On Tue, 18 Aug 2015 05:45:09 -0700, trader_4 wrote:

You have a logic problem if you believe that the above statement
means that estimation is the only issue with the data. The disclaimer
does not say that. It points out that the data are estimated and then
it says that year to year comparisons should be made with caution.

I think you have a problem with large numbers.

If the accident rates, given the tens of thousands of accidents yearly,
aren't changing, then it would take a stupendously stupifyingly
coincidental alignment of the stars to then make the accident rates
exactly cancel out the *entire effect* of millions upon millions
of cellphones being owned (and presumably used) by almost every
person of driving age in the United States.

That your *entire argument* is based on refuting yearly accident
rate figures based on a minor estimation detail, is unbelievable.

Do you realize how MANY cellphones there are owned by people in
the USA?

If those cellphones were being used, while driving, and if they were
causing accidents, no amount of fudging of the data would show what
the data actually shows.

There is a paradox, to be sure, but the answer is never going to be
found in the puny numbers associated with *estimation errors* that
you want it to.

You're grasping at straws if you truly feel that the *estimation
errors* exactly cancel out the absolutely stupendous effect we
presume cellphone ownership to have on accident rates, in both
timing and in number.

It's just not possible,and, it's a bit scary that you believe it
is. Does anyone else believe that the answer to the paradox is
simply that estimation errors have skyrocketed, and then plateaued
at exactly the same rate as cellphone ownership has?