Small case-"control" studies tend to predict significant benefit from
helmet use.  This is the sort of study that predicted "85% benefit."
Note, I put "control" in quotes because it's been shown that such
studies do NOT use proper "control."  In perhaps the most famous
example, when Dr. Dorothy Robinson examined the data set from the "85%"
study, she found that the helmets in that study also prevented
something like 75% of leg injuries.

Population-wide studies that ignore long term trends and give _all_
benefit to helmets tend to predict moderate benefit from helmet use.
IOW, some pro-helmet teams in New Zealand and Australia compared pre-
and post- helmet wearing data, but ignored the fact that
hospitalizations for head injuries had been falling gradually for
decades.  They purposely ignored that trend, and pretended all the
benefit happened at once, when helmet use was forced to jump.

Population-wide studies that ignore the drop in cycling when MHLs are
imposed show moderate benefit from helmet use.  The problem is, a 30%
drop in HI accompanied by a 35% drop in cycling fools only those who
don't know about the cycling reduction.  Still, many helmet promoters
will tout the drop in HI without mentioning it actually conceals a rise
in HI per remaining cyclist.

Population-wide studies that compute rate of HI per remaining rider
tend to show no benefit from helmet use.  In fact, they show a
detriment regarding cyclist safety.

Population-wide studies that compare cylcist data with pedestrian data
show no comparative benefit per cyclist when helmet use jumps.

Population-wide studies that take time trends into account show no
benefit from jumps in helmet use.

In other words, the studies which are least biased are those which show
least benefit.

Please keep in mind, this issue should _not_ be treated like religion -
"He says his is better, she says hers is better, and there's no way to
tell."  Instead, we are talking facts, data and analysis.  We _can_
analyze, look for strengths and weaknesses and reach logical
conclusions.

Most of the helmet skeptic people adopted that position after doing
exactly that.  I know I certainly did.

- Frank Krygowski