When I run new samples I use the method of trial and error to get the perfect column to separate the 2 enantiomers of a compound(for racemic case).How does one narrow in on a range of columns which are best fit for that particular sample?Is there a thumb rule which can provide this information with even a fair degree of accuracy?