Nomadsfest Sox Fans

A forum for the old AOL board Sox fans and others.


    Giolito

    Share
    avatar
    rmapasad
    Roof Shot

    Posts : 2489
    Join date : 2009-04-06
    Location : Northridge, CA

    Re: Giolito

    Post by rmapasad on Thu Sep 28, 2017 12:13 pm

    alohafri wrote:You two statheads are giving me a migraineTwisted Evil

    I wouldn't expect anyone else to read a word of this stuff.  Sorta like the lengthy discussions in Blago's Cell a while ago.  A couple posters get caught up in one topic.
    avatar
    Soxillinirob
    Chairman Reinsdorf

    Posts : 7401
    Join date : 2009-04-05
    Age : 51
    Location : St. Charles, IL

    Re: Giolito

    Post by Soxillinirob on Thu Sep 28, 2017 12:22 pm

    I definitely read and usually understand, but I do it while shaking my head in horror.  I'm a math nerd, a math major, and a former teacher of various levels of high school math, but you guys are in some serious weeds and there will never be a meeting of the minds (which I think you know).  You are wasting your breath.
    avatar
    Deplorable Mark
    Roof Shot

    Posts : 1847
    Join date : 2016-09-16

    Re: Giolito

    Post by Deplorable Mark on Thu Sep 28, 2017 1:13 pm

    Soxillinirob wrote:I definitely read and usually understand, but I do it while shaking my head in horror.  I'm a math nerd, a math major, and a former teacher of various levels of high school math, but you guys are in some serious weeds and there will never be a meeting of the minds (which I think you know).  You are wasting your breath.


    Spare me RobZ.


    Any math nerd reading these should quickly realize this is nothing more than stats 101.  You don't need a PhD to understand the concept of independent variables.  OBP and SLG are not independent of each other.  You don't need to be a stat geek to figure out that base hits comprises over half the value of each statistics.  Since these two are not independent of each other, they should not be in the same linear regression.  However, you DO have sabermeticians putting OBP and SLG in the same regression.  The result is a coefficient for OBP that is about 1.8 greater than the coefficient for SLG.  This lead some sabermeticians to conclude OBP is 80% more valauble than SLG.  This is a false conclusion.  What they are really measuring is multicolinearity.  When I finally dug into the data, the multicolinearity practically jumped off the screen.  In any given year, the amount of extra bases on hits is about 70% - 80% more frequent than walks.  These items just happen to be the non base hit portions of OBP and SLG.  I highly doubt its a coincidence.


    I have explained this over and over and over again over the past decade.  The real reason there will never be a meeting of the minds is that one person here is completely ignoring the very sound math in the paragraph above.  Instead he counters with lies and insults.  Claiming Tom Tango is bushels smarter than the KARK is not a serious counter.  It is actually quite childish.  I will spare you the laundry list of absolute crap Roger has made up over the past decade.  Although I will remind you that when all this started 10 years ago, Roger claimed it was PhD's producing it and I quickly discovered one of the guys he was touting, a man by the name of Gleeman, was actually a journalism drop out.  After mocking him, Roger claim Gleeman had to be smarter than me because he's on TV.  I know I can get very vulgar at times, but Roger was not only insulting me, but lying to me as well from the beginning.


    What really gets me pig biting mad is how a former math teacher did not see my point right away.  If one of your students turned in a regression assignment that showed RUNS = a + b*OBP + c*SLG; I am certain you would take points off for using variables that were not independent of each other.  The fact that you never pointed this out on board is just another reason I believe you deliberately disagree with me.


    If this board wants this topic to end, all Roger has to do is admit my first paragraph is 100% correct.  I'll make the apology for a decade of lies and insults optional.


    XOXOXOXOOX
    THE KARK, a CPA with a minor in math from the U of I Urbana campus.  Into SABRmetrics since 1983
    avatar
    Deplorable Mark
    Roof Shot

    Posts : 1847
    Join date : 2016-09-16

    Re: Giolito

    Post by Deplorable Mark on Thu Sep 28, 2017 1:20 pm

    rmapasad wrote:Here is the problem that all these "run-predicting" systems have. They can't forecast how events are sequenced, which matters a whole lot.  
    Inning 1:   Fly out - Triple - BB - BB - K - Fly out   =    0 runs with .250 BA/.500 OBP/.750 Slug
    Inning 2:   K - BB - BB - Triple -Sac Fly Out - Fly out = 3 runs with .250 BA/.500 OBP/.750 Slug 


    Yeah, this is a problem, but it has zero to do with 1.8


    BTW, whats the answer?  Is the 1st inning triple worthless?!?!?  Is it fair to treat it differently than the 2nd inning triple?!?!?!?  And if you do treat the 1st inning triple as less than the 2nd inning triple, that will mushroom into the premise that great players on crap teams are less valuable than very very very good players on great teams.


    I'm sure the converse also presents in own list of problems
    avatar
    Deplorable Mark
    Roof Shot

    Posts : 1847
    Join date : 2016-09-16

    Re: Giolito

    Post by Deplorable Mark on Thu Sep 28, 2017 1:31 pm

    rmapasad wrote:
    Deplorable Mark wrote:THAT DOESN"T MEAN OBP IS MORE VALUABLE THAN SLUG OR WALKS ARE BETTER THAN DOUBLES.   Geesh - stop your endless streams of distortions.  


    *************************


    From Fangraphs


    https://www.fangraphs.com/library/offense/woba/



    wOBA is based on a simple concept: Not all hits are created equal. Batting average assumes that they are. On-base percentage does too, but does one better by including other ways of reaching base such as walking or being hit by a pitch. Slugging percentage weights hits, but not accurately (Is a double worth twice as much as a single? In short, no) and again ignores other ways of reaching base. On-base plus slugging (OPS) does attempt to combine the different aspects of hitting into one metric, but it assumes that one percentage point of SLG is the same as that of OBP. In reality, a handy estimate is that OBP is around twice as valuable than SLG (the exact ratio is x1.8). In short, OPS is asking the right question, but we can arrive at a more accurate number quite easily.


    Can someone please explain how the KARK distorted the bolded red portion?!?!?


    Also, if OBP is 1.8 SLG, then why does Bill James basic runs created formula which can be rewritten as AB times OBP times SLG so accurate?  I ran this formula and the 5 White Sox division winners, 1983,1993,2000,2005,2008 and the different between estimated and actual was about 2%.


    It doesn't take a CPA with a 135 IQ to figure out that 98% accuracy would be impossible if there truly was an 80% different between OBP and SLG.


    PS, somebody give Kevin an asipirin


    Yes, ALL these different calculations work basically.  Depending on the scoring era or season, some work better than others.  Oddly enough, in the low-scoring era of the 1960's, OBP*1.8 + Slug produced the best correlation to run-scoring of any system - w/ OBA, straight OPS, your James formula.

    I WOULD LOVE TO SEE THE LINK TO THIS AS WELL AS THE ACTUAL FORMULA USED.  MY GUESS, MORE MAKE BELIEVE AND MORE DISTORTION

    But the differences were small.   I ran it again on 2017,which is both the highest HR and highest K season of any in history.  All four systems produced virtually the same level of correlation to Runs Scored.  Meaning that THERE IS NOTHING WRONG WITH 1.8 * OPS nor was there anything that wrong with the James formula or even straight OPS.  Whatever refinements the new-fangled 1.8 * OBP or w/OBA provided were not huge, major improvements.  
    BTW, let me address this multicolinearity stuff once and for all.  There was a reason that no researcher of any substance raised that objection to these studies.  The way they were done "interrelated variables" were not a factor. Had you read those studies you would have seen why.

    DON'T NEED TO READ THE STUDIES.  THE DEFINITION OF MULTICOLINEARITY ACTUALLY STATES THAT THE ERRORS ON THE INDIVIDUAL COEEFICIENTS WOULD CANCEL EACH OTHER OUT.  BUT AGAIN, YOU LIE AND DISTORT.  OBP IS NOT 80% MORE VALUABLE THAN SLG.  

      You add a BB which only changes OBP but not Slug.  Conversely you change a single to a double or HR and that only changes Slug but not OBP.  But even when OBP and SLug are intertwined, as happens with all hits, you can still separate out the Slug component.

    The 1.8 stuff was a big tempest in a teapot, as it added something to the party but not a whole lot.

    AGAIN, PURE BULLSHIT.
      However the bigger tempest came in this folder by someone who continues to spout his professional credentials and IQ numbers in some ego-flexing exercise to prove that he knows more than very smart guys who studied this for several years.  Bottom line: he doesn't.


    GO FUCK YOURSELF.  IF ANYBODY IS TRYING TO MAKE THEMSELVES LOOK SMART ITS YOUR TERM PAPER WRITING ASS AND YOUR RETARDED EXAMPLES.  MAKE SOME MORE SHIT UP LIAR.  AGAIN, I'M NOT ARGUING WITH TOM TANGO.  I AM EXPOSING A DUMBFUCK WHO IS TRYING TO HIDE BEHIND IS SKIRT
    avatar
    Deplorable Mark
    Roof Shot

    Posts : 1847
    Join date : 2016-09-16

    Re: Giolito

    Post by Deplorable Mark on Thu Sep 28, 2017 1:52 pm

    https://www.stat.berkeley.edu/~aldous/157/Old_Projects/moy.pdf



    READ PAGE 21!!!!!!


    AND HONORS STUDENT AT BERKLEY!!!!!!!


    WHAT A HACK!!!!!!


    I HOPE THE PROF FAILED HIM BECAUSE HE DID NOT USE INDEPENDENT VARIABLES


    ONLY AN OVER EDUCATED IDIOT THINKS ABOUT 1 POINT OF OBP VERSUS 1 POINT OF SLG AS IF THEY ARE TWO DIFFERENT THINGS!!!!!


    IF ITS NOT A WALK, YOU A JUST MEASURING THE SAME THING IN TWO DIFFERENT WAYS!!!!!


    I WONDER TOM TANGO EVER HEARD OF OCCAM' RAZOR?!?!?!?!?  BASEBALL ISN'T CALCULUS.  A MANAGER DOESN'T TELL HIS PLAYER TO TRY AND HIT AN OBP INSTEAD OF AN SLG.  YET THE TERM PAPER WRITER FROM CALIFORNIA SWEARS TO ALL THIS OBVIOUS NONSENSE.  THIS CLOWN CAN'T THINK OUTSIDE THE BOX SO HE FALLS FOR CIRCULAR REASONING.
    avatar
    Deplorable Mark
    Roof Shot

    Posts : 1847
    Join date : 2016-09-16

    Re: Giolito

    Post by Deplorable Mark on Thu Sep 28, 2017 2:44 pm

    BTW, let me address this multicolinearity stuff once and for all.  There was a reason that no researcher of any substance raised that objection to these studies. 


    ************************************************


    AND THIS IS WHY I GET SO PISSED AT ROGER!!!!!!


    Not only is the above line phrased in an incredibly biased and self serving way, but where is the proof?!?!?!?!?


    AND THIS IS WHY I CONSTANTLY ACCUSE ROGER OF MAKING STUFF UP AND LYING!!!!!


    I know enough about math to know bullshit when I see it.  Independence is critical in statistical research.  Yet here Roger is claiming that a whole bunch of experts just ignored it without a second thought!!!!!!  That makes ZERO sense.  So either Roger is lying his ass off, or his experts are a bunch of dumb fucks.  


    Then is ignorant arrogant liar is going to besmirch my reputation?!?!?!
    avatar
    Deplorable Mark
    Roof Shot

    Posts : 1847
    Join date : 2016-09-16

    Re: Giolito

    Post by Deplorable Mark on Thu Sep 28, 2017 3:44 pm

    AND ANOTHER!!!!!


    AS A CPA, IT IS MY UNDERSTANDING THAT DATA ANALYTICS IS SUPPOSE TO PROVIDE REAL WORLD SOLUTIONS


    WHAT SOLUTIONS IS THIS 1.8 CRAP PRODUCING?!?!?!?!


    Game 7, bottom of the 9th, tie game.  Does the manage order his hitter to hit an OBP or and SLG?  With the world championship in balance, Tom Tango feverishly punchs away on his HP 12c to get the answer.


    Meanwhile, dumbfuck coach KARK is telling the guy to look for a pitch to drive, if not take a walk.


    Hey, when Kirk Gibson took Dennis Eckersley deep, was it the OBP or the SLG that really won the game?!?!?  Can somebody get Tom Tango on the phone.  this is important stuff
    avatar
    Deplorable Mark
    Roof Shot

    Posts : 1847
    Join date : 2016-09-16

    Re: Giolito

    Post by Deplorable Mark on Thu Sep 28, 2017 4:04 pm

    rmapasad wrote:
    alohafri wrote:You two statheads are giving me a migraineTwisted Evil

    I wouldn't expect anyone else to read a word of this stuff.  Sorta like the lengthy discussions in Blago's Cell a while ago.  A couple posters get caught up in one topic.


    SINCE YOUR THE EXPERT, FIND A FORUM WHERE KNOWEDGLEABLE PEOPLE DISCUSS SABR TOPICS AND I WILL GLADLY PUT YOU TO SHAME THERE AS WELL


    ALONG WITH ANY OTHER HACK THAT BELIEVES YOUR. NOT TOM TANGO'S , BUT YOUR NONSENSE.


    MY GUESS IS THAT YOU'LL BE LAUGHED OFF SUCH A BOARD AS THE KARK DAZZLES EVERYBODY WITH HIS BRILLANCE AS ONLY THE KARK CAN
    avatar
    rmapasad
    Roof Shot

    Posts : 2489
    Join date : 2009-04-06
    Location : Northridge, CA

    Re: Giolito

    Post by rmapasad on Thu Sep 28, 2017 4:44 pm

    I WOULD LOVE TO SEE THE LINK TO THIS AS WELL AS THE ACTUAL FORMULA USED.  MY GUESS, MORE MAKE BELIEVE AND MORE DISTORTION>>
    You can run and cite the White Sox numbers in their pennant years but yet I pull stats from Fangraphs, run numbers with Excel (which I've used forever), do standard deviations and it's "lying" and fabrication.  Go fuck yourself. Maybe you want to bring back up my pre-season "lies" about the chances of guys like Avi Garcia to have some sort of career uptick. 
    BTW, you never have the definition of "lying" straight. "Lying" is not rendering an opinion you disagree with. 

    DON'T NEED TO READ THE STUDIES.  THE DEFINITION OF MULTICOLINEARITY ACTUALLY STATES THAT THE ERRORS ON THE INDIVIDUAL COEEFICIENTS WOULD CANCEL EACH OTHER OUT.  >>
    This is the essence of the problem.  You don't need to read the methodology used so you ASS-U-ME what they did is automatically invalid. I told you that one study that I remember did the following:
    you add a BB which only changes OBP but not Slug.  Conversely you change a single to a double or HR and that only changes Slug but not OBP.  
    [size=18]And here was your well-documented, well-explained, well-justified answer

    AGAIN, PURE BULLSHIT.[/size]

     IF one of your students turned in a regression assignment that showed RUNS = a + b*OBP + c*SLG; I am certain you would take points off for using variables that were not independent of each other
    Isn't that the same problem that Bill James' Slug * OBP /(Plate Appearances had ?
    How is it unique to 1.8 * OBP + Slug ? 

    When I finally dug into the data, the multicolinearity practically jumped off the screen.  In any given year, the amount of extra bases on hits is about 70% - 80% more frequent than walks.  These items just happen to be the non base hit portions of OBP and SLG.  I highly doubt its a coincidence. >>
    OK, from 2010-2017 all teams in baseball had a ratio of 1.67 (TB-hits)/BB.  The ratio of Slug to OBP was 1.26.  You aren't explaining your point worth a shit.  

    Interesting discovery in this process.  If you take Net Bases Gained/4, that's a pretty good indicator of Runs.  It's TB+ BB's+HBP+SF+SH+SB - CS-GDP/4.  The average # of runs scored by teams in this period was 5589 and using this formula the aggregate of average teams NBG/4 was 5593.  So for simplicity sake every base gain or lost is .25 runs.  Not saying it's perfect, just close enough for usefulness.

    So simple example. 10 PA-1 BB- 2 singles- 1 double = .333/.400/.444  = 5 total bases gained
    Next PA is 0 on base, 0 outs.
    Next PA is BB thus:  11 PA- =.333/.454/.444 = 6 bases gained 54 pt OBP gain per 1 base gained
    IF instead it's a double, then: 11 PA - 400/.454/.600=  7 bases G  156 Slug/2 BG= 78 Slug / 1 BG 
    It's even more so if it's a HR which is .800 slug = 356 gain in Slug/ 4 BG = 89 Slug per 1 BG

    OBP jumps 13.5% for 1 BG, Slug jumps 35% for 2 BG,  80 % for 4 BG.  Do the math and you see that Slug % gives disproportionate rewards  per base gained than OBP.  This has zero to do with multicolinearity as these distortions do not come from any interaction of the two variables (OBP/Slug).  They come from the nature of the one variable itself (Slug)..

    MAKE SOME MORE SHIT UP LIAR.  AGAIN, I'M NOT ARGUING WITH TOM TANGO.  I AM EXPOSING A DUMBFUCK WHO IS TRYING TO HIDE BEHIND IS SKIRT>
    Of course you're arguing with Tango's conclusions- just look at this sentence of yours :

    BUT AGAIN, YOU LIE AND DISTORT. OBP IS NOT 80% MORE VALUABLE THAN SLG.  >>I did not say that I believe with absolute certainty that 1.8 is the perfect adjustment to OBP  But Tango sure as hell did and you used this very quote of his:  a handy estimate is that OBP is around twice as valuable than SLG (the exact ratio is x1.8). 

    So email Tango if you think his conclusions are flawed by "multicolinearity", and see what he says.  Pure OBP+ Slug  distorts the value of Slug as against OBP.  
    THIS IS THE REASON ALL THESE BASEBALL RESEARCHERS HAVE COME UP WITH ALL TYPES OF NEW SYSTEMS to enhance plain OPS and why w/OBP is now what is used.  But feel free to yearn for the good old days of the 1970's and Bill James' models and  then explain why his system doesn't suffer from the same multicolinearity problem.
    (in plain English not mumbo-jumbo).
    avatar
    rmapasad
    Roof Shot

    Posts : 2489
    Join date : 2009-04-06
    Location : Northridge, CA

    Re: Giolito

    Post by rmapasad on Thu Sep 28, 2017 4:54 pm

    << SINCE YOUR THE EXPERT, FIND A FORUM WHERE KNOWEDGLEABLE PEOPLE DISCUSS SABR TOPICS AND I WILL GLADLY PUT YOU TO SHAME THERE AS WELL
    ALONG WITH ANY OTHER HACK THAT BELIEVES YOUR. NOT TOM TANGO'S , BUT YOUR NONSENSE. 
    MY GUESS IS THAT YOU'LL BE LAUGHED OFF SUCH A BOARD AS THE KARK DAZZLES EVERYBODY WITH HIS BRILLANCE AS ONLY THE KARK CAN >>

    As I said before, feel free to email Tango about his huge multicolinearity flaw and please post his response.  
    Also if you go onto one of those boards be sure to post the following statements of yours.  I'm sure they'll all be impressed as hell with you.  ROFLMAO 

    <
    Game 7, bottom of the 9th, tie game.  Does the manage order his hitter to hit an OBP or and SLG?  With the world championship in balance, Tom Tango feverishly punchs away on his HP 12c to get the answer. Meanwhile, dumbfuck coach KARK is telling the guy to look for a pitch to drive, if not take a walk.
    Hey, when Kirk Gibson took Dennis Eckersley deep, was it the OBP or the SLG that really won the game?!?!?  Can somebody get Tom Tango on the phone.  this is important stuff>>
    avatar
    rmapasad
    Roof Shot

    Posts : 2489
    Join date : 2009-04-06
    Location : Northridge, CA

    Re: Giolito

    Post by rmapasad on Thu Sep 28, 2017 5:10 pm

    ************************************************ wrote:AND THIS IS WHY I GET SO PISSED AT ROGER!!!!!!
    Not only is the above line phrased in an incredibly biased and self serving way, but where is the proof?!?!?!?!?
    AND THIS IS WHY I CONSTANTLY ACCUSE ROGER OF MAKING STUFF UP AND LYING!!!!!>
    I gave you the proof, and as usual you ignore it.  I noted that one researcher simply controlled out OBP in one example and Slug in the other. What about your proof ?  The usual "I'm a CPA with a 135 IQ", "I've studied math and statistics" "no one in his right mind would think this"
    Typical "I'm this, I'm that","these guys know nothing" which is Kark type proof as you amply demonstrate:
    I know enough about math to know bullshit when I see it.  Independence is critical in statistical research.  Yet here Roger is claiming that a whole bunch of experts just ignored it without a second thought!!!!!!  That makes ZERO sense.  So either Roger is lying his ass off, or his experts are a bunch of dumb fucks.  
    Then is ignorant arrogant liar is going to besmirch my reputation?!?!?!

    You besmirch your own reputation by using insults as arguments, constantly brandishing your ego and your credentials, and rudeness/nastiness to others.

    Sponsored content

    Re: Giolito

    Post by Sponsored content


      Current date/time is Tue Oct 17, 2017 2:43 am