Some family traditions are more true than others. The problem is proving them. My usual approach for dealing with family traditions is to break them into the things that (if true) might have created a record or document and those things that probably did not leave any record.Then I focus on trying to prove or disprove the parts of the story that I can.
Often the family tradition is mostly false, but sometimes there is a grain of truth in there somewhere. I find that breaking it up into "possibly provable" and "probably not provable" a good way to start working on the tradition.
There is an old article on the Ancestry.com site I wrote on this topic.