On Wed, 20 Jan 2016 19:28:52 -0500, Ed Pawlowski wrote:

True, but the ones on phones make more mistakes.

If there is a 10% chance that any particular new mistake will
result in an accident, then there should be a corresponding
number of additional new accidents if people are making that
new mistake.

If there is only a 0.000001% chance that any particular new
mistake will result in an accident, then there should be
corresponding number of new accidents if people are making
that new mistake.

The answer as to which is which, is clearly shown in the
reliable overall accident rate data already.