Loading ...
Sorry, an error occurred while loading the content.
 

Re: The Race Thing

Expand Messages
  • timmermant
    ... In that case, we should probably get rid of Errors and the distinction between Wild Pitches and Passed Balls (these are all the subjective judgments of
    Message 1 of 18 , Jan 9, 2002
      > The problem is that someone's race isn't a simple fact. It
      > is a subjective label based on subjective definitions. [snip]
      > If we had a table with race, the values that would be
      > included would just be someone's opinion.

      In that case, we should probably get rid of Errors and the
      distinction between Wild Pitches and Passed Balls (these are all the
      subjective judgments of scorekeepers). Also balls and strikes (except
      clear swings) are just the opinions of the umpires.

      > But I haven't heard anything to convince me that player race is
      > actually a piece of data.

      But don't the books/articles you cited verify that race is a piece of
      data worth examining?

      TT
    • tmasc
      French hockey players feel discriminated against, because a team, given the choice between 2 equal players, will choose the player who can speak the language.
      Message 2 of 18 , Jan 9, 2002
        French hockey players feel discriminated against, because a team,
        given the choice between 2 equal players, will choose the player who
        can speak the language. This is why the majority of European players
        are skill players. Why invest money and time for a 4-th line player
        from Europe when you can get him cheap and easy in Canada?

        Does this happen in baseball too? Maybe not now, but at some point
        sure. At what point was the average MLB white player at the same
        performance level as the MLB black player? 1970? 1980? 1990? At
        what point did racism stop being an issue?
      • Sean Lahman
        ... An error is defined as a play which the official scorer calls an error. We re not reporting errors based on what you think or what I think. But that s
        Message 3 of 18 , Jan 9, 2002
          timmermant wrote:
          > > The problem is that someone's race isn't a simple fact. It
          > > is a subjective label based on subjective definitions.
          > [snip]
          > > If we had a table with race, the values that would be
          > > included would just be someone's opinion.
          >
          > In that case, we should probably get rid of Errors and the
          > distinction between Wild Pitches and Passed Balls (these are
          > all the subjective judgments of scorekeepers). Also balls and
          > strikes (except clear swings) are just the opinions of the
          > umpires.

          An error is defined as a play which the official scorer calls an error.
          We're not reporting errors based on what you think or what I think. But
          that's exactly what is being proposed for players' race. If the
          Sporting News included player race in its annual guides, I wouldn't
          object to including it. I don't want to include a table based on one
          guy going through a bunch of old baseball cards and coming up with
          labels.

          >But don't the books/articles you cited verify that race is a
          >piece of data worth examining?

          Sure, but they all observe that you get different results depending on
          how you define it. TMASC raised an anaolgy in his email about French
          hockey players. The same issue could be raised about Spanish speaking
          ballplayers, some of whom could be considered white, some black, and
          some latino. It's a quagmire of definitional uncertainty.

          I've raised some players' names as rhetorical examples, but let's hear
          some answers. How would you guys label Davey Lopes, Derek Jeter, David
          Justice, Tino Martinez, Rod Carew, Randall Simon, and Bernie Williams?
          If there's unanimity on those seven players, then maybe I'm wrong.

          --S
        • micke.hovmoller@stockholmsborsen.se
          ... As someone else said, biology is more and more avoiding the term race becuase of the scientific difficulties of separating races (whatever that is,
          Message 4 of 18 , Jan 9, 2002
            On 2002-01-09 18:02:26 tmasc wrote:

            >Let's talk about race. From what I remember reading in high school,
            >there are 3 races of people: black, white, and asian (I think the
            >technical terms were negroids, caucasoids, and ???).

            As someone else said, biology is more and more avoiding the term "race" becuase
            of the scientific difficulties of separating "races" (whatever that is, anyway,
            and they deal mostly with plants and animals and not humans, so it's not a PC
            thing, either). And if biologists doesn't want to use it because it lacks
            scientific meaning, maybe that is a hint to the rest of us.

            Still, I understand what you are trying to get at. There were loads of rumors
            when black players entered the majors that they didn't have the "toughness" of
            whites, so pitching them inside or hitting them with pitches would scare them.
            So, great study idea: where black players more hit by pitches or is this just a
            myth?

            The problem here is that the prejudice that was prevalent was based on
            perception, not on race per se (since that is unknown/undefined in a large
            number of cases). So what you really want is a "perception meter", really. And
            how do you do that? Scan a zillion baseball cards and give everyone a skin tone
            score? Ask the scouts or other players?

            /Micke
          • micke.hovmoller@stockholmsborsen.se
            ... But the problem isn t how interesting these reports would be, but rather their correctness. How do you classify each of the holy trinity of Nomar, ARod and
            Message 5 of 18 , Jan 9, 2002
              On 2002-01-09 17:14:55 tmasc wrote:

              >> Usually, I fall firmly in the camp of "more data." However, I don't
              >> really see the need for race in the database.
              >
              >I see the need, because there are at least 3 research reports I'd
              >like to run.

              But the problem isn't how interesting these reports would be, but rather their
              correctness. How do you classify each of the holy trinity of Nomar, ARod and
              Jeter? Why?

              The problem other people stated wasn't primarily one of political correctness,
              but one of actual correctness.

              >I know Americans
              >have no problems with language, but Canadians do. Canadians have no
              >problem with race, but Americans shy from talking about it. (I
              >know, I know, not ALL americans and not ALL Canadians.) Every
              >country has some cultural taboo. We don't have to shy away from this
              >as objective analysts.

              In Sweden we have a truly hard time translating "hispanics" since that is not
              usually a denomination/classification when we look at races. (Usually it is
              turned into "spansktalande" (spanish speaking) which I find awfully misleading.)

              My point, then, is this: what categories would you use? Why? I'd assume, from
              the preceding discussion something along the lines of
              white/black/hispanic/asian. If so, why lumping all white and not having "irish",
              "italian" etc? And where would that lead us?

              >We shouldn't decide on a piece of data makinig the database on
              >whether it will offend some group of person somewhere in the world.
              >We are census takers.

              True. As census takers, isn't it our first task to decide exactly what should be
              counted and how?

              /Micke
            • tmasc
              ... what should be ... Yes, no question. But half the responses here have been oh, let s not bother, cause it s too hard . In response to Sean L, and his
              Message 6 of 18 , Jan 9, 2002
                > True. As census takers, isn't it our first task to decide exactly
                what should be
                > counted and how?

                Yes, no question. But half the responses here have been "oh, let's
                not bother, cause it's too hard".

                In response to Sean L, and his list of Tiger-type players: I don't
                know. However, I would classify them in any way that any definition
                says that they should be classified. If the 1998 US Census
                definition says that Jeter is white, then he is white. If he is race
                17, then he is race 17. If the Swedish Census Bureau says that Jeter
                is black, then he is black. If they say that he is 50% white, 25%
                black, and 25% asian, and that makes him race 23, then he is race 23.

                I really don't care how we classify the player. The important thing
                is that we follow a definition, and that this defintion can be
                applied to everyone.

                The point is how far do we want to go? I don't think anyone here
                wants to go that far. I want to take the first step. Black, White,
                Latin, undetermined. Can't we do that?
              • timmermant
                ... error. But what s the correlation between officially called errors and the MLB definition of an error (An error shall be charged for each misplay
                Message 7 of 18 , Jan 9, 2002
                  >An error is defined as a play which the official scorer calls an
                  error.

                  But what's the correlation between "officially called errors" and the
                  MLB definition of an error (An error shall be charged for each
                  misplay (fumble, muff or wild throw) which prolongs the time at bat
                  of a batter or which prolongs the life of a runner, or which permits
                  a runner to advance one or more bases.)? If it's not a perfect
                  correlation, then we have the same problem as the measurement of
                  race. It doesn't matter who does the measurement, it's still a
                  judgment call.

                  Now we're getting at the heart of the issue (and my own reservations
                  about including it in the database). What is the _purpose_ of
                  recording errors? It is supposed to be a measure of fielding screw-
                  ups. It is measured by the subjective judgment of the official
                  scorer. Most plays are clearly errors or clearly non-errors. There
                  are a few close plays; on these, the scorekeeper may occasionally
                  call a "MLB" error an "Official" non-error, or call a "MLB" non-error
                  an "Official" error. When we use "Official" errors in statistical
                  analyses, we have to trust that these mistakes balance out.
                  (Incidentally, the net effect of these types of mistakes is to
                  decrease the size of the correlation between two variables; or make
                  it _less_ likely that we'll actually find relationships*).

                  The definition of race really depends on your purpose for measuring
                  it. I'm a social scientist, so I'm interested in the _social_ aspects
                  of race. In other words, I'm interested in other people's reactions
                  to Tiger Woods. If most people identify him (and react to him) as
                  Black, then he's Black in my database. In my HBP study, my theory was
                  that Blacks would be hit more often (not because they ARE Black, but
                  because they are PERCEIVED as Black by pitchers). Back to the
                  measurement issue: Most players could be easily identified as White,
                  Black, or Latino. On a few close calls, I may have identified someone
                  like Jeter as White, when most people would identify him as Black. I
                  may have identifed someone like Tino as Latino when others would say
                  he looks White. Just like with "official errors", however, I would
                  guess that these mistakes balance out when aggregated.

                  If we wanted to measure (perceived) race, the best way to do it would
                  be to have more than one person assign labels; then check for
                  disagreements; try and resolve them; and then label any further
                  disagreements as "unknown." I'd bet that, across the entire database,
                  the percentage agreement would be pretty high. I think this is the
                  procedure used in the Racial Report Card. We could also use other
                  sources (Hard Road to Glory by Ashe & Ashe has a list of Black
                  players; Beisbol by Oleksak and Oleksak has a list of Latino players
                  as does www.latinobaseball.com [which lists Tino Martinez as Latino]).

                  But, again, this assumes that the purpose of measurement is social in
                  nature. If someone were interested in some sort of biologically-based
                  performance differences, my measurement scheme would probably still
                  be the most practical way to do it, but there would also probably be
                  a greater amount of measurement error.

                  TT

                  *classes start tomorrow, so I'm trying to practice sounding smart.
                • tmasc
                  I second everything timmermant said.
                  Message 8 of 18 , Jan 9, 2002
                    I second everything timmermant said.
                  • Holmes, Dan
                    i think we are really getting way off subject as the debate spirals into a conversation about errors and real errors, etc. the point made (and i agree) is that
                    Message 9 of 18 , Jan 9, 2002
                      i think we are really getting way off subject as the debate spirals into a
                      conversation about errors and real errors, etc.

                      the point made (and i agree) is that the data would be BAD. not that it
                      wouldn't be politically-correct, but that it would be bad data. garbage in,
                      garbage out.

                      if every player filled out a form with the players union or MLB, and on that
                      form their was a check mark for Black, White, Asian, Latin, Other, and we
                      HAD that data - I'd accept it as something we should consider putting in the
                      database. But we don't have that.

                      the Hall of Fame has contract cards for most players prior to the 1980s, and
                      some data on ethnicity or heritage, but nothing that is available for every
                      player.

                      i can see the merit in doing research on this issue (i'd like to know if
                      black players have had as much success as ML pitchers, or is that a myth?),
                      but that would require too much conjecture, IMO.

                      -----Original Message-----
                      From: timmermant [mailto:timmermant@...]
                      Sent: Wednesday, January 09, 2002 2:43 PM
                      To: baseball-databank@yahoogroups.com
                      Subject: [baseball-databank] Re: The Race Thing


                      >An error is defined as a play which the official scorer calls an
                      error.

                      But what's the correlation between "officially called errors" and the
                      MLB definition of an error (An error shall be charged for each
                      misplay (fumble, muff or wild throw) which prolongs the time at bat
                      of a batter or which prolongs the life of a runner, or which permits
                      a runner to advance one or more bases.)? If it's not a perfect
                      correlation, then we have the same problem as the measurement of
                      race. It doesn't matter who does the measurement, it's still a
                      judgment call.

                      Now we're getting at the heart of the issue (and my own reservations
                      about including it in the database). What is the _purpose_ of
                      recording errors? It is supposed to be a measure of fielding screw-
                      ups. It is measured by the subjective judgment of the official
                      scorer. Most plays are clearly errors or clearly non-errors. There
                      are a few close plays; on these, the scorekeeper may occasionally
                      call a "MLB" error an "Official" non-error, or call a "MLB" non-error
                      an "Official" error. When we use "Official" errors in statistical
                      analyses, we have to trust that these mistakes balance out.
                      (Incidentally, the net effect of these types of mistakes is to
                      decrease the size of the correlation between two variables; or make
                      it _less_ likely that we'll actually find relationships*).

                      The definition of race really depends on your purpose for measuring
                      it. I'm a social scientist, so I'm interested in the _social_ aspects
                      of race. In other words, I'm interested in other people's reactions
                      to Tiger Woods. If most people identify him (and react to him) as
                      Black, then he's Black in my database. In my HBP study, my theory was
                      that Blacks would be hit more often (not because they ARE Black, but
                      because they are PERCEIVED as Black by pitchers). Back to the
                      measurement issue: Most players could be easily identified as White,
                      Black, or Latino. On a few close calls, I may have identified someone
                      like Jeter as White, when most people would identify him as Black. I
                      may have identifed someone like Tino as Latino when others would say
                      he looks White. Just like with "official errors", however, I would
                      guess that these mistakes balance out when aggregated.

                      If we wanted to measure (perceived) race, the best way to do it would
                      be to have more than one person assign labels; then check for
                      disagreements; try and resolve them; and then label any further
                      disagreements as "unknown." I'd bet that, across the entire database,
                      the percentage agreement would be pretty high. I think this is the
                      procedure used in the Racial Report Card. We could also use other
                      sources (Hard Road to Glory by Ashe & Ashe has a list of Black
                      players; Beisbol by Oleksak and Oleksak has a list of Latino players
                      as does www.latinobaseball.com [which lists Tino Martinez as Latino]).

                      But, again, this assumes that the purpose of measurement is social in
                      nature. If someone were interested in some sort of biologically-based
                      performance differences, my measurement scheme would probably still
                      be the most practical way to do it, but there would also probably be
                      a greater amount of measurement error.

                      TT

                      *classes start tomorrow, so I'm trying to practice sounding smart.




                      http://www.baseball-databank.org/

                      To unsubscribe from this group, send an email to:
                      baseball-databank-unsubscribe@yahoogroups.com



                      Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
                    • tmasc
                      The data wouldn t be bad . We d get it right in 95% of the cases, at least. I like timmer s viewpoint: is this guy PERCEIVED as black? . After all, these
                      Message 10 of 18 , Jan 9, 2002
                        The data wouldn't be "bad". We'd get it right in 95% of the cases,
                        at least.

                        I like timmer's viewpoint: "is this guy PERCEIVED as black?". After
                        all, these are the some questions I am trying to answer. "Would this
                        player be allowed to play in MLB prior to 1947?" So, make this a
                        yes/no field. Would Jeter have been allowed to play? I don't know,
                        I'm not an historian. But that's one way to do this.
                      • rockymtnsabr@aol.com
                        ... We have nationality data, which serves my purposes just fine, although POB info can be deceptive in some cases, ie Danny Graves. I can deal with that. The
                        Message 11 of 18 , Jan 13, 2002
                          In a message dated 1/13/02, baseball-databank@yahoogroups.com writes:

                          a race table could increase the likelihood that future studies are all using the same
                          labels. Therefore if these future studies yield different results, it couldn't be because of the differences in the way that the researchers measured race.  
                           

                          We have nationality data, which serves my purposes just fine, although POB info can be deceptive in some cases, ie Danny Graves. I can deal with that.   The five groups in Billet and Formwalt's study don't include any designation for the multitude of Asian ballplayers, and in order to avoid WWIII, I would recommend not lumping them all into a single classification.   And let's not forget Australia which is already represented in the MLB population, but we'll want to account for the Anglos and the Aborigines.  This just in: a well placed source in international scouting tells informs me that the next great untapped source of baseball talent is presently being scouted in Africa, so don't forget to account for that, noting tribal origins and dialect, of course.  And surely you're aware of the amazing talent on display annually at the Midnight Sun Tournament in the Alaskan League.  Is it correct to consider Aleuts as Native Americans or Asians or Russians?  There's that whole Bering Strait land mass theory...   Alaska Baseball League -- Midnight Sun Game http://www.geocities.com/shoeless_60067/alaska.html

                          Q. Was Scott Gomez the first Latin American or Alaskan to make the NHL? 
                          A. Neither!  He's the 7314th American to make the NHL...  Enough already....

                          But we haven't even begun to consider the variations according to religious persuasion, although we've all seen lists of Jewish ballplayers, which are Orthodox and which are not?  (Maybe we should study the effects of circumcision on OPS.)   Of all the great Irish ballplayers in the 19th century, which were of Catholic and which were Protestant descent?  There's a place to start, and course we all know, it will never end... Interested in more contemporary demographic research?  Well, regarding the HBP study, what portion of the black-on-black violence can be attributed to Crips and/or Bloods affiliations?  And how do we handle the black kid born a Baptist, which has adopted Islam??  And I don't even want to think what we'll do if some scout finds a Saudi al Quayda rebel being incarcerated Guadalcanal who developed a wicked knuckleball hurling grenades in Afghanistan...

                          And just as I was about to close, I almost forgot.... Canadians..  What about the Canucks?  We always forget the Canucks.. Of course, there's the Quebecois and..

                          My two cents: there are really only two kinds of people in the world...

                          Those that separate things into two categories and those that don't.

                          Rod Nelson  - still trying to identify the ever elusive PTBNL
                        Your message has been successfully submitted and would be delivered to recipients shortly.