Nomadsfest Sox Fans

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


    Jose Abreau

    Share
    avatar
    rmapasad
    Chairman Reinsdorf

    Posts : 2645
    Join date : 2009-04-06
    Location : Northridge, CA

    Re: Jose Abreau

    Post by rmapasad on Fri Dec 01, 2017 12:28 pm

    For shits and giggles, I decided to plug this into the basic runs created formula.  Player A creates 110  Player B creates 103.. so I guess that proves another one of my points that its more accurate to multiply than to add OBP and SLG
    Which reminds me of another one of your lies when you completely fabricated that no serious researcher would do such a thing.  Well no serious mathematician would put OBP and SLG in the same linear regression.  And no serious analyst would claim a point of OBP is 1.8 more valuable than a point of SLG.>>
    That doesn't prove that "it's more accurate to multiply than to add OBP + Slug" because you didn't compare that same result to 1.8 * OBP + Slug.  You would find that the guy with the higher OBP also came out more valuable using that method. 
    I will take your word for it that you got that RC result.  So now it is time for you to take my word for something.  I pulled the league numbers of last 20 years from BR, put them into Excel, then calculated the 20 year average on Runs, OBP, Slug, OPS, OBP*1.8+ Slug, and OBP* Slug. Calculated each season's STDEV for all categories against their 20 yr Avg. and then looked to see which of the categories STDEV most closely resembled Runs STDEV in every season.   Bottom line:  
    the winner of the most years' resemblance was 1.8*OBP.   It better resembled Runs in 7 seasons, followed by raw OBP at 5 seasons, OPS at 4 seasons, raw Slug at 3 seasons, then OBP*Slug 1 season. So much for "multiplying is superior to adding.". 
    However, to be completely fair when the entire 20 season STDEV's were averaged for all categories 1.8 * OBP and OBP* Slug finished in a virtual tie at .115 average deviation from Runs.  Raw OPS came in at .14 and raw OBP at .26, Slug at .29. While OBP and Slug were better correlators in some seasons, they went completely off the rails in the other seasons.  2017 being the perfect example where OBP had a .74 correlation and Slug .55.

    I have had enough of this discussion and so has the Board. I have shown that the people who came up with the 1.8 theory are NOT full of shit as you keep asserting. Both in looking at Run Expectancy proof and 
    now actual seasons' results.  
    I am no longer interested in running numbers ad infinitum in an effort to convince someone who never was going to be objective about this subject anyway because his only interest is trying to "embarrass" or "expose" me by whatever contortions possible.  There is a reason that at least a dozen research papers have confirmed that the 1.7  or 1.8 theory is valid. Because it is.

    But I will end by conceding one important thing to you though. While the 1.8 theory is valid it didn't add much to the party.  It is not significantly more accurate than raw OPS on a season by season basis.  While 1.8 did kick OBP*Slug's ass by beating it 13 to 7 seasons in head to head comparisons, on the whole there was hardly any difference over the entire 20 seasons.
    This is why, like so many other things including James' old RC, 1.8 theory yielded to a slightly better method and that today is w/OBA.   
    I see no reason to continue this unless you have a substantive real point to make.  If it's just about more insults or accusations, then go pound sand. 
    avatar
    rmapasad
    Chairman Reinsdorf

    Posts : 2645
    Join date : 2009-04-06
    Location : Northridge, CA

    Re: Jose Abreau

    Post by rmapasad on Fri Dec 01, 2017 12:33 pm

    alohafri wrote:I am. Don't you know that a HR is a rally killer?

    Not sure where you;'re going with that Kevin.  HR's are great.  HR's are greatest when guys are aboard though, and that's why OBP means something. Outs are the rally killers. 
    avatar
    Soxillinirob
    Chairman Reinsdorf

    Posts : 7649
    Join date : 2009-04-05
    Age : 52
    Location : St. Charles, IL

    Re: Jose Abreau

    Post by Soxillinirob on Fri Dec 01, 2017 12:41 pm

    Lol...Hawk used say HRs were rally killers, Roger. That is where Aloha is going with it.
    avatar
    rmapasad
    Chairman Reinsdorf

    Posts : 2645
    Join date : 2009-04-06
    Location : Northridge, CA

    Re: Jose Abreau

    Post by rmapasad on Fri Dec 01, 2017 1:00 pm

    So what this suggests is that when there are more outs and more runners on base, slugging is actually more valuable than simply getting on base.>
    OK, this is a substantive point and you are correct. With runners in SP , first base unoccupied Slug trumps pure OBP in RE value PER PT OF INCREASE..  But also note from the frequency tables, those are much less frequent situations.  O on, 0 out happens most often (25%).  57% of all AB's are with no one on base.  The next most frequent is runner on 1st (18%) but even in those cases, usually only  HR's or Triples with 1 or 2 outs mean that Slug on a per pt. basis will trump OBP ..Unless doubles score the runner from 1st.
    Remember this isn't about whether an extra base hit is better than a BB.
    It always is of course. It's whether the RE value to Slug on a per pt. basis for the EB hit is out of proportion to the reward for an extra pt. of OBP in that particular scenario.  In the most frequent scenarios it is, sometimes by as much as 2-3 times. When you combine these RE values with the frequencies of each situation, that is where the 1.7 or 1.8 number comes out.
    avatar
    rmapasad
    Chairman Reinsdorf

    Posts : 2645
    Join date : 2009-04-06
    Location : Northridge, CA

    Re: Jose Abreau

    Post by rmapasad on Fri Dec 01, 2017 1:18 pm

    Unless the Sox either a) throw $ 65-70 million at Lorenzo Cain in hopes that he will still be decent at ages 34-35 when the Sox contend which is a bad gamble  or  b) pull off a trade to get a CFer who will still be under age 32 by 2020-2021, then they will be starting a guy in CF in 2018 that no contender would start.  
    The only reason Leury would be better than Engel is that Leury is out of options where Engel might not be.  Whether better defense might be better for the younger pitchers is open to discussion.  In one sense it might not be.  I'm not saying this to "deliberately disagree" but just looking at it practically. Encouraging young pitchers to leave balls out over the plate because the CFer will chase down the long flies anyway isn't a good thing.  Especially if the great defense CFer won't be permanent because he can't hit enough.  The ballclub eventually has to find a CFer who can contribute most overall not just catch the most flyballs.  And pitchers have to realize that is the CFer they have to live with.  
    avatar
    Deplorable Mark
    Roof Shot

    Posts : 2357
    Join date : 2016-09-16

    Re: Jose Abreau

    Post by Deplorable Mark on Fri Dec 01, 2017 3:57 pm

    rmapasad wrote:For shits and giggles, I decided to plug this into the basic runs created formula.  Player A creates 110  Player B creates 103.. so I guess that proves another one of my points that its more accurate to multiply than to add OBP and SLG
    Which reminds me of another one of your lies when you completely fabricated that no serious researcher would do such a thing.  Well no serious mathematician would put OBP and SLG in the same linear regression.  And no serious analyst would claim a point of OBP is 1.8 more valuable than a point of SLG.>>
    That doesn't prove that "it's more accurate to multiply than to add OBP + Slug" because you didn't compare that same result to 1.8 * OBP + Slug.  You would find that the guy with the higher OBP also came out more valuable using that method. 
    I will take your word for it that you got that RC result.  So now it is time for you to take my word for something.  I pulled the league numbers of last 20 years from BR, put them into Excel, then calculated the 20 year average on Runs, OBP, Slug, OPS, OBP*1.8+ Slug, and OBP* Slug. Calculated each season's STDEV for all categories against their 20 yr Avg. and then looked to see which of the categories STDEV most closely resembled Runs STDEV in every season.   Bottom line:  
    the winner of the most years' resemblance was 1.8*OBP.   It better resembled Runs in 7 seasons, followed by raw OBP at 5 seasons, OPS at 4 seasons, raw Slug at 3 seasons, then OBP*Slug 1 season. So much for "multiplying is superior to adding.". 
    However, to be completely fair when the entire 20 season STDEV's were averaged for all categories 1.8 * OBP and OBP* Slug finished in a virtual tie at .115 average deviation from Runs.  Raw OPS came in at .14 and raw OBP at .26, Slug at .29. While OBP and Slug were better correlators in some seasons, they went completely off the rails in the other seasons.  2017 being the perfect example where OBP had a .74 correlation and Slug .55.

    I have had enough of this discussion and so has the Board. I have shown that the people who came up with the 1.8 theory are NOT full of shit as you keep asserting.

    and I have shown the people have used linear regression to come up with 1.8 and then I showed how to reproduce the multicolinearity and your responds was to insult me

    Both in looking at Run Expectancy proof and 
    now actual seasons' results.  

    You have shown nothing of the sort.  You pointed out one scenario out of 24 then jump to a conclusion that nobody can see.  If you have it on a spreadhseet, then email it to Karko05@aol.com.  But again, nowhere in the decade have you shown where 1.8 comes from.  all you examples and links that attempted it, show its all over the place.

    BTW, runs created shows they are the same.  Your claim runs created has been discredited is false and without merit.  While some may have developed other formulas for runs which are supposedly more accurate, it would not account for an 80% discrepency.


    I am no longer interested in running numbers ad infinitum in an effort to convince someone who never was going to be objective about this subject anyway

    I am very objective.  Again, what you have posted has never ever shown a clear path to 1.8.  At best, you give part of the equation which has everybody scratching their heads then jump to a conclusion completely disconnected from your example

    because his only interest is trying to "embarrass" or "expose" me by whatever contortions possible.  There is a reason that at least a dozen research papers have confirmed that the 1.7  or 1.8 theory is valid. Because it is.

    Wrong again.  It proves nothing.  My purpose is to expose bad math because bad math ruins the legitimacy of the movement.  My research has shown that OBP is more or less equal to SLG.  Research confirmed by the works of Pete Palmer and Bill James among others.  I then showed the 1.8 is actually the colinearity between OBP and SLG.  It is you that refused to objectively review that because it is you who are determined to prove me wrong because you are unable to admit I'm right.

    But I will end by conceding one important thing to you though. While the 1.8 theory is valid it didn't add much to the party. 

    duh

    It is not significantly more accurate than raw OPS on a season by season basis.  While 1.8 did kick OBP*Slug's ass by beating it 13 to 7 seasons in head to head comparisons, on the whole there was hardly any difference over the entire 20 seasons.

    and just how in the hell does 1.8 kick the ass of OBP * SLG?!?!?

    This is why, like so many other things including James' old RC, 1.8 theory yielded to a slightly better method and that today is w/OBA.

    Really?!?!?!?!!?  Here is wOBA





      
    Please show how 1.8 influenced this.  I don't see 1.8, I don't see SLG, I don't see OBP.  Email a link to Karko05@aol.com if its easier

    I see no reason to continue this unless you have a substantive real point to make.  If it's just about more insults or accusations, then go pound sand. 


    and as long as you continue to be a smug SOB who clearly invents things that I know are patently false, I will continue to nail your ass to the wall on it.


    BTW, I was saving this for the SABR you were suppose to find, but since you are too much of a coward to do that, this is what is really happening with 1.8


    2017, 42,215 Hits, 70,517 total Bases, the difference is 28,302 extra bases on hits. 15,829 walks, 1763 HBP, the sum is 17,592 bases without a hit.  Divide 28,302 by 17,592 you get 1.61.  Ten years ago, you claimed the range was 1.4 - 1.8.  So 2017 is smack dab in the middle.  So there is your multicolinearity.


    This is why OBP is roughly the same as SLG.  When you try to force these two into the same linear equation, which many of your researchers clearly did, you need to weight OBP by 1.4 to 1.8 in order to get it to equal SLG in order to account for the difference in the under lying bases.  That's bases as in 1st, 2nd 3rd, not me misspelling basis.


    This also explains why your 1.8 didn't add much to the party as you have just admitted.


    Now can you reply to this in an adult like manner, or will you resort to you usual insults and fabrications
    avatar
    Deplorable Mark
    Roof Shot

    Posts : 2357
    Join date : 2016-09-16

    Re: Jose Abreau

    Post by Deplorable Mark on Fri Dec 01, 2017 4:10 pm

    rmapasad wrote:Unless the Sox either a) throw $ 65-70 million at Lorenzo Cain in hopes that he will still be decent at ages 34-35 when the Sox contend which is a bad gamble  or  b) pull off a trade to get a CFer who will still be under age 32 by 2020-2021, then they will be starting a guy in CF in 2018 that no contender would start.  
    The only reason Leury would be better than Engel is that Leury is out of options where Engel might not be.  Whether better defense might be better for the younger pitchers is open to discussion.  In one sense it might not be.  I'm not saying this to "deliberately disagree" but just looking at it practically. Encouraging young pitchers to leave balls out over the plate because the CFer will chase down the long flies anyway isn't a good thing.  Especially if the great defense CFer won't be permanent because he can't hit enough.  The ballclub eventually has to find a CFer who can contribute most overall not just catch the most flyballs.  And pitchers have to realize that is the CFer they have to live with.  


    In one sense it might not be.  I'm not saying this to "deliberately disagree" but just looking at it practically. Encouraging young pitchers to leave balls out over the plate because the CFer will chase down the long flies anyway isn't a good thing.


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


    Excuse, but you do realize its a team game dontcha?!?!?!?!


    and having a young pitcher think he has to strike out a hitter everytime he is in a pinch because he can't count on his fielders is an even worse thing.


    Like you said, no real contender would be starting Leury or Engel.  the way the team is currently constructed, Engel makes more sense because of his clearly superior defense.


    And what's the point of gambling on Leury being the the 2018 Garcia when you probably ended up calling him a BABIP fluke if he is.


    A fluke 2018 for Leury does nothing for 2020.  Getting Engel at least 700 IP in CF definitely helps develop the pitching staff because contrary to the stats, pitching and fielding are suppose to work together, no not independently


    Last edited by Deplorable Mark on Fri Dec 01, 2017 4:12 pm; edited 1 time in total
    avatar
    Deplorable Mark
    Roof Shot

    Posts : 2357
    Join date : 2016-09-16

    Re: Jose Abreau

    Post by Deplorable Mark on Fri Dec 01, 2017 4:11 pm

    Soxillinirob wrote:Lol...Hawk used say HRs were rally killers, Roger.  That is where Aloha is going with it.


    Only the Chicago White Sox could turn a home run into a bad thing

    No
    avatar
    rmapasad
    Chairman Reinsdorf

    Posts : 2645
    Join date : 2009-04-06
    Location : Northridge, CA

    Re: Jose Abreau

    Post by rmapasad on Fri Dec 01, 2017 6:22 pm

    2017, 42,215 Hits, 70,517 total Bases, the difference is 28,302 extra bases on hits. 15,829 walks, 1763 HBP, the sum is 17,592 bases without a hit.  Divide 28,302 by 17,592 you get 1.61.  Ten years ago, you claimed the range was 1.4 - 1.8.  So 2017 is smack dab in the middle.  So there is your multicolinearity.


    This is why OBP is roughly the same as SLG.  When you try to force these two into the same linear equation, which many of your researchers clearly did, you need to weight OBP by 1.4 to 1.8 in order to get it to equal SLG in order to account for the difference in the under lying bases.  That's bases as in 1st, 2nd 3rd, not me misspelling basis.
    This also explains why your 1.8 didn't add much to the party as you have just admitted.
    Now can you reply to this in an adult like manner, or will you resort to you usual insults and fabrications>>
    OK, I just emailed you a spreadsheet and you can see for yourself.

    First of all, as to your (TB-hit)/(BB+HBP) formula, the results from 2017-2016 were in 1.60 range but over a 20 year cycle the average was 1.48. 
    Second, I guess you have to be saying after removing the multicolinearity (which would be hits or BA portion portion shared by both OBP and Slug) then the  "independent" part of Slug (extra bases from EB hits) and the "independent" portion of OBP (BB+HBP) have a 1.6 (or actually 1.48 ratio).  OK, but that only accounts for the bases gained comp between EB hits v. BB/HBP. 
    An important component that can be gleaned by using Run Expectancy tables is the out avoidance aspect of OBP.   The cost of an out (when GDP is included) is about .30 runs.   So add .30 to 1.48 = 1.78.   I'm not saying this is another validation of the theory because it might be co-incidence. But it is some food for thought.

    Third, as you can see on the spreadsheet, I've also included your category of (TB-hits)/(BB+HBP) and comped that to runs.  There is no correlation.  BUT there is an equal correlation of Runs between OBP*1.8  and  OBP * OPS over the 20 year span.  Their variances to Runs parallel each other closely year after year. 
    Here's the thing though.    OBP *1.8 wins out in far more seasons ( 7 to 1) in correlating to Runs than OBP*OPS even though in many cases OBP*1.8 wins by a narrow margin.  So maybe this is one reason the 1.8 theory got so much currency for a while.  It was more accurate on a year to year basis.
    In any event, I think this horse has been beaten to 12 deaths by now.  I am always open to hearing whatever flaw I've had in reasoning/calculation but like anyone else (including yourself) it's best digested when delivered calmly and without undeserved personal insult.   
    avatar
    rmapasad
    Chairman Reinsdorf

    Posts : 2645
    Join date : 2009-04-06
    Location : Northridge, CA

    Re: Jose Abreau

    Post by rmapasad on Fri Dec 01, 2017 6:41 pm

    Really?!?!?!?!!?  Here is wOBA
    Please show how 1.8 influenced this.  I don't see 1.8, I don't see SLG, I don't see OBP.>
    [size=16]One more for the road.    [/size]
    "Well, unless you having nothing else available or need to calculate something quickly yourself, I don’t see the point in using OPS. (You can actually get an approximation of wOBA out of the three slash with by doing ((OBP x 1.7) + SLG)/3.) "
    Here's the link to the article
    https://www.fangraphs.com/blogs/avgobpslg-in-an-age-of-woba/

    avatar
    rmapasad
    Chairman Reinsdorf

    Posts : 2645
    Join date : 2009-04-06
    Location : Northridge, CA

    Re: Jose Abreau

    Post by rmapasad on Fri Dec 01, 2017 7:37 pm

    Deplorable Mark wrote:
    And what's the point of gambling on Leury being the the 2018 Garcia when you probably ended up calling him a BABIP fluke if he is.
    A fluke 2018 for Leury does nothing for 2020.  Getting Engel at least 700 IP in CF definitely helps develop the pitching staff because contrary to the stats, pitching and fielding are suppose to work together, no not independently

    Leury's fluke was his power in 2017.  Or was it a fluke ? Maybe he keeps up a 150-160 ISO and possibly turns in a .270/.320/.425 season by cutting the K's a bit and upping the BB's.
    Which could make him a decent reasonably priced utility guy who can play multiple positions. That's a very handy utility guy who can play multiple spots to have around for the next 3-5 years to sub for regulars who get hurt or need rest.  Not saying Leury will keep his power jump or that he will improve.  He could easily be a low-mid 600's guy who's a dime a dozen.  Then cast him loose after 2018.  But in the absence of better options, no harm in seeing what he can do in 2018.

    As to Engel, he really can't play in the majors no matter how good his glove if he struggles to hit his weight.  Out of 59 CFers last year he ranked 57th and that was with his defense included.
    Pitchers have to live with the fact that their lives will not be made easier by having a bunch of fielding geniuses behind them if those geniuses can't hit for shit.
    avatar
    Deplorable Mark
    Roof Shot

    Posts : 2357
    Join date : 2016-09-16

    Re: Jose Abreau

    Post by Deplorable Mark on Tue Dec 05, 2017 10:21 am

    First of all, as to your (TB-hit)/(BB+HBP) formula, the results from 2017-2016 were in 1.60 range but over a 20 year cycle the average was 1.48. 
    Second, I guess you have to be saying after removing the multicolinearity (which would be hits or BA portion portion shared by both OBP and Slug) then the  "independent" part of Slug (extra bases from EB hits) and the "independent" portion of OBP (BB+HBP) have a 1.6 (or actually 1.48 ratio).  OK, but that only accounts for the bases gained comp between EB hits v. BB/HBP. 
    An important component that can be gleaned by using Run Expectancy tables is the out avoidance aspect of OBP.   The cost of an out (when GDP is included) is about .30 runs.   So add .30 to 1.48 = 1.78.   I'm not saying this is another validation of the theory because it might be co-incidence. But it is some food for thought.


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


    WTF are you talking about




    Your run expectancy tables is expressed completely in runs and shows nothing in terms of OBP or SLG




    and your new obsession of yours with calling (TB - H)/(BB + HBP) my formula suggests to me that you are making it up as you go along.  It is not my formula.  It is not meant to be used as some kind of player evaluator or run estimator.  It simply shows the multicolinearity.




    As for your proclamation about wOBA can be approxomately estimated with 1.7OBP * SLG/3 also proves nothing.  OBP*SLG/2 does the same thing.  Not only that, wOBA does not give you total runs.  AB * OBP * SLG does.  In the past 20 years, this formula overstates runs by 2%, implying its 98% accurate.




    what is the wOBA figure?!?!?




    BTW, if wOBA is more accurate its due to the formula changing every year.  Like Bill James stated long ago, if I adjusted RC every year it would be as accurate as hell.  so 98% on a simple and consistant formula is damn good and certainly should not be discredited as you falsely claimed this summer

    Sponsored content

    Re: Jose Abreau

    Post by Sponsored content


      Current date/time is Sun Feb 18, 2018 10:13 pm