I wouldn't say MMM's strategy is baseless, or dumb. In fact, from a statistical standpoint, it probably makes sense.

At least conceptually, your insurance company is aware of the odds of loss. They annualized those odds times the average lawsuit size to get an annual "loss." Then they add their administrative expenses, and add a projected profit margin to get your rate. They take that rate and invest it, creating returns greater than the "annualized" loss, and presto they make money. If you can eliminate their administrative expenses and projected profit margin, that's money in your pocket.

Let me start off by pulling some numbers out of my a**.

Looking at my state, NC has a population of approximately 10,150,000.

According to the US census bureau from 2010, the average US household size was 2.58. That makes 3,934,108 households in NC.

I would estimate there are about 100 liability "lawsuits" or other claims made per county in NC per year (total number out of my a**, but a place to start).

There are 100 counties in NC.

Which makes 10,000 "lawsuits" or other claims made in NC per year.

If you own a home in NC, that means you have a 0.254% chance that this year you'll have a claim made against you (10,000/3,934,108).

Each of these claims will probably run anywhere between $1,000 (pay an attorney to tell the other side to p*ss off) to $100,000 (for considerable medical or property damage). But lets say it averages out to $20,000 per incident (in both attorneys fees on your end and medical bills).

Annualized, that makes $50.83 in liability per household per year ($20,000*0.254%).

Now that doesn't include the million dollar weirdo lawsuit (where you find someone with an Eggshell Skull (

https://en.wikipedia.org/wiki/Eggshell_skull)). Lets say one of those happens per year in NC. That makes your odds of liability 0.00002%, or annualized at $0.20 per year ($1,000,000*.00002%).

So total annualized liability equals $51.03. If you're paying more than that, statistically speaking, you're overpaying for insurance. My annual premium is about $500.

Now that doesn't count property damage to your home (hail, tree falls) or your neighbor's home (your tree falls on your neighbor's home) but you get the idea. And yes, all those numbers are completely made up, but you can play with them as you see fit to see where you think it should be.

Essentially though, MMM is calculating is projected liability as:

A*B*C*D

Where:

A = statistical odds that he'll be negligent

B = statistical odds that he'll invite a "dickhead" into his home

C = statistical odds that his attorney won't get him out of it anyway

D = statistical odds that the judge or jury won't protect him

MMM assumes that C and D are fairly low odds. He also believes that, based on his own actions, A and B are fairly low too. Which means his projected liability is less than average. How much less? I don't know, maybe half or a quarter? But if average liability is equal to $51.03 per year, and his is a third of it, his projected liability is $17.01.

But that's all playing the odds and statistics. Statistically speaking, more sex is safer sex

https://www.amazon.com/More-Sex-Safer-Unconventional-Economics/dp/1416532226 which we all know doesn't exactly work quite right.

Having seen horror stories of "dickheads" wondering onto your property (to sell bibles or some nonsense), horror stories of negligence by fairly astute people, horror stories of judges or juries that screw someone's life up, I'll pay my $500 per year and not worry about it. Based on the 4% rule, I need $12,500 in the bank to pay my insurance, which is less than the average "claim" incident based on my math. So I'm happy with it, knowing if I have one incident in my life I've saved money by having the insurance. But that's me.