Some runs with the NCAA data.

> fsr(x=ncaa2[,1:19],y=ncaa2[,20],B=500,seed=278)->out.fsr
   var   pval  pvmax
1    2 0.0000 0.0000
2    3 0.0001 0.0001
3    5 0.0116 0.0116
4    4 0.0053 0.0116
5    7 0.0025 0.0116
6   17 0.0433 0.0433
7   15 0.0527 0.0527
8    6 0.1056 0.1056
9    9 0.0826 0.1056
10   8 0.0536 0.1056
11  12 0.2350 0.2350
12  10 0.2864 0.2864
13  13 0.3163 0.3163
14  18 0.2697 0.3163
15  11 0.4953 0.4953
16   1 0.6326 0.6326
17  14 0.7056 0.7056
18  19 0.8605 0.8605
19  16 0.9032 0.9032
   m  n gam0 amax   B seed alphahat.RE alphahat.ER
1 19 94 0.05  0.3 500  278       0.034       0.034

Variables Selected for alphahat.ER = x2 x3 x5 x4 x7


To illustrate the possible variation, 8 more runs:

   m  n gam0 amax   B seed alphahat.RE alphahat.ER
1 19 94 0.05  0.3 500  279       0.045       0.039

   m  n gam0 amax   B seed alphahat.RE alphahat.ER
1 19 94 0.05  0.3 500  280       0.046       0.045

   m  n gam0 amax   B seed alphahat.RE alphahat.ER
1 19 94 0.05  0.3 500  281       0.047       0.046

   m  n gam0 amax   B seed alphahat.RE alphahat.ER
1 19 94 0.05  0.3 500  282       0.053       0.053

   m  n gam0 amax   B seed alphahat.RE alphahat.ER
1 19 94 0.05  0.3 500  283       0.053       0.053

   m  n gam0 amax   B seed alphahat.RE alphahat.ER
1 19 94 0.05  0.3 500  284       0.055       0.055

   m  n gam0 amax   B seed alphahat.RE alphahat.ER
1 19 94 0.05  0.3 500  285       0.035       0.035

   m  n gam0 amax   B seed alphahat.RE alphahat.ER
1 19 94 0.05  0.3 500  286       0.053       0.053

                          av. =  0.047       0.046 
                             (se=.003)     (se=.003)

A few B=5000 runs:

   m  n gam0 amax    B seed alphahat.RE alphahat.ER
1 19 94 0.05  0.3 5000  287       0.044       0.038

   m  n gam0 amax    B seed alphahat.RE alphahat.ER
1 19 94 0.05  0.3 5000  288       0.045       0.044

   m  n gam0 amax    B seed alphahat.RE alphahat.ER
1 19 94 0.05  0.3 5000  289       0.044       0.037

   m  n gam0 amax    B seed alphahat.RE alphahat.ER
1 19 94 0.05  0.3 5000  290       0.045       0.044


Finally, a B=50000 run:

   m  n gam0 amax     B seed alphahat.RE alphahat.ER
1 19 94 0.05  0.3 50000  291       0.045       0.044

This data set is pretty sensitive to alpha-to-enter, probably 
because of the non-monotone p-values to enter.  Typically, B=500 
would be large enough, B=5000 to be sure.  Here, even after B=50000
it's not totally clear whether variable 17 should be included
or not.  The results above support the model: x2,x3,x5,x4,x7,x17
using alpha to enter=.044-.046.