Originally Posted by RobertHurst

It is hardly just this data set which refutes Forester's claims about the magical safety properties of rule-following...



I looked at Failures-to-Yield alone because they are unambiguous in terms of fault, which party caused the wreck. Many of the crash types listed are ambiguous in terms of fault.

I'm pretty sure a complete crunching of the numbers, looking at all crash types and determining at-fault for each, would reveal the same phenomenon: kids are different and they blow up the stats.

We need to stop pretending that looking at statistics for the entire bicycling population gives meaningful results, even if it makes our pet traffic safety dogma look good.