Thanks - I still like the saying, though I'm not sure how to correct it now.

To me multiple anecdotes signify a small uncontrolled, non random sample set, that you really shouldn't use to draw conclusions.