[Yxven]

Your speculation is mostly right. Our brains are pretty good at automating repetitive tasks such as driving, so most drivers regularly don't pay complete attention.

The trouble is our brains primarily adapt to repeated stimuli eg driving in good conditions. Since no one spends most of their time driving narrowly avoiding accidents, that activity isn't one the brain can learn to automate ( well, maybe it's possible with a realistic driving simulator), so dangerous driving requires everyone's full attention.

Since we don't usually have advanced warning of when dangerous driving will happen, the best we can do is spend as much time as possible paying attention to road conditions.

Studies have found hands-free phones are just as dangerous. http://www.cbsnews.com/news/hands-free-phones-just-as-risky-...

[mrexroad]

this may get me down-voted, but i tend to drive faster and more aggressively than, uh say, my parent's would like me to [1]. when i'm driving this way, i'm fully engaged and almost always "in the zone." i constantly work towards a better position and continuously work to predict actions of all cars around me and the consequent reactions of other cars to that action, as well as any reactions i'd make at any point. my mental model of traffic is largely based on fluid dynamics, w/ hooke's law (springs) thrown in for stop/go traffic. i often joke w/ my wife that if there were an olympic event for navigating through traffic, i'd at least qualify for the team. while risks are greater than going 65 in a 65 and not changing lanes, i'm never ever reckless. some may disagree, but b/c of the heightened level of engagement, i do consider my self far safer than other cars on the road who commute home, thinking about work and largely disengaged. day dreaming? can't. texting, etc? cant. drinking coffee? can't. fwiw, i've always driven a car with a manual transmission, and i do feel it's a variable in lower levels of driver disengagement. i've never had an accident or ticket while driving in this manner.

driver disengagement is the problem, whether caused by a physical, visual, or cognitive distraction. the goal shouldn't be weighting how dangerous these various devices are, it should be on how to ensure drivers are engaged to a point where they can take over from auto-pilot (repetitive aspect) when something goes wrong.

[1] i'm not a kid, i have many kids of my own, just trying to avoid quantification of driving.