The previous work, Accidents in North American Mountaineering, did not answer the question of how much risk there is in climbing. To measure the risk, we need to know the number of people climbing in addition to the number of accidents.
The American Alpine Club provides an estimate for the number of climbing accidents, but getting a reasonable estimate for the number of climbers is more difficult. As a first approximation, I used the number of 'tics' per state in the MountainProject.com data as a measure of the relative amount of climbing occuring in each state or Canada. I then scaled those numbers by an estimate for the total number of actual climbers in the US provided by the Outdoor Foundation. This provided a rough (very rough!) estimate of the number of climbers in each state. So then if we divide the number of accidents in each state by the number of climbers we get an estimate for the accident rate in each state. Of course, there are biases and innacuracies in both the accident and participation numbers, but it seemed like a reasonable starting point.