If you go through all the efforts to randomize participants and groups, then why even consider using a pretest? What statistical technique would be helpful when you would like to equalize groups on...

If you go through all the efforts to randomize participants and groups, then why even consider using a pretest?

What statistical technique would be helpful when you would like to equalize groups on Variable A, which is related to the dependent variable, but are unable to use the matching technique?