Statistics for An Alternative Subset Approach to the Ranking and Selection Problem