It's immensely frustrating when someone is discussing a well conducted study showing an interesting result to have another person say "This feels wrong. I had $ANECDOTE, so maybe they researchers are missing $CONCLUSION_FROM_ANECDOTE?", which has been covered in the study and in the HN comments and is just a weird point anyway.
This is true, but there is also the reverse situation - sometimes there is no hard data, in which anecdotes are the nearest thing that we have. I can't help feeling that "the plural of anecdote is not data" should require a link to the relevant data that the parent post had apparently missed. This would be an excellent rebuttal to the anecdote and advance the conversation more than the one-liner...
Most of the conclusions I find in comments on HN, are rarely conclusive. I've always struggled with comments that provide one data point (or better yet, a generalization of data points) that the user declares a conclusion from. I can't tell if I'm just being pessimistic when I detract these poorly thought out conclusions or if they are genuinely good for the discussion.
Because comments probably aren't the best place to talk about conclusive anything: it's a discussion. If you've done the leg work to get something solid it can at least get the discussion it deserves if submitted to HN.
