Political Theory

Re: Defence in anarchy

Malachi — Tue, 07 May 2013 00:07:22 GMT

I can understand why you might want this discussion to disppear down the memory hole.

Re: Defence in anarchy

ToxicAssets — Mon, 06 May 2013 09:17:18 GMT

I've split the debate about rational expectations to another thread, you might want to check it out:

http://mises.org/community/forums/t/33439.aspx

Re: Defence in anarchy

Malachi — Sun, 05 May 2013 14:12:58 GMT

It would be cool if a moderator reading this could split this whole discussion of expectation and probability to another thread because it's kinda of a side issue here,

actually rational expectations is central to questions of defense.

and it's an interesting point that many people get confused about, like malachi.

do you even know what mathematical expectation is?

Re: Defence in anarchy

Malachi — Sun, 05 May 2013 14:10:17 GMT

The whole point with rational expectations theory is that the decision maker will take a gamble if it maximizes his expected utility, and not necesseraly his expected money gains.

My statement is still meta to yours. if you reduce the set of all opportunities to one, and the marginal unit becomes $50k, and you only give the guy the choice between no mathematical expectation, and death, or negative mathematical expectation and the minor chance of living, then he is going to choose the least worst option. whats your point?

that will have positive expectation in utility terms.

the negative mathematical expectation still mens that people who make this decision will suffer a loss over the long run, which means they will be outcompeted by people who make bets with positive mathematical expectation.

Re: Defence in anarchy

ToxicAssets — Sun, 05 May 2013 14:10:07 GMT

It would be cool if a moderator reading this could split this whole discussion of expectation and probability to another thread because it's kinda of a side issue here, and it's an interesting point that many people get confused about, like malachi.

It could start at my post just after the "markmerich" post (which by the way is a scam-bot).

Re: Defence in anarchy

ToxicAssets — Sun, 05 May 2013 14:04:27 GMT

The whole point with rational expectations theory is that the decision maker will take a gamble if it maximizes his expected utility, and not necesseraly his expected money gains.

If his utility function is convex, he'll take gambles with negative expected monetary returns, that will have positive expectation in utility terms.

And this is the case of certain people facing urgent problems.

Re: Defence in anarchy

ToxicAssets — Sun, 05 May 2013 14:00:15 GMT

Malachi:

I think we are done here, unless you have something else to say.

I see,

The mathematically savvy reader probably has figured it out already, but the last post was a bit too technical, with convexity and utility functions. When I have more time in my hands I'll comeback and make a more detailed explanation for lay people and you will understand what's going on.

Don't give up though, you're showing some progress...

Re: Defence in anarchy

Malachi — Sun, 05 May 2013 13:52:22 GMT

I think we are done here, unless you have something else to say.

Re: Defence in anarchy

ToxicAssets — Sun, 05 May 2013 13:50:16 GMT

rational actors consider alternative courses of action. the entire point of the odds is to compare different courses of action and their likely effects. the entire point of our discussion is a comparison of different courses of action, and you concede that for a rational actor it comes down to what means is right for his ends, and you suddenly dont want to compare means?

It's not that I don't want to compare means, it's that I don't want to overcomplicate the scenario unnecessarily.

My point was only about someone contemplating the decision of entering a game of roulette or not entering a game of roulette.

And if this guy has some urgent need for fast cash, and if he has no alternative way of raising cash at a lower risk, it might be rational for him to enter.

Even if his expectation of wining in the game is negative (like all roulette games), he's not better of if he doesn't play.

Say he's in debt with mobsters, for 10 grand, and he has only 5. Say the mobsters will kill him if he shows up with less than the 10.

If he puts 5 on red he has a probability of slightly less than 50% of making 10 grand, and slightly more than 50% of loosing it all.

Say the only decision available is to play or not to play. He doesn't have a lot of options, he doesn't know how to play the other games nor how to raise 5 grand overnight by any other way.

In this scenario, he has interest in playing the game.

His situation is atypical, he has a convex utility function. He values having 10 grand much more than twice the value he gives for having only 5, which would be the case if he didn't play the game.

If he wins, he can pay the mobsters and live. If he loses, he's done. And that's the same result if he'll get if he doesn't play the game.

So the gamble is justified, in terms of expected utility, but not of expected money figure.

Re: Defence in anarchy

Malachi — Sun, 05 May 2013 13:45:22 GMT

just to clarify, if someone needs $50k or they will suffer unfortunate consequences, the marginal unit is $50k/{all opportunity spaces}, that is "Fifty thousand dollars divided by the set of all opportunity spaces" meaning that if someone has fifty opportunities to play a game, he would rationally only choose games that have a mathematical expectation equal to or greater than one thousand dollars per opportunity to bet. rationally speaking, a lesser mathematical expectation is not worth the opportunity cost.

Re: Defence in anarchy

Malachi — Sun, 05 May 2013 12:42:08 GMT

But you're making unnecessary over complications when you start considering other games.

rational actors consider alternative courses of action. the entire point of the odds is to compare different courses of action and their likely effects. the entire point of our discussion is a comparison of different courses of action, and you concede that for a rational actor it comes down to what means is right for his ends, and you suddenly dont want to compare means?

The point is how people rationally take decisions using the implied odds of that decision within their situations, and not the abstract generic odds such a decision might have per se, which give only a baseline for the computation of the implied odds.

this seems like an obtuse way of obfuscating the fact that a rational actor will use the mathematical expectation of roulette to compare its relative merits to other games of chance. obviously the mathematical expectation of a skill-based game will have a subject-dependent component (in the same way a professional cat burglar might have a positive expectation for a robbery that an amateur would have a negative expectation for).

Re: Defence in anarchy

ToxicAssets — Sun, 05 May 2013 12:38:53 GMT

*final betting rounds

Re: Defence in anarchy

ToxicAssets — Sun, 05 May 2013 12:33:21 GMT

I agree, the baseline odds of the roulette game are important when you're calculating your implied odds that determine the benefits of entering the game or not.

But you're making unnecessary over complications when you start considering other games.

I'm not saying that roulette will be necessarily the best solution for every guy in such a situation. If the guy happens to be a professional poker player, he's probably better of trying to win a smalltime poker tournament with lots of fish, if there's one available. Same thing if he can count cards in twenty-one without being caught.

But since most guys are probably fish, odds will be even more skewed against them when playing these "skill" games, so most guys in need of fast cash would probably opt for the roulette (or similar things, like crabs).

In any case, that's largely besides the point here.

The point is how people rationally take decisions using the implied odds of that decision within their situations, and not the abstract generic odds such a decision might have per se, which give only a baseline for the computation of the implied odds.

Poker players need to understand this implied odds when they calculate their calls and raises.

Some pot-odds will not look very rewarding if you have few possible outs in a hand, but these outs might give you a concealed nuts which will allow you to rip off the other guy in the last betting rounds.

Re: Defence in anarchy

Malachi — Sun, 05 May 2013 12:02:22 GMT

Correct, but it makes the decision of playing the roulette by that particular individual at that particular moment a rational decision.

The regular odds of the abstract game of roulette does not change, but the "implied odds" of that particular instance of roulette are favorable.

no, because the mathematical expectation of roulette is what is rationally used to compare roulette against other games. the need for $50,000 doesnt make it more beneficial to play a game with negative mathematical expectation, it just establishes the marginal unit at $50k/(time until deadline).

Re: Defence in anarchy

Malachi — Sun, 05 May 2013 11:59:12 GMT

ToxicAssets was kind enough to elaborate my two main points in his own words, so heres a brief summary of my position on this subject:

I've never said that overt violent actions don't have likely consequences, but the decision takers who rationally opt to engage in violence also know that, and do their cost-effectiveness analysis considering those likely consequences.

So, as you can see, this mathematical expectation thing is not wrong, but only after you have considered the specific opportunity costs people are facing when taking decisions.