Before you say anything - I'm aware that a study is fair by plucking randoms of certain groups - but it would answer the question I've had rammed down my throat for 11 pages of why I wear a helmet whilst cycling and not walking if statistically, healthy, smart, sober, young, non-drug taking able-bodied mid-twenties men are safer walking down the road than those stats suggest.
Your questions about stats have been answered several times, it is just that you have not read the answers!
Your determination to abuse them and select the safest possible pedestrian for comparison with the average cyclist is laughable, but then again if that is what you need to support your beliefs then fine.
It is called "Bias" in peer reviewed literature and would have any paper submitted thrown out from any respectable publication.
The original studies (as pointed out several times before) take in all the cohort (usually hospital admissions whether they be good, bad, sober, drunk, trained, untrained, old, young etc.
You want to cherry pick and find a nice safe low risk pedestrian, yet hypocritically don't do the same and compare with a nice safe low risk cyclist.
Lets take the above example... do you or do you not agree with the following:
Why wear a cycle helmet if statistically, healthy, smart, sober, young, non-drug taking able-bodied mid-twenties men are safer cycling down the road than those stats suggest.
This is an equivocal statement to yours. Surely if a pedestrian and cyclist are both safer than average then neither would need to wear a helmet, or if there is still a risk then both should wear one.
Now the big question is.... do you now understand what bias is and why choosing truly comparative cohorts is necessary?