What's up with Corey Hart? Latest: Hart wins arbitration hearing (post # 190)

[rluzinski]

The projection line really seems to forget one basic principle of math... that averages are greatly skewed by one number that is significantly higher or lower than the rest. This is why the housing market uses median numbers rather that average numbers. One million dollar house sold skews the average price if all the other houses are $65,000. In Hart's case, one .895 OPS season skews the fact that he's been a .750 OPS guy for the past two years.

 

 

Basic principle of math? People who suggest throwing data out of a sample because it's an "obvious outlier" are breaking one of the most basic rules of statistical analysis. There are situations where the median can tell you something the mean can't but I don't think this is one of them.

 

Hart's 2007 season probably makes up only about 20% of his projection anyway, so it doesn't make much of a difference either way.

[Funketown]

-The projection line really seems to forget one basic principle of math... that averages are greatly skewed by one number that is significantly higher or lower than the rest. This is why the housing market uses median numbers rather that average numbers. One million dollar house sold skews the average price if all the other houses are $65,000. In Hart's case, one .895 OPS season skews the fact that he's been a .750 OPS guy for the past two years.

I've never heard this point referred to as a basic principle of math...Just because a data point skews the mean of the data set does not mean this point should not be included.

[Milltownhooker]

Corey's days as a brewer are numbered.

[monty57]

Okay, maybe not "Basic pricipal of math," but "basic mathematical knowledge." Taking an average of three numbers is going to be extremely volatile, and likely to be skewed if any of the three numbers is far off from the other two (around here it's known as small sample size). I wouldn't say the 2007 numbers shouldn't be included, but there needs to be an asterisk next to them. I agree that this is not a situation where median numbers should be used, but I thought the real estate example could show how basic averages can be thrown way off. Basically, all I'm trying to say is that what works in some situations doesn't work in all situations. I never said to throw a year out as an "obvious outlier," I'd say that current statistical analysis uses far too many aggregates and tries to apply them to every individual situation, leading to many cases (I believe this is one) where the charts are wrong.

 

In Hart's case, there seem to be circumstances that have changed from point A in 2007 to points B & C in 2008-9. Most likely the circumstance is that baseball consists of many moving parts, each acting on the other (correlations). Pitchers have figured something out about Hart, and he is struggling mightily to correct it. He may be able to, he may not. Some players adapt more readily than others. Since this is probably a mental challenge, and therefore not quantifiable, statistical analysis has to assume it doesn't exist, or at the very least, take multiple years to determine that a problem indeed exists. If Hart has another .750 OPS season, the previously referenced projections will put Hart's projections at .750, but as long as the .895 season is in the three-year range, they will project him higher. Hart could lose a leg tomorrow, and these projections would still show him as a .785 OPS player for 2010, and unfortunately, I believe there are people who would argue that he's likely to do it based on these projections.

 

Insurance actuarial tables are probably as close as anything to what SABRmetrics is trying to achieve. If someone is trying to buy life insurance, and he's right smack in the middle of the "healthy" charts, he'll get a certain rate. If he comes back two years later and there is a significant negative material change in his physiology which began two years ago, he will not get the same rate. Something has changed and the actuaries account for it. If they don't consider this material fact, the insurance companies will lose a lot of money.

 

MLB teams have to consider material changes. If they don't, they will lose a lot of money. People trying to make a name for themselves on public domain websites have nothing to lose by putting out a formula that disregards non-quantifiable attributes. Heck, they may even get people quoting them. I think actuaries would put a lot of "risk premium" on Corey Hart regaining his old form. I believe most MLB GM's will do the same. Fortunately for Corey, the arbitration system virtually assures him a pay raise (if he's offered arby) and his first arby year was during a period where a reasonable GM could still believe that he just had a down 1/2 year. I'd bet that if he has another .750 OPS season, he won't get offered arby next year, and he'll be paid less in 2011 than he will be this year.

 

Instead of using aggregate stats of every player from the dawn of time, is there somewhere that compares "like players?" In other words, can you find other players who have followed similar career paths to Corey Hart to see how their careers turned out? To me, that seems a more logical path to take. Again, more numbers are not always better.

[Ennder]

In Hart's case, there seem to be circumstances that have changed from point A in 2007 to points B & C in 2008-9

 

But again this really isn't the case. Hart's 2008 was very similar to his 2007 until the last month of the season where he fell off the cliff. He has been bad for about 1 single season now, the 3/4 of a season he played in 2009 and the last month of 2008.

 

I'm not sure there is a meaningful pattern in there or not though I'd obviously rather have his last 162 games be better than worse.

[rluzinski]

People trying to make a name for themselves on public domain websites have nothing to lose by putting out a formula that disregards non-quantifiable attributes.

 

First of all, several of those guys work for MLB teams, so lets not pretend this is just bunch of amateurs playing around. Second of all, no one is suggesting that there isn't a lot of other information that can and should be considered beyond whatever a canned projection system does. Injuries is a huge one, for instance. But what are you considering that is non-quantifiable? You aren't adding any extra information that I see, just blind speculation. You might be right but you haven't offered up any evidence to support your theory that pitchers "have figured him out".

[trwi7]

In what way was Hart's 2008 season similar to his 2007 season? He had an .892 OPS in 2007. Heading into September of 2008 he had an .817 OPS.

[monty57]

But what are you considering that is non-quantifiable? You aren't adding any extra information that I see, just blind speculation. You might be right but you haven't offered up any evidence to support your theory that pitchers "have figured him out".

 

Because the evidence you seek is what is what's non-quantifiable. The only evidence you believe to be pertinant to any discussion is information that can be viewed in a stat line. Seeing every pitcher work him the same way, with the low & away slider strike three as the knockout punch, showed me that the pitchers had figured out how to get him out. The Brewers coaches of course realized this, because they have eyes, and anyone who watched the game could see what the pitchers were doing to Hart, even if Microsoft Excel had not yet figured it out. During last offseason, it was widely reported (and Hart said this himself) that the coaching staff told him to take more pitches to try to lay off the low & away slider. The changes made Corey late on fastballs (again, he said this himself in interviews). The pitchers realized this and blew fastballs by him. I know, that doesn't show up on a stat line, so it has to be "blind speculation," but MLB pitchers are going to blindly speculate Hart right out of the majors.

 

I don't know what proof you would require. Pitchers all work Corey the same way, and he himself has admitted he lost power because taking more pitches has caused him to get tentative and behind on the fastball. Maybe this isn't purely non-quantifiable... Hart's had two straight .750ish OPS seasons. Or, if we want to use Ennder's arguement, he was so ridiculously God-awful at the end of the 2008 season after the pitchers figured him out that he turned a decent season into a nightmare in a month.

 

I really hope I'm not right. I hope Corey figures it out. Macha's answer has been to let him swing away this season. I guess his thought is that someone who gets on base around 30% of the time, but can hit 20 HR is better than someone who gets on base at a 10% better rate with 60% the HR. I worry we'll see Sept 2008 again, but that's just speculation.

[Ennder]

In what way was Hart's 2008 season similar to his 2007 season? He had an .892 OPS in 2007. Heading into September of 2008 he had an .817 OPS.

The difference between these two things is pretty small. On sept 1st of 2008 Hart had an .822 OPS. Hart had an OPS as low as .828 on August 17th of 2007. The difference between these two seasons was pretty much one months production. That is the point here, when you are keying in on a single season the difference between a STUD and a DFA can be as little as one month. You just need to look at a larger sample in general.

[rluzinski]

The only evidence you believe to be pertinant to any discussion is information that can be viewed in a stat line.

 

No need to be so melodramatic. Evidence can take many forms. From aggregate stats to Pitch F/X, just about everything is recorded these days.

 

We all know Hart is a freeswinger who doesn't have much plate discipline. He never has. The question is whether pitchers are exploiting that weakness better over the last two years than they did in 2007. It's quite possible that they are. But again, most projection systems aren't weighing Hart's 2007 stats very much anyway.

[thebruce44]

We all know Hart is a freeswinger who doesn't have much plate discipline. He never has. The question is whether pitchers are exploiting that weakness better over the last two years than they did in 2007.

 

I think its pretty obvious Corey hasn't been seeing as many fastballs as he did in 2007. Fangraphs backs that up. The number of sliders he has seen has dramatically risen as well.

[rluzinski]

His run value vs. sliders (wSL) was terrible as well:

 

http://www.fangraphs.com/...layerid=1945&position=OF

 

I'm concerned that Hart has lost all this weight to try and turn into some bunting speedster. I hope I'm wrong since I don't think he has the skill set to pull it off.

[berniebrewer4life]

We all know Hart is a freeswinger who doesn't have much plate discipline. He never has. The question is whether pitchers are exploiting that weakness better over the last two years than they did in 2007.

 

I think its pretty obvious Corey hasn't been seeing as many fastballs as he did in 2007. Fangraphs backs that up. The number of sliders he has seen has dramatically risen as well.

And that is going to be the key, if he is going to have any success this year. He needs to either hit, or at least recognize the sliders when they come.

[thebruce44]

Because the evidence you seek is what is what's non-quantifiable. The only evidence you believe to be pertinant to any discussion is information that can be viewed in a stat line. Seeing every pitcher work him the same way, with the low & away slider strike three as the knockout punch, showed me that the pitchers had figured out how to get him out.

 

Anyway, I just wanted to show that this observation, which I have also had, can in fact be quantified and backed up with stats. I am also sure Corey is aware of this issue. He needs to make adjustments to be successful again. He didn't make them last year, so I really don't think a light bulb is going to go on at this point or it already would have.

[chadomac]

when do we find out if Hart won or lost his arbitration case this morning?

[Invader3K]

Hart's hearing isn't until this Thursday, according to the JSOnline Brewers blog.

[Milltownhooker]

PProbably will get a ruling on Friday. Things may get interesting regarding Hart after this hearing. Just a hunch

[berniebrewer4life]

PProbably will get a ruling on Friday. Things may get interesting regarding Hart after this hearing. Just a hunch

i agree. Hart strikes me as a guy who takes these kinds of things personally, and not as just business.

[TooLiveBrew]

Why? What is that kind of statement based on? Because he turned down an extension offer a while ago?

[berniebrewer4life]

Why? What is that kind of statement based on? Because he turned down an extension offer a while ago?

It's mostly from his reaction to others getting deals done ahead of him(the 'i wish they would have offered me that kind of deal' comment he made a while ago, amongst other things), and just a general feeling i get from him.

[BREWCREW5]

What is everyone's educated guess? I dont think he wins his case.

[thebruce44]

I would be shocked if he won, but stranger things have happened.

 

Add me to the list of people who think he may struggle with taking the criticism. I have heard these hearings can be pretty nasty for the player to sit in on. I am glad the Brewers hired a law firm to present the case instead of Doug or Ash doing it.

[berniebrewer4life]

. I have heard these hearings can be pretty nasty for the player to sit in on. I am glad the Brewers hired a law firm to present the case instead of Doug or Ash doing it.

I am shocked the players sit in on it. I know i wouldn't, but that's mostly because i know i don't take criticism very well.

[young guns]

The players must attend the hearing.

[monty57]

Yes, it's kind of an odd situation. The team has to explain why the player doesn't deserve more money, which means they have to rip into the player's bad points. Pretty much the opposite of what a team wants to do with a player they've spent years developing.

 

I know I've been down on Hart in this thread. I don't think he'll win the case, but I met his family in Arizona during Spring Training and they were very nice people. I don't wish ill on Corey. I hope he takes this experience and uses it as motivation to play very well and regain his past form.