<p>“Your analysis assumes that each $1 has equal value.”
Most economics works this way.</p>
<p>“If you didn’t work, you couldn’t feed yourself, and you would die anyway.”</p>
<p>I think you’re missing the point I’m making. The point is that when you try to introduce reasons such as “blank is worth infinitely much to me, and therefore I must to whatever is possible to ensure I get blank” in order to justify your logic, you actually create logical conundrums like this.</p>
<p>In other words, <em>you</em> are the one who introduced such silly reasoning into the discussion, not me. </p>
<p>“The answer to the dilemma is when people “think” they have some effect on the odds. I can’t control if you will shoot me, but I can keep myself from getting hit by a car.”</p>
<p>The answer to the dilemma is for each person to estimate, or guess wildly, about what the odds of specific things occurring are, to analyze the expected value of a certain course of action, and act accordingly.</p>
<p>It may seem like I’m agreeing with <em>you</em>, but this is exactly what <em>I</em> said to do. And you never have perfect control over a situation… you can prevent me from shooting you, to an extent, and you can’t prevent yourself from getting by a car, to a certain extent. There is inherent uncertainty in everything you could possibly bet on.</p>
<p>When you have imperfect information, you have to make (educated?) guesses and act accordingly. How people act is a reflection on how they weigh various risks. If you’re 100% sure that nothing will happen to you, insurance is a total waste of money. If you’ree 100% sure something will happen to you, you’d be an idiot not to have insurance.</p>
<p>If you’re 99% sure nothing will happen to you, well, then you need to start calculating the expected value of the <em>investment</em> which is insurance. Say insurance costs $1000 per year and you’re 99% sure that in any given year nothing will happen to you. Well, the expected value of your insurance is (payout)(0.01) + (0)(0.99). Assume payout is $1,000,000. Then your expected payout is $10,000, and you should get the insurance. If, on the other hand, the payout is more reasonable… say, $100,000, then it’s a wash as the expected value is $1,000. If you expect that any insurance costs you incur will certainly be under $100,000 (99% sure that you won’t have any claims, 1% sure that you will have a $100,000 claim) then you shouldn’t get the insurance.</p>
<p>A more realistic model would involve varied probabilities of incurring such and such amount of insurance payout. For instance, maybe…</p>
<p>50% chance of paying $0
25% chance of paying $100
20% chance of paying $1000
5% chance of paying $10,000</p>
<p>Then the expected value is $25+$200+$500 = $725, and you’d still be a sucker to pay for insurance.</p>
<p>Now, the trick is coming up with the right percentages. To do this, you can use available statistics about this sort of thing, you can use your past history as an indicator, and you can round up in places just to be sure.</p>
<p>But the idea is right, and all that’s left to worry about is finding the percentages. It’s a very mathematical exercise which can (and does) have huge benefits for those who know how to exploit it.</p>