I agree that if I have to guess who's right about this, Henry17 and wild will or the company that is investing many millions of dollars in it and has studied accident rates for decades because it's their business and they employ hundreds of actuaries whose job it is to work these things out, I'm going with Progressive. Not 100% by any means — there are a lot of stupid companies that have made a lot of mistakes about a lot of things — but that's where I'd put my money.

Win-win transactions are not only possible but describe virtually all voluntary transactions.

If you know your behavior can reduce your rates you'll be more likely to behave safely. That lets the company offer lower rates and incur less costs than their competition.

But that's just in theory. In reality driving, insurance and government are terrible and I avoid them all.

This gives me doubt as to your ability to recognize safe driving habits.

Insurance models have limitations -- there is an obvious difference between speeding in a residential area, speeding on a highway within city limits, and speeding on a highway in the middle of nowhere at 4am when it is basically empty yet insurance models treat all of these activities in exactly the same way.

In my experience a lot of near accidents are caused by people who are not comfortable driving but the official fault would be assigned to someone else. My city seems obsessed with introducing roundabouts which I find work great in Europe but no one here knows how to use them. Stopping on on-ramps because a nit freaks out and can't merge with traffic is also common as is pulling onto the highway while going about half the speed of the traffic you are merging into. Something I also run into a lot are people who go to the end of a lane that ends and they can get ahead by just speeding up and merge and you yield to them but they instead slow down because even though you are clearly letting them in they are too afraid to take the opportunity. Eventually you conclude they are not going to go and return to your normal speed only to have them now decide to pull out into your lane. All of these behaviours require good drivers to react and if in that defensive reaction they cause an accident it will be their fault even though it was the nit's behaviour that created the situation.

There is also the assumption that insurance models are designed to model accident occurrence rather than maximize profits. It doesn't matter if X doesn't actually correlate with an increase in claims so long as X can be used to justify an increase in rates. Take for example lending your car to an insured driver who has an accident -- at least in my jurisdiction that does not lead to an increase in my insurance rates -- his do go up. Now say in three years time that individual decides to not have insurance for whatever reason -- all of a sudden my insurance rate will be higher. How exactly does an individual-- who I may not even know at this point-- making the decision to not have insurance impact the probability that I will make a claim?

You're doing it wrong. Do not slow down to allow them to merge. Please maintain your speed.

The merging vehicle adjusts their speed to match the speed of traffic. If you change your speed to allow them to merge, you make it more difficult.

Also, would not allow insurance company to put tracking device on my vehicle. Creepy.

You don't really have a choice. The way this usually develops is that they have plenty of room to merge and they start to but then abort and go back into their lane. Meanwhile you'd slowed down in anticipation of them cutting in front at about half your speed because you know the lane ends and you saw them moving over. If you choose to not slow down and they end up not aborting you'll have to slam on the brakes because they are going so slow.

Very little in your car insurance payment is about how you drive or where you put your car when you are not driving it.

Virtually all of it is

1. Experience (which they base overwhelmingly on age)
2. How much you drive (more miles more likelihood of a claim)
3. Where you drive (more densely populated more likelihood of a claim)
4. Traffic violations (OK this is a slight nod to how you drive but they don't really care if you regularly drive 40 in a 30 limit or 30 because statistically the difference in claims made is virtually zero)

Having talked to the people designing the system, I can confirm this. All they're doing is taking the old models and data sets and giving themselves an advantage of a gigantic pile of more data with which to determine current and future individual and estimated new customer cost to the company. They can then use their (in theory way better than industry standard because of their data advantage) model of these costs to price more effectively and make more money. What their precise strategy about which segment, if any, of the driving population they really have the goal of dominating, I am not privy to.

Now that that's out of the way, I've got to add that neither anecdotally nor from any data I've ever seen is Henry correct in his statements about aggressive drivers.

I can confirm this as well. Information is incredibly valuable in our industry. The amount of metrics that are taken into calculation by a modern insurance company in regards to setting a specific individuals rates is baffling. It goes waaaaaaaay beyond the 5 items that were posted a couple replies above.

Ultimately the goal is to use the information gathered to set rates correctly.

as a fellow AC'er... i disagree in regards to insurance

unless you are talking about health insurance... but that's the govt's fault as you already know

How is 12am to 4am high risk? More drunk or tired people out on the roads? I'm of the impression that on empty roads, you're less likely to hit someone, tailgate, etc.

It's obviously drunks, and tired people. Drunkedness has to increase a hundred fold in those hours.

I think driving tired is worse imo.