[jim-greer]

You need to do the stats per ride, not per rider. If you are biking to work everyday in San Francisco, that helmet is much more likely to save your head than it is for the average weekend rider.

[Prrometheus]

Yep, and the site didn't provide those stats, so all its data is essentially useless for ascertaining the safety effects of wearing a helmet.

[divia]

This page has a ton of stats: http://www.bhsi.org/stats.htm

[Prrometheus]

What I'd really like, though, is deaths per mile with and without a helmet.