Nomadsfest Sox Fans

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


    MEET TOM TANGO

    Share
    avatar
    Son of Ron
    Al Smith's Beer Bath

    Posts : 181
    Join date : 2018-10-17

    MEET TOM TANGO

    Post by Son of Ron on Tue Nov 27, 2018 1:01 pm

    https://sabr.org/analytics





    HE WAS AT THE 2018 CONFERENCE!!!!


    But I am only assuming he will be at the 2019.


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


    TWO WEEKS TO WRITE UP A RESEARCH PIECE THAT WILL DESTROY 1.8 FOREVER!!!!!


    But do I will want to piss away $1000 to shut up roger once and for all....


    Not to mention wasting precious PTO days to frollick with a bunch of amateurs and hacks
    avatar
    rmapasad
    Chairman Reinsdorf

    Posts : 2851
    Join date : 2009-04-06
    Location : Northridge, CA

    Re: MEET TOM TANGO

    Post by rmapasad on Tue Nov 27, 2018 1:37 pm

    Son of Ron wrote:https://sabr.org/analytics





    HE WAS AT THE 2018 CONFERENCE!!!!


    But I am only assuming he will be at the 2019.


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


    TWO WEEKS TO WRITE UP A RESEARCH PIECE THAT WILL DESTROY 1.8 FOREVER!!!!!


    But do I will want to piss away $1000 to shut up roger once and for all....


    Not to mention wasting precious PTO days to frollick with a bunch of amateurs and hacks


    Knock yourself out.   Be sure to tell all of them how full of shit the 1.8 * OBP theory is due to multicollinearity !!  LMAO...
    avatar
    alohafri
    Pope Malort I

    Posts : 7656
    Join date : 2009-04-03
    Age : 51
    Location : Southwest Suburbs

    Re: MEET TOM TANGO

    Post by alohafri on Tue Nov 27, 2018 8:39 pm

    Freaking geeks!  Twisted Evil Laughing bom
    avatar
    Jack Brickhouse
    Silent Hot Dog Vendor

    Posts : 1044
    Join date : 2014-04-03

    Re: MEET TOM TANGO

    Post by Jack Brickhouse on Wed Nov 28, 2018 5:17 am

    Was Murph there?
    avatar
    Son of Ron
    Al Smith's Beer Bath

    Posts : 181
    Join date : 2018-10-17

    Re: MEET TOM TANGO

    Post by Son of Ron on Wed Nov 28, 2018 6:49 am

    rmapasad wrote:
    Son of Ron wrote:https://sabr.org/analytics





    HE WAS AT THE 2018 CONFERENCE!!!!


    But I am only assuming he will be at the 2019.


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


    TWO WEEKS TO WRITE UP A RESEARCH PIECE THAT WILL DESTROY 1.8 FOREVER!!!!!


    But do I will want to piss away $1000 to shut up roger once and for all....


    Not to mention wasting precious PTO days to frollick with a bunch of amateurs and hacks


    Knock yourself out.   Be sure to tell all of them how full of shit the 1.8 * OBP theory is due to multicollinearity !!  LMAO...

    This coming from the retard that invents his own math and after a decade still can’t demonstrate where 1.8 comes from.  Your bullshit spreadsheet never showed how it was developed.  Yesterday I looked at the posts in the gaming section and it obvious to anyone with a math background that you are posting out your ass and making it as you go along

    PS, I bet even Tango would wonder WTF you are babbling about because you butchered his work so badly.  Too bad your ego has made you so delusional that you consider yourself Tango’s representive.  When the truth is you don’t know the man or what he has or hasn’t considered when conducting his research.  So again, argument wasn’t with Tango, it was with your lying stupid ass
    avatar
    Son of Ron
    Al Smith's Beer Bath

    Posts : 181
    Join date : 2018-10-17

    Re: MEET TOM TANGO

    Post by Son of Ron on Wed Nov 28, 2018 7:29 am

    For the adults in the room, which excludes the egotistical liar that drove his wife away because he makes shit up instead of admitting when he is clearly wrong

    My question to Tango would be

    How do you reconcile OBP is 1.8 of SLG when the Bill James runs created clearly demonstrates they are roughly equal

    tangotiger
    Joe Borchard's Teammate

    Posts : 10
    Join date : 2018-11-29

    Re: MEET TOM TANGO

    Post by tangotiger on Thu Nov 29, 2018 3:15 pm

    I'd be happy to answer whatever questions you have. However, based on the tone in this thread, I would simply ask that if you guys can holster your weapons while I'm here, then I can stick around as long as you want me here.  I'm not interested in anything other than advancing baseball knowledge.

    Agreed?
    avatar
    Soxillinirob
    Chairman Reinsdorf

    Posts : 8013
    Join date : 2009-04-05
    Age : 52
    Location : St. Charles, IL

    Re: MEET TOM TANGO

    Post by Soxillinirob on Thu Nov 29, 2018 9:00 pm

    tangotiger wrote:I'd be happy to answer whatever questions you have. However, based on the tone in this thread, I would simply ask that if you guys can holster your weapons while I'm here, then I can stick around as long as you want me here.  I'm not interested in anything other than advancing baseball knowledge.

    Agreed?
     
    OMG....please be the actual Tom Tango!  I would love to see the analysis.  Please proceed.  If you meet any unfriendly resistance it'll only be from one person in particular.

    tangotiger
    Joe Borchard's Teammate

    Posts : 10
    Join date : 2018-11-29

    Re: MEET TOM TANGO

    Post by tangotiger on Thu Nov 29, 2018 9:30 pm

    Perusing through here, it seems there are two claims or questions being made:
    1. Why 1.8
    2. Is the 1.8 inconsistent when you compare to Bill's RC

    First point: 1.7 is described here:
    www
    .
    insidethebook
    .
    com/ee/index
    .
    php/site/comments/why_does_17obpslg_make_sense/

    (Note: I had to split the above because this site is blocking newbies from posting links.)

    As for second point: I don't see any evidence  this is true.  PROBABLY RC is consistent with a 1.3 or so multiplier. But that would just point out the shortcomings of RC.  That's  because the walk is underweighted in OPS and to a lesser degree in RC.

    ***

    In  any case, no need to use any formulation of OPS anyway.  Just use wOBA.  OPS is just a bridge.

    ***

    Let me know if you have any other questions.
    avatar
    Son of Ron
    Al Smith's Beer Bath

    Posts : 181
    Join date : 2018-10-17

    Re: MEET TOM TANGO

    Post by Son of Ron on Fri Nov 30, 2018 8:08 am

    tangotiger wrote:Perusing through here, it seems there are two claims or questions being made:
    1. Why 1.8
    2. Is the 1.8 inconsistent when you compare to Bill's RC

    First point: 1.7 is described here:
    www
    .
    insidethebook
    .
    com/ee/index
    .
    php/site/comments/why_does_17obpslg_make_sense/

    (Note: I had to split the above because this site is blocking newbies from posting links.)

    As for second point: I don't see any evidence  this is true.  PROBABLY RC is consistent with a 1.3 or so multiplier. But that would just point out the shortcomings of RC.  That's  because the walk is underweighted in OPS and to a lesser degree in RC.

    ***

    In  any case, no need to use any formulation of OPS anyway.  Just use wOBA.  OPS is just a bridge.

    ***

    Let me know if you have any other questions.



    Mr Tango,


    Here are my questions.
     
    1) - In simple layman's terms, how was the 1.7 or 1.8 determined?  According to my research, this number has been calculated by comparing the coefficients of a linear regression where OBP and SLG are the independent variables.
    2) - Assuming #1 is true, which I believe is a safe assumption based on a 2003 internet conversation you had where you specifically used such a linear model, what about multicolinearity?
    3) - In that same conversation, your also ran a regression that added batting average along with OBP and SLG.  The coefficent for batting average came up as negative.  What was your basis for dismissing that negative coeefficient as offensive context?  My math background leads me to believe that it was multicolinearity, just like the others in your 2003 internet conversation
    4) - In your reply above, you claim you see NO EVIDENCE that 1.8 is inconsistant with Bill James RC.  However, based on another internet conversation you have had, I believe in 2007, you should be well aware that the basic Bill James RC formula can be rewritten as AB * OBP * SLG.  I am sure you are also quite aware that this simple formula is quite accurate regardless of the season.  For 2018, the biggest misses at */- 5%.  Most are within 1 or 2%.  So how can OBP be 1.8 of SLG when Bill James RC clearly demonstrates they are roughly equal?
    5) - Considering that hits are about 80% of OBP and about 60% of SLG, what is the point of examining a point of OBP vs a point of SLG when the underlying source is usually the same event?  In other words, what is the practical application to this excercise?
    6) - I know Tom Tango is a pseudonym design to protect your privacy.  Can you at least tell us your highest level of education as well your major in college and any post graduate work?


    Thank you in advance for taking the time to come to this board and answering our questions.


    Sincerely,
    Mark the CPA
    Husband to Achara
    Father of Stanley and Megan
    Owner of Patsy the Dog
    Class of 1987 from the University of Illinois at Champaign Urbana
    avatar
    Son of Ron
    Al Smith's Beer Bath

    Posts : 181
    Join date : 2018-10-17

    Re: MEET TOM TANGO

    Post by Son of Ron on Fri Nov 30, 2018 8:12 am

    Soxillinirob wrote:
    tangotiger wrote:I'd be happy to answer whatever questions you have. However, based on the tone in this thread, I would simply ask that if you guys can holster your weapons while I'm here, then I can stick around as long as you want me here.  I'm not interested in anything other than advancing baseball knowledge.

    Agreed?
     
    OMG....please be the actual Tom Tango!  I would love to see the analysis.  Please proceed.  If you meet any unfriendly resistance it'll only be from one person in particular.

    QUICK


    SOMEBODY BLOCK ROSEBOWL!!!!!
    avatar
    Son of Ron
    Al Smith's Beer Bath

    Posts : 181
    Join date : 2018-10-17

    Re: MEET TOM TANGO

    Post by Son of Ron on Fri Nov 30, 2018 8:44 am

    tangotiger wrote:I'd be happy to answer whatever questions you have. However, based on the tone in this thread, I would simply ask that if you guys can holster your weapons while I'm here, then I can stick around as long as you want me here.  I'm not interested in anything other than advancing baseball knowledge.

    Agreed?


    Mr Tango,


    In the interest of advancing baseball knowledge, are you familiar with the research of Craig Wright, specifically the topics covered in his brilliant book "Diamond Appraised"  If so, are you aware of any other researcher that would have confirmed or denied his research?  As a researcher yourself, I'm sure you are quite aware of the importance of replication.


    On a similiar note, I recall John Dewan doing a pitching study for HardBall Times where he concluded that light pitching workloads early in the season have a negative effect on pitching performance later in the season.  Presumably due lack of endurance.  Have you done any research in the area of workloads and pitching performance?


    The reason I ask, OUR White Sox lost Michael Kopech to an arm injury.  To me, Kopech's 2018 workload in AAA and the MLB was as light as can be.  Yet he still got hurt.  As a 53 year old live long baseball fan, and an avid reader of Bill James during the 1980's and 90's, it seems to me that light workloads do nothing to prevent arm injuries.  If the above mention Dewan study is accurate, I would think they might even be detrimental.  so again, any thoughts or research on a topic that I would think would be of vital interest to every MLB ballclub.


    Thanks in advance for your input


    Mark the CPA

    tangotiger
    Joe Borchard's Teammate

    Posts : 10
    Join date : 2018-11-29

    Re: MEET TOM TANGO

    Post by tangotiger on Fri Nov 30, 2018 9:19 am

    1) - In simple layman's terms, how was the 1.7 or 1.8 determined?  According to my research, this number has been calculated by comparing the coefficients of a linear regression where OBP and SLG are the independent variables.


    I provided a link that went through it step by step.  Please read that first.


    2) - Assuming #1 is true, which I believe is a safe assumption based on a 2003 internet conversation you had where you specifically used such a linear model, what about multicolinearity?


    It's not that OBP and SLG are independent, but each of  the components are independent.  Again, please see above.




    3) - In that same conversation, your also ran a regression that added batting average along with OBP and SLG.  The coefficent for batting average came up as negative.  What was your basis for dismissing that negative coeefficient as offensive context?  My math background leads me to believe that it was multicolinearity, just like the others in your 2003 internet conversation



    The  batting average has a negative or positive relationship based on the profile of your  OBP and SLG.  So if  you want to create a metric where the coefficient of the BA changes (and is plus or minus) based on OBP and SLG, you won't find that appeal  to anyone.




    But, more importantly, there is NO REASON to rely on OBP and SLG, since you can rework  the components that make up OBP and SLG into wOBA. 




    So, I have moved on from OBP and SLG about 13-15 years ago.  We don't need it.




    4) - In your reply above, you claim you see NO EVIDENCE that 1.8 is inconsistant with Bill James RC.  However, based on another internet conversation you have had, I believe in 2007, you should be well aware that the basic Bill James RC formula can be rewritten as AB * OBP * SLG.  I am sure you are also quite aware that this simple formula is quite accurate regardless of the season.  For 2018, the biggest misses at */- 5%.  Most are within 1 or 2%.  So how can OBP be 1.8 of SLG when Bill James RC clearly demonstrates they are roughly equal?




    What applies for OBP PLUS SLG is not the same thing as OBP TIMES SLG.  Bill's RC is closer to 1.3*OBP + SLG.




    5) - Considering that hits are about 80% of OBP and about 60% of SLG, what is the point of examining a point of OBP vs a point of SLG when the underlying source is usually the same event?  In other words, what is the practical application to this excercise?



    There isn't any!  I am begging people every chance  I get to stop using OPS in any form whatsoever.  In The Book, I specifically said after we devoted one paragraphs to OPS and  1.8OBP+SLG that we would never talk about OPS in The Book. And we didn't.




    6) - I know Tom Tango is a pseudonym design to protect your privacy.  Can you at least tell us your highest level of education as well your major in college and any post graduate work?

    I let the work speak for me.

    tangotiger
    Joe Borchard's Teammate

    Posts : 10
    Join date : 2018-11-29

    Re: MEET TOM TANGO

    Post by tangotiger on Fri Nov 30, 2018 9:24 am

    With  regards to workload: we did something about workloads in The Book, both for starting pitchers and relievers.  But our focus was on careers rather than year to year.


    Yes, I read Craig's book, and I don't remember what John wrote.


    Regardless, there is plenty of room for aspiring saberists to research this issue, especially as "expected" workloads change year to year.


    There is the idea that you pitch relative to what you train for.  So, pitching on 3 days rest would be terrible today, but ok 50 years ago, because that's how they create their schedule.  In other words, you have to be really careful how you interpret whatever results you end up finding.

    tangotiger
    Joe Borchard's Teammate

    Posts : 10
    Join date : 2018-11-29

    Re: MEET TOM TANGO

    Post by tangotiger on Fri Nov 30, 2018 9:26 am

    It  would be super helpful if I can provide links, or, if one of you guys can simply create a link for this to start with (just put everything in one line):


    www
    .
    insidethebook
    .
    com/ee/index
    .
    php/site/comments/why_does_17obpslg_make_sense/



    Or, change the three ! into a .

    www!insidethebook!com/ee/index!php/site/comments/why_does_17obpslg_make_sense/
    avatar
    Son of Ron
    Al Smith's Beer Bath

    Posts : 181
    Join date : 2018-10-17

    Re: MEET TOM TANGO

    Post by Son of Ron on Fri Nov 30, 2018 10:41 am

    Mr Tango,


    I assume this is the link you are trying to post.


    http://www.insidethebook.com/ee/index.php/site/comments/why_does_17obpslg_make_sense/




    I must apologize for wasting your time.  I do not know how you found us.  But here is a brief history of this board's 1.8 debate.  One person here refuses to acknowledges concepts taught in a typical Stats 101 course.  That same person than completely butchered explaining your research, as well as the research of other sabermeticians.  After I displayed all the flaws in his explanations, well things just got nasty.


    There are a couple of things I still do not understand:


    1)  You are telling me to read something titled "Why 1.7 OBP + SLG makes Sense"  but now you are telling me to forget OBP and SLG all together.  That seems a bit contradictory.


    2)  After reading the link, it appears 1.7 is the result of comparing the regression coefficients of OBP and SLG.  However, this result cannot be trusted due to multicolinearity.  You state you avoid this problem by by using the individual components of hits and walks and homers etc etc etc...   I agree is this the correct why to run a regression, however by doing so, you eliminate the coeefficients for OBP and SLG therefore you have nothing to compare, so 1.7 disappears.  Is that not correct.


    Don't feel obliged to answer.  After all, you just told me you abandoned this line of thinking over a decade ago.  So why dig up a dead horse just to beat it.  I'm sure you have better things to do than teach Stats 101 to a pair of middle aged men.


    Thank You for gracing this board with your presence


    Sincerely,
    Mark the CPA

    tangotiger
    Joe Borchard's Teammate

    Posts : 10
    Join date : 2018-11-29

    Re: MEET TOM TANGO

    Post by tangotiger on Fri Nov 30, 2018 11:01 am

    you have nothing to compare


    Actually, in that post I show I am comparing to Linear Weights.



    but now you are telling me to forget OBP and SLG all together.  That seems a bit contradictory.

    For those who INSIST on using some form of OPS, then 1.7 or  1.8 * OBP plus SLG is necessary to move one step forward,  and put OPS without any coefficient in their rear-view mirror.


    But once you realize how limiting it is to use OBP and SLG in ANY form (added, multiplied, coefficient, whatever), then save yourself, and everyone around you, by no longer talking about it!  That's why we  have wOBA.
    avatar
    Son of Ron
    Al Smith's Beer Bath

    Posts : 181
    Join date : 2018-10-17

    Re: MEET TOM TANGO

    Post by Son of Ron on Fri Nov 30, 2018 11:10 am

    tangotiger wrote:With  regards to workload: we did something about workloads in The Book, both for starting pitchers and relievers.  But our focus was on careers rather than year to year.


    Yes, I read Craig's book, and I don't remember what John wrote.


    Regardless, there is plenty of room for aspiring saberists to research this issue, especially as "expected" workloads change year to year.


    There is the idea that you pitch relative to what you train for.  So, pitching on 3 days rest would be terrible today, but ok 50 years ago, because that's how they create their schedule.  In other words, you have to be really careful how you interpret whatever results you end up finding.


    So are you implying that getting back to a 4 man rotation on 3 days rest is possible if the MLB would make a conscience effort as a whole to gradually increase the training and the workload over what ever years necessary?


    To me, this makes more sense than the alternative.


    And why hasn't MLB expanded the rosters?  Are they waiting for the only man on the bench to be the backup catcher?

    tangotiger
    Joe Borchard's Teammate

    Posts : 10
    Join date : 2018-11-29

    Re: MEET TOM TANGO

    Post by tangotiger on Fri Nov 30, 2018 11:24 am

    Right, I would say it would at least be a consideration for any study.  But given where we are, and how effective relief pitchers are, the trend is clear as to direction for workload.


    As for roster-size: I don't have any opinion on that.
    avatar
    Son of Ron
    Al Smith's Beer Bath

    Posts : 181
    Join date : 2018-10-17

    Re: MEET TOM TANGO

    Post by Son of Ron on Fri Nov 30, 2018 11:37 am

    tangotiger wrote:
    you have nothing to compare


    Actually, in that post I show I am comparing to Linear Weights.



    but now you are telling me to forget OBP and SLG all together.  That seems a bit contradictory.

    For those who INSIST on using some form of OPS, then 1.7 or  1.8 * OBP plus SLG is necessary to move one step forward,  and put OPS without any coefficient in their rear-view mirror.


    But once you realize how limiting it is to use OBP and SLG in ANY form (added, multiplied, coefficient, whatever), then save yourself, and everyone around you, by no longer talking about it!  That's why we  have wOBA.


    I disagree.


    The classic Bill James runs created formula as expressed by Runs = AB * OBP * SLG works well enough.  Besides, this is baseball.  Not putting a man on Mars.  So 98% accuracy is more than good enough.  Plus the formula never changes.  So I don't have to rerun coefficients constantly throughout the season.  Plus what conclusions can be made with wOBA that can't be made with RC?  RC has Avisail Garcia creating 43.8 runs in 2018.  While some metric that gets tweeked every year to gain an extra point of accuracy may produce a different result, I doubt that is would be so different as to change the opinion of somebody judging him on the RC result.




    Anotherthing, According to Fangraphs,


    wOBA = (0.690×uBB + 0.722×HBP + 0.888×1B + 1.271×2B + 1.616×3B +
    2.101×HR) / (AB + BB – IBB + SF + HBP)



    I don't see any adjustment for park effect our league context.  So wOBA has the same significant shortcoming as RC.  However, I would think easier to adjust RC.  Simply multiply it by the park factor.  Not sure if it is that simple with wOBA.


    PS, classic runs created doesn't use coefficients so they are already in the rear view mirror.
    avatar
    Jack Brickhouse
    Silent Hot Dog Vendor

    Posts : 1044
    Join date : 2014-04-03

    Re: MEET TOM TANGO

    Post by Jack Brickhouse on Fri Nov 30, 2018 11:48 am

    What does NVICH think??

    tangotiger
    Joe Borchard's Teammate

    Posts : 10
    Join date : 2018-11-29

    Re: MEET TOM TANGO

    Post by tangotiger on Fri Nov 30, 2018 12:27 pm

    Not looking to argue here, nor is my goal to change anyone's mind.  I'm expressing what I've done, and what I've learned.  If that helps someone, great.  If not, that's fine.  We're all on the same team.


    If you have a question, I'm happy to help.
    avatar
    Son of Ron
    Al Smith's Beer Bath

    Posts : 181
    Join date : 2018-10-17

    Re: MEET TOM TANGO

    Post by Son of Ron on Fri Nov 30, 2018 12:51 pm

    Mr Tango,


    On behalf of this board, I must again apologize for wasting your time.  What you are unaware of is that one of the board members, I assume the one that invited you, has been misrepresenting your work as well as the work of others for years.  Not only that, that person doesn't understand basic statistics.  He cannot calculate a simply correlation coefficient.  The concept of multicolinearity is clearly beyond his comprehension.  Instead, he acts like these problems don't exist because he read your book.  He fails to understand your book is about baseball, its not a Stats 101 textbook.  It is unfair to you to expect a free Stats 101 tutorial him.  Also my questions to you were unfair because they were really meant for another person.  To put it another way, Ralph Kramden dragged you into a pissing contest hoping you'll tell Alice what a dingbat her mother is, and vice versa.  Google Honeymooners if you don't get the reference.


    I'm sure you have better things to do with your time then explain research that was abandoned over a decade ago to two guys that are determined to prove who has the bigger slide rule.


    Thank You again for visiting our board.  And thank you for making me realize I'd be wasting my time writing an article for the SABR research journal about multicolinearity since it is now obvious that it is only my fellow board member that doesn't get it.


    Enjoy Arizona this March.  Once the kids are grown, I hope to catch one of these conferences as well as some White Sox games.


    Sincerely,
    Mark the CPA

    tangotiger
    Joe Borchard's Teammate

    Posts : 10
    Join date : 2018-11-29

    Re: MEET TOM TANGO

    Post by tangotiger on Fri Nov 30, 2018 1:59 pm

    Again, not really interested in all the background and trying to prove a winner.  I'd rather that no one uses me or my work as support for their position.  If both sides don't win, then whatever I've done is a waste.  What I do is about moving forward together,  and not leaving people behind.  (Unless they are determined to go their own way.)


    And no one invited me, I just stumbled onto this.


    Have fun guys!
    avatar
    Son of Ron
    Al Smith's Beer Bath

    Posts : 181
    Join date : 2018-10-17

    Re: MEET TOM TANGO

    Post by Son of Ron on Fri Nov 30, 2018 2:48 pm

    tangotiger wrote:Again, not really interested in all the background and trying to prove a winner.  I'd rather that no one uses me or my work as support for their position.  If both sides don't win, then whatever I've done is a waste.  What I do is about moving forward together,  and not leaving people behind.  (Unless they are determined to go their own way.)


    And no one invited me, I just stumbled onto this.


    Have fun guys!


    Really?!?!?!?!?!?


    That's amazing that someone of your statue found our obscure little board.


    If only you had found us 10 years ago....


    Anyway, incredibly cool of you to go out of your way to try and help.

    Sponsored content

    Re: MEET TOM TANGO

    Post by Sponsored content


      Current date/time is Mon Dec 10, 2018 12:11 am