Maybe this is not possible to quantize accurately, but might it be the case that
Q. Approximately 2% of those who view a posting on MO vote +/- on the question?
That's approximately true for my latest, "How to explain the concentration-of-measure phenomenon intuitively?," and I think (without careful data) consistent with experience.