Saturday, December 28, 2013

Hall Of Fame Voting And Race

This is based on an earlier post called My Predictions For The Hall Of Fame Vote. You need to go there to understand the variables and the model.

Here I simply added a single dummy variable, Nonwhite (1 for non-whites, 0 for whites). This is a very simple way to do it. It does not take into account the slopes of the other variables regarding race. And it just lumps blacks and Hispanics into one group.

Anyway, the results were about the same in terms of r-squared and the coefficient values. Below are the new coefficient values. Notice that Nonwhite, although negative, has an extremely small value (it also was not statistically significant with a t-value of -.125). It means that being non-white would lower your predicted vote by .136 percentage points. So where a white guy would get 50.136%, a non-white would get 50%.

Variable            b Coeff.  
    Constant        -0.01223  
   WSAS/1000         0.00112  
   GGAS/1000         0.06243  
         MVP         0.04821  
    3000 HIT         0.45176  
      500 HR         0.16757  
      ASSQ10         0.00216  
       GGSQ7        -0.00122  
       500SB         0.04949  
WSIMPSQ50/10        -0.01188  
     10000PA         0.09924  
    Nonwhite        -0.00136  

No comments: