I do wonder how much to trust averages on these statistics. I observe that I am much more risk averse than the average cyclist in my city. Perhaps my risk is really much lower, conditional on that knowledge?