Greinke: someone talk me off the ledge

[gregmag]

I've intentionally avoided rehashing 4-5 years worth of debates regarding certain metrics with the usual suspects in this thread but I do have a couple of comments.

 

I like stuff as much as anyone, and who has better raw stuff in MLB than Edwin Jackson? The problem is he's never been able to locate consistently, there are plenty of pitchers who never reach their full potential. We have a LHP on our DL that can hit 95 MPH... you can count those guys on one hand, and he didn't have success. There's much more to baseball than pure statistical analysis.

 

I love to talk about peripherals, in fact on the minor league forum you'll see me posting about raw stuff and peripherals all the time, but at some point the actual results have to count for something.

 

If you look at Greinke's career as a whole, what year is the outlier?

 

Greinke's fantastic K rate this season has relatively little to do with his FB location, he's getting those Ks off his secondary pitches. The poorly located FBs and the occasional hanging slider are the pitches that are getting rocked.

 

You will find an occassional pitcher's pitch that is hit for a HR, or some lucky garbage like golfing one out off of one knee, but those are events that happen a couple of times a season, they are extremely rare. How many hitters do the Brewer's have that can hit a pitcher's pitch consistently? I'll submit there's 1... Braun. Morgan has slapped a couple of hit of hits on them as well, but I haven't seen enough to make a judgement about him yet. Everyone else on the team is a mistake/fastball hitter.

 

If people refuse to believe that pitching is all about location (which includes pitch selection), then there really isn't anything to discuss ever. The game isn't played in metrics, there is no "cause" in any metric, they only measure results. In baseball the power lies with the pitchers and by extension the cause mostly rests in their ability to execute a given pitch.

 

If less than 50% of HRs can be classified, then GB, LD, and FB% are all pretty meaningless aren't they? If we can't classify what happens on the best possible outcome for a hitter, how can we classify the rest? Wait that's right, we don't care about HRs because BABIP doesn't include them in the calculation.

 

We've reached a stagnant place where a poster can't say a player has done X over Y number of games without the same tired stastical analysis regarding sample size, sometimes being horrible or good over a small stretch is the point of the discussion. Something causes those stretches other than luck, I'd love to see many of you coach... don't worry about it son, your BABIP is just low, you'll be fine, the hitch in your switch has nothing to do with it.

First of all, I'm not sure all of your complaints are really about "statistical analysis." Peripherals and BABIP, yes; stuff, I thought, was a scouting concept in the great divide. Of course, the divide is converging, but that fact just makes your objections more confusing.

 

What are you really saying? I don't take you to argue that statistical analysis is without value (or that scouting analysis of stuff, etc. is without value). You're saying that these things are overused. I have two problems with that argument. First, it's imprecise. I don't think anybody would disagree that "results" should figure heavily in assessing players. In fact, "results" are what statistics represent. So are you saying something specific about how and when we should place more emphasis on what you're calling "results," or are you just unhappy that people agree with your somewhat negative assessment of Greinke? Second, your last paragraph just strikes me as myopic. Do you really think both sides of this argument don't get aired? What you're saying is no more or less "tired" than the responses you're complaining about.

[rluzinski]

Some only care about results, some attempt to infer the cause of good or bad performance, so attempt to estimate future performance.

 

This is where people tend to talk past each other. I think there is some confusion as to which each metric is doing. One thing I tend to see with Greinke's K rate is it gets used as a stat that tells us he was doing fine when in fact it is more of a predictive stat to tell us he will be fine in the future. K rates tend to get used as the definition of doing well. That to me is the wrong way to use it. The only reason it was ever determined to be a good indicator of performance was because players with good K rates had good results. It wasn't the k rates that was inherently good but the indicator of such.

I agree that some people are attempting to answer different questions. There are two general questions I see here:

1. How has Greinke's performance (not results) been in 2011?

2. How much should that performance cause us to adjust our expectations for him going forward?

I disagree with the usefulness of using K rate. First of all, if you want to look at K rate, the denominator should be batters faced, not IP. Using IP inflates the K rate of a pitcher who is allowing a lot of base runners (since he faces more overall batters as a result). Not sure who even has K/BF though.

K rate is SO important because:

1. A strikeout is an out 99.9% of the time.

2. Strikeout rates for a pitcher stabilizes quicker than other batting-result stats. What I mean by that is K rate is more representative of a pitcher's true skill with a smaller sample than some other stats (HR rate and BABIP for instance). If we are trying to estimate pitcher performance based on batting-results only and we only have 60 IP, I might only look at strikeouts and walks. That's not to say that the pitcher has no effect on anything else; it's just that with such a small sample, statistical noise, defense, etc... drowns out that "pitcher performance" part of it.

Now, if we want to start looking at pitch-by-pitch data, that's entirely different. I have no confidence than anyone can just watch the games and tell me if a pitcher's command has been good over 60 IP, 250 batters faced and over 1,000 pitches. Any opinion they have will be heavily biased by the pitcher's aggregate results, IMO. But as Crew pointed out, just looking at spray charts isn't the whole picture either. Pitch f/x data is the way to go but I don't think we are there yet in terms of knowing how to exactly evaluate that data to determine pitcher performance.... well at least nothing publicly available at least.

 

[1992casey]

I disagree with the usefulness of using K rate. First of all, if you want to look at K rate, the denominator should be batters faced, not IP. Using IP inflates the K rate of a pitcher who is allowing a lot of base runners (since he faces more overall batters as a result). Not sure who even has K/BF though.

Yeah, I mentioned that above.

 

This is getting really technical, but buried somewhere in the Statistical Analysis forum, one of our stat-oriented members said that K/AB is likely to be a better measure than K/BF (or K/PA). I tried searching for that post, but I couldn't find it. I'd guess that the difference between the two would be really minimal, though, and that either one would be definitely be better than K/9 innings.

[Thurston Fluff]

I disagree with the usefulness of using K rate.

 

Do you disagree with using K rate to forecast the future or that it is incorrect to use it to say the player's results have been good? I agree K rates/ab tell us what the success rate should be going forward. The correlation has been well established. I disagree that it tells us he's had good results. I don't know how to explain it in mathematical terms but I can in a logic format.

 

If k rates are indicators of a pitchers ability to be successful then that ability has to be defined by another standard than simply K rates. Otherwise you would be saying: If K then K. While that is wonderful example of a-priori knowledge it is also useless in this discussion.

 

What I think was meant is: If K then S. With Success being a set standard of measurement other than K itself. That set standard of results is what determines if he was successful to that point not the K rate. In other words; If S then S.

 

Now if we accept S, and not K, as the set standard of how well he has done and accept the: If K then S, argument as true then it should always equate to success regardless of sample size. I don't think it does and I 'm sure you would agree. Thus I think there is a correction that needs to be made to the syllogism.

I suggest; If K then FS. With FS meaning future success.

Those are two separate arguments. One I agree with the other I do not.

K=S is not true.

K=FS is true.

Does that make sense?

[1992casey]

Greinke tries to pitch in a way that gives him the best FIP.

Quoth Zack:

---

“That’s pretty much how I pitch, to try to keep my FIP as low as possible,” Greinke said.

He also tries to adjust for ballparks and team defense.

Tyler Kepner wrote (in the article linked to above):

---

(Brian) Bannister said Greinke has learned to adjust his pitching based on the advanced defensive statistics. Because of the size of the outfield at Kauffman Stadium and the strength of the Royals’ outfielders, relative to their infielders, it sometimes made more sense to induce fly balls.

[monty57]

Because of the size of the outfield at Kauffman Stadium and the strength of the Royals’ outfielders, relative to their infielders, it sometimes made more sense to induce fly balls.

 

So in realizing that the Brewers have no strengths on defense, he's just trying to strike everyone out. He does need to realize he's not in Kauffman Stadium anymore... in Miller Park those long fly balls are called home runs.

[monty57]

Because of the size of the outfield at Kauffman Stadium and the strength of the Royals’ outfielders, relative to their infielders, it sometimes made more sense to induce fly balls.

 

So in realizing that the Brewers have no strengths on defense, he's just trying to strike everyone out. He does need to realize he's not in Kauffman Stadium anymore... in Miller Park those long fly balls are called home runs.

[Oxy]

Zack's pitching to lower his FIP is very interesting and could perhaps be telling. It would support the theory that he isn't careful enough with guys after he is behind in the count, since a walk would hurt his FIP a lot more than a double, even though a double would hurt the team a lot more than a walk.

 

The goal should be to win the game, not lower your FIP. Sometimes you need to be more aggressive with batters and sometimes you need to be more careful with batters in order to maximize your chances of winning the game. Fielding is a large part of playing winning baseball, so ignoring that within the context of a single game just doesn't make sense.

 

I'm probably reading too much into those comments, but I still think it's pretty interesting.

[Thurston Fluff]

That isn't something I want to hear from a pitcher. They should pitch to get guys out and limit runs and the stats will take care of themselves.

[TheCrew]

I'd take any Greinke quote with a grain of salt - I could easily see him yanking a reporters/teammates chain on something like that. It's an isolated quote from 2 years ago....... now if Greinke made references to it multiple times, or something like that I'd be more of a believer.

[And That]

I

disagree with the usefulness of using K rate. First of all, if you

want to look at K rate, the denominator should be batters faced, not IP.

Using IP inflates the K rate of a pitcher who is allowing a lot of

base runners (since he faces more overall batters as a result). Not

sure who even has K/BF though.

Yeah, I mentioned that above.

OK, so we have SO% from Baseball-Reference, which is the percentage of all plate appearances ending in a strikeout...

 

Zack Greinke has a 31.9% SO%, which would lead all of qualified pitchers if he had 15 more innings pitched.

 

Once again, MLB qualified leaders from the past five seasons, this time using SO%:

 

2010
Jon Lester: 26.1 SO%, 3.25 ERA

Tim Lincecum: 25.8 SO%, 3.43 ERA

Jered Weaver: 25.8 SO%, 3.01 ERA

Mat Latos: 25.3 SO%, 2.92 ERA

Jonathan Sanchez: 25.3 SO%, 3.07 ERA

 

2009

Tim Lincecum: 28.8 SO%, 2.48 ERA

Justin Verlander: 27.4 SO%, 3.45 ERA

Javier Vasquez: 27.3%, 2.87 ERA

Jon Lester: 26.7 SO%, 3.41 ERA

Zack Greinke: 26.5 SO%, 2.16 ERA

 

2008
Tim Lincecum: 28.6 SO%, 2.62 ERA

Edinson Volquez: 24.6 SO%, 3.21 ERA

CC Sabathia: 24.5 SO%, 2.70 ERA

A.J. Burnett: 24.1 SO%, 4.07 ERA

Ervin Santana: 23.9 SO%, 3.49 ERA

 

2007
Erik Bedard: 30.2 SO%, 3.16 ERA

Scott Kazmir: 26.9 SO%, 3.48 ERA

Johan Santana: 26.8 SO%, 3.33 ERA

Jake Peavy: 26.7 SO%, 2.54 ERA

A.J. Burnett: 25.5 SO%, 3.75 ERA

 

2006
Johan Santana: 26.5 SO%, 2.77 ERA

Jake Peavy: 25.4 SO%, 4.09 ERA

Orlando Hernandez: 23.5 SO%, 4.66 ERA

Carlos Zambrano: 22.9 SO%, 3.41 ERA

Brett Myers: 22.7 SO%, 3.91 ERA

 

Not surprisingly at all, we reach about the same set of statistics as we did with K/9, a bit better for ERA, actually

 

Median=3.25; Mid=3.25; Max=4.66; Min=2.16; Standard deviation= .57

[1992casey]

That isn't something I want to hear from a pitcher. They should pitch to get guys out and limit runs and the stats will take care of themselves.

If you're going to pitch to a stat, FIP is a good choice. It covers the things that a pitcher can control. As noted above, he isn't just pitching to FIP, which attempts to rule out defense. He's also attempting to take defensive factors into account.

 

If he were pitching to ERA, then we'd have a problem. There are times when it's better to let a run score and attempt to induce a double play, for instance.

[Thurston Fluff]

I'm not talking about FIP's value as much as I would rather a player not worry about stats at all. They have a pretty simple job. Help win as many games as possible. I tend to worry about players who look at stats simply because I think that can lead to losing sight of that main objective. If that is his way of obtaining the main objective fine. It's really not that big of a deal. It only concerns me if his focus on FIP trumps his focus on helping the team win as many games as possible. I know they tend to go hand in hand but I think you know what I'm getting at.

[Oxy]

If you're going to pitch to a stat, FIP is a good choice. It covers the things that a pitcher can control. As noted above, he isn't just pitching to FIP, which attempts to rule out defense. He's also attempting to take defensive factors into account.

 

If he were pitching to ERA, then we'd have a problem. There are times when it's better to let a run score and attempt to induce a double play, for instance.

He does contradict himself when he brings up pitching to lower his FIP yet also pitching to the strengths of his defense. I don't really think his comments should be taken that seriously since they were initiated by a reporter who was obviously doing a story about advanced pitching stats. Still, I think this is an interesting discussion.

 

The relative value of an out and a run depends on the score and the number of outs and innings remaining in the game. Sometimes you need to cut down that run at the plate, sometimes you need to take the double play. Sometimes an outfielder should catch the foul ball with a guy on 3rd and less that 2 outs and sometimes he should let it drop (although this situation is pretty rare these days with the new stadiums). Still, these are fielding strategy questions, not pitching strategy.

[monty57]

Doing something like throwing high fastballs could lead to more K's and more flyballs... flyballs which may be outs in KC but HRs in Milwaukee. Plus, we don't have great OF defense, especially when Kotsay and {cough, cough} Almonte were out there. Earlier posters said it was his fastball that was the problem this season, and maybe this is part of it. A hard hit ball to a good OF in a spacious stadium leads to an out. A hard hit ball to Mark Kotsay in Miller Park is either a HR or extra bases.

[rluzinski]

I disagree with the usefulness of using K rate.

 

Do you disagree with using K rate to forecast the future or that it is incorrect to use it to say the player's results have been good? I agree K rates/ab tell us what the success rate should be going forward. The correlation has been well established. I disagree that it tells us he's had good results. I don't know how to explain it in mathematical terms but I can in a logic format.

 

If k rates are indicators of a pitchers ability to be successful then that ability has to be defined by another standard than simply K rates. Otherwise you would be saying: If K then K. While that is wonderful example of a-priori knowledge it is also useless in this discussion.

 

What I think was meant is: If K then S. With Success being a set standard of measurement other than K itself. That set standard of results is what determines if he was successful to that point not the K rate. In other words; If S then S.

 

Now if we accept S, and not K, as the set standard of how well he has done and accept the: If K then S, argument as true then it should always equate to success regardless of sample size. I don't think it does and I 'm sure you would agree. Thus I think there is a correction that needs to be made to the syllogism.

I suggest; If K then FS. With FS meaning future success.

Those are two separate arguments. One I agree with the other I do not.

K=S is not true.

K=FS is true.

Does that make sense?

If success is measured by runs given up, of course a high strikeout rate does not equal success. A high strikeout rate suggests the pitcher is performing well in general but it doesn't guarantee it either.

[Thurston Fluff]

If success is measured by runs given up, of course a high strikeout rate does not equal success. A high strikeout rate suggests the pitcher is performing well in general but it doesn't guarantee it either.

 

Exactly. K/AB rates will tell us he is bound to get better at preventing runs but that is not the same as his K rate means he was preventing runs. Since preventing runs is ultimately what determines if he was successful that is the only stat that can tell us if he was successful. Now if someone wants to say it was just bad luck or his defense let him down as reasons for lack of success we can have that discussion. But that discussion is about why he did lack success not that he was successful even if it didn't show.

 

For my part I think he is going to be fine. I think so for the same reason's you do. His predictive stats look promising. AS to why he didn't succeed to the level I thought he would until now I think he just missed his spots a little more than usual and maybe one or two of his pitches weren't as sharp as they could be. I'm basing that mostly on comments by RRR after a pretty good outing by Greinke that he can be even better. Also after his last outing he made a comment somethign to the effect of his change-up working well so when he made a mistake with a pitch they hitter was off balance and it didn't get tagged. That would lead me to believe, though his mistakes were few, they were getting hit not by bad luck but by a lack of effectiveness prior to that mistake pitch. I'm not one to just assign luck as a reason until other factors are eliminated as possibilities.

[Oxy]

If success is measured by runs given up, of course a high strikeout rate does not equal success. A high strikeout rate suggests the pitcher is performing well in general but it doesn't guarantee it either.

 

Exactly. K/AB rates will tell us he is bound to get better at preventing runs but that is not the same as his K rate means he was preventing runs. Since preventing runs is ultimately what determines if he was successful that is the only stat that can tell us if he was successful. Now if someone wants to say it was just bad luck or his defense let him down as reasons for lack of success we can have that discussion. But that discussion is about why he did lack success not that he was successful even if it didn't show.

 

For my part I think he is going to be fine. I think so for the same reason's you do. His predictive stats look promising. AS to why he didn't succeed to the level I thought he would until now I think he just missed his spots a little more than usual and maybe one or two of his pitches weren't as sharp as they could be. I'm basing that mostly on comments by RRR after a pretty good outing by Greinke that he can be even better. Also after his last outing he made a comment somethign to the effect of his change-up working well so when he made a mistake with a pitch they hitter was off balance and it didn't get tagged. That would lead me to believe, though his mistakes were few, they were getting hit not by bad luck but by a lack of effectiveness prior to that mistake pitch. I'm not one to just assign luck as a reason until other factors are eliminated as possibilities.

I agree with you for the most part, but it's not like managers and even players themselves are totally immune to the introduction of biases from positive results. Perhaps even moreso since it is probably psychological helpful to maintain the illusion that they have more control over what happens than they actually do.

[reillymcshane]

i thought wins were the determining factor if a pitcher is pitching well http://forum.brewerfan.net/images/smilies/smile.gif

[umphrey]

After Nyjer Morgan and Casey McGehee's brilliant defense it was obvious Greinke started throwing garbage

[theavrock]

After Nyjer Morgan and Casey McGehee's brilliant defense it was obvious Greinke started throwing garbage

LOL WUT???

[markedman5]

At least he isn't hurt. RR said he pulled him "because I didn't want him to give up 9 runs".......

[Thurston Fluff]

After Nyjer Morgan and Casey McGehee's brilliant defense it was obvious Greinke started throwing garbage

LOL WUT???

It was either a joke or frustration. But I think it's clear he's performed below expectations so far. I can buy him not being this bad going forward due to some stat or another that is supposed to tell me everything I need to know but I really hope no one is going to try to say he's been fine. That he's been unlucky or the defense has let him down, his BAPIP is unusually high, his FIP says he's doing well, he's in early season form, sample sizes are too small or any other excuse. Bottom line is he hasn't pitched as well as he should and danged well better turn it around in a quick hurry.

[theavrock]

Agreed. He pitched like garbage tonight. The 1st inning was not his fault, but the 2nd was all his fault.

 

Just wasn't sure what the poster was saying. I laughed and then questioned. Hence, the LOL WUT???

[FVBrewerFan]

ZG is the #4 SP on the staff as of today. That's the only stat that matters. The Brewers gave up a lot to get him, and they didn't make that trade for a #4 pitcher. If you're looking for stats, you can find them on either side to support your opinion. You can focus on this year, or on his career. The good news is whatever he has or hasn't done so far this year can turn around quickly in the 2nd half. If we get the ace ZG in September, all is forgiven.