PDA

View Full Version : Good Article on Floyd/Danks in 2009 + Cool New (?) Formula/Stat...


WhiteSoxFan84
12-15-2008, 01:32 AM
Link to Article (http://www.examiner.com/x-425-Chicago-White-Sox-Examiner~y2008m12d13-Can-John-Danks-and-Gavin-Floyd-repeat-their-breakout-seasons)

It's an article by a student from the University of Mizzou and it's pretty decent but the one thing that caught my attention is the "Fielding Independent Pitching (FIP)" statistic that supposedly "is an excellent statistic for measuring a pitcher's true ability". The formula is a bit scary for those who fear numbers...

The formula is (HR*13+(BB+HBP-IBB)*3-K*2)/IP, plus a league-specific factor (usually around 3.2) to round out the number to an equivalent ERA number.

...according to the author, this is a formula that helps determine how well a pitcher pitched regardless of the defense behind him. If his ERA and FIP are close enough, then it means the pitcher had a legit season (legit meaning if he did good he was good, if he did bad he was bad). But if his ERA is a lot lower than his FIP, it means his season was a "mirage".

Danks' ERA in 2008 was 3.32 and his FIP was 3.44 which is good news.
Floyd's ERA was 3.84 and his FIP was 4.77, very bad news.

guillensdisciple
12-15-2008, 02:12 AM
I have read many things like this about Floyd, stating that he didn't have much to back up his numbers, and many believe that he is due for a drop-off because of this.
However, I believe that a sign of a player that isn't ready or had a fluke year, is one that pitches in a situation that isn't pressured or stressful in any way. Floyd pitched the Sox out of some very important situations, even when he looked bad, he would come out of it just fine.

Add the fact that "miracle" seasons usually happen to older people (Loaiza, Javy), there is nothing that should be alarming about Floyd and his ability should either stay the same or improve as his career progresses.

Jimmy John
12-15-2008, 02:36 AM
I never thought we would ever get what we got out of Gavin last season, so I'm satisfied. The dude flirted with two no-hitters, for god sakes.

Konerko05
12-15-2008, 04:00 AM
I can see where people are coming from with the "FIP," and "BABIP" thing.
It makes sense in Garland's case because he already set a norm for his pitching ability. 2005 was an anomaly.

In Floyd's case, it doesn't really do much for me. This was Floyd's first full season in the MLB. Why does everyone assume these stats mean Floyd is going to regress next season? Why is it out of the realm of possiblity that Floyd can actually improve on his peripheral numbers next season?

Maybe Floyd was just really good at missing centers of bats last season. Next year he could possibly improve to the point of missing entire bats.

DumpJerry
12-15-2008, 07:13 AM
Oh goody. Another reason to not waste my time going to ballgames. Just feed me the numbers......

Craig Grebeck
12-15-2008, 07:29 AM
Oh goody. Another reason to not waste my time going to ballgames. Just feed me the numbers......
Really though, is that the intent? No. The numbers are simply used to analyze a pitcher's ability to control things independent of the defense.

On Floyd, I'd argue that his FIP in the second half was much, much better, IIRC. He was not very good early on, but his core numbers started to right themselves in the second half.

whitesox901
12-15-2008, 08:54 AM
I have read many things like this about Floyd, stating that he didn't have much to back up his numbers, and many believe that he is due for a drop-off because of this.
However, I believe that a sign of a player that isn't ready or had a fluke year, is one that pitches in a situation that isn't pressured or stressful in any way. Floyd pitched the Sox out of some very important situations, even when he looked bad, he would come out of it just fine.


Agreed, I think Gavin can repeat what he did. I doubt he had a fluke year & I believe he is the real deal.

doublem23
12-15-2008, 09:32 AM
Defense Independent Pitching Stats, or DIPS, have been around for almost a decade now.

That specific formula is for the formula for DICE (Defense Independent Component ERA), and that's been around since 2000 or so.

hi im skot
12-15-2008, 09:53 AM
http://savetherobot.files.wordpress.com/2007/11/535full.jpg

-Daver

:redneck

guillen4life13
12-15-2008, 09:58 AM
FYI, the author of the article is a frequent poster on WSI.

Zisk77
12-15-2008, 10:01 AM
I suppose Floyd's curveball and slider are mirages too :scratch:

WhiteSox5187
12-15-2008, 10:07 AM
I suppose Floyd's curveball and slider are mirages too :scratch:
Yep, who cares that he won 17 games, threw 200 innings, had an ERA in the threes and a WHIP of 1.2, his FIP is bad. Gotta get him out of here. :rolleyes:

Craig Grebeck
12-15-2008, 10:22 AM
Yep, who cares that he won 17 games, threw 200 innings, had an ERA in the threes and a WHIP of 1.2, his FIP is bad. Gotta get him out of here. :rolleyes:
Again, not really the point of FIP or DIPS. They're meant to examine the possibility of a pitcher's ERA not telling the full story. Of course, Floyd was successful last season, but FIP and DIPS are meant to show how sustainable that performance is. In Danks' case, his performance is more sustainable because of a solid K rate and good control, while Floyd's K rate (mainly in the first half), when compared to his BB rate, was mediocre. In the second half, he was much better, which showed progression through the course of the season.

Lip Man 1
12-15-2008, 11:06 AM
After trying to read and understand (even a little bit) of this "Newtonian mathematics', my only comment is that my brain hurts.

Lip

doublem23
12-15-2008, 11:43 AM
After trying to read and understand (even a little bit) of this "Newtonian mathematics', my only comment is that my brain hurts.

Lip

The formula posted doesn't require anything beyond elementary mathematics (adding, subtracting, multiplying, and dividing).

CashMan
12-15-2008, 12:25 PM
The formula posted doesn't require anything beyond elementary mathematics (adding, subtracting, multiplying, and dividing).


He probably means, how they come up with these formulas, not the math.

guillen4life13
12-15-2008, 12:31 PM
He probably means, how they come up with these formulas, not the math.

I would imagine they're a combination of trial and making the numbers fit.

I happen to be in the camp that says Floyd either repeats or improves on 2008. I expect that, like just about every pitcher who works under Coop for a year or more, that he will add a cut fastball to compliment his slider and throw hitters off.

Control and confidence are Floyd's biggest issues. 2008 gave him confidence. Now lets see him build on it.

forte
12-15-2008, 12:49 PM
The formula posted doesn't require anything beyond elementary mathematics (adding, subtracting, multiplying, and dividing).

Keep in mind that this is the same poster that thinks that advanced calculus is required to calculate OBP. Math must not have been his strong subject when he was in school.

Lip Man 1
12-15-2008, 01:32 PM
Forte:

In fact it wasn't but the point is that this is baseball... something to enjoy. You shouldn't need to have to try to understand all these bizarre, formulas that some folks have come up with to try to "predict" what's going to happen, then when it completely and totally falls apart and they are dead wrong, you never see or hear "I (or we) were wrong...)

I'm looking at you egg-heads from Baseball Prospectus.

Lip

infohawk
12-15-2008, 01:38 PM
I never thought we would ever get what we got out of Gavin last season, so I'm satisfied. The dude flirted with two no-hitters, for god sakes.
Gavin Floyd will succeed or fail based on his curveball. When he gets it over, he's very tough, and that won't change. One would hope that his consistency with the pitch will continue to improve with experience and confidence.

CashMan
12-15-2008, 01:54 PM
Forte:

In fact it wasn't but the point is that this is baseball... something to enjoy. You shouldn't need to have to try to understand all these bizarre, formulas that some folks have come up with to try to "predict" what's going to happen, then when it completely and totally falls apart and they are dead wrong, you never see or hear "I (or we) were wrong...)

I'm looking at you egg-heads from Baseball Prospectus.

Lip


I think yes and no. Is Floyd going to repeat a 17 win under 4 ERA, maybe. It is kinda cool looking at some of these formulas and seeing if they are right. I would have to say, how accurate is this equation?

Eddo144
12-15-2008, 03:03 PM
Forte:

In fact it wasn't but the point is that this is baseball... something to enjoy. You shouldn't need to have to try to understand all these bizarre, formulas that some folks have come up with to try to "predict" what's going to happen, then when it completely and totally falls apart and they are dead wrong, you never see or hear "I (or we) were wrong...)

I'm looking at you egg-heads from Baseball Prospectus.

Lip
Yeah, but what mainstream writers do you hear "I was wrong" from? It's bad for business.

And you don't need to understand formulas. Some of us, however, enjoy these "formulas". Why should we be subject to the insults of those who don't share our enjoyment?

RockyMtnSoxFan
12-16-2008, 10:36 AM
That formula looks nice, but what happens if you change some of the coefficients? Why are home runs multiplied by 13 and not 12? I think that empirical formulas like this one can be misleading because they have to be tuned (the coefficients have to be chosen manually) in order to make them appear to match some set of data, which was chosen arbitrarily. Also, I think there is a tendency in the sabermetrics community to attribute anything for which there is no stat to luck. If a pitcher's BABIP is lower than .300, he must have been lucky. Personally, I think that BABIP might be useful when a pitcher has a season that is out of family compared to his previous record, but comparing against other pitchers with different styles can be erroneous.

I wouldn't be surprised to see Gavin drop off a bit this year. But as someone said earlier, if he can continue to improve his command of the curve, he should see success. Seventeen wins might not happen in 2009, but he could still be a good #3.

Lip Man 1
12-16-2008, 10:54 AM
Eddo:

Because some of you (and you know who you are both nationally and locally) act like these bizarre convolutions of the imagination are "the absolute truth" and need to be reminded that they aren't

No personal offense to you but honestly, will MLB cease to be or the world stop spinning if fans don't know what Paul Konerko hit in the 8th inning or later against left handed pitchers who wear their cap off center with last names of less then six letters on every other Friday night at home?

And before you say 'well that's an extreme example' that's my point. Today fans are getting hammered with extreme, bizarre, made up numbers from everybody and his brother, who considers themselves a baseball "expert."

I venture to say 90% of these "experts" never played the game beyond Little League.

Lip

jabrch
12-16-2008, 10:56 AM
Why should we be subject to the insults of those who don't share our enjoyment?

Why do you give a ****? If Lip (or anyone else) thinks they are stupid and only a propeller head would pay any attention to it, why do you care so much? You aren't subject to anything. Change the channel. Ignore him? Who cares?

If you think 13*x+YZ^2 increases your enjoyment, and that the accuracy of that is significant - then enjoy it. If Lip thinks watching baseball in more enjoyable than "newtonian mathematics" then he should enjoy it.

I still don't get why some people get so defensive about their own ethos and why they care so much about the opinions of strangers. If you don't like it, the ignore feature is your friend.

DumpJerry
12-16-2008, 10:57 AM
Really though, is that the intent? No. The numbers are simply used to analyze a pitcher's ability to control things independent of the defense.

On Floyd, I'd argue that his FIP in the second half was much, much better, IIRC. He was not very good early on, but his core numbers started to right themselves in the second half.
Yeah, he had only two near no-hitters in the first half of the season in 2008.

voodoochile
12-16-2008, 11:01 AM
FYI, the author of the article is a frequent poster on WSI.

The stat FIP was created by a guy who has posted here in the past too. Mostly he's had a few surveys. I think he's posted one a year for the last few years. I would expect he'll be back again.

voodoochile
12-16-2008, 11:06 AM
Yeah, but what mainstream writers do you hear "I was wrong" from? It's bad for business.

And you don't need to understand formulas. Some of us, however, enjoy these "formulas". Why should we be subject to the insults of those who don't share our enjoyment?

I'd like to see a rationale for the numbers being used in the formula before I would be willing to swallow it blindly.

BABIP is at least something anyone can understand and choose to accept it or not. I mean it's basically the batting average for balls that are put in play by any given batter or batting averaqe that doesn't include walks and K's. That's not hard to understand and it might actually have some merit if they can do some studies on it long term.

But again, who knows where TangoTiger got the numbers he used in that forumla or why we should accept them as valid?

jabrch
12-16-2008, 11:10 AM
I venture to say 90% of these "experts" never played the game beyond Little League.

I would say their playing experience is relatively insignificant. I totally agree with you on the silliness of some of these statistics, and the gerryrigging that is needed to make them "work" is of questionable value. The coefficients and the factors used to adjust the value of an actual measureable number that then creates a made-up statistic is silly. Just look at this formula - (HR*13+(BB+HBP-IBB)*3-K*2)/IP and some of it is difficult to explain. It all derrives from dERA which is very complicated. Tom Tango came up with FIP, which is similar in results to dERA and DICE. The question is why 13? why 3? Why 2? Why not 11.25, 1.86 and 2.45? I get it - but they are weighting the importance of different things - and that determines the output. It also fails to factor in the fact that there are many ways to skin a cat. There are a lot of GREAT pitchers who do depend on their defense. There is absolutely nothing wrong with that. In fact, many of them don't go so deep in counts, go more innings, and are more effective than their strikeout focused peers.

Bottom line - these composite numbers tell you a small piece of the puzzle. They are not nearly as accurate as some (many who don't understand them - but cite them freely) want to believe. I'm withyou on this Lip. These defensive independent statistics are stupid - until we go to a version of baseball with no defense. I don't believe you can determine how good a pitcher is without a defense because that devalues the guys who strategicly and effectively rely on their defenses.

voodoochile
12-16-2008, 11:14 AM
I think yes and no. Is Floyd going to repeat a 17 win under 4 ERA, maybe. It is kinda cool looking at some of these formulas and seeing if they are right. I would have to say, how accurate is this equation?

Exactly, but even if Floyd had a FIP of 4.77 - how horrible is that in his first full season? I mean lets say that this formula is correct, and Floyd's ERA should have been 4.77 last year. Most people would say that's a fine first full season in the majors and please keep giving him the ball. Heck, Danks posted an actual ERA of 5.5 the year before and everyone wanted to see him back.

Oh and just for ****s and giggles...

Danks FIP from 2007: (28*13 + (54+4-4)*3 - 109*2)/139 = 2.22

2.22+3.2=5.42 FIP for Danks in 2007. Heck, based on this, we should expect good things from Floyd next year...

jabrch
12-16-2008, 11:14 AM
I'd like to see a rationale for the numbers being used in the formula before I would be willing to swallow it blindly.

BABIP is at least something anyone can understand and choose to accept it or not. I mean it's basically the batting average for balls that are put in play by any given batter or batting averaqe that doesn't include walks and K's. That's not hard to understand and it might actually have some merit if they can do some studies on it long term.

But again, who knows where TangoTiger got the numbers he used in that forumla or why we should accept them as valid?

The difference between something like this and something like BABIP is that BABIP is just doing figuring out an average (BA) of a specific set of data (BIP). At the end of the day, it doesn't profess to measure who is better or worse - just tells you what happened. FIP has 3 constants being used to convert a mathematical function of 6 different measureables. FIP professes to tell you who would be better if Fielding were not a factor.

voodoochile
12-16-2008, 11:20 AM
Good point about valuing K's and HR more than other factors. For example Mark Buehrle's Career FIP is 4.21 - about .4 higher than his career ERA. Guys like him and Floyd are going to be undervalued by FIP...

spiffie
12-16-2008, 11:22 AM
Here's the formula I use to tell me Gavin Floyd is a good pitcher:

Eye 1 + eye 2 + brain = Gavin Floyd rocks.

Here's the formula I use to tell me what to think of a stat:

numbers involved in figuring out stat * time stat creator spends playing D&D in basement / girls kissed by stat creator = how many better things there are to think about than useless mental masturbation.

Seriously, baseball is a fun game. Watch it once in a while.

voodoochile
12-16-2008, 11:26 AM
Here's the formula I use to tell me Gavin Floyd is a good pitcher:

Eye 1 + eye 2 + brain = Gavin Floyd rocks.

Here's the formula I use to tell me what to think of a stat:

numbers involved in figuring out stat * time stat creator spends playing D&D in basement / girls kissed by stat creator = how many better things there are to think about than useless mental masturbation.

Seriously, baseball is a fun game. Watch it once in a while.

Personally, I'm more interested in a pitchers FIP after a night when they spent the first 8 hours pounding shots and the final 5 hours locked in a hotel room with 3 baseball groupies and then had to start an afternoon game.

:hawk:
"Don't tell me what your FIP is, tell me when you FIP it..."

itsnotrequired
12-16-2008, 11:43 AM
Seriously, baseball is a fun game. Watch it once in a while.

i refuse

RockyMtnSoxFan
12-16-2008, 11:49 AM
I would say their playing experience is relatively insignificant. I totally agree with you on the silliness of some of these statistics, and the gerryrigging that is needed to make them "work" is of questionable value. The coefficients and the factors used to adjust the value of an actual measureable number that then creates a made-up statistic is silly. Just look at this formula - (HR*13+(BB+HBP-IBB)*3-K*2)/IP and some of it is difficult to explain. It all derrives from dERA which is very complicated. Tom Tango came up with FIP, which is similar in results to dERA and DICE. The question is why 13? why 3? Why 2? Why not 11.25, 1.86 and 2.45? I get it - but they are weighting the importance of different things - and that determines the output. It also fails to factor in the fact that there are many ways to skin a cat. There are a lot of GREAT pitchers who do depend on their defense. There is absolutely nothing wrong with that. In fact, many of them don't go so deep in counts, go more innings, and are more effective than their strikeout focused peers.

Bottom line - these composite numbers tell you a small piece of the puzzle. They are not nearly as accurate as some (many who don't understand them - but cite them freely) want to believe. I'm withyou on this Lip. These defensive independent statistics are stupid - until we go to a version of baseball with no defense. I don't believe you can determine how good a pitcher is without a defense because that devalues the guys who strategicly and effectively rely on their defenses.

I agree with you jabrch. This formula was manually adjusted to match an arbitrarily chosen set of data. And I think the reason it focuses heavily on Ks, BBs, and HRs is that those are easy things to measure. They haven't figured out how to measure when a guy is getting weak off-balance swings vs. having a lot of fly balls die just short of the wall. I think stats can be insightful, but this is a case where they definitely misleading.

jabrch
12-16-2008, 12:03 PM
I agree with you jabrch. This formula was manually adjusted to match an arbitrarily chosen set of data. And I think the reason it focuses heavily on Ks, BBs, and HRs is that those are easy things to measure. They haven't figured out how to measure when a guy is getting weak off-balance swings vs. having a lot of fly balls die just short of the wall. I think stats can be insightful, but this is a case where they definitely misleading.

And there are some that measure this too. They measure the quality of the hit, the placement, etc.

At that point, however, you are back to subjective and to non-quantitative analysis - or to trying to put a quantitative value on a non-quantifiable item.

Statistics have their place in baseball...I don't think anyone disagrees. I think the question is where that place is. And I think the answer is that each person is entitled to use/ignore them at their leisure.

wsgdf
12-16-2008, 06:07 PM
It's all about probability based on historical data. All FIP is saying is that Danks is more likely to repeat than Floyd... because pitchers like Danks tend to be better historically than pitchers like Floyd.

Obviously there are pitchers that don't fit that mold (Buehrle has confounded DIPS and FIP forever).

By the way - FIP doesn't in any way say Floyd sucks or sucked last year... just that he profiles as more of a league average type of guy... that's not a bad thing.

It would be interesting to see a first half and second half FIP, because I agree with some here who think it was much better in the second half. I don't think anyone ever suggested that a pitcher's FIP can't improve from year to year.

wsgdf
12-16-2008, 06:16 PM
Eddo:

Because some of you (and you know who you are both nationally and locally) act like these bizarre convolutions of the imagination are "the absolute truth" and need to be reminded that they aren't

No personal offense to you but honestly, will MLB cease to be or the world stop spinning if fans don't know what Paul Konerko hit in the 8th inning or later against left handed pitchers who wear their cap off center with last names of less then six letters on every other Friday night at home?

And before you say 'well that's an extreme example' that's my point. Today fans are getting hammered with extreme, bizarre, made up numbers from everybody and his brother, who considers themselves a baseball "expert."

I venture to say 90% of these "experts" never played the game beyond Little League.

Lip


That's not an extreme example it's just a really poor example. Statheads (sabermetric) are the last fans that would start bringing up 8th inning/left handed/blah blah blah... in fact... those are exactly the kind of stats that suck.

As far as who played the game and who hasn't - what a bunch of BS.

Lip Man 1
12-16-2008, 09:36 PM
WSQ:

If my example of particular stats "suck" then what do you call the convoluted stat that started this thread?

Talk about overkill, bizarre-ness etc. Yet this is typical of what we find from those you think the game, a human game can be reduced to cold, calculated equations, that are often more incorrect then correct.

Lip

Lukin13
12-16-2008, 09:57 PM
Floyd's ERA was 3.84 and his FIP was 4.77, very bad news.

Yep, that Gavin Floyd sure is untouchable, no chance in hell he will regress in '09... won't happen.

wsgdf
12-16-2008, 10:18 PM
WSQ:

Yet this is typical of what we find from those you think the game, a human game can be reduced to cold, calculated equations, that are often more incorrect then correct.

Lip


There are many ways to enjoy a game Lip and many ways to understand it... and to think that someone can't both enjoy the game and also try to understand it better by doing a little math... is making a pretty huge underestimation of what humans are capable of.

WhiteSoxFan84
12-16-2008, 11:59 PM
Here's the formula I use to tell me Gavin Floyd is a good pitcher:

Eye 1 + eye 2 + brain = Gavin Floyd rocks.

Here's the formula I use to tell me what to think of a stat:

numbers involved in figuring out stat * time stat creator spends playing D&D in basement / girls kissed by stat creator = how many better things there are to think about than useless mental masturbation.

Seriously, baseball is a fun game. Watch it once in a while.


It's a very fun game but it's also the game that numbers mean the most. A lot of coaches live by stats to make decisions about which pitcher to bring in, whether to pinch hit or not, etc.

I love the game. But I love being able to mess with numbers and working on the next new, useful stat. Unfortunately I'm not smart enough to come up with things like FIP :(:

I have come up with cool stats before though ... I just can't remember them...



WSQ:

If my example of particular stats "suck" then what do you call the convoluted stat that started this thread?

Talk about overkill, bizarre-ness etc. Yet this is typical of what we find from those you think the game, a human game can be reduced to cold, calculated equations, that are often more incorrect then correct.

Lip

Overkill? Bizarre-ness? No one is being forced to use that formula or even understand it. For those of us who don't fear numbers and like "breakthrough" measurements of performance, we love things like FIP and whatever else we find/come up with.

Some of the most successful sports bettors have either came up with their own system for selecting plays or follow a system they came across that involves complicated formulas. It can be VERY rewarding to be a "nerd" or a "D&D" freak lol

Lip Man 1
12-17-2008, 08:44 AM
Fan 84:

You hit the nail on the head when you used the word "betting."

That and the insane fantasy crap is what started all this garbage in the first place.

Daver would probably be able to say it more eloquently then I did but you get my point.

"Breakthrough" stat? That's funny. Just on this thread folks who do place some value in numbers have shown that it's arbitrary and perhaps menaingless because of it.

But feel free to "breakthrough" as much as your heart desires. As I said originally, just reading all this gobbledygook, makes my brain hurt.

Just play the damn game.

Lip

Craig Grebeck
12-17-2008, 08:56 AM
WSQ:

If my example of particular stats "suck" then what do you call the convoluted stat that started this thread?

Talk about overkill, bizarre-ness etc. Yet this is typical of what we find from those you think the game, a human game can be reduced to cold, calculated equations, that are often more incorrect then correct.

Lip
No one believes the game can be reduced to "cold, calculated equations." No one has ever said this, here or anywhere, to my knowledge. It's a way of studying the game, Lip, and if these sixth grade level equations are so mind-blowing, I suggest you go back to grammar school.

The stat is hardly convoluted, and it's often used in conjunction with mainstream stats like ERA. Now, with Gavin, it's important to note he surrendered an inordinate amount of unearned runs -- something that isn't necessarily excusable, despite the arbitrary definition of an earned run. That's something that could bite him in the ass next season and in the future, because it deflated his ERA a great deal.

Fan 84:

You hit the nail on the head when you used the word "betting."

That and the insane fantasy crap is what started all this garbage in the first place.

Daver would probably be able to say it more eloquently then I did but you get my point.

"Breakthrough" stat? That's funny. Just on this thread folks who do place some value in numbers have shown that it's arbitrary and perhaps menaingless because of it.

But feel free to "breakthrough" as much as your heart desires. As I said originally, just reading all this gobbledygook, makes my brain hurt.

Just play the damn game.

Lip
Everyone wants to play the game Lip. Everyone.

downstairs
12-17-2008, 09:13 AM
Oh goody. Another reason to not waste my time going to ballgames. Just feed me the numbers......

I'll never understand this interpretation of people who enjoy numbers.

I enjoy the numbers aspect of baseball. First of all, its analysis of things that HAPPEN IN THE BASEBALL GAME. So the game itself is still just as important to me as to you. I just look at it through more filters than you. I enjoy it as a game: the drama, randomness, all that. Like you. I also happen to enjoy it as a science. You don't. Who cares. Neither affects the other.

Secondly, everyone I've met that enjoys baseball uses numbers to express themselves. What- you don't get on a Paul Konerko because he has a .100 batting average through May? You don't consider Javy Lopez a bust because he couldn't win games (a number)?

So I mention VORP or FIP and I'm a stathead ruining the game, but you mention batting average and ERA and that's all good?

For once and for all: the game is both a dramatic, random art and a hard-numbers science.

downstairs
12-17-2008, 09:19 AM
Forte:

In fact it wasn't but the point is that this is baseball... something to enjoy. You shouldn't need to have to try to understand all these bizarre, formulas that some folks have come up with to try to "predict" what's going to happen, then when it completely and totally falls apart and they are dead wrong, you never see or hear "I (or we) were wrong...)

I'm looking at you egg-heads from Baseball Prospectus.

Lip

What's being misunderstood here is what "prediction" means. If I said last September that the Red Sox have a 60% chance to make the World Series and the Tampa Bay Rays have a 10% chance- I wasn't wrong. There is no wrong. They defied the odds. But its still nice (at least for me) to have their chances summed up in a simple number in September.

If I say Danks is a better pitcher than Floyd based off some complex formula- why is that different than whatever you use to rank pitchers?

No offense, I just think both styles of observing the game can co-exist just fine.

When I watch a game at a bar, I'm not thinking numbers. I'm into the game like any non-stathead. But on the side I think baseball is a great experiment for applying numbers to.

Frater Perdurabo
12-17-2008, 09:21 AM
Statistics are a pretty good means of analyzing what has happened in the past.

Taken by themselves, statistics are not a reliable predictor of what will happen in the future.

jabrch
12-17-2008, 09:27 AM
But on the side I think baseball is a great experiment for applying numbers to.

It would be - if there were only a small number of independent variables and you could, reliably, make some of the predictions people make.

Given that the game is a series of millions of independent variables, and that you can not reliably take them all out and roll this into single measures or metrics of good/better/best like some try to do.

FIP is ridiculous. The concept behind it is stupid, and the mathematcis behind it are not truly indicative of reality. 13? 2? 3? Those are selected numbers with too little science behind them. And anyhow, you can't take out FIELDING because there are pitchers who rely on their fielding. The nature of that formula will tell you that high K, low BB pitchers are great and that a guy who pitches to contact is less great.

For all those who complain that wins is a bad way to measure a pitcher (and I agree with the principle behind it - that wins alone tell you very little) FIP is equally weak because it fails to tell you a big piece of the picture.

jabrch
12-17-2008, 09:36 AM
Statistics are a pretty good means of analyzing what has happened in the past.

Taken by themselves, statistics are not a reliable predictor of what will happen in the future.


I'd preface what you say with, "in baseball". Because in the real world, statistics can be a tremendous predictor of what will happen. When I was in manufacturing, we used statistics to tell us exactly how many widgets we would produce in X time, with X$ and X people. We knew if we stopped spending on maintenence, that the # of widgets would go down, and the cost (waste) would go up. We had good models that predict how, over time, a machine would be less efficient/effective and what inputs we can change to make it more or less effective.

Here's the differnce - a machine in a manufacturing environment, where you have a highly automated process, has many fewer variables, almost all of which are controllable. In baseball, there are almost no controllable variables so all you have to do is draw a trend line and assume it will look similar to other trend lines - which is not a reliable prediction methodology.

wsgdf
12-17-2008, 09:58 AM
Statistics are a pretty good means of analyzing what has happened in the past.

Taken by themselves, statistics are not a reliable predictor of what will happen in the future.


Batter A played 162 games last year, batted 500 times and hit .330 with 50 homers.

Batter B played 162 games last year, batted 500 times and hit .210 with 1 homer.

Taken by themselves, could those statistics help you predict who's going to have a better year in 2009?

Lip Man 1
12-17-2008, 10:39 AM
WSG:

Not necessarily. What if batter B played with an illness (say mono or something) that sapped his strength just enough so that while he could go out on the field, he wasn't able to perform to his capabilities. If healthy he could have a better season than player A. What if player A was on steroids for a season?

Like Jab said ANYONE who tries to predict anything in baseball with any amount of "certainty" (ahhhhh I'm thinking of the good people at BP) are being extremely foolish (well that description certainly fits the doofus at BP)

Just play the damn game.

Lip

wsgdf
12-17-2008, 10:43 AM
WSG:

Not necessarily. What if batter B played with an illness (say mono or something) that sapped his strength just enough so that while he could go out on the field, he wasn't able to perform to his capabilities. If healthy he could have a better season than player A. What if player A was on steroids for a season?

Like Jab said ANYONE who tries to predict anything in baseball with any amount of "certainty" (ahhhhh I'm thinking of the good people at BP) are being extremely foolish (well that description certainly fits the doofus at BP)

Just play the damn game.

Lip

Of course all of that could happen. We're talking about likelihood.

Probability.

If there are 500 Player A's and 500 Player B's... is it more likely that a given Player A will outperform a given Player B?

As far as your repeated BP slams... you clearly haven't read much of their material if you think they claim certainty.

btrain929
12-17-2008, 10:47 AM
FIP is ridiculous. The concept behind it is stupid, and the mathematcis behind it are not truly indicative of reality. 13? 2? 3? Those are selected numbers with too little science behind them. And anyhow, you can't take out FIELDING because there are pitchers who rely on their fielding. The nature of that formula will tell you that high K, low BB pitchers are great and that a guy who pitches to contact is less great.

So is Javy Vazquez a God according to this formula?

jabrch
12-17-2008, 10:49 AM
Batter A played 162 games last year, batted 500 times and hit .330 with 50 homers.

Batter B played 162 games last year, batted 500 times and hit .210 with 1 homer.

Taken by themselves, could those statistics help you predict who's going to have a better year in 2009?

If all the comparisons of two players were this simple, the accuracy would be much more correct.

It isn't. Your "example" is unrealistic. Nobody would compare those two players and project B to be better. Find me anyone arguing about the quality of Alex Rodriguez vs Nick Swisher in 2008. Nobody is projecting Swish to outperform A-Rod next year. But that is not the level of detail we are looking at.

jabrch
12-17-2008, 10:56 AM
So is Javy Vazquez a God according to this formula?

The HRs likely don't help Javy. I don't have the FIP list in front of me, but from what I know, I imagine Halliday, Beckett and Lee come up very high. Daniel Cabrerra, Vicente Padilla and Gavin look worse. (HR and Walks are higher than Ks - relatively)

FIP has lots of holes. And if you ask the question, name pitcher who don't give up HRs and walks, but have lots of Ks, you could do that without the "newtonian math" or the creation of a formula that has constants (13, 3 and 2) that get added, subtracted, multiplied and divided to create some set of results that are not necesarily an arithmatic ranking.

wsgdf
12-17-2008, 11:26 AM
If all the comparisons of two players were this simple, the accuracy would be much more correct.

It isn't. Your "example" is unrealistic. Nobody would compare those two players and project B to be better. Find me anyone arguing about the quality of Alex Rodriguez vs Nick Swisher in 2008. Nobody is projecting Swish to outperform A-Rod next year. But that is not the level of detail we are looking at.


The reason that 'nobody is projecting Swish to outperform A-Rod next year' is exactly because of their statistics.

As the players being compared become closer in quality, it becomes trickier to pick who will outperform who... which is why people look for better evaluation methods.

OBVIOUSLY, no evaluation method is perfect - which seems to be the major criticism here - "if you can't predict every player's 2009 numbers perfectly then your numbers are garbage and you hate the game".

jabrch
12-17-2008, 11:48 AM
The reason that 'nobody is projecting Swish to outperform A-Rod next year' is exactly because of their statistics.

Nobody needs to look at ANY statisics to do that - in particular, nobody needs to look at anything other than acutal measurements of results. It takes no sophistication, projections, complications or any other detail not available in a box score, not actually COUNTED, to do that.

This other crap is unnecesary if that's what you need it to prove.

As the players being compared become closer in quality, it becomes trickier to pick who will outperform who... which is why people look for better evaluation methods.

Let me know when they find one. This is not any better when it gets closer because you start nitpicking. And you can't nitpick when there are so many variables that you can not control. So this **** is less helpful as we get closer - and it wasn't overly helpful when we were not close at all.

OBVIOUSLY, no evaluation method is perfect - which seems to be the major criticism here

No - that's not the major criticism. It is that something like this is nearly useless because it fails to fairly compare different types of players and fails to accomdate for the old addage that there is more than one way to skin a cat.

Penalizing a pitcher who is skilled at getting outs in a way other than how this formula gives credit is...well...silly.

- "if you can't predict every player's 2009 numbers perfectly then your numbers are garbage and you hate the game".

Bull**** again. that's what weak minded people say to respond to criticism of their precious formulae. It's not about predicting near perfectly. It's about not predicting what is nearly unpredictable. Too many variables you can't control. So prediction (which FIP doesn't do anyhow) is silly. You can straight line a prediction, do an average, or divide numbers by the ratio of a players ass to his armpits and it still doesn't matter because....drumroll please...in baseball - tomorrow is virtually independent of today - and moreso of yesterday. And don't even start comparing tomorrow to 3 days ago...

wsgdf
12-17-2008, 12:47 PM
Nobody needs to look at ANY statisics to do that - in particular, nobody needs to look at anything other than acutal measurements of results.

Statistics are the measurements of results.


Let me know when they find one.

Why should I bother? MLB teams believe they've found some. Use Google.


No - that's not the major criticism. It is that something like this is nearly useless because it fails to fairly compare different types of players and fails to accomdate for the old addage that there is more than one way to skin a cat.

I'm not trying to be the great defender of FIP... I just don't get the knee-jerk blasting of anything that involves math. FIP just seems to be just a restatement of what Voros said a long time ago. You had it right in an earlier post - give up less homers, less walks and K more guys and you're probably a better pitcher. PROBABLY. In the history of baseball - guys who have given up fewer homers, fewer walks and struck out more hitters are generally better than guys who haven't - so looking forward it's safe to say that baseball will continue to work that way.


Penalizing a pitcher who is skilled at getting outs in a way other than how this formula gives credit is...well...silly.


A pitcher can't be skilled at having a CF who can run down balls in the gap better than Rob Macowiack. A pitcher can't be skilled at having a 3B who doesn't let anything get down the line. Rob Mackowiack absolutely effected the ERAs of the White Sox pitching staff in 2006.


Bull**** again. that's what weak minded people say to respond to criticism of their precious formulae.


It's not my formula and it's certainly not precious. Go ahead and name call, kick and scream, type harder and harder... whatever.

Lip said that people who enjoy stats have never played baseball past Little League and can't just sit back and enjoy 'the game'. That's what I was referring to. I played high school and college ball. Billy Beane (and I'm not a Beane fanatic so don't start) played pro-ball.


in baseball - tomorrow is virtually independent of today - and moreso of yesterday. And don't even start comparing tomorrow to 3 days ago...


Right - so we're back to Player A and Player B. You don't know if ARod is better than Swisher.

jabrch
12-17-2008, 01:40 PM
Statistics are the measurements of results.

That depends. Some are not. FIP measures no countable result. It only mesures the output if its own formula. You can't count your FIP. You convolute it.


I'm not trying to be the great defender of FIP... I just don't get the knee-jerk blasting of anything that involves math.

That's not what you are seeing here. You and your ilk have this wild fantasy that it is just math-haters. It isn't. It's the fact that, in this particular case, the number in question is bull****. Don't make this out to be a persecution of the science of mathematics. It isn't.


FIP just seems to be just a restatement of what Voros said a long time ago. You had it right in an earlier post - give up less homers, less walks and K more guys and you're probably a better pitcher.

I didn't say that. I said that was what FIP said. There are more than one way to skin a cat and trying to come up with one way to do it is Vorostupid.

A pitcher can't be skilled at having a CF who can run down balls in the gap better than Rob Macowiack. A pitcher can't be skilled at having a 3B who doesn't let anything get down the line. Rob Mackowiack absolutely effected the ERAs of the White Sox pitching staff in 2006.

And FIP does nothing to factor this in. You can't make FIP into more than it is.


Lip said that people who enjoy stats have never played baseball past Little League and can't just sit back and enjoy 'the game'.


I'm not in the business of defending Lip. But that is NOT what he said. Go back and re-read it.


That's what I was referring to. I played high school and college ball. Billy Beane (and I'm not a Beane fanatic so don't start) played pro-ball.

And again - nobody said that 100% of all people citing stats have never worn a jockstrap. Lip said that many didn't - and that's true. You may be an exception. Billy Beane clearly is. But again - that's not the point either that Lip is making, or that I am.

The point is that FIP is not a measure of what it professes to measure. It doesn't tell you what a Pitcher will do without defense. It tells you what the result of http://upload.wikimedia.org/math/6/4/c/64cce9b9812b7fbc1dabe63152818531.png . Those numbers nowhere quantify the value of Rob Mackowiak - nowhere...


Right - so we're back to Player A and Player B. You don't know if ARod is better than Swisher.

Wrong

I do. I am 100% sure. If you aren't, there's an issue. But I am not sure because of any calculation you make. I don't need that to be 100% sure.

mikedaspike
12-17-2008, 02:30 PM
Irregardless of what the numbers say, Gavin just seems like a winner. True, this was his first full major league season, and he could just as easily fall off the face of the earth, but Gavin could be the type of guy who posts a decent ERA with around 12-14/15 wins per year. Hopefully I'm not being too optimistic...

doublem23
12-17-2008, 02:57 PM
The point is that FIP is not a measure of what it professes to measure. It doesn't tell you what a Pitcher will do without defense. It tells you what the result of http://upload.wikimedia.org/math/6/4/c/64cce9b9812b7fbc1dabe63152818531.png . Those numbers nowhere quantify the value of Rob Mackowiak - nowhere...


That's not what FIP is supposed to measure. Not saying that I like DIPS, but the stats are supposed to equalize ERA based on elements a pitcher can control. It's not designed to tell you whether or not a pitcher will succeed with a crappy CF playing behind him.

Please guys, if you're going to rip on these statistics, at least understand what they mean first (it's still tremendously easy to do after that).

jabrch
12-17-2008, 03:03 PM
That's not what FIP is supposed to measure. Not saying that I like DIPS, but the stats are supposed to equalize ERA based on elements a pitcher can control. It's not designed to tell you whether or not a pitcher will succeed with a crappy CF playing behind him.

Please guys, if you're going to rip on these statistics, at least understand what they mean first (it's still tremendously easy to do after that).

I was repsonding to the specific claim of the poster that it does measure how skillful a pitcher is in having Mackowiak play CF.

I'm fully aware of what it means.

And it means very little because it only takes 3 of the many components that make a pitcher effective, and weights them using certain coefficients.

wsgdf
12-17-2008, 03:16 PM
I was repsonding to the specific claim of the poster that it does measure how skillful a pitcher is in having Mackowiak play CF.



No. My claim was exactly the opposite of that.

jabrch
12-17-2008, 03:20 PM
No. My claim was exactly the opposite of that.


A pitcher can't be skilled at having a CF who can run down balls in the gap better than Rob Macowiack. A pitcher can't be skilled at having a 3B who doesn't let anything get down the line. Rob Mackowiack absolutely effected the ERAs of the White Sox pitching staff in 2006.

You are saying that defense has nothing to do with a pitcher's skill, and FIP measures that. We all know the first half is true. But I completely disagree that FIP measures that. Show me where in the FIP it takes into account Rob Mackowiak. Or any other defender for that matter.

It doesn't.

FIP assigns positive and negative value to 3 measureables. That's not an effective way to accomplish exactly what you claim it does.

wsgdf
12-17-2008, 03:40 PM
You are saying that defense has nothing to do with a pitcher's skill, and FIP measures that. We all know the first half is true. But I completely disagree that FIP measures that. Show me where in the FIP it takes into account Rob Mackowiak. Or any other defender for that matter.

It doesn't.

FIP assigns positive and negative value to 3 measureables. That's not an effective way to accomplish exactly what you claim it does.


LOL - Why do you keep saying that? My point is that FIP is taking Mackowiack OUT of the equation - where ERA leaves him in.

jabrch
12-17-2008, 04:01 PM
LOL - Why do you keep saying that? My point is that FIP is taking Mackowiack OUT of the equation - where ERA leaves him in.

It does not do what you say it does. It doesn't take take Mack out of the equation. It creates a totally different equation that doesn't measure ANYTHING except Ks, BBs, HRs and it assigns a value to them. That doesn't measure how effective a pitcher is. Nowhere does it give credit for the littany of things a pitcher can do that are effective.

wsgdf
12-17-2008, 05:46 PM
It does not do what you say it does. It doesn't take take Mack out of the equation. It creates a totally different equation that doesn't measure ANYTHING except Ks, BBs, HRs and it assigns a value to them.

Creating a different equation that doesn't measure ANYTHING except Ks, BBs, and HRs is the whole point and yes, doing that takes Mack out of the equation.

Again - I'm really not here to defend FIP, I've probably seen the equation twice in my life... what bugs me is that people are so quick to blow something off that they've never seen before, not try to understand it, and then start calling people names, accusing them of never playing baseball or not enjoying the game as much as themselves.

That's all a big load of crap and a sign of massive amounts of intellectual laziness.

jabrch
12-17-2008, 09:01 PM
That's all a big load of crap and a sign of massive amounts of intellectual laziness.

And it all ends up where it usually does...with the stathead professing their intellectual superiority...thanks pal..gnight

Craig Grebeck
12-17-2008, 10:34 PM
And it all ends up where it usually does...with the stathead professing their intellectual superiority...thanks pal..gnight
Jabrch,

You have been so horribly off base in this thread. Anyone with an ounce of brainpower that uses FIP would do so with the caveat that, at times, a pitcher's skill-set (first and foremost the ability to induce ground balls) can make FIP less meaningful. But, in looking at a guy like Gavin, who had a breakout season that didn't match his previous campaigns, it's important to look at the things he can control (BB/K/HR) and see if he did a good job of suppressing the negatives.

The point of this thread, and how FIP applies to Gavin in general, I guess, is that if he doesn't improve in these areas it is unlikely that he will replicate his 2008 numbers. Now, his second half showed progression, which makes me hopeful.

Daver
12-17-2008, 10:58 PM
Jabrch,

You have been so horribly off base in this thread. Anyone with an ounce of brainpower that uses FIP would do so with the caveat that, at times, a pitcher's skill-set (first and foremost the ability to induce ground balls) can make FIP less meaningful. But, in looking at a guy like Gavin, who had a breakout season that didn't match his previous campaigns, it's important to look at the things he can control (BB/K/HR) and see if he did a good job of suppressing the negatives.

The point of this thread, and how FIP applies to Gavin in general, I guess, is that if he doesn't improve in these areas it is unlikely that he will replicate his 2008 numbers. Now, his second half showed progression, which makes me hopeful.

When a stat is measuring anything that is based on a variable it becomes worthless, you may as well measure the intensity of moonbeams. It becomes the dividing line between clear statisical analysis and pure mental masturbation. It is the reason that range factor is a worthless stat, as is the hold and the save.

Craig Grebeck
12-17-2008, 11:11 PM
When a stat is measuring anything that is based on a variable it becomes worthless, you may as well measure the intensity of moonbeams. It becomes the dividing line between clear statisical analysis and pure mental masturbation. It is the reason that range factor is a worthless stat, as is the hold and the save.
Your statement doesn't really make sense. FIP is just magnifying three of the most important aspects of a pitcher's job (hold down HR and BB while putting up Ks).

wsgdf
12-17-2008, 11:21 PM
And it all ends up where it usually does...with the stathead professing their intellectual superiority...thanks pal..gnight

Hate to have to do this, but for the third post in a row... no, that's not what I said... I said lazy.

After two pages of "weak-minded", "those of your ilk", and defending Lip's claims that statheads have never played baseball and don't appreciate the game, it's interesting how quick you are to be offended.

jabrch
12-18-2008, 09:24 AM
When a stat is measuring anything that is based on a variable it becomes worthless, you may as well measure the intensity of moonbeams. It becomes the dividing line between clear statisical analysis and pure mental masturbation. It is the reason that range factor is a worthless stat, as is the hold and the save.


Get outs. Don't give up runs. It's a lot more simple than most want to make it. Anything else slants analysis one direction or another based on what constants/variables you use, and what inputs you use or discard.

wsgdf
12-18-2008, 10:55 AM
Get outs. Don't give up runs.

How you get outs matters.

If one pitcher get 27 outs in a row on Ks and another pitcher gets 27 outs in a row on Aaron Rowand face plants, I'll take my chances on the guy with the 27 Ks, even if it looks like the other guy throws the perfect Aaron Rowand face plant pitch.

Lip Man 1
12-18-2008, 11:15 AM
Jab:

Your last two posts have been well said.

The comments in Bull Durham say it best "you hit the ball, you throw the ball, you catch the ball...."

See? It's that simple folks! No advanced degree in physics, quantum mechanics or astrophysics required.

Lip

voodoochile
12-18-2008, 11:21 AM
How you get outs matters.

If one pitcher get 27 outs in a row on Ks and another pitcher gets 27 outs in a row on Aaron Rowand face plants, I'll take my chances on the guy with the 27 Ks, even if it looks like the other guy throws the perfect Aaron Rowand face plant pitch.

Yeah, except for that part about wear and tear on the arm. The guy with the K's threw at least 81 pitches. The guy with the ARFPP may have only thrown 27. Over the years they both might have success but the guy with the ARFPP will probably have a longer career.

For example: WHo would you rather have on your team? 1998 Kerry Wood or 2000 Mark Buehrle?

Marqhead
12-18-2008, 11:22 AM
Yeah, except for that part about wear and tear on the arm. The guy with the K's threw at least 81 pitches. The guy with the ARFPP may have only thrown 27. Over the years they both might have success but the guy with the ARFPP will probably have a longer career.

Motion to declare ARFPP an official WSI acronym.

All in favor, say I.

jabrch
12-18-2008, 11:23 AM
How you get outs matters.

If one pitcher get 27 outs in a row on Ks and another pitcher gets 27 outs in a row on Aaron Rowand face plants, I'll take my chances on the guy with the 27 Ks, even if it looks like the other guy throws the perfect Aaron Rowand face plant pitch.

If you want to live in the world of fantasy, where it is so binary that all outs are this or that, I'll take the guy who gets 27 first pitch popups over the guy who gets 27 12 pitch strikeouts, and assume my guy is more able to go every 3rd day and make 45 starts since he is thowing 27 pitchers per game over your guy who just took 324 pitches and likely has shredded his elbow.

Now my fantasy scenario is as silly, theoretical, and unrealistic as yours.

Now lets talk about REALITY. An occasional great play will happen even with the worst defender out there. But if a guy shows the ability to consistently get outs, how he gets them doesn't matter. Unless you prove to me that there is this guy out there, who consistently has Rowand Faceplants that change the face of his numbers, I'll assume that +/- an acceptable margin of error (in a game that already has millions of variables that are not being accounted for in your 3 variable forumula), that a guy who consistently get enough outs to not give up a lot of runs will continue to do that even if it isn't "by striking out 27 guys".

Get out of hyperbole and unrealistic scenarios of makebelieve fantasy. All that matters is that you get your outs, or that you score runs. There is not style points involved. If I wanted style points, I'd watch gymnastics.

wsgdf
12-18-2008, 11:32 AM
motion to declare arfpp an official wsi acronym.

All in favor, say i.

i

FedEx227
12-18-2008, 12:18 PM
Really though, is that the intent? No. The numbers are simply used to analyze a pitcher's ability to control things independent of the defense.

On Floyd, I'd argue that his FIP in the second half was much, much better, IIRC. He was not very good early on, but his core numbers started to right themselves in the second half.

And FIP has been around since at least 1999-2000, so it's not like this is a new NERD tool because clearly if you look at numbers you never actually watch real baseball, you couldn't possibly do both :o::o::o:

doublem23
12-18-2008, 12:50 PM
If you want to live in the world of fantasy, where it is so binary that all outs are this or that, I'll take the guy who gets 27 first pitch popups over the guy who gets 27 12 pitch strikeouts, and assume my guy is more able to go every 3rd day and make 45 starts since he is thowing 27 pitchers per game over your guy who just took 324 pitches and likely has shredded his elbow.

Now my fantasy scenario is as silly, theoretical, and unrealistic as yours.

Now lets talk about REALITY. An occasional great play will happen even with the worst defender out there. But if a guy shows the ability to consistently get outs, how he gets them doesn't matter. Unless you prove to me that there is this guy out there, who consistently has Rowand Faceplants that change the face of his numbers, I'll assume that +/- an acceptable margin of error (in a game that already has millions of variables that are not being accounted for in your 3 variable forumula), that a guy who consistently get enough outs to not give up a lot of runs will continue to do that even if it isn't "by striking out 27 guys".

Get out of hyperbole and unrealistic scenarios of makebelieve fantasy. All that matters is that you get your outs, or that you score runs. There is not style points involved. If I wanted style points, I'd watch gymnastics.

Guys that gets outs by strikeout are usually more reliable than guys that get outs in the field. Not a revolutionary concept.

doublem23
12-18-2008, 12:51 PM
Jab:

Your last two posts have been well said.

The comments in Bull Durham say it best "you hit the ball, you throw the ball, you catch the ball...."

See? It's that simple folks! No advanced degree in physics, quantum mechanics or astrophysics required.

Lip

<head slap>

And I don't even like sabermetrics that much.

ode to veeck
12-18-2008, 01:28 PM
For example: WHo would you rather have on your team? 1998 Kerry Wood or 2000 Mark Buehrle?

whichever one doesn't have Dusty managing his younger years as a starter

ode to veeck
12-18-2008, 01:29 PM
Guys that gets outs by strikeout are usually more reliable than guys that get outs in the field. Not a revolutionary concept.

I like the ground ball pitchers myself

ode to veeck
12-18-2008, 01:30 PM
Please guys, if you're going to rip on these statistics, at least understand what they mean first (it's still tremendously easy to do after that).

you mean we can't just out and out hate sabremetrics without having to understand what the **** they're supposed to be?!

jabrch
12-18-2008, 01:30 PM
Guys that gets outs by strikeout are usually more reliable than guys that get outs in the field. Not a revolutionary concept.

Define reliable - because I am not so sure I agree with you.

Javy is an example... I'll take Mark Buehrle over Javy - quite easily.

ode to veeck
12-18-2008, 01:32 PM
When a stat is measuring anything that is based on a variable it becomes worthless, you may as well measure the intensity of moonbeams. It becomes the dividing line between clear statisical analysis and pure mental masturbation. It is the reason that range factor is a worthless stat, as is the hold and the save.

same can be said for errors the way they are scored today

ode to veeck
12-18-2008, 01:39 PM
funniest thread I've read in months

jabrch
12-18-2008, 01:41 PM
you mean we can't just out and out hate sabremetrics without having to understand what the **** they're supposed to be?!

We can - that's our choice.

But the truth of the matter is that you can't make the assumption that those who think FIP is horsecrap don't understand what FIP is. Manipulating Ks, BBs and HRs by some static factors is horsecrap. It doesn't measure what it professes to measure.


same can be said for errors the way they are scored today

Who is professing errors are a great way to evaluate defense?

wsgdf
12-18-2008, 01:44 PM
Define reliable - because I am not so sure I agree with you.

Javy is an example... I'll take Mark Buehrle over Javy - quite easily.

Javy and Buerhle are perfect examples for your point, absolutely.

I think it can be shown that they're both exceptions to the rule.

Can Floyd be another exception? Sure. It's just more likely that he's not.

jabrch
12-18-2008, 01:56 PM
I think it can be shown that they're both exceptions to the rule.

Can Floyd be another exception? Sure. It's just more likely that he's not.


Or maybe your rule is wrong?

http://upload.wikimedia.org/math/6/4/c/64cce9b9812b7fbc1dabe63152818531.png tries to boil pitching to 3 variables and 3 constants. That's an intellectually lazy way to look at pitching, which has millions of variables and unmeasurable constants.

wsgdf
12-18-2008, 02:25 PM
Or maybe your rule is wrong?

http://upload.wikimedia.org/math/6/4/c/64cce9b9812b7fbc1dabe63152818531.png tries to boil pitching to 3 variables and 3 constants. That's an intellectually lazy way to look at pitching, which has millions of variables and unmeasurable constants.


Or maybe it's not right or wrong. Maybe it's just better than ERA at predicting the next year's ERA. That's the only claim.

Craig Grebeck
12-18-2008, 06:41 PM
Or maybe your rule is wrong?

http://upload.wikimedia.org/math/6/4/c/64cce9b9812b7fbc1dabe63152818531.png tries to boil pitching to 3 variables and 3 constants. That's an intellectually lazy way to look at pitching, which has millions of variables and unmeasurable constants.
Dude, seriously, millions? C'mon. You know as well as I do that throughout the history of the game, pitchers who excel at limiting HR and BB while racking up decent to good K totals will more often than not be successful.

ode to veeck
12-19-2008, 10:01 AM
We can - that's our choice.

But the truth of the matter is that you can't make the assumption that those who think FIP is horsecrap don't understand what FIP is. Manipulating Ks, BBs and HRs by some static factors is horsecrap. It doesn't measure what it professes to measure.



I'll admit I think FIP is horsecrap and that I don't even want to ****ing understand what FIP is

spiffie
12-20-2008, 01:09 AM
I'll admit I think FIP is horsecrap and that I don't even want to ****ing understand what FIP is
You'll never win your fantasy league with an attitude like that.

Here's my prediction. Good pitchers with smarts like Danks, Floyd and Buehrle will most likely do well next year. Pitchers with lots of K's and no brains like Javy will not do as well.

spiffie
12-20-2008, 01:37 AM
Or maybe it's not right or wrong. Maybe it's just better than ERA at predicting the next year's ERA. That's the only claim.
Just running a quick check of our 2007 starters, and what the difference between their 2007 ERA and 2008 ERA vs. the difference between 2007 FIP and 2008 ERA. I used a factor of 3.2 for both.

Jon Garland:
ERA vs ERA: 0.67 FIP vs. ERA: 1.49
Javier Vazquez:
ERA vs. ERA: 0.93 FIP vs. ERA: 3.18
Gavin Floyd:
ERA vs. ERA: -1.43 FIP vs. ERA: -4.39
John Danks:
ERA vs. ERA: -2.18 FIP vs. ERA: 3.77
Mark Buehrle:
ERA vs. ERA: 0.16 FIP vs. ERA: 0.75

So for all the 2007 starters from the Sox their 2007 ERA was a much better predictor of their 2008 ERA than their FIP was.

Konerko05
12-20-2008, 02:12 AM
Just running a quick check of our 2007 starters, and what the difference between their 2007 ERA and 2008 ERA vs. the difference between 2007 FIP and 2008 ERA. I used a factor of 3.2 for both.

Jon Garland:
ERA vs ERA: 0.67 FIP vs. ERA: 1.49
Javier Vazquez:
ERA vs. ERA: 0.93 FIP vs. ERA: 3.18
Gavin Floyd:
ERA vs. ERA: -1.43 FIP vs. ERA: -4.39
John Danks:
ERA vs. ERA: -2.18 FIP vs. ERA: 3.77
Mark Buehrle:
ERA vs. ERA: 0.16 FIP vs. ERA: 0.75

So for all the 2007 starters from the Sox their 2007 ERA was a much better predictor of their 2008 ERA than their FIP was.

I'm kind of confused about what you are doing here. FIP and ERA are two completely different statistics.

wsgdf
12-20-2008, 01:42 PM
I'm kind of confused about what you are doing here. FIP and ERA are two completely different statistics.

He's adding 3.2 to the FIP calculation to project next season's ERA.

Thanks for doing that spiffie. It's an extremely small sample size obviously. I believe the FIP vs ERA test has been run for every starting pitcher since 1994 and FIP came ahead significantly.

Also... I ran them myself quickly to double check.

I came up with:

Danks: 2007 FIP 5.42
2007 ERA 5.50
2008 ERA 3.32

FIP would win that one.

Garland: 2007 FIP 4.27
2007 ERA 4.23
2008 ERA 4.90

FIP would win that one.

Vazquez 2007 FIP 3.67
2007 ERA 3.74
2008 ERA 4.67

ERA would win that one .

Buehrle 2007 FIP 4.15
2007 ERA 3.63
2008 ERA 3.79

ERA would win that one.

Floyd only pitched 70 innings in 2007, so I won't check that one... it's really supposed to be used for starters... even Danks' 139 IP are on the low side.

Maybe my math is wrong (like I said I did them quickly) - but it looks like 2-2 to me, and I already mentioned that Vazquez and Buehrle are exceptions, so that was really expected.

voodoochile
12-20-2008, 02:50 PM
Maybe my math is wrong (like I said I did them quickly) - but it looks like 2-2 to me, and I already mentioned that Vazquez and Buehrle are exceptions, so that was really expected.

What percentage of the pitching population qualifies as "exceptions"? I'd definitely want to see that answered before I'd accept anything that FIP predicts/quantifies.

So far 50% of the pitchers used as examples are "exceptions". Yes, it's a small sample, but it's the only sample you've shown.

Any stat that fails 50% of the time as a predictive tool isn't much of a predictive tool.

Reminds me of the group who claims they were able to cut violent crime in Philly by thinking positive thoughts as a group for 6 weeks. They actually managed to do it after they threw out the deadliest weekend of the 6 weeks (27 killed) because it was clearly an "outlier" (read: exception). Add the 27 murders back in and the rate actually went up for those 6 weeks (so much for happy thoughts).

What percentage of the pitching population is this an accurate tool for? Are Buehrle and Vazquez the only exceptions? Don't manipulate the data to make it say what you want it to say, present the data and see how it stands up to peer review...

jabrch
12-20-2008, 03:04 PM
:cleo

I see the future....

http://imgsrv.670thescore.com/image/wscr/UserFiles/Image/Time_Feature_On_Air/Email-Murph.jpg


"I predict...."

http://upload.wikimedia.org/math/6/4/c/64cce9b9812b7fbc1dabe63152818531.png
"If you disregard the millions of thing that happen to impact a baseball game, and stick with my 3, and my choice of how to maniupulate them, then I can predict..."

Frater Perdurabo
12-20-2008, 06:05 PM
:tomatoaward:

wsgdf
12-21-2008, 12:25 AM
What percentage of the pitching population qualifies as "exceptions"? I'd definitely want to see that answered before I'd accept anything that FIP predicts/quantifies.


Javy Vazquez is bizarre.

Mark Buehrle is less bizarre, but still rare.


So far 50% of the pitchers used as examples are "exceptions". Yes, it's a small sample, but it's the only sample you've shown.

It wasn't my sample.


Any stat that fails 50% of the time as a predictive tool isn't much of a predictive tool.

It doesn't.


Reminds me of the group who claims they were able to cut violent crime in Philly by thinking positive thoughts as a group for 6 weeks. They actually managed to do it after they threw out the deadliest weekend of the 6 weeks (27 killed) because it was clearly an "outlier" (read: exception). Add the 27 murders back in and the rate actually went up for those 6 weeks (so much for happy thoughts).

It's not at all like that. Like I said, they ran this against every starting pitcher since 1994. If you're actually interested enough, it's pretty easy to find. I'd do it but it's been a long fun night and I'm going to bed.


What percentage of the pitching population is this an accurate tool for? Are Buehrle and Vazquez the only exceptions? Don't manipulate the data to make it say what you want it to say, present the data and see how it stands up to peer review...

It works for starting pitchers in general. I didn't manipulate anything, what are you talking about? I don't have any data to present. I'm NOT THE FIP GUY. Seriously, do a Google search on Defense Indepentent Pitching Stats. There are plenty of articles out there for peers to review. Bill James has his own version, Tango Tiger (it's hard to take Tango seriously because he doesn't use his real name - and Tango Tiger sounds stupid), Voros - each one of them has a slightly diferent version.

voodoochile
12-21-2008, 10:42 AM
It works for starting pitchers in general. I didn't manipulate anything, what are you talking about? I don't have any data to present. I'm NOT THE FIP GUY. Seriously, do a Google search on Defense Indepentent Pitching Stats. There are plenty of articles out there for peers to review. Bill James has his own version, Tango Tiger (it's hard to take Tango seriously because he doesn't use his real name - and Tango Tiger sounds stupid), Voros - each one of them has a slightly diferent version.

Sorry, I wasn't attempting to single you out. I was more using the generic "you". Just wanted to clear that up. I'll do a search some day, but right now it simply isn't that big of an issue.

EndemicSox
12-21-2008, 02:00 PM
Forte:

In fact it wasn't but the point is that this is baseball... something to enjoy. You shouldn't need to have to try to understand all these bizarre, formulas that some folks have come up with to try to "predict" what's going to happen, then when it completely and totally falls apart and they are dead wrong, you never see or hear "I (or we) were wrong...)

I'm looking at you egg-heads from Baseball Prospectus.

Lip

No, the average fan shouldn't bother with these formulas...if you work for an MLB team, you had better know them inside and out.