http://www.sabernomics.com/sabernomics/index.php/2008/09/my-response-to-the-mitchell-report-study-response/ Economic Thinking about Baseball Sun, 09 Jan 2011 17:16:18 +0000 hourly 1 http://wordpress.org/?v=3.3.1 http://www.sabernomics.com/sabernomics/index.php/2008/09/my-response-to-the-mitchell-report-study-response/comment-page-1/#comment-104591 Millsy Mon, 08 Sep 2008 21:13:34 +0000 http://www.sabernomics.com/sabernomics/?p=971#comment-104591 I apologize for the addendum to the post, but after more thought, it seems that there could be real value in a tightly defined population such as this (as the title states specifically ''Mitchell Report Batters").  A question to ask of this group may be ''what were the gains of taking steroids given the ultimate mention of a given player in the report".  Though the study attempts to calculate this in definite performance increases based on steroid use, what about the actual player 'salary' increases (we may assume these are a function of one another), and ultimately, the detrimental effects on a player's well being (be it salary, career length, sponsorship, etc.) from being mentioned <em>in the Mitchell Report?  </em>Difficulties arise, however, based on the fact that many players had retired or become irrelevant before the release of the Mitchell Report...but I suspect there may be something there?  Anyone have ideas on this? I apologize for the addendum to the post, but after more thought, it seems that there could be real value in a tightly defined population such as this (as the title states specifically ”Mitchell Report Batters”).  A question to ask of this group may be ”what were the gains of taking steroids given the ultimate mention of a given player in the report”.  Though the study attempts to calculate this in definite performance increases based on steroid use, what about the actual player ‘salary’ increases (we may assume these are a function of one another), and ultimately, the detrimental effects on a player’s well being (be it salary, career length, sponsorship, etc.) from being mentioned in the Mitchell Report?  Difficulties arise, however, based on the fact that many players had retired or become irrelevant before the release of the Mitchell Report…but I suspect there may be something there?  Anyone have ideas on this?

]]> http://www.sabernomics.com/sabernomics/index.php/2008/09/my-response-to-the-mitchell-report-study-response/comment-page-1/#comment-104585 Millsy Mon, 08 Sep 2008 19:54:57 +0000 http://www.sabernomics.com/sabernomics/?p=971#comment-104585 I understand the differences in opinion on coding...and possibly the worries of how the coding was done by the researchers.  I have only briefly read the study, so I can't really comment on this.  However, putting aside the flip-flopping on the p-value issue, it seems that the problem is differences in defining the population.  Since statistics are readily available for all players (more or less) for the time period of 1995-2007, wouldn't this analysis be taking into account an entire population (as defined by all players playing within this time period with 50+ at bats).  In this case, we have all needed information on the actual population, as well as all needed information on the population of "those players mentioned in the report."  It seems that this would be a true difference between the ''mitchell report population" vs.  the entire major leagues.  I think that this makes the data fragile in its generalizations, as not all steroid users (most likely) were mentioned within the report...leaving for something to be desired in the definition of a population of steroid users.  <em>Please correct me if I am totally missing something here</em>...my point is that the real difficulty in studies such as this (or any study implicating steroid use and performance) is agreeing on who took steroids and who did not. I understand the differences in opinion on coding…and possibly the worries of how the coding was done by the researchers.  I have only briefly read the study, so I can’t really comment on this.  However, putting aside the flip-flopping on the p-value issue, it seems that the problem is differences in defining the population.  Since statistics are readily available for all players (more or less) for the time period of 1995-2007, wouldn’t this analysis be taking into account an entire population (as defined by all players playing within this time period with 50+ at bats).  In this case, we have all needed information on the actual population, as well as all needed information on the population of “those players mentioned in the report.”  It seems that this would be a true difference between the ”mitchell report population” vs.  the entire major leagues.  I think that this makes the data fragile in its generalizations, as not all steroid users (most likely) were mentioned within the report…leaving for something to be desired in the definition of a population of steroid users.  Please correct me if I am totally missing something here…my point is that the real difficulty in studies such as this (or any study implicating steroid use and performance) is agreeing on who took steroids and who did not.

]]>