaches what Lohse got.

Hunter also suffers from what I'll call the "Elaine Boosler" syndrome -like she said in the famous deodorant ad - never follow a better comedian. Puckett's shoes have proved too big for Torii to fill. As much as he wanted to, the Twins paid him to do and the fans longed for him to do, he was not and will never be Kirby Puckett. It would seem that he has one more crack at acheiving some Puckett-esque accolades by having a monster season and by being part of a historic third world championship run. But I fear his career as a Twin will end before the completion of this season and he'll be bolted on to some other team as they make a run for post-season glory.

I really can't say I disagree with anything you wrote. Your feature does read like Hunter's career - a maddening series of ups and downs.

My main issue with Hunter's performance is $$$ that were paid for it. I guess $$$ goes hand in hand with expectations, though.

I was surprised that in a writeup this long, you didn't mention that Torii was always running at the mouth available to the media for a comment.

The Twins seemed to own the White Sox and Hunter’s running over of Burke was symbolic of the tire tracks that the White Sox had on their backs for the last four seasons.

Or, maybe it was this:

Barry got a rare pitch to hit and he uncorked a long drive to right centerfield. Hunter and Ichiro! raced back and Hunter got to the wall, leaped and made what the baseball world thought was a spectacular catch. And certainly it was. While it might have been his most famous catch, it certainly wasn’t out of the ordinary — Twins fans were accustomed to seeing such plays.

You'll note, too, that I have foreshadowed disappointment with a Twin whose contribution is far greater than Hunter's. But, I know you don't agree with me on Hunter and hey, that's fine by me. But, comparing me to a Dick Morris hatchet job on the Clintons... I haven't seen anything that harsh at this page since I said on Tuesday that Earl Battey will be forgotten in Joe Mauer's shadow.

During his peak, if you consider Hunter to be an elite CF, he was probably worth something like 20-25 runs a season above average with his glove. Now, I have to think that the toll the turf and his injuries have taken him down to the 10-15 run range, at best.

All things considered, Hunter looks to me to be about a 3-win player after this year. (I wouldn't be surprised if he's a bit more valuable this year with money on his mind, though.) In today's free agent market, a 3-win player can be worth $9-$10M/year. Hunter would likely get a lot more than that, though. I'm going to miss him roaming CF when he's gone, though. Heck, I missed him this year when he was a shadow of what he used to be in the field while playing at 60-70% after his injury.

I don't buy it either, but I can't find much in the way of an objective record to support my observations. Tangotiger's UZR had Hunter at just 5 runs above average per 162 games from 2000 to 2003, which is probably Hunter's peak as a fielder.

If he's merely an average defensive CF, though, he puts on quite the facade. The only explanation I could possibly think up to bridge the gap between how good Hunter looks and how average his numbers are is positioning. It has been said that perhaps he likes to shade back a bit so he can make those spectacular leaping grabs at the wall. It's possible (though I personally find it unlikely) that he's giving up a lot of outs in front of him by playing back.

As an extreme example of this effect, I've played in softball leagues without fences for as long about the last seven years. It took some getting used to on my part. Eventually, after trying about every possible way to position myself in the outfield, I've just decided to play shallow. Maybe once or even twice a game I'll get burned on something over my head, but it's totally worth the extra two or three outs I can get on short fly balls/soft liners just over the infield.

Certainly it's nothing like that drastic of a trade-off for Hunter, but I'd guess that the frequency of balls hit in front of him is much greater than the frequency of balls hit to the wall. The balls hit to the wall are potentially more valuable to the offense, so that's what I'm guessing makes this whole issue (mostly) a wash.

At any rate, if Hunter's positioned poorly, during his peak that had to be his only defensive liability. He was fast, ran good routes, and had good instincts. His numbers probably suffer somewhat from having a solid infield defense through most of '01-'03 and having pretty rangy corner outfielders flanking him for much of his time in Minnesota. It's a lot easier to look good by the numbers when your teammates are letting balls drop so you get more opportunities to make outs in subsequent at-bats.