Didnt James himself say WS is weak for one year things like this? esp. player to player one yr comps?

Why are people insistent in not considering Zambrano to be the Cubs' number one pitcher?

Actually, it's tough to blame them as Chicago doesn't seem to understand this fact either.

Why are people insistent in not considering Zambrano to be the Cubs' number one pitcher?

He's not as glamourous as the other two (I don't think anyone considers Maddux to be their best pitcher). Prior has the awesome body, the fluid mechanics and the stuff. Kerry Wood throws hard.

Carlos Zambrano is fat, really ugly, and just breaks bats with basically one pitch.

To be fair, the guy rated the pitchers rotation slots by GS and IP..hence Pedro a #2 to i dont know who

And Ben Sheets a #4

Jeez, if the DBacks had only had a league average fourth starter (average for the slot, not the whole league i.e. 6 win shares), they would have been in the top 5 rotation wise.

Although I'm sure just about any team could say this about one pitcher, I'm also sure that those pitchers weren't multi-million dollar offseason signings this year. On the postive side, I didn't realize that by win shares B. Webb had the fifth best season in the national league. Not bad for the twenty-six year old whom BPro pegged as being on the verge of complete collapse.

In response to #1, I don't recall that. I remember him saying that mid year WS would be pretty funky though.

That's an odd grading system. Like any such system we get somewhat silly distinctions ... a difference of 1 WS is a B+ vs a B as if 1 WS is actually meaningful. But there's also some odd inconsistencies. For example 14 WS earns Beckett only a B as a #3 but is also good enough for Vazquez to get a B as a #2 and Glavine to get a B as a #1. The B range for a #1 is 14-18, for a #2 it's 14-16 (at most). Burnett gets a C+ as a #2 with 12 WS, while Capuano and Marquis get Cs with 12 WS. (That's probably just a typo)

Not that it matters a lot, but turning a continuous measure into a categorical one is generally only useful if there are clear distinctions between categories -- i.e. that the difference between two grades isn't something as meaningless as a single win share. Not that this has ever stopped teachers (including myself).

I find this a ridiculous system. It artificially implants the '1-5 starter paradigm', so that Oliver Perez is apparently Pittsburgh's number 5 starter, and Zach Duke's work receives no credit in their rotational standings. Who consider Greg Maddux to be the Cubs #1?

I can see giving the Cardinals an A at #5 (or can I?), but I can't see giving Morris an A for 7 WS. IOW rating staffs may be a different exercise requiring a different methodology than rating the individual pitchers (or it may not). Yes, the whole thing feels a bit odd.

I ignored the grades, as they seemed to be generated in relation to whether the pitcher was "#1" or "#2" etc. Which means if you just give it the once over it makes little sense. Especially as the way 1/2 etc. was defined seemed... arbitary, with little relation to actual position on the staffs. Greg Maddux as a #1 being a prime example.

Why use GS/IP rather than....Win Shares! to determine #1, #2, etc?

I guess I don't understand how the grades work. Does the grade in each slot depend on the team's pitcher's WS relative to other pitchers in that slot? And if so, what's the breakout based on - standard deviations from the mean? Is this a system they've used before that's explained elsewhere in more detail?

And if he recognizes the inherent problem of using total win shares vs pitching win shares b/c of the batting component, then why not do the easy thing and just compare them by pitching win shares? I'd learn a lot more from that than I do now, in part because I don't know which of these guys (besides Carpenter) is getting dinged for his lack of batting prowess (and by how much) and in part because I don't care what kind of hitters they are.

I guess I don't understand how the grades work. Does the grade in each slot depend on the team's pitcher's WS relative to other pitchers in that slot? And if so, what's the breakout based on - standard deviations from the mean? Is this a system they've used before that's explained elsewhere in more detail?

I'm fairly certain the grading is relative to the other pitchers in the slot, but whether the grading was on STDEV from the mean... I don't know. But it's as good a supposition as any.

Didnt James himself say WS is weak for one year things like this? esp. player to player one yr comps?

I don't believe he said that. It's true of course, but I don't think he said it.

I believe Win Shares and Warp and XYZ Uberstat all have a place, but it's place isn't in doing player to player single year comparisions. It's useful when looking at careers in aggregate, and looking at large groups of players.

When doing specific player by player comparisions you get a much better understanding of the players worth by actually examining the specific of his relative stat line rather than rolling it up into one uber metric.

Any system that gives Chris Carpenter a B for 2005 is so f*cked up that it can't be taken seriously.

That should have been a red flag to the author that something is flawed here.

Any system that gives Chris Carpenter a B for 2005 is so f*cked up that it can't be taken seriously.

That should have been a red flag to the author that something is flawed here.

- Absolutely....Chris Carpenter won the CY YOUNG award and went 21-5, which clearly, without any possible modicum of doubt, makes him the best pitcher in the National League, and this guy is saying he is a "Grade B" pitcher. This would make Dontrelle Willis a B- pitcher (after all Willis finished 2nd in Cy Young voting and lost 10 games!) and would probably make Clemens, and his wimpy 13-8 record a C- or D+ ..... I'm curious how his system ranks Bartolo Colon, who was the AL's best pitcher by a long shot this year.

Thanks for the feedback, everyone. Just to clarify: Yes, I know that win shares is flawed and/or isn't necessarily the greatest for single-season analysis. And yes, trying to group starters by flight is a bit of a contrivance. And yes, it's crazy that the CY got a B. I'm with you.

Here's a little bit of explanation, and I'm open to hearing about ways to improve upon what I did or how you might've done it differently:

1. Win shares as grading tool: It was easy, it was fun and win shares are useful in their own right. If anything, my method may have been helpful in confirming some of your opinions about just how useful WS are.

2. Determining "flights": I almost didn't even try to rank starters individually for the very reason that it was going to be a contrivance figuring out who was "#1," "#2," etc. As I said in the article, I'm not aware of any official designations, so I went with hard data. It's difficult, even if I had polled each manager, because what exactly IS a #1 starter? A #4 starter? What about a rookie like Duke (or Prior previously) who comes on the scene and gets a few "back-of-the-rotation" starts, but excels? That's partly why I didn't use WS to flight the pitchers (I considered it). Also, there's no physical law that says the guy a team considers their "ace" is not going to fare poorly. The Pirates, for example, had Oliver Perez start opening day. Does that make him their #1? Probably not, since he started only 20 games. Again, if anyone has any good hard data that would give a better picture, I'm open to it.

3. Grading method: I graded on a curve, using the average as the break-even point between B and C, and one standard deviation for each letter grade away from that.

I've been thinking of reworking Ted Oliver's old Wins Above Team stat to compare teams records by starter to the records of a teams spot (replacement) starters. It's not very accurate for a single season, to say the least. Houston's record was 15-17 in Clemen's starts. WHat did his bulpen have against him ;)? That's 2-9 in games where he was decisionless.

WHat did his bulpen have against him ;)?

The same thing his Red Sox bullpen had against him...35 blown saves in Clemens' games from 1984 to 1996.

Thank you Jeff Reardon...

I always liked Jeff Russell myself, but thats what happens when your first exposed to the 93-94 edition of the Sox.

I always liked Jeff Russell myself, but thats what happens when your first exposed to the 93-94 edition of the Sox.

Man, that Hobson era was disappointing. I had already been exposed to the Sox for @ 18 years by then.

I wasn't even really into the Sox until 1994, and the strike didn't make sense to an 8 year old child. At that point of course, into the Sox meant loving John Valentin and Tim Naehring and Mo Vaughn.

I know about Jeff Russell and the 1993 team more from the videogames that were released in 1994-1995 than from the actual games on television. My initial baseball mind database came from videogames, heh.

Anybody but Dusty Baker Posted: November 17, 2005 at 09:53 AM (#1735471)
Any system that gives Chris Carpenter a B for 2005 is so f*cked up that it can't be taken seriously.

That should have been a red flag to the author that something is flawed here.

- Absolutely....Chris Carpenter won the CY YOUNG award and went 21-5, which clearly, without any possible modicum of doubt, makes him the best pitcher in the National League, and this guy is saying he is a "Grade B" pitcher. This would make Dontrelle Willis a B- pitcher (after all Willis finished 2nd in Cy Young voting and lost 10 games!) and would probably make Clemens, and his wimpy 13-8 record a C- or D+ ..... I'm curious how his system ranks Bartolo Colon, who was the AL's best pitcher by a long shot this year.

I love it when a cub fans obtusely(which is normal for them of course) take a comment that is factually correct, then apply their own logic and interpretation of the comment to it to try and make a point that had nothing to do with the original comment.

name the five most successful pitchers in the nl last year.... If carpenter isn't on the list you are being a homer more than likely. especially when you go by a combination of rate/playing time. I don't think the guy that questioned carpenters ranking that you originally responded to, claimed that carpenter was the best, but at no point in time can you look at the guy that was top three in pretty much every meaningful pitching category and not think he was pretty good.

sure he wasn't as dominant and great as kerry wood and the great prior, but really, who is?