PDA

View Full Version : Reds Should Trade Stanton to Mets or Red Sox



Kc61
09-20-2007, 11:11 PM
Mets bullpen is getting destroyed these days. They have ten games left and can use any reinforcements they can get. If Stanton has cleared waivers, as I assume he has (or would), trade him to the Mets for a minor leaguer. Stanton wouldn't be eligible for post-season, but Mets probably won't care at this point.

Another possibility is Boston, which is suffering from the same bullpen woes.

Matt700wlw
09-21-2007, 12:14 AM
Trade him anywhere....for anything.

Even if it's nothing.

RedsManRick
09-21-2007, 12:19 AM
If you can get somebody to assume his 2008 salary, then by all means.

Team Clark
09-21-2007, 12:59 AM
Which GM do you want to see fired first? That would be the team he should go to.

RedsManRick
09-21-2007, 01:16 AM
Which GM do you want to see fired first? That would be the team he should go to.

I hear Houston has a new GM with a thing for overrated veteran middle relievers.

blumj
09-21-2007, 07:15 AM
The Red Sox already did that once, they traded for Mike Stanton with just a few games left in the '05 season. Of course, he was not 40+ and under contract for '06 at the time. But, you never know, maybe they want to break the record for most 40+ year old pitchers on a staff, since they already have 3.

jojo
09-21-2007, 07:30 AM
So the Mets and BoSox have pens that are on fire and the Reds come to them offering a tanker of gasoline. Even at a discounted price, it seems a little unlikely that either would jump at the offer. Besides Rick White is a free agent again... Gasoline is freely available when it comes to powering major league bullpens.

redsmetz
09-21-2007, 08:20 AM
If the receiving team is taking on his contract next year, I say do it. I've always felt folks on RZ were too hard on Stanton. He's had a number of outings in which he stunk up the universe - take away five outings in which he was awful and he's got a decent 3.10 ERA. I think he can help one of these two teams.

jojo
09-21-2007, 08:58 AM
If the receiving team is taking on his contract next year, I say do it. I've always felt folks on RZ were too hard on Stanton. He's had a number of outings in which he stunk up the universe - take away five outings in which he was awful and he's got a decent 3.10 ERA. I think he can help one of these two teams.

Stanton at this point in his career is just an arm. His peripherals suggest middle reliever. He hasn't had platoon splits to suggest he'd be good as a situational guy for a couple of years. Whatever club that would trade for him would be on the hook for a minimum of $3.5M. They could expect to get similar production from several arms in their own system at the cost of league minimum. Neither the Mets nor the Redsox should consider Stanton an answer.

nate
09-21-2007, 09:18 AM
I don't think Stanton is a good pitcher but it seems like when he gets into trouble, he gets doinked, blipped and flared to death. His BABIP (.362) kind of bears that out. Of course, its his fault when hitters are batting .317 against him.

By all means, trade him if you can but we all know the minimum number of participants required to tango.

RFS62
09-21-2007, 09:22 AM
What have you got against the Mets and the Red Sox?

Kc61
09-21-2007, 09:37 AM
What have you got against the Mets and the Red Sox?


Funny.

PuffyPig
09-21-2007, 10:55 AM
I don't think Stanton is a good pitcher but it seems like when he gets into trouble, he gets doinked, blipped and flared to death. His BABIP (.362) kind of bears that out. Of course, its his fault when hitters are batting .317 against him.



I'm guessing that the .362 BABIP number has alot to do with that .317 BAA. If his BABIP was a more normal .306, his BAA would be about .247.


His DIPS ERA is a more reasonable 4.25.

He's had bad luck to be sure. He may not be great, but he's also not as bad as he has looked.

redsmetz
09-21-2007, 11:07 AM
I'm guessing that the .362 BABIP number has alot to do with that .317 BAA. If his BABIP was a more normal .306, his BAA would be about .247.


His DIPS ERA is a more reasonable 4.25.

He's had bad luck to be sure. He may not be great, but he's also not as bad as he has looked.

As I noted earlier, if you look at his stats without his five worst outings (3-4 runs allowed in each), his ERA is a respectable 3.10 - that's out of a total of 66 appearances. I can recall a few times he let up some dink hits that fell in. It's what makes baseball both fascinating and maddening. I'm not saying the guy's a superstar, but he's not the horror story some paint him to be.

jojo
09-21-2007, 11:55 AM
As I noted earlier, if you look at his stats without his five worst outings (3-4 runs allowed in each), his ERA is a respectable 3.10 - that's out of a total of 66 appearances. I can recall a few times he let up some dink hits that fell in. It's what makes baseball both fascinating and maddening. I'm not saying the guy's a superstar, but he's not the horror story some paint him to be.

If you look at Todd Coffey's stats minus his five worst outings he's made 48 appearances with an ERA of 3.13.

M2
09-21-2007, 12:03 PM
Trade him to the Devil Rays for Jason Pridie.

He'd give that team some veteran presence in the bullpen, Pridie would give the Reds a potential CF.

redsmetz
09-21-2007, 12:48 PM
Trade him to the Devil Rays for Jason Pridie.

He'd give that team some veteran presence in the bullpen, Pridie would give the Reds a potential CF.

We don't have a potential center fielder in Jay Bruce?

M2
09-21-2007, 01:02 PM
We don't have a potential center fielder in Jay Bruce?

Bruce's optimal position, and I think the bevy of Larry Walker comparisons bears this out, is probably RF. Can Jay Bruce play CF? Yes. Would he be my nominal CF for this club next year given who's in town at the moment? Yes. Yet that doesn't mean he's necessarily going to be stationed in CF.

None of us know what the talent shuffle this winter will be like. Pridie's a legit CF in a system that's clogged with CFs. My general take is it's good to have up-the-middle talent on hand. Pride might just be a 4th OF type who can step in when injuries hit (like now). Main thing is he'd create options and that's a good thing.

blumj
09-21-2007, 01:16 PM
BTW, the Red Sox apparently still control Mike "The Torch" Burns, too.

RedsManRick
09-21-2007, 01:26 PM
Hamilton for Jacoby Ellsbury?

jojo
09-21-2007, 01:52 PM
Trade him to the Devil Rays for Jason Pridie.

He'd give that team some veteran presence in the bullpen, Pridie would give the Reds a potential CF.

I'm wondering why Tampa would want Stanton.

Raisor
09-21-2007, 01:56 PM
While we're at it, lets trade Castro to Anaheim.

Tom Servo
09-21-2007, 02:16 PM
How about Dumatrait to the Tigers?

M2
09-21-2007, 02:59 PM
I'm wondering why Tampa would want Stanton.

They're insane? That's my story and I'm sticking to it. If there's a team out there that's queued up for some desperate moves, it's the D-Rays. They've just become the first franchise in baseball history lose 90+ games 10 years in a row. The pre-Whiz Kids Philadelphia Phillies never did it, the St. Louis Browns never did it, the turn of the century Boston Braves never did it, the post-Ruth Red Sox never did it, the Philadelphia A's in the waning years of Connie Mack never did it. Plumb the depths of their desperation.

CaiGuy
09-21-2007, 03:31 PM
Stanton would actually be one of the best relievers in their bullpen. They seem to make the Reds look good.

GAC
09-21-2007, 03:37 PM
I'll pack his bags and drive him to the airport. ;)

jojo
09-21-2007, 03:56 PM
They're insane? That's my story and I'm sticking to it. If there's a team out there that's queued up for some desperate moves, it's the D-Rays. They've just become the first franchise in baseball history lose 90+ games 10 years in a row. The pre-Whiz Kids Philadelphia Phillies never did it, the St. Louis Browns never did it, the turn of the century Boston Braves never did it, the post-Ruth Red Sox never did it, the Philadelphia A's in the waning years of Connie Mack never did it. Plumb the depths of their desperation.

Tampa Bay has become one of the better run organisations in the majors. You're not going to see too many desperate moves by those guys even if on the superficial surface one may seem weird at first. They have a plan, intelligent people running the show, and frankly they're deep in talent and are posturing themselves for a serious run in a shorter time span than many people might realize.

M2
09-21-2007, 05:18 PM
Tampa Bay has become one of the better run organisations in the majors.

I've got 10 straight 90+ loss seasons that say otherwise. They have some talent, but they've still got the worst overall pitching and defense in the majors. They've supposedly been turning the corner for the past five years and it still hasn't happened.

They are literally the opposite of a well-run organization, perpetually unable to fill their needs with their excess.


You're not going to see too many desperate moves by those guys even if on the superficial surface one may seem weird at first. They have a plan, intelligent people running the show, and frankly they're deep in talent and are posturing themselves for a serious run in a shorter time span than many people might realize.

They've yet to show even the vaguest recognition of what they need. The plan, apparently, is to hope that continued inactivity is emblematic of a chrysalis state and not organizational rigormortis. It's a franchise enamored with favorable press fluff, slavish to conventional wisdom and immune to success.

Guaranteed the vigil of waiting for Longoria, Brignac, Niemann, Davis, McGee, Howell, Hammel, Sonsatine and Jaso will continue through next year (as even those who arrive will have their fair share of adaptation issues). Meanwhile Carlos Pena is a free agent. Scott Kazmir's about to hit arbitration (and you can bet Scott Boras would like to get him away from the penurious D-Rays). Carl Crawford's price tag will double by 2009. Fans have now been conditioned to stay away from the execrable franchise, which means revenue will always be short, so my guess is the future for this club will be that of a nouveau Kansas City A's, where young talent gets shipped out once it starts to cost some money and success is always just over a never-ending horizon.

RedsManRick
09-21-2007, 05:35 PM
Meanwhile Carlos Pena is a free agent.

Actually he's arbitration eligible and won't hit FA until 2009.

The management group in place in Tampa Bay is not responsible for the crappy record they've had since their inception any more than the next president of the US will be accountable for Iraq.

It's stupid to use the present status of their major league record and the actions of a management team no longer in place to evaluate the competency of the team that is.

I'd bet you 50 bucks that within the next 2 years, Tampa Bay finishes 3rd or better in the AL East.

M2
09-21-2007, 07:38 PM
Actually he's arbitration eligible and won't hit FA until 2009.

The management group in place in Tampa Bay is not responsible for the crappy record they've had since their inception any more than the next president of the US will be accountable for Iraq.

It's stupid to use the present status of their major league record and the actions of a management team no longer in place to evaluate the competency of the team that is.

I'd bet you 50 bucks that within the next 2 years, Tampa Bay finishes 3rd or better in the AL East.

They could lose 90 and still finish 3rd in the AL East. It's a two team division. I'll bet you 50 bucks they could finish 3rd and no one would care.

Does "until 2009" before or after 2009? Either way, I'm guessing Adam Dunn's $7.5M price tag in 2006 will be a pretty decent comp for Pena. Tampa can probably squeeze Carlos in next year, but in 2009 Pena, Crawford and Kazmir could cost more than the team paid in salary in 2007 with Upton and Shields hitting arbitration. Just you watch 'em squirm when players start escalating in price.

And it's stupid not to recognize that a change in regime hasn't netted a change in behavior. They promoted from within and they're still following the Chuck Lamar roadmap - don't spend, draft, wait. About the only thing Silverman's done differently is collect some extra prospect arms from other organizations (Jackson, Howell). Plus ca change ...

The club's had a top 10 ranked farm system five years running. It's been flush with highly-rated prospects since it snagged Matt White and Bobby Seay in 1996.

Through the years we've heard fish stories about Dewon Brazleton, Seth McClung, Jesus Colome, Doug Waechter and Jason Standridge. The Durham Bulls won back-to-back titles in 2002 and 2003 and that was supposed to be the harbinger of a turnaround that never happened.

This isn't like the Braves, a club which amassed talent, added key veterans and had the resources to keep it all together. This is a franchise that's going to be constantly dismantling itself and the folks in charge today haven't shown so much as an inkling of being able to find a veteran arm, doesn't help when you've got a thrift store budget.

SteelSD
09-21-2007, 08:56 PM
I've got 10 straight 90+ loss seasons that say otherwise. They have some talent, but they've still got the worst overall pitching and defense in the majors. They've supposedly been turning the corner for the past five years and it still hasn't happened.

They are literally the opposite of a well-run organization, perpetually unable to fill their needs with their excess.

I consider Tampa Bay the dumbest franchise in Major League Baseball. They can't figure out how to deal promising but potentially-pyrite low-IsoD toolsy offensive prospects for pitching. They can't figure out who to play where and, damn, have they drafted some Milos. They want to rely on speed while fielding a team that can't translate such into defensive range as their highest DER finish over the past three years has been 24th.

After so many putrid seasons, the D-Rays should been able by now to produce something resembling respectability given their consistent high draft positions, but we've seen nothing but underachievement. From upper management down to the scouts, the organization has simply been inept.

But a goodly portion of their players look good in jeans. At least that's something?

blumj
09-21-2007, 11:08 PM
Since the topic of the Rays came up here, I was wondering, is the fact that the Rays otherwise not so great pitching staff has piled up the Ks this season indicate that they could have a much improved pitching staff next season, or does it not work that way?

OnBaseMachine
09-21-2007, 11:32 PM
The DRays have some nice young arms coming up through the system. Obviously Scott Kazmir and James Shields are already very good pitchers at the major league level but they also have some fine arms with high potential in the minors. Wade Davis, Jeff Niemann, Jacob McGee, and #1 overall pick David Price are all guys who are expected to be above average major league pitchers. I would love to see the Reds get their hands on one of those guys.

WVRedsFan
09-21-2007, 11:41 PM
Mets bullpen is getting destroyed these days. They have ten games left and can use any reinforcements they can get. If Stanton has cleared waivers, as I assume he has (or would), trade him to the Mets for a minor leaguer. Stanton wouldn't be eligible for post-season, but Mets probably won't care at this point.

Another possibility is Boston, which is suffering from the same bullpen woes.

And everyone else. What in the world happened to pitching in MLB?

Patrick Bateman
09-21-2007, 11:42 PM
The DRays have some nice young arms coming up through the system. Obviously Scott Kazmir and James Shields are already very good pitchers at the major league level but they also have some fine arms with high potential in the minors. Wade Davis, Jeff Niemann, Jacob McGee, and #1 overall pick David Price are all guys who are expected to be above average major league pitchers. I would love to see the Reds get their hands on one of those guys.

All 4 won't be above average major leaguers. The odds are stacked against any pitching prospect. Even David Price is no sure lock. By the time any of those guys are good Kazmir/Shields will be pricing them out of Tampa (Kazmir anyways).

No questioning their young talent, but until they find a way to keep the young talent they do produce after they have success in the majors, they will never compete for higher than third.

OnBaseMachine
09-21-2007, 11:44 PM
All 4 won't be above average major leaguers. The odds are stacked against any pitching prospect. Even David Price is no sure lock. By the time any of those guys are good Kazmir/Shields will be pricing them out of Tampa (Kazmir anyways).

No questioning their young talent, but until they find a way to keep the young talent they do produce after they have success in the majors, they will never compete for higher than third.

I know that. I'm just saying all four have ceilings as above average major league pitchers. Will they all four reach that potential? Highly doubtful.

Patrick Bateman
09-22-2007, 12:02 AM
I know that. I'm just saying all four have ceilings as above average major league pitchers. Will they all four reach that potential? Highly doubtful.

They obviouslyhave a great stable of prospects, but the problem is that the D-Rays have such a low payroll that they can't afford to keep more than a couple higher priced players.

So basically to compete, they are going to need 2 of those pitching prospects to develop fast enough that they can be in the same rotation with Kazmir and Shields. If they can manage to that, maybe they can string together enough talent to make a run for the wildcard. But pitching (no matter how good the talent) isn't going to develop over night.

A complete reliance on the farm just isn't going to work. You need to get some talent from other areas, and Tampa has showed virtually no ability to do that. They still can't find a reliever, and until their defense is better than horrid their pitching is going to struggle.

I like that Friedman seems to be able to identify talent (at leats better than LaMar) but the same overrall plan is in place. Maybe you can blame that on the market in TBay, but that's not going to change. If their plan is just to develop all of their players and get rid of them when they become a little pricey, then there's no hope for them to compete with the 2 big guns.

OnBaseMachine
09-22-2007, 12:37 AM
They obviouslyhave a great stable of prospects, but the problem is that the D-Rays have such a low payroll that they can't afford to keep more than a couple higher priced players.

So basically to compete, they are going to need 2 of those pitching prospects to develop fast enough that they can be in the same rotation with Kazmir and Shields. If they can manage to that, maybe they can string together enough talent to make a run for the wildcard. But pitching (no matter how good the talent) isn't going to develop over night.

A complete reliance on the farm just isn't going to work. You need to get some talent from other areas, and Tampa has showed virtually no ability to do that. They still can't find a reliever, and until their defense is better than horrid their pitching is going to struggle.

I like that Friedman seems to be able to identify talent (at leats better than LaMar) but the same overrall plan is in place. Maybe you can blame that on the market in TBay, but that's not going to change. If their plan is just to develop all of their players and get rid of them when they become a little pricey, then there's no hope for them to compete with the 2 big guns.

They need a reliever? I'd gladly give them Weathers for McGee. :D

But seriously, I'd offer up a reliever like Weathers or even Bray plus a toss in like a Guevara or for one of their pitching prospects just to see if they are dumb enough to do it.

toledodan
09-22-2007, 12:45 AM
If Mike Stanton Is A Red Next Season I'll Give Up All Hope!

jojo
09-22-2007, 06:14 AM
Actually he's arbitration eligible and won't hit FA until 2009.

The management group in place in Tampa Bay is not responsible for the crappy record they've had since their inception any more than the next president of the US will be accountable for Iraq.

It's stupid to use the present status of their major league record and the actions of a management team no longer in place to evaluate the competency of the team that is.

I'd bet you 50 bucks that within the next 2 years, Tampa Bay finishes 3rd or better in the AL East.

Yep. Just as the legend of a great player lives past his actual ability so too do the sins of previous regimes.

jojo
09-22-2007, 06:18 AM
So basically to compete, they are going to need 2 of those pitching prospects to develop fast enough that they can be in the same rotation with Kazmir and Shields. If they can manage to that, maybe they can string together enough talent to make a run for the wildcard. But pitching (no matter how good the talent) isn't going to develop over night.

They've already got one of the best rotations in the AL. What they really need to do is learn to catch. That's where the "guranteed vigil" will help nicely.

RedsBaron
09-22-2007, 07:07 AM
What have you got against the Mets and the Red Sox?

I agree. I believe the Reds should give Stanton to the Yankees.

KronoRed
09-22-2007, 11:43 AM
I agree. I believe the Reds should give Stanton to the Yankees.

Amen.

corkedbat
09-22-2007, 12:29 PM
If the Reds deal any player of any value in the off season, the other team should be required to take Stanton also.

Patrick Bateman
09-22-2007, 01:11 PM
They've already got one of the best rotations in the AL. What they really need to do is learn to catch. That's where the "guranteed vigil" will help nicely.

You know what, I think I underestimated Sonnanstine and Howell. After exploring their stats, I have to admit that they have pitched pretty well. They just need better environments.

Their offense isn't very good which is hard to forgive considering how putrid their fielding is. They need to make some massive changes there if they want to make use of their pitching. I find that's pretty hard to do overnight, which is basically what they are going to have to.

But I admit, they do have some things going right for them right now. If they can manage to turn the defense around without hurting the offense a lot, you could see Tampa make some noise. But that still leaves the problem of the bullpen (from top to bottom) and their offense. But they do have some offensive talent around, so I could see how they would improve that without giving up much talent.

M2
09-22-2007, 05:26 PM
They've already got one of the best rotations in the AL.

The starter's ERA for the Rays is 5.12, good for 12th in the AL. How you translate that to "one of the best rotations in the AL" is beyond me. What the Rays have is two good pitchers and three awful ones and the awful is currently heavier than the good. Maybe some of the young guys will step up, then again maybe Kazmir and Shields will be spending next year recovering from their 200+ IP efforts in 2007.

Patrick Bateman
09-22-2007, 06:39 PM
The starter's ERA for the Rays is 5.12, good for 12th in the AL. How you translate that to "one of the best rotations in the AL" is beyond me. What the Rays have is two good pitchers and three awful ones and the awful is currently heavier than the good. Maybe some of the young guys will step up, then again maybe Kazmir and Shields will be spending next year recovering from their 200+ IP efforts in 2007.

Now I wouldn't go as far as saying that Tampa's has one of the best rotations in the league, but I think there is some merit to what Jojo is saying.

Sonnastine has been better than the numbers indicate. The bad defense is one of the reasons, and he seems to give up his runs in bunches. I think he could surprise some people next year. He has some traits similar to Shields.

Howell has been mostly undone by a .400 BAPIP. He has been pretty decent albeit the small sample size. OF course, he does have an ERA north of 6.00, so I think it's pretty easy to see what argument you will come back with. But he also was really good in AAA, so he's another guy that I could see having success in the right environment.

Anyways, I think these 2 guys may actually be pretty good pitchers. It's easy to completely dismiss this argument, but they do have a lot of things going for them, and their peripherals suggest they have been pretty decent (and I'm a pretty big believer on peripherals).

Obviously, if these guys are indeed good, then it's really quite disgusting what the defense is making them out to be, and either way, it shows how ridiculously far away Tampa is away from having a good team. But I think if there is one place where Tampa may be having some success is their starters. If I'm them, I look to make a complete and utter overhaul on the offensive side of the diamond. Because right now the D-Rays have below average offensive players that are playing awful defense, and that's a big problem. I think if they can put together a half way decent defense, that their current starters can succeed. Again, a massive way to go, but I can see why Tampa could leapfrog Baltimore and the Jays in quick fashion next year.

jojo
09-22-2007, 11:09 PM
The starter's ERA for the Rays is 5.12, good for 12th in the AL. How you translate that to "one of the best rotations in the AL" is beyond me. What the Rays have is two good pitchers and three awful ones and the awful is currently heavier than the good. Maybe some of the young guys will step up, then again maybe Kazmir and Shields will be spending next year recovering from their 200+ IP efforts in 2007.

ERA is a lousy way to evaluate a starting pitcher especially when Tampa's defense is so atrocious. Truthfully ERA is worse than lousy. Ranking rotations based upon FIP (a defense independent metric), Tampa's rotation is top 5 in the AL supporting the notion that they have one of the best in the AL. Here's a breakdown:

Kazmir: 200 IP; FIP: 3.52;
Shields: 215 IP; FIP: 3.82;
Sonnanstine: 120 IP; FIP: 4.05;
Hammel: 57 IP; FIP: 5.15;
Jackson: 148 IP; FIP: 4.90;
Howell: 46 IP; FIP: 4.24;

Their #3-6 arms in the rotation have combined for 372 IP as starters for a FIP of 4.50. American League average FIP for starters is 4.43. Essentially Tampa's #3 thru #6 gave them league average. How many organisations can not only boast that-which is impressive in and of itself- but also that their rotation is 6 deep with 5 of the arms controlled for the next half decade while yet another arm is essentially major league ready in Durham (Niemann). Oh ya, they control their ace for another three years too.

Basically their #6 would probably be a #3 for the Reds.

Tampa's rotation is one of the best kept secrets amongst fans but it's a certainty that any team needing starters (and that's roughly 28 or so) will give Friedman a call this off season.

M2
09-23-2007, 01:55 AM
ERA is a lousy way to evaluate a starting pitcher especially when Tampa's defense is so atrocious. Truthfully ERA is worse than lousy. Ranking rotations based upon FIP (a defense independent metric), Tampa's rotation is top 5 in the AL supporting the notion that they have one of the best in the AL. Here's a breakdown:

Kazmir: 200 IP; FIP: 3.52;
Shields: 215 IP; FIP: 3.82;
Sonnanstine: 120 IP; FIP: 4.05;
Hammel: 57 IP; FIP: 5.15;
Jackson: 148 IP; FIP: 4.90;
Howell: 46 IP; FIP: 4.24;

Their #3-6 arms in the rotation have combined for 372 IP as starters for a FIP of 4.50. American League average FIP for starters is 4.43. Essentially Tampa's #3 thru #6 gave them league average. How many organisations can not only boast that-which is impressive in and of itself- but also that their rotation is 6 deep with 5 of the arms controlled for the next half decade while yet another arm is essentially major league ready in Durham (Niemann). Oh ya, they control their ace for another three years too.

Basically their #6 would probably be a #3 for the Reds.

Tampa's rotation is one of the best kept secrets amongst fans but it's a certainty that any team needing starters (and that's roughly 28 or so) will give Friedman a call this off season.

When they start playing defense independent baseball you let me know. FWIW, I think FIP is a grandiose failure of a statistic. It divorces out some stuff (e.g. hits) that might very well reflect the pitcher's quality. And why do you multiply HR by 13 and then BB+HBP by 3 and Ks by 2? Did Tangotiger prove that those are the relative weights of those events? No he didn't. What he did was monkey around with multipliers until he got something that spit out something that looked like an ERA stat. That's right, he completely ignored hits and the quality of the hits the pitchers allow and then came up with a fairly arbitrary set of multipliers all in the name of generating a doppleganger of a statistic you just called "worse than lousy." Talk about crap math.

As a fun little side project, do note that HR*13 and K*2 in the NL the past five years have been roughly equivalent (not so in the AL, but that's a DH league). So what FIP is measuring for NL pitchers, speaking in large terms, is (BB+HBP)*3/IP. Please explain how that gives us a superior measurement of pitcher quality.

Plus, it's a lousy indicator of future performance. I remember conversations earlier this season where folks insisted that Adam Wainwright wasn't going to improve because he had a poor FIP. If I worshipped at the foot of FIP, I'd have thought Chris Young would suck this year and that Zach Duke would be the Pirates' breakout kid. Jeremy Bonderman would be a Cy Young contender and he'd be carrying Justin Verlander on his back. I've actually seen arguments like this made in cyberspace over the past year with FIP used as the proof point. Fortunately, I always knew better.

Anyway, after tonight, Sonnanstine's ISOp is going to be close to .200. Hammel's is .200 even. That's a problem. Edwin Jackson is a toxic mixture of poor control and ripable stuff. There's a reason these kids are flirting with 6.00 ERAs and it's not bad luck. Of the yet-to-click group I like Howell, who's got a K pitch and he keeps it on the ground. He wasn't near ready to pitch in the majors the past three years, but I don't hold being grossly overpromoted against him.

But feel free to cling to the delusion that there's such an animal as a pitcher with an above 5.00 ERA who's pitching well. Folks on this board have been making that claim constantly throughout the 21st century and all it's ever done is make for a nice egg facial.

Truthfully, ERA tells us these kids are vomiting up runs. They've got the same defense behind them as Kazmir and Shields (whose ERAs happen to match their FIPs, which should be a statistical impossibility given the putridity of the Devil Rays defense).

Tampa Bay's rotation is not some insider secret. It's a unit with a bunch of kids, two of whom are pitching well and the rest of whom are not. And the organization hasn't done some miraculous makeover the past two seasons. It's still the cheapest franchise in the business and it's still working off the same blueprint as the previous regime. On top of that, the new regime is on its way to finishing with the worst record in baseball for the second straight year. Feel free to keep making stuff up about how great things are in Tampa Bay, but reality is going to prove a hurdle you can't clear.

As for Kazmir, let's see if the money-starved D-Rays control him for another three years. My guess, and it's just a guess, is he never pitches a sixth year with the club ... provided he stays healthy. In the relatively near future I expect a complete lack of revenue and Scott Boras to be controlling the D-Rays when it comes to Kazmir.

Patrick Bateman
09-23-2007, 02:44 AM
If I worshipped at the foot of FIP, I'd have thought Chris Young would suck this year and that Zach Duke would be the Pirates' breakout kid. Jeremy Bonderman would be a Cy Young contender and he'd be carrying Justin Verlander on his back. I've actually seen arguments like this made in cyberspace over the past year with FIP used as the proof point. Fortunately, I always knew better.


I'm going o have to comment on a couple of these because I have publically not been high on Chris Young, and have been on Bonderman (but not to the degree of him carrying Verlander in any sense).

In regards to Young, I still consider him to have some pretty big flaws. His K/BB is 2.45 and he has a GB/FB of 0.55. Very few pitchers have been able to maintain sustained success with numbers like these. To be successful here you basically have to pitch in a park that suppresses homers. You toss him in GABP and the HR would fly out like crazy. Guys who give up that many flyballs should be giving up more than 9 homers in a year.

Also, Young has clearly been getting an artificial boost from his fielding. It wasn't until he came to San Diego that he ever started posting posting favourable BAPIP's. This isn's an inherent skill that he possesses.

His K/BB and groundball rates basically haven't changed much at all since he moved from Texas to SanDiego. The only things that have changed have been the things independent of him. This isn't to say that he's a stinky pitcher, but take away the great defense and Petco park and I think you are left with a good, but not great pitcher, similar to that in Texas where he posted an ERA of 4.26. That's pretty darned good considering the environments in Texas, and it was insane for them to trade him for Eaton, but he's not the 2.80 ERA pitcher he's currently being made out to be.

As for Bonderman, he continues to confuse me. He posts very solid K and BB numbers, he gets a solid amount of grounders, he pitches in a favourable park, and he plays in front of a fine defensive ballclub (and last year it was elite). Usually that's a formula for a great ERA.

In reality, his lowest ERA to date is 4.08, and this season he has an ERA over 5.00. He constantly posts BAPIP's above his teammates. It's to the point where a trend appears to be forming. Is this going to continue in the future? I would bet against it (like I was the last few seasons, but look how good that turned out). Bonderman has so many reasons on why he should be successful and he continues to confuse me even though his peripherals continue to be quite good.

What I do like is that last season he did manage a 4.08 ERA even though he had a .328 BAPIP. That shows he must have been doing something right. It would be quite an anomaly if he continued posting such high BAPIP's even with a great defense. To the best of my knowledge I have never really seen anything like it over much of a sample size. So I have trouble seeing it continue. Of course his health has now become the issue that many always thought it would, so I may not be able to really see the end of this issue, which is too bad, because I think the numbers indicate success was going to happen sooner or later.


I have admittedly only been looking at this type of stuff really since last season, but I can say that looking at BAPIP/FIP/etc. can be extremely useful. I don't think it should be used as an end all to an argument in any way, but I can tell you that I am right mor often then I used to be. I'd like to see how Chris Young would do right now in an environment that isn't tailored to his skillset so perfectly. But I don't think he's changed that much since Texas, so I still think that his numbers there are more indicative of his true abilities.

M2
09-23-2007, 11:47 AM
AK, FIP likes Young in 2007 and it liked him in 2005. In 2006 he gave up a few too many longballs so his FIP went up. Did I need a statistical formula to tell me that ideally a pitcher shouldn't give up homers at the rate he did in 2006? No, I don't. Yet Young really isn't the issue for me here. It's that FIP failed to do what it purports to do - attempt to reveal the real quality of the pitcher. It's over its head. It can't even begin to sum up the complexity of a good season. The advantage ERA has over it is that it's based on outcome. ERA doesn't pretend to reveal the nuance of what makes an effective pitcher, but it does tell when someone's been an effective pitcher better than anything else we've got.

On Bonderman, my take is:

- he's always been overrated
- he's got a 10-cent head and doesn't make adjustments
- his vulnerability against LHBs creates a glass ceiling for him

I'm a big proponent of looking at BABIP (I believe I was one of the first on this board to bring it into the discussion). I also think you need to look at HR, BB and Ks, but FIP mixes them into a patently useless jumble.

Patrick Bateman
09-23-2007, 01:03 PM
Did I need a statistical formula to tell me that ideally a pitcher shouldn't give up homers at the rate he did in 2006? No, I don't.

I agree to an extent here. Pitching in Petco made his 1.41 HR/9 basically a feat impossible to match. I think in a neutral park, Young's major flyball tendencies very well could lead to that many homers. But as far as 2007, it doesn't surprise me a whole lot that his homer rate wen't down for that reason.



On Bonderman, my take is:

- he's always been overrated
- he's got a 10-cent head and doesn't make adjustments
- his vulnerability against LHBs creates a glass ceiling for him


These could very well be true. I have never really seen him as Cy Young calibre, but based on his K, BB, and HR numbers you could come to the conclusion that the guy was making serious progress. And in 2006 he basically did excepting the BAPIP (which IMO left him from being a sub 4.00 pitcher). This season again he put up solid peripherals that would suggest an ERA around 4.00, and yet, he posts an ERA above 5.00. Even with the reasons you suggested that seems quite odd.



I'm a big proponent of looking at BABIP (I believe I was one of the first on this board to bring it into the discussion). I also think you need to look at HR, BB and Ks, but FIP mixes them into a patently useless jumble.

At times I find this to be true. I find it does a pretty terrible job on the homer rates to start off. I generally find that pitchers who control solid K, BB, and GB/FB rates will do well in just about any surrounding (Bonderman would be an example of someone who has done otherwise).

Using HR rates rather than GB/FB I find is more useful to predict 'actual' success, but not as great of an indicator when pitchers switch parks.

jojo
09-23-2007, 10:28 PM
This is kind of long but I offer it in the spirit of discussion-not as part of a pissing contest or snark fest.


When they start playing defense independent baseball you let me know.

We are talking about how to best evaluate a pitcher and which method best predicts future performance aren't we? I'll assume you're trying to be snarky rather than obtuse.


FWIW, I think FIP is a grandiose failure of a statistic. It divorces out some stuff (e.g. hits) that might very well reflect the pitcher's quality. And why do you multiply HR by 13 and then BB+HBP by 3 and Ks by 2? Did Tangotiger prove that those are the relative weights of those events? No he didn't. What he did was monkey around with multipliers until he got something that spit out something that looked like an ERA stat. That's right, he completely ignored hits and the quality of the hits the pitchers allow and then came up with a fairly arbitrary set of multipliers all in the name of generating a doppleganger of a statistic you just called "worse than lousy." Talk about crap math.

Wow. It would be hard to describe the derivation of FIP more erroneously-your criticisms simply aren’t factually correct. However, based upon an apparent distrust of those who monkey around with multipliers, I’m guessing you don’t have a lot of respect for Bill James and you really, really, really, hate winshares.

First, the issue of hits probably should be addressed. While there has been some recent evidence that certain pitchers can affect batted ball velocity more than other pitchers (which in turn can influence the odds a ball in play becomes a hit or not), in general pitchers have much, much less ability to control the fate of a batted ball than your argument is suggesting. A consensus of those who are actively researching such questions suggests you are dramatically overstating the influence of pitchers on hits. In other words, the influence a pitcher exerts on batted balls, while measurable, is very minimal in the grand scheme of things. It’s very appropriate for hit information to be absent from a defense-independent pitching metric.

So then how were the weights derived in the FIP formula ? Describing the process in a general sense, Tom Tango basically categorized a defensive responsibility spectrum (i.e. all of the ways to get on base or make an out) and focused only on those that were solely influenced by the pitcher (SO, BB, HBP, HR) to essentially reduce the outcomes into a defense- independent equation that looked something like this : FIP= ((x*HR+ y*(BB+HBP-IBB)-z*SO))/IP. Then with a huge pitcher database at hand, Tango, using a combination of linear weights and regression analysis, basically solved the x,y, and z constants so that the function best predicted ERA. So, there wasn’t any monkey business involved with determination of the weights in the FIP formula.

If you’re interested in the gory details, here’s a link to a detailed discussion of how FIP was derived by Kevin Harlow:

http://members.cox.net/~harlowk22/DIPS-GS.html

It’s not exactly how Tom Tango did it but it’s similar enough for the purposes of this discussion.


As a fun little side project, do note that HR*13 and K*2 in the NL the past five years have been roughly equivalent (not so in the AL, but that's a DH league). So what FIP is measuring for NL pitchers, speaking in large terms, is (BB+HBP)*3/IP. Please explain how that gives us a superior measurement of pitcher quality.

Since command is a first prerequisite for success at the major league level for a pitcher, I’d suggest “that” informs quite a lot actually.


Plus, it's a lousy indicator of future performance.

This statement simply isn’t true. By extension you’re arguing that FIP is a worse indicator of future performance than ERA. That has been proven false by numerous studies. Here’s just one of many that refute your assertion:

http://www.hardballtimes.com/main/article/fip-and-the-long-ball/

A quick google search can provide many more as well as some tremendous discussion on this issue.


I remember conversations earlier this season where folks insisted that Adam Wainwright wasn't going to improve because he had a poor FIP. If I worshipped at the foot of FIP, I'd have thought Chris Young would suck this year and that Zach Duke would be the Pirates' breakout kid. Jeremy Bonderman would be a Cy Young contender and he'd be carrying Justin Verlander on his back. I've actually seen arguments like this made in cyberspace over the past year with FIP used as the proof point. Fortunately, I always knew better.

I have no idea who you’ve been arguing with so I can’t really speak to your disdain for FIP relative to cyberspace. Also, no one here is arguing that FIP predicts exact ERA with 100% accuracy. Rather, FIP predicts a pitcher's future performance a lot better than ERA. I frankly don’t see anything all that eye popping with regards to Wainwright and Bonderman other than they are both guys I’d like to see in a Reds uniform. Regarding Duke, you’re not arguing that FIP should have predicted that he’d have significant injury problems this season are you?

Ignoring for a moment that cherry picking a few individual pitchers to support your argument isn't very compelling, let’s focus on Chris Young. I’ve made an argument against the Reds acquiring him earlier this season that had nothing to do with his past FIPs. Here's a link: http://www.redszone.com/forums/showpost.php?p=1370784&postcount=6

It's important to realize that the biggest difference between Young in '06 and '07 is that he is benefitting from an extremely low and unsustainable HR/FB this season. Basically I’d argue there are certain environments which dramatically mitigate his effectiveness. If you examine his FIP normalized for HR rate (xFIP), you'll notice it is in line with his career xFIPs. While I think there are rare pitchers who can consistently post lower than league average HR/FB rates (especially some relievers), generally this is not a repeatable skill for pitchers and the highly variable HR/FB rate Young has shown during his career suggests that it's not a repeatable skill for him either. While I don't have his batted ball data for home/away to calculate his HR/FB for the split, it's pretty clear his HR/G is dramatically different at Petco than on the road.

Anyway, this all deflects from the real issue though. When looking at groups of pitchers, FIP is a better predictor of next season's ERA than the pitcher's actual ERA and when HR/FB rates are controlled (i.e. xFIP), ERA is moved from the backseat to the trunk as a predictor of pitcher performance.


But feel free to cling to the delusion that there's such an animal as a pitcher with an above 5.00 ERA who's pitching well. Folks on this board have been making that claim constantly throughout the 21st century and all it's ever done is make for a nice egg facial.

To begin with, that’s a strawman argument and really, what is that tone adding to the discussion?

I actually argued that defense-independent metrics suggest the Rays’ #3 thru #6 starters were essentially league average based upon their peripherals (i.e. their true skill sets) and this coupled with their two aces was why their rotation was one of the top 5 in the AL.


Truthfully, ERA tells us these kids are vomiting up runs.

No. It tells you that the Devil Rays vomited up runs when they pitched. It doesn’t actually indicate how to accurately spread the blame. That’s two dramatically different things and really since it’s at the heart of the discussion, it’s not a question that should be begged.


They've got the same defense behind them as Kazmir and Shields (whose ERAs happen to match their FIPs, which should be a statistical impossibility given the putridity of the Devil Rays defense).

FIP is an idealized ERA based upon a pitcher’s peripherals so a pitcher’s ERA should roughly approximate his ERA all things being equal. But as already discussed, ERA is really the culmination of many factors out of the pitcher’s control not the least of which is simple randomness, effect of environment and the defense behind him so all things aren’t equal. Kazmir has an insane Krate. Shields has an above average Krate and an insane BB rate. Both basically have neutral batted ball tendencies. So the skill sets of both can help the defense out a lot. In any event, the fact that their ERAs and FIPs approximate one another in no way supports your conclusion that FIP is a contrived, useless stat that is inferior to ERA.


Tampa Bay's rotation is not some insider secret. It's a unit with a bunch of kids, two of whom are pitching well and the rest of whom are not.

Using ERA as a tool to evaluate pitcher performance is an approach that has been debunked over and over again. Looking at the peripherals of pitchers is a vastly superior method for predicting future performance. Metrics which do just that support a dramatically different conclusion than yours.


And the organization hasn't done some miraculous makeover the past two seasons.

Except that it has in recent years of course. Their ownership and FO have underwent a complete transformation with a new managing partner, new GM, new second in command , new director of player personnel, new director of international scouting, and a new manager. It’s important to note that through these changes, Tampa’s FO has become much more “new school” as it's been populated with some of the brightest up and coming minds in the game who generally are highly respected by others in the industry. Tampa has one of the deepest and best stocked farms in the game and the major criticism is they haven’t cashed in some of their chips for pitching? I’m thinking they don’t really need pitching and I shudder at the notion of retooling a bullpen by shipping out valuable prospects. Tampa needs just what their FO is giving them -patience so that they can replace their porous defense with several soon to arrive prospects that can both hit and field.


Feel free to keep making stuff up about how great things are in Tampa Bay, but reality is going to prove a hurdle you can't clear.

I’ve made rational, well supported arguments and I’ve tried to do so in a respectful manner. The reality is that there are a lot of reasons to be optimistic about Tampa Bay’s very near future.

We've recently had a similar discussion in tone about PBP-based defensive metrics. Another reality is there are now much better ways to evaluate defense and pitching than things like fielding percentage, range factor, RZR, WHIP and ERA. Frankly, I don’t get the willingness to continually reject superior metrics in favor of ones that have been proven to be much more severely flawed. Do these better metrics also have flaws? Of course they do. However, rejecting a better metric because it’s not absolutely perfect is tantamount to choosing to race the Indy 500 with a Yugo because it doesn’t use as much gas. To me the best approach to player evaluation is to recognize the flaws in all metrics so that they can be weighted accordingly when using a spectrum-based approach to evaluate talent while also being ready to reject an intuitively held belief when there is good cause for pause.

M2
09-24-2007, 03:03 AM
We are talking about how to best evaluate a pitcher and which method best predicts future performance aren't we? I'll assume you're trying to be snarky rather than obtuse.

I wasn't trying to do either, just noting your particular slavishness to statistical dead ends.


Wow. It would be hard to describe the derivation of FIP more erroneously-your criticisms simply aren’t factually correct. However, based upon an apparent distrust of those who monkey around with multipliers, I’m guessing you don’t have a lot of respect for Bill James and you really, really, really, hate winshares.

First, the issue of hits probably should be addressed. While there has been some recent evidence that certain pitchers can affect batted ball velocity more than other pitchers (which in turn can influence the odds a ball in play becomes a hit or not), in general pitchers have much, much less ability to control the fate of a batted ball than your argument is suggesting. A consensus of those who are actively researching such questions suggests you are dramatically overstating the influence of pitchers on hits. In other words, the influence a pitcher exerts on batted balls, while measurable, is very minimal in the grand scheme of things. It’s very appropriate for hit information to be absent from a defense-independent pitching metric.

So then how were the weights derived in the FIP formula ? Describing the process in a general sense, Tom Tango basically categorized a defensive responsibility spectrum (i.e. all of the ways to get on base or make an out) and focused only on those that were solely influenced by the pitcher (SO, BB, HBP, HR) to essentially reduce the outcomes into a defense- independent equation that looked something like this : FIP= ((x*HR+ y*(BB+HBP-IBB)-z*SO))/IP. Then with a huge pitcher database at hand, Tango, using a combination of linear weights and regression analysis, basically solved the x,y, and z constants so that the function best predicted ERA. So, there wasn’t any monkey business involved with determination of the weights in the FIP formula.

If you’re interested in the gory details, here’s a link to a detailed discussion of how FIP was derived by Kevin Harlow:

http://members.cox.net/~harlowk22/DIPS-GS.html

It’s not exactly how Tom Tango did it but it’s similar enough for the purposes of this discussion.

Golly, let's pretend I knew about Voros McCracken and how FIP got created long before you ever posted in these parts. In fact, let's put that front and center because what we've got here is you looking in the wrong direction. I know the background (and for the record x=13, y=3 and z=2 and I was 100% factually correct (http://www.hardballtimes.com/main/statpages/glossary/) about that and don't take the least bit kindly to you saying otherwise). The issue here, for me, is the unthinking credulity you've chosen to lend this statistic.

First off, walks, strikeouts and homers aren't solely influenced by the pitcher. There are batters involved and that's not an insignificant point. Pitchers are part of an interactive game and that's why I commented about defense independent baseball. Tango started from a fatally flawed concept with FIP. Beyond that, even if the defensive responsibility for non-homer hits can't be wholly lumped onto pitchers it's still a massive part of the pitcher's performance. In fact if you take hits, homers, BBs and Ks, hit's is the largest event group. Trying to come up with an evaluation of pitcher performance that ignores the plurality of events in that pitcher's performance is an exercise in futility. Using that same logic, we could apply the same basic formula, with minor modifications, to batters, who also don't completely control the fate of a batted ball. I mean if balls in play are uncontrollable random events, why count 'em anywhere if they aren't really telling us how "good" somebody really is?

As for Tango's methodology, he solved for x, y an z using run value system that ultimately is tied up with ... wait for it ... hits. It's the dopiest math ever. Let me say it clearly one more time so that it resonates, Tango tried to come up with linear weights for the run value of non-hit events using a system where the largest contributor to runs is hits. The thing he's trying to move out of the equation is still running wild under the covers. Beyond that, the relation of all those events to each other is not a constant. If you don't think 13, 3 and 2 were selected because ultimately it represented the best way to create a static ERA-like equation then I'd say you've missed a fairly large boat.


Since command is a first prerequisite for success at the major league level for a pitcher, I’d suggest “that” informs quite a lot actually.

You know, you didn't have to actively sidestep that one. I always knew you weren't going to have an answer. We both know you won't be using BB rate as the true estimate of pitcher value at any point in the future.


This statement simply isn’t true. By extension you’re arguing that FIP is a worse indicator of future performance than ERA. That has been proven false by numerous studies. Here’s just one of many that refute your assertion:

http://www.hardballtimes.com/main/article/fip-and-the-long-ball/

Can it be that you didn't read to the end of that piece? Can that be true? I'm going to re-post it because it's a laugh riot.

Here's the list of guys Studeman predicted to pitch a lot better in 2006:

Vazquez, Javier 12
Lowe, Derek 12
Ortiz, Ramon 11
Ramirez, Horacio 11
Myers, Brett 10
Perez, Oliver 9
Maroth, Mike 9
Robertson, Nate 8
Chen, Bruce 8
Astacio, Ezequiel 8
Marquis, Jason 8
Lieber, Jon 8
Webb, Brandon 8

Exactly two, Robertson and Webb, actually had notably better seasons. Most guys on the list pitched worse in terms of adjusted ERA. I could have a monkey throw darts at names on a corkboard and have a better success rate than this guy has with a mathematical system.

Meanwhile he cast a suspicious eye toward Kenny Rogers, Chris Young, Dontrelle Willis, Tom Glavine, Bronson Arroyo, Roger Clemens, Scott Kazmir, John Lackey and Roy Oswalt. Go ahead and check out how those guys fared in 2006. Predictive value indeed.

And once again, it is high comedy when you get an argument that claims ERA doesn't matter and then uses ERA as the final test point. Trust me when I say that when I make a claim that a statistic is lousy, you won't find me using it as a proof point or under the covers to derive my supposedly better system.

Also, what are we looking for with pitchers? Insignificant ERA nudges or ways to identify guys who should pitch well and those who probably won't? I submit it's the latter. For instance, I don't care if Scott Kazmir's ERA has gone from 3.77 to 3.24 to 3.54. That's all good as far as I'm concerned. What I'm looking for is Scott Kazmir and not Horacio Ramirez. Studeman strikes me as a guy who's thoroughly lost the plot.


To begin with, that’s a strawman argument and really, what is that tone adding to the discussion?

I actually argued that defense-independent metrics suggest the Rays’ #3 thru #6 starters were essentially league average based upon their peripherals (i.e. their true skill sets) and this coupled with their two aces was why their rotation was one of the top 5 in the AL.

I don't care a whit about your tone opinions. As for what you actually argued, the answer is you actually argued yourself into a concrete wall. That you based it upon your defense independence fallacy does nothing to diminish the severity of the collision. My strawman's got brain enough to avoid that crash.


No. It tells you that the Devil Rays vomited up runs when they pitched. It doesn’t actually indicate how to accurately spread the blame. That’s two dramatically different things and really since it’s at the heart of the discussion, it’s not a question that should be begged.

How did you get so incurious? Seriously. You've made these broad assumptions (actually you haven't made them, you've just taken them on faith best as I can determine) and even when a rotation like Tampa Bay's lays an egg you cling to them. I think there's a lot of questions you could stand to beg in this area.


FIP is an idealized ERA based upon a pitcher’s peripherals so a pitcher’s ERA should roughly approximate his ERA all things being equal. But as already discussed, ERA is really the culmination of many factors out of the pitcher’s control not the least of which is simple randomness, effect of environment and the defense behind him so all things aren’t equal. Kazmir has an insane Krate. Shields has an above average Krate and an insane BB rate. Both basically have neutral batted ball tendencies. So the skill sets of both can help the defense out a lot. In any event, the fact that their ERAs and FIPs approximate one another in no way supports your conclusion that FIP is a contrived, useless stat that is inferior to ERA.

I already know that Kazmir's got a high K rate and low HR rate to offset his high BB rate and that Shields has a high K rate and low BB rate to offset his high HR rate. Thank you for stating the obvious, but that only explains their FIPs. What you've sidestepped is that all things supposedly aren't equal with them, that the Rays defense is supposedly masking how well other pitchers are doing yet it's having no effect on Kazmir and Shields when it comes to ERA. They're not the only ones with attractive K rates on that team. Sonnanstine is also pretty stingy with the BB. You can't have your cake and eat it too. If the Tampa Bay defense is undermining the rest of the starting staff then it would be undermining Kazmir and Shields too, and don't try to hide behind the randomness barrier on this one. It's such a flimsy defense in this case.


Using ERA as a tool to evaluate pitcher performance is an approach that has been debunked over and over again. Looking at the peripherals of pitchers is a vastly superior method for predicting future performance. Metrics which do just that support a dramatically different conclusion than yours.

I feel like I'm arguing with a Platonic ethicist here. Clearly you've got an order to the universe into which everything neatly fits, no matter how fictional the construct may be. Good for you. In the meantime there is that small, teensy tiny matter of how it translates to actual runs. I know that's messy, but I'm not trying to pretend we live in anything but a messy universe.

You keep acting like I've got something against peripherals. I don't. I'm just stating that FIP does a poor job of relating their weight and that what FIP ignores matters quite a bit. As for projecting pitcher performance, I've been around these parts a long time and I've got a track record on these things, which I will immodestly point out is pretty damn good. No one's 100% and I don't claim to have any special sixth sense on these matters, but I'm generally ahead of the curve with pitchers. You want to talk about using peripherals to estimate who'll be an effective pitcher? We haven't even mentioned things like age, experience and usage (with only passing reference to geographic location). Oh, and hits and hit quality, that needs to be in there if you're going to pretend to making a thorough evaluation. Everything fits together differently for each pitcher (FIP would tell you Tom Glavine's never been all that good) and once again, what's the aim here? To figure out which awful pitchers might only be highly objectionable next year or to find guys that can be good year-in, year-out? Seems to me your interest lies in tweaking insignificant margins.


Except that it has in recent years of course. Their ownership and FO have underwent a complete transformation with a new managing partner, new GM, new second in command , new director of player personnel, new director of international scouting, and a new manager. It’s important to note that through these changes, Tampa’s FO has become much more “new school” as it's been populated with some of the brightest up and coming minds in the game who generally are highly respected by others in the industry. Tampa has one of the deepest and best stocked farms in the game and the major criticism is they haven’t cashed in some of their chips for pitching? I’m thinking they don’t really need pitching and I shudder at the notion of retooling a bullpen by shipping out valuable prospects. Tampa needs just what their FO is giving them -patience so that they can replace their porous defense with several soon to arrive prospects that can both hit and field.

Thank you for writing a D-Rays press release. I'm aware they've got an "under new management" sign on the front lawn (though a lot of those young guys were in-house promotions). Yet those "brightest up and coming minds" are still waiting on Chuck LaMar's picks to save them. I mean, once we get past the blather you just typed, that's what they're doing. If these guys are so bright then why can't they find a single veteran pitcher who does anything but suck? Forget about trading away prospects for a moment, that doesn't preclude a team from evaluating talent and making some astute pick ups. The next good pitcher this regime finds will be its first. The bullpen is beyond bad and there really isn't much hope for improvement from the current cast of jokers they've got out there. They inherited a bad defense that's actually gotten worse. Meanwhile there's a glut of outfielders and seemingly no plan to parlay that into needed talent in other areas. Worst record in baseball two years running pretty much speaks for itself. Before we start handing out genius medals maybe we ought to wait until the club doesn't finish 30th. J.P. Ricciardi's "new school" too and just ask him about how brilliant he is. What's he done in the AL East? I'd say the Rays will be hard-pressed to match him given their current situation.


I’ve made rational, well supported arguments

Wow, you agree with yourself? Who'd have thunk it? I agree with myself too.


We've recently had a similar discussion in tone about PBP-based defensive metrics. Another reality is there are now much better ways to evaluate defense and pitching than things like fielding percentage, range factor, RZR, WHIP and ERA. Frankly, I don’t get the willingness to continually reject superior metrics in favor of ones that have been proven to be much more severely flawed. Do these better metrics also have flaws? Of course they do. However, rejecting a better metric because it’s not absolutely perfect is tantamount to choosing to race the Indy 500 with a Yugo because it doesn’t use as much gas. To me the best approach to player evaluation is to recognize the flaws in all metrics so that they can be weighted accordingly when using a spectrum-based approach to evaluate talent while also being ready to reject an intuitively held belief when there is good cause for pause.

My contention is you're overly credulous about supposedly "better" metrics and that you aren't recognizing their flaws. You buy into them and then try to force everything through those filters, ignoring the spectrum and putting all your eggs in shoddy basketry. I suggest you start paying attention to a broader spectrum of inputs when you're evaluating talent and not fall into the trap of there being one stat to rule them all.

RedsManRick
09-24-2007, 01:14 PM
Without diving in to the blow-by-blow (which, while unnecessarily acerbic, has been very informative and interesting), FIP does conceptually include hit events.

From my understanding, the premise is that the rate of balls in play is implied as the remainder of all events not captured in FIP. However, since the assumption is that the proportion of BIP events which should become hits is constant, it is not necessary to include it in the formula. It is essentially factored in to the weights that are present.

He has to exclude hit rate from the equation because the entire premise is that H/BIP is not a function of the pitcher. While BIP/PA is within the control of the pitcher, it is already accounted for in the formula inherently, isn't it?

Also, Jojo, isn't the concept of FIP just that, "how would this guy pitch with a neutral defense". Its correlation to future observed ERA is really irrelevant, given that year-to-year defensive performance is not random. If a guy has a poor ERA this year, relative to his FIP, and his defense doesn't change, we wouldn't expect his ERA to decrease beyond some basic regression to the mean. Thus, the real value of FIP is as a comparative measure of pitcher quality, not as a projection tool.

The point to be gleaned from the Rays pitchers' FIP is that the addition of pitchers with ERAs less than the current batch of pitchers may not really improve the ERA of the staff. Those new pitchers would be adversely affected by the existing poor defense and their ERAs would rise accordingly, relative to the ones which they put up prior. Thus, rather than pursue additional pitchers, the Rays would be better served improving their defense, which would improve all of their pitchers.

The biggest problem we have with all of these new metrics is that, due in part to their complexity, it's difficult to use them properly. We have enough problems with simple things like ERA, a measurement of the earned runs allowed while the pitcher was pitching and not the runs "he al'owed", or batting average, the frequency with which a player gets a "hit", not a measurement of how good a hitter is at doing his job, namely, producing runs. It behooves us all to be very careful in using any metric solely within the limits of it's definition.

Patrick Bateman
09-24-2007, 01:23 PM
Also, Jojo, isn't the concept of FIP just that, "how would this guy pitch with a neutral defense". It's correlation to future observed ERA is really irrelevant, given that year-to-year defensive performance is not random. If a guy has a poor ERA this year, relative to his FIP, and his defense doesn't change, we wouldn't expect his ERA to decrease beyond some basic regression to the mean. Thus, the real value of FIP is as a comparative measure of pitcher quality, not as a projection tool.

I agree with this for sure. Good points.

M2
09-24-2007, 02:26 PM
Without diving in to the blow-by-blow (which, while unnecessarily acerbic, has been very informative and interesting), FIP does conceptually include hit events.

It doesn't count them and yet it does. On one hand it packages up HR, BB and Ks in a tidy formula that ignores hits. Yet the run values assigned HR, BB and Ks are derived from a world in which hits are the key determinant in run values. What are the run values of HR, BB and Ks when they aren't caught in a current driven by hits? We don't know and the mathematical convenience that, voila, the hits disappeared doesn't cut mustard. That is particularly so because we're learning that many good pitchers do better on hit rate and/or quality over time than their bad pitching counterparts. No one's come up with a Rosetta Stone to translate that difference, but we're finding it.


From my understanding, the premise is that the rate of balls in play is implied as the remainder of all events not captured in FIP. However, since the assumption is that the proportion of BIP events which should become hits is constant, it is not necessary to include it in the formula. It is essentially factored in to the weights that are present.

He has to exclude hit rate from the equation because the entire premise is that H/BIP is not a function of the pitcher. While BIP/PA is within the control of the pitcher, it is already accounted for in the formula inherently, isn't it?

Also, Jojo, isn't the concept of FIP just that, "how would this guy pitch with a neutral defense". It's correlation to future observed ERA is really irrelevant, given that year-to-year defensive performance is not random. If a guy has a poor ERA this year, relative to his FIP, and his defense doesn't change, we wouldn't expect his ERA to decrease beyond some basic regression to the mean. Thus, the real value of FIP is as a comparative measure of pitcher quality, not as a projection tool.

I think "comparative measure of quality" is solid wording. FIP seeks to assign a run value to your three true outcomes. I'll guarantee you better formulae will be devised to assign that value in the future, we've already got xFIP. The relationship of those three true outcomes to each other and to the overall question of pitcher performance is far from a settled matter. We should be congnizant that ERA will still be ERA when we're kicking around FIP v5.

In the meantime, I suggest the application of sensibility. For instance, maybe the D-Rays kids have untapped upside (most do, though tapping it is often painfully delayed). I don't have a problem with taking the position that it's a good idea to collect pitchers who are tidy with their peripherals. I'm for that, but tidy with your peripherals doesn't necessarily translate one-to-one to good pitching. The 3-4-5 pitchers in the Devil Rays rotation are taking a beating this season. I think, sensibly speaking, the Devil Rays have to be aware that some slick SS defense isn't going to magically turn Sonnanstine, Jackson, Howell and Hammel into a fearsome foursome.

Patrick Bateman
09-24-2007, 02:36 PM
I think, sensibly speaking, the Devil Rays have to be aware that some slick SS defense isn't going to magically turn Sonnanstine, Jackson, Howell and Hammel into a fearsome foursome.

I agree, but I think some decent defense could at least turn Sonnanstine and Howell into legitimate rotation starters. Nothing earth shattering, but I think those two guys in particular are a lot closer to being solid pitchers than their ERA would indicate. Both guys tore up AAA and have good peripherals in the majors. I think both of those signs indicate that some reasonable success should be in the near future. If I'm the Reds, those are two guys that I would target as potential undervalued options that could step in right now at the back of the rotation, and realistically, could become the good pitchers we need in reasonably short order (by comparison to the usual young pitchers you encounter).

Jackson and Hammel still have a lot of lumps to get through before they can be counted on for a whole lot, but I would say that Tampa's rotation could be pretty underrated next season with some astute changes on defense.

RedsManRick
09-24-2007, 02:44 PM
But the question then remains M2, what is the solution to the Rays run prevention problem? Jojo is making the case that obtaining higher quality starting pitchers would not improve the team as much as one might suspect, because the starters are not as at fault for the run preventon failures as ERA suggests.

I would say that the real failure is similar to that of the Reds. Namely, that they have too much offensive production tied up in negative defensive production, preventing the run prevention from ever catching up to run creation.

They need to rework the offensive mix so that they can continue to score runs while allowing their pitching talent to work more effectively. The IF defense has been particularly horrid, but should improve with Longoria at 3B, Iwamura/Harris at 2B, and a new SS. Brignac is probably comparable to Harris. The loss of Wigginton and moving Aki off 3B will help quite a bit.

M2
09-24-2007, 03:30 PM
But the question then remains M2, what is the solution to the Rays run prevention problem? Jojo is making the case that obtaining higher quality starting pitchers would not improve the team as much as one might suspect, because the starters are not as at fault for the run preventon failures as ERA suggests.

Higher quality pitchers seems to work for Kazmir and Shields (though I'm guessing Shields falls back to the other side of 4.00 next season). The Rays certainly don't have enough pitchers at the moment. They need a bullpen cattle call (I figure that's Edwin Jackson's ultimate destination, with the team hoping he can be a Francisco Cordero type). Sonnanstine's ISOp concerns me. I'm thinking he's going to spend 2008 getting victimized by big hits a little too often. Looks like he's got problems pitching from the stretch to me.

Aside from that, I'd think the D-Rays, if they could, would package up some young players for a steady veteran arm. I don't think they can because of finances, so it becomes a moot point. Yet because it's a franchise that likely can't do what it would do, that opens the doors to some potentially desperate acts. For instance, Mike Stanton's got a 4.38 FIP. Maybe he can be a steady veteran influence in a young pen and do it at a price the D-Rays can afford. I swear I didn't intend for this thread to go down an FIP rabbit hole when I first mentioned pawning off Stanton on the D-Rays, but how convenient is that?


I would say that the real failure is similar to that of the Reds. Namely, that they have too much offensive production tied up in negative defensive production, preventing the run prevention from ever catching up to run creation.

It'll be interesting to see how the D-Rays rate in individual defensive metrics after this season. My guess is they're worse collectively than individually. In fact, aside from finding a guy who can play SS well, I doubt the club will be looking at many obvious answers. I'd also love to know what their home DER is. Is this Nu-Turf related? Or maybe it's a matter of the team in the field needing time to jell. There's also the shooting gallery theory -- young, inconsistent pitchers are making it impossible for players and coaches to get a handle on tendencies for where the ball might be headed in a given AB. I'm not pretending to have the answer here, but I suspect it's not simple.

jojo
09-25-2007, 12:22 AM
My contention is you're overly credulous about supposedly "better" metrics and that you aren't recognizing their flaws. You buy into them and then try to force everything through those filters, ignoring the spectrum and putting all your eggs in shoddy basketry. I suggest you start paying attention to a broader spectrum of inputs when you're evaluating talent and not fall into the trap of there being one stat to rule them all.

Please, enough with the straw man arguments (http://www-personal.umich.edu/~lilyth/strawman.html). Really, I’m asking nicely for the sake of the discussion.

This issue with FIP started when it was suggested Tampa has one of the top 5 rotations in the AL, a fact that is obscured by a precursory examination of their ERA because Tampa’s defense is so atrocious. I supported this statement by looking at the FIPs of the Tampa rotation in the context of the league. I used FIP mainly because it was handy to do so. In other words, it was easy to compile FIPs for the league using excel as opposed to the extra tedium of tabulating BIP data for all 14 AL teams in order to calculate xFIP for example. Really for the point I was making, FIP is good enough because regardless of the defense-independent metric used (i.e. FIP, xFIP, DIPs ERA etc.) the conclusion would essentially be the same. The fact that we happen to be discussing FIP in no way implies what you’re suggesting (i.e. the false notion that I’ve argued FIP and only FIP should solely be used when evaluating pitchers).

To frame a counterargument that implies I advocate a single metric as the end all without the foggiest consideration of its flaws (and by extension therefore my argument should be ignored) is intellectually dishonest especially in light of multiple posts I’ve authored in the past on this very subject (similar in nature to this link (http://www.redszone.com/forums/showpost.php?p=1296318&postcount=21)):


ERA basically just tells you how many runs were scored while a pitcher was on the mound-there's alot of reasons why those runs may have scored that are completely unrelated to how effective the pitcher actually was.

FIP basically evaluates a pitcher based upon the counting stats that are only influenced by himself (i.e. walk, strikeout, and home run rates). It has its flaws-namely it assumes a pitcher can control whether a flyball is a homerun or not and it assumes a constant LOB% (neither of which are true as pitchers that deviate will regress back to the mean except of course in the case where a pitcher is toast hence these are sometimes considered luck factors). xFIP basically corrects FIP for the HR rate. The great thing about FIP is that since it is based upon things that are related to the pitcher's actual skill set, it's a much better predictor of future performance than a blunt metric like ERA which is influenced by luck, the defense behind him etc.

If you look at a pitcher's peripherals (K%, BB%, K/BB, GB%, HR/FB%, LOB%, and BABIP) in conjunction with a superior metric like FIP you really can get a very complete picture of why a pitcher performed the way he did and a really good feel for what you should expect he'll do in the future.

and statements in this very thread like this:


To me the best approach to player evaluation is to recognize the flaws in all metrics so that they can be weighted accordingly when using a spectrum-based approach to evaluate talent while also being ready to reject an intuitively held belief when there is good cause for pause.

Really the actual issue at the heart of this discussion is “what is the best way to evaluate a pitcher”? Clearly as indicated by the above post (and many others just like it) when I advocate a spectrum-based approach to evaluating a pitcher, I’m arguing that many, many factors should be considered. Essentially, I come from the position that evaluating a pitcher based upon things he actually controls (i.e. rate stats, FIP etc) while considering his luck/unluck metrics is a superior approach to evaluating him based upon metrics that are so dramatically influenced by factors out of the pitcher's control (which in essence dilute out the pitcher’s contribution to run prevention with tons of noise i.e. ERA or w-l record). I guess you can call that a bias but it’s one I’ve arrived at after considerable thought.

So rather than to continue to refute straw man after straw man, I offer this:

Intuitively, a metric which captures a pitcher’s repeatable skills would be a superior metric to use when evaluating a pitcher compared to a metric that contains information which is heavily influenced by factors that are out of a pitcher’s control.

This notion was a primary motivation behind defense-independent metrics like FIP et al and it really suggests that metrics which measure a pitcher’s repeatable skills should correlate more highly to his future performance than ones that are heavily influenced by non-repeatable events. This has been proven true by several studies where things like K rate and BB rate predicted future ERA better than did past ERA. The point isn’t to predict exact ERA per se but rather that ERA is a flawed metric that doesn’t necessarily reflect the pitcher’s true skill.

If K rate and BB rate are “better” than ERA then metrics based upon them should also better than ERA. Below is a quick analysis that tests this prediction. The data set is comprised of all starting pitchers in 2005 and 2006 that both pitched for the same team (i.e. thus controlling park and league effects) and pitched enough innings to be qualified both seasons (i.e in order to avoid the influence of injury). Here are the results of correlating pitcher’s 2005 ERA, FIP, and xFIP with their 2006 performances:

http://img150.imageshack.us/img150/5121/qualifiedpitchers200520ka9.jpg (http://imageshack.us)

Immediately it is obvious that while 2005 ERA did show a significant correlation to 2006 ERA, the relationship is much weaker than that of the defense independent metrics, FIP and xFIP. In fact, 2005 ERA really wasn’t strongly correlated with anything associated with 2006. On the other hand, FIP and xFIP –which focus on indices of repeatable skill in an attempt to eliminate the noise associated with the effects of things that are largely out of the pitchers control by – not only strongly correlate with ERA but also with future FIP and xFIP as well.

If defense-independent metrics like FIP and xFIP are more reflective of a pitcher’s future ability to prevent runs, then shouldn’t they also be considered more reflective of a pitcher’s true ability to prevent runs in the present/past as well for the same reasons?


I wasn't trying to do either, just noting your particular slavishness to statistical dead ends.

Really, it seems to me that if defense-independent metrics are statistical dead ends in reference to pitcher evaluation, then using ERA is the statistical equivalent to sailing off of the edge of the world.

Lets shift focus by summarizing some of the criticisms of FIP that have been argued in this thread:

1.The derivation of the weights used in the FIP formula was criticized as being based upon unsound methodology:


And why do you multiply HR by 13 and then BB+HBP by 3 and Ks by 2? Did Tangotiger prove that those are the relative weights of those events? No he didn't. What he did was monkey around with multipliers until he got something that spit out something that looked like an ERA stat. That's right, he completely ignored hits and the quality of the hits the pitchers allow and then came up with a fairly arbitrary set of multipliers all in the name of generating a doppleganger of a statistic you just called "worse than lousy." Talk about crap math.

This assessment is simply patently false as shown earlier (it’s not a matter of whether 13, 2, and 3 were factually correct-I obviously never suggested you had that wrong. Rather your criticism centering upon how they were derived is specious). There was no “monkeying around” with multipliers unless using time-honored statistical methods like probability theory and regression analysis to fit a function to a massive data set via a predetermined rational is considered “monkey business”. I’m guessing if that is “monkey business” or “crap math”, the disciplines of science, medicine, and mathematics have some explaining to do and we should stop pretending that we're enjoying the benefits of their break throughs. Tom Tango and others applied accepted statistical methodologies to a baseball-related problem.

2. FIP was criticized because it eliminates the direct impact of hits from the pitcher’s responsibility. While the compromise to eliminate hits from the equation probably does reduce the power of defense-independent metrics, as a flaw it is minimal relative to ERA’s blunt inclusion of the defense. There is a large body of literature that suggests in general a pitcher’s ability to control whether a ball in play becomes a hit or an out is measurable but minimal. Obviously there will be statistical anomalies (which can be easily dealt with using other metrics as suggested above) but focusing upon the exclusion of hits grossly overestimates the influence a pitcher has on that outcome.

3. Another criticism has centered upon linear weights (LWTs) and the assertion that base/out stats and resulting run expectancy can’t be divorced from a hit-containing environment as embodied by statements such as these:


As for Tango's methodology, he solved for x, y an z using run value system that ultimately is tied up with ... wait for it ... hits. It's the dopiest math ever. Let me say it clearly one more time so that it resonates, Tango tried to come up with linear weights for the run value of non-hit events using a system where the largest contributor to runs is hits. The thing he's trying to move out of the equation is still running wild under the covers. Beyond that, the relation of all those events to each other is not a constant. If you don't think 13, 3 and 2 were selected because ultimately it represented the best way to create a static ERA-like equation then I'd say you've missed a fairly large boat.


Trying to come up with an evaluation of pitcher performance that ignores the plurality of events in that pitcher's performance is an exercise in futility.

While I understand you’re troubled by this issue, really to continually argue this is a fatal flaw (especially in light of the abundant evidence that FIP actually does effectively do what it attempts to do) you really should demonstrate it empirically.

(FYI-for the uninitiated, base/out states are really just contexts that have associated run expectancies based upon past run totals. In layman’s terms, LWTs are really just the average value of the event combined across all base/out states of the run expectancy table. There is are excellent discussions of these issues in The Book and in Total Baseball)


Seems to me your interest lies in tweaking insignificant margins.

Actually your criticisms of FIP essentially peck away at the periphery. Really, the flaws derived from the compromises made in order to eliminate as much of the effect of defense as possible from the equation pales in comparison to the effects of just the defense alone that are contained in ERA. It’s the white elephant in the closet that much of your argument against FIP fails to acknowledge.

As an aside:


We both know you won't be using BB rate as the true estimate of pitcher value at any point in the future.

Personally, I think command is a very important component of pitcher value as threads such as these indicate:

http://www.redszone.com/forums/showthread.php?t=57849&highlight=Bailey+Livingston

http://www.redszone.com/forums/showthread.php?t=58578&highlight=Bailey+Livingston

Why are these issues important? Rather than being about FIP per se (or any one metric), they speak to the heart of how best to evaluate a pitcher's true talent and why some find statements like the following incredulous:


Aside from that, I'd think the D-Rays, if they could, would package up some young players for a steady veteran arm.

pedro
09-25-2007, 12:40 AM
One Stat to rule them all, One Stat to find them,: One Stat to bring them all and in the darkness bind them

Ltlabner
09-25-2007, 08:41 AM
I'm going on record to say that is the most redzoniest (and boring) pissing match ever.

I'm sure there's something interesting in there somewhere. But my eyes glazed over and I started drooling all over my keyboard.

Geez....at least argue over something intresting like strikeouts, or girls or something.

:p:

Kc61
09-25-2007, 09:51 AM
I'm going on record to say that is the most redzoniest (and boring) pissing match ever.

I'm sure there's something interesting in there somewhere. But my eyes glazed over and I started drooling all over my keyboard.

Geez....at least argue over something intresting like strikeouts, or girls or something.

:p:

I thought I started a thread about trading Stanton! Oh well, there are still six days left in the season, let the discussion continue.

M2
09-25-2007, 11:30 AM
Please, enough with the straw man arguments ((http://www-personal.umich.edu/~lilyth/strawman.html)). Really, I’m asking nicely for the sake of the discussion.

This issue with FIP started when it was suggested Tampa has one of the top 5 rotations in the AL, a fact that is obscured by a precursory examination of their ERA because Tampa’s defense is so atrocious. I supported this statement by looking at the FIPs of the Tampa rotation in the context of the league. I used FIP mainly because it was handy to do so. In other words, it was easy to compile FIPs for the league using excel as opposed to the extra tedium of tabulating BIP data for all 14 AL teams in order to calculate xFIP for example. Really for the point I was making, FIP is good enough because regardless of the defense-independent metric used (i.e. FIP, xFIP, DIPs ERA etc.) the conclusion would essentially be the same. The fact that we happen to be discussing FIP in no way implies what you’re suggesting (i.e. the false notion that I’ve argued FIP and only FIP should solely be used when evaluating pitchers).

Oh, you're precious. You used FIP as a proof point to try to establish the Rays have a top 5 rotation in the AL and now you're going to play this game about you never having put too much stock in it? Stop being evasionary. Studies for the past five years keeping turning up that, lo and behold, pitchers do have an effect on balls in play (which really should have elicited a "no duh" response).

You can keep trying to make the claim that taking a partial look at pitcher performance is the way to go, but you're always going to be handicapped by that small world view. It's just going to keep running into inconsistencies like Tampa Bay's defense not being a problem for some guys, but apparently crippling to others, and this without hit luck involved. PRC (which I'd suggest if you're going for a new age pitching performance measurement) suggests you're full of it too.

That you can't understand the mathematical imperative to not use the thing that you claim distorts performance data as part of the equation to establish your linear weights speaks for itself. That's just bad algebra.

And I'm getting a tremendous kick out of your "I think command is a very important component of pitcher value" dance. Um, everybody thinks that. I suggest you check out that strawman link you posted because that's what you've got right there. Once again, what you're not doing, because it would be high stakes insanity to do so, is boiling pitcher performance down to BB rate (something that FIP does at the league level in the NL). Paul Byrd is not the best pitcher in the AL. Dave Bush and Matt Belisle aren't the 2nd and 3rd best pitchers in the NL.

Though let's wrap back to FIP, the D-Rays and Mike Stanton. Stanton has a 4.39 FIP on the heels of consistently good FIPs in preceeding seasons. That would make him one of the better pitchers in the Tampa Bay bullpen by your reckoning. Sonnanstine, whom you want to praise, has a 4.31 FIP. I understand Stanton's age is an issue, but given the volatility of young arms, you'd think the Rays would welcome a steady performer or two, particularly in a bullpen that ranks 14th in the AL in FIP? So, if defense independence really is the bee's knees, then what's so wrong with Mike Stanton? And while we're on that point would you like to now go on record in support of the season that Gary Majewski (4.17 FIP) is having?

M2
09-25-2007, 01:45 PM
If defense-independent metrics like FIP and xFIP are more reflective of a pitcher’s future ability to prevent runs, then shouldn’t they also be considered more reflective of a pitcher’s true ability to prevent runs in the present/past as well for the same reasons?

No, in fact that's highly tortured logic, but you put work into the chart and I appreciate that so I thought I should address this separately.

First let's start with the test you devised. You took a group of pitchers who were generally effective two seasons in a row. This gets back to the insignificant margins I mentioned. I already know Brandon Webb is a good pitcher. That he gives up a few more or less runs in a season isn't all that revelatory.

And, since you're big on strawmen, let's remove the gigantic one you've created here. No one has ever pretended ERA is predictive. In fact there is nothing in ERA that would even intimate predictive abilities. It measures present and historical runs allowed and that's it. When stuff happens in the future it will measure that, but it will have to happen first. FIP does make that pretense of predictability.

What your study actually shows is that even when cherrypicking a highly correlated group -- pitchers who stayed healthy and generally pitched well enough to qualify for an ERA title two seasons in a row -- FIP had a low correlation (.531). That literally is proof of non-predictability, even for the low margin predictions at stake in this study.

More interesting, to me, were the dozen pitchers who posted 4.50+ ERAs in 2006. These were the guys you'd have been wise to stay away from and FIP really didn't deliver that message in any substantive way. ERA didn't either, but, as mentioned, ERA doesn't pretend to offer up future insight.

Yet that dirty dozen is the tip of the iceberg. What about all the pitchers who flopped in 2006? Or all the pitchers who flourished? I'm not going to run an FIP study on those constituencies because I can already pretty much assure you I'd find an incredibly weak correlation, weak enough where anyone would have to admit that doing an estimate based on that data would be pure hit and miss, which is what .531 was. Correct me if I'm wrong here, but RBIs correlates a whole lot better to runs scored than that, doesn't it?

Now onto a second strawman, nowhere did anyone say that low HR and BB totals and high K totals aren't good things to do or that pitchers who can repeat those skills year after year won't be good. Of course those things are important, I pay a ton of attention to them in fact. Yet it is a Snake Canyon rocket car launch to go from there to FIP. Just because those things are important does not mean FIP has skewered their relative importance or that they are the only things that are important when it comes to pitcher evaluation.

Per your study, FIP doesn't correlate all that well with itself either (.610). So even pitchers who do well with HR, BB and Ks experience volatility in those select areas of performance. It's not like that performance is a fixed quantity that replicates itself each year. Some pitchers repeat it better than others, but when you get down to predictive ability at the individual level FIP isn't a beacon that anyone ought to rely upon.

And that brings me to the following - there is a universe of difference between current state and predictive ability. They are totally different animals. I think we'd both agree that run prevention is what we care about most when it comes to pitching and, IMO, it speaks volumes that you went back to ERA as your barometer in the above study. ERA does what it does, it tells you the current state of runs allowed. FIP tells me the current state of HR, BB and Ks (though, as mentioned, I fully expect it to go through years of revisions before it reaches a final state). I care more about the former than the latter. When it comes to predictive ability, ERA doesn't feign it and FIP is no better than quasi-science in that area. Just because you might use the components in FIP to aid in a pitcher projection does not mean they add up to any sort of current measure of quality.

So, no, one non-predictive measurement is not a better measurement of current state than another non-predictive measurement as neither acts as a reasonable summation of "true ability." When it comes to current state, I'm going to prioritize runs or something that correlates to runs (PRC) over peripherals. You can always argue that maybe it would roll up differently if you did it again, but they don't do it all over again and that is how it rolled up.

SteelSD
09-27-2007, 12:41 AM
Per your study, FIP doesn't correlate all that well with itself either (.610). So even pitchers who do well with HR, BB and Ks experience volatility in those select areas of performance. It's not like that performance is a fixed quantity that replicates itself each year. Some pitchers repeat it better than others, but when you get down to predictive ability at the individual level FIP isn't a beacon that anyone ought to rely upon.

You're correct, M2. DIPS and FIP are a starting point suggesting that a pitcher may have produced worse than expected results, but it's only a partial predictor. I've run the 2005 v 2006 numbers for 2005 ERA qualifiers and I can tell you that xFIP is a "junk" statistic as a predictor for the 2006 season. It correlates lower than either DIPS or FIP even though THT theorizes that their "experimental" stat it's a better predictor. But that's THT in a nutshell- "theory" and "experimental".

Projecting pitching is an art form. It can't be limited to simple defense-independent numbers because we need to know whether or not those peripherals are likely to be matched by a pitcher from season to season.

For example, Carlos Silva might look like an attractive free agent pickup to some team who thinks that his 2007 4.19 DIPS rate is repeatable. The problem is that his 5.50 DIPS rate from 2006 and his 4.33 DIPS rate from 2005 suggest it's not repeatable as his 2007 HR rate is a complete aberration and his 2005 BB rate is non-repeatable. Some team might pay over 5M for Silva's services and I'd suggest that team is dumb. Overall, Silva's DIPS ERA tells us that if he produces outlier rates in some categories versus his norm, he'll be a decent value. The problem is that such an assumption depends on a repeat of virtually non-repeatable events given his norms.

Now, I think ERA sucks big donkey balls as a predictor, but defense-independent numbers need to be analyzed more closely and expectations need to be adjusted based on the probability that a pitcher can reproduce defense-independent peripherals.

To this point, I'd consider xFIP to be a complete failure as a predictive indicator and both DIPS and FIP, while suggestive, don't go far enough to equalize based on pitcher trends. Are both a reasonably accurate estimate of what a pitcher "should" have done based on their defense-independent statistics? Sure. But neither is something that should be relied on as demonstrably predictive without further research into the player's probability to repeat their defense-independent numbers.

Ltlabner
09-27-2007, 07:27 AM
Now, I think ERA sucks big donkey balls as a predictor.

Steel,

A well written, well stated and very interesting post.

This is the only thing I learned from it.

:D